Mono 108

SOME DRUGS AND
HERBAL PRODUCTS
VOLUME 108
IARC MONOGRAPHS
ON THE EVALUATION
OF CARCINOGENIC RISKS
TO HUMANS
SOME DRUGS AND
HERBAL PRODUCTS
VOLUME 108
This publication represents the views and expert

opinions of an IARC Working Group on the
Evaluation of Carcinogenic Risks to Humans,
which met in Lyon, 411 June 2013
Lyon, France - 2016
IARC MONOGRAPHS
ON THE EVALUATION
OF CARCINOGENIC RISKS
TO HUMANS
IARC MONOGRAPHS
In 1969, the International Agency for Research on Cancer (IARC) initiated a programme on the evaluation of the
carcinogenic risk of chemicals to humans involving the production of critically evaluated monographs on individual chemicals.
The programme was subsequently expanded to include evaluations of carcinogenic risks associated with exposures to complex
mixtures, lifestyle factors and biological and physical agents, as well as those in specific occupations. The objective of the
programme is to elaborate and publish in the form of monographs critical reviews of data on carcinogenicity for agents to
which humans are known to be exposed and on specific exposure situations; to evaluate these data in terms of human risk
with the help of international working groups of experts in carcinogenesis and related fields; and to indicate where additional
research efforts are needed. The lists of IARC evaluations are regularly updated and are available on the Internet at http://
monographs.iarc.fr/.
This programme has been supported since 1982 by Cooperative Agreement U01 CA33193 with the United States
National Cancer Institute, Department of Health and Human Services. Additional support has been provided since 1986 by
the European Commission Directorate-General for Employment, Social Affairs, and Inclusion, initially by the Unit of Health,
Safety and Hygiene at Work, and since 2014 by the European Union Programme for Employment and Social Innovation EaSI
(20142020) (for further information please consult: http://ec.europa.eu/social/easi). Support has also been provided since
1992 by the United States National Institute of Environmental Health Sciences, Department of Health and Human Services.
The contents of this volume are solely the responsibility of the Working Group and do not necessarily represent the official
views of the United States National Cancer Institute, the United States National Institute of Environmental Health Sciences, the
United States Department of Health and Human Services, or the European Commission.
Published by the International Agency for Research on Cancer,
150 cours Albert Thomas, 69372 Lyon Cedex 08, France

International Agency for Research on Cancer, 2016
On-line publication, 15 September 2015 (see Corrigenda)
Distributed by WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland
(tel.: +41 22 791 3264; fax: +41 22 791 4857; email: bookorders@who.int).
Publications of the World Health Organization enjoy copyright protection in accordance with the provisions
of Protocol 2 of the Universal Copyright Convention. All rights reserved.
Corrigenda to the IARC Monographs are published online at http://monographs.iarc.fr/ENG/Publications/corrigenda.php
To report an error, please contact: editimo@iarc.fr

Co-funded by the European Union
The International Agency for Research on Cancer welcomes requests for permission to reproduce or translate its
publications, in part or in full. Requests for permission to reproduce or translate IARC publications whether for sale or for
non-commercial distribution should be addressed to the IARC Communications Group at: publications@iarc.fr.
The designations employed and the presentation of the material in this publication do not imply the expression of any
opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any country,
territory, city, or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.
The mention of specific companies or of certain manufacturers products does not imply that they are endorsed or
recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and
omissions excepted, the names of proprietary products are distinguished by initial capital letters.
The IARC Monographs Working Group alone is responsible for the views expressed in this publication.
IARC Library Cataloguing in Publication Data
Some drugs and herbal products / IARC Working Group on the Evaluation of Carcinogenic Risks to Humans
(2013: Lyon, France)
(IARC monographs on the evaluation of carcinogenic risks to humans ; volume 108)
1. Carcinogens pharmacology 2. Drug-Related Side Effects and Adverse Reactions drug effects 3. Plant Preparations
adverse effects 4. Pharmaceutical Preparations adverse effects 5. Plants, Medicinal adverse effects
6. Cell Transformation, Neoplastic drug effects
I. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans
II. Series
ISBN 978 92 832 0174 8 (NLM Classification: W1)

ISSN 1017-1606
CONTENTS
NOTE TO THE READER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
LIST OF PARTICIPANTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
PREAMBLE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
A. GENERAL PRINCIPLES AND PROCEDURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1. Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2. Objective and scope. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3. Selection of agents for review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4. Data for the Monographs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
5. Meeting participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
6. Working procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
B. SCIENTIFIC REVIEW AND EVALUATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1. Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2. Studies of cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3. Studies of cancer in experimental animals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4. Mechanistic and other relevant data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5. Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
6. Evaluation and rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
GENERAL REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
ALOE VERA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
1.1 Identification of the agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
1.3 Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.4 Production, sales, and consumption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
1.5 Occupational exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
1.6 Regulations and guidelines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
III
IARC MONOGRAPHS - 108
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3. Cancer in Experimental Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.1 Studies of carcinogenicity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2 Photo-co-carcinogenicity studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4. Mechanistic and Other Relevant Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.1 Absorption, distribution, metabolism, and excretion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2 Genetic and related effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.3 Toxic effects and other mechanistic data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.4 Other mechanistic data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.5 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.6 Mechanistic considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5. Summary of Data Reported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.2 Human carcinogenicity data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.3 Animal carcinogenicity data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.4 Mechanistic and other relevant data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.2 Cancer in experimental animals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
6.3 Overall evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
GOLDENSEAL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
1.1 Identification of the agent and its major constituents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
1.3 Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3. Cancer in Experimental Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4. Mechanistic and Other Relevant Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.1 Absorption, distribution, metabolism, and excretion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.2 Genetic and related effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.3 Other mechanistic data relevant to carcinogenicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.5 Mechanistic considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5. Summary of Data Reported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.2 Human carcinogenicity data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.3 Animal carcinogenicity data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.4 Mechanistic and other relevant data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
IV
Contents
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.2 Cancer in experimental animals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.3 Overall evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
GINKGO BILOBA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
1.1 Identification of the agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
1.3 Uses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2.2 Randomized controlled trial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
2.3 Casecontrol studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3. Cancer in Experimental Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4. Mechanistic and Other Relevant Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4.1 Absorption, distribution, metabolism, and excretion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4.2 Genetic and related effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.3 Other mechanistic data relevant to carcinogenicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4.5 Mechanistic considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5. Summary of Data Reported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.2 Human carcinogenicity data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.3 Animal carcinogenicity data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
5.4 Mechanistic and other relevant data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.2 Cancer in experimental animals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.3 Overall evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
KAVA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
1.1 Identification of the agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
1.3 Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
1.4 Production, sales, and consumption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
1.5 Occupational exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
1.6 Regulations and guidelines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
V
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4.3 Other mechanistic data relevant to carcinogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
PULEGONE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
1.2 Production and use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
1.3 Occurrence and exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
4.3 Other mechanistic data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
VI
Contents
METHYLENE BLUE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

1. Exposure Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
1.1 Chemical and physical data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
4.3 Other relevant mechanisms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
PRIMIDONE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
2.1 Cohort studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
2.2 Nested casecontrol studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
4.3 Other mechanistic data relevant to carcinogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
VII

5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
SULFASALAZINE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
1.2 Production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
2.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
2.2 Longitudinal, cohort, and nested casecontrol studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
2.3 Casecontrol studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
2.4 Meta-analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
3.3 Studies of co-carcinogenicity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
VIII
Contents
PENTOSAN POLYSULFATE SODIUM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
1.3 Production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
TRIAMTERENE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
2.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
2.2 Triamterene and cancer of the breast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
2.3 Combined use of triamterene and hydrochlorothiazide, and cancer of the lip . . . . . . . . . . . . . . . . 271
2.4 The drug class including triamterene, and cancer of the breast or colorectum. . . . . . . . . . . . . . . . 271
3.1 Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
3.2 Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
IX

4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
HYDROCHLOROTHIAZIDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
2.1 Cancers of the lip and skin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
2.2 Renal cell carcinoma. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
2.3 Other cancers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
3.1 Oral administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
3.2 Coexposure with modifying agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
X
Contents
PIOGLITAZONE AND ROSIGLITAZONE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
1.1 Chemical and physical data on pioglitazone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
1.2 Analysis of pioglitazone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
1.3 Production and use of pioglitazone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
1.4 Occurrence and exposure to pioglitazone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
1.5 Regulations and guidelines for pioglitazone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
1.6 Chemical and physical data on rosiglitazone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
1.7 Analysis of rosiglitazone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
1.8 Production and use of rosiglitazone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
1.9 Occurrence and exposure to rosiglitazone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
1.10 Regulations and guidelines for rosiglitazone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
2.1 Cancer of the bladder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
2.2 Cancer of the liver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
2.3 Cancer of the colorectum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
2.4 Cancer of the lung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
2.5 Cancer of the prostate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
2.6 Cancer of the breast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
2.7 Other site-specific cancers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
3.1 Pioglitazone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
3.2 Rosiglitazone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
4.1 Absorption, distribution, metabolism, and excretion of pioglitazone. . . . . . . . . . . . . . . . . . . . . . . . . 361
4.2 Absorption, distribution, metabolism, and excretion of rosiglitazone. . . . . . . . . . . . . . . . . . . . . . . . 362
4.4 Other mechanistic data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
4.5 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
XI
DIGOXIN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
1. Exposure Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
1.2 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
2.1 Cancer of the breast. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
2.2 Cancers of the uterus and ovary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
2.3 Cancer of the prostate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
2.4 Non-Hodgkin lymphoma. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
2.5 Other cancer sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
4.4 Susceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
5.1 Exposure data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
6. Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
6.1 Cancer in humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
LIST OF ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
XII
NOTE TO THE READER
The term carcinogenic risk in the IARC Monographs series is taken to mean that an agent is
capable of causing cancer. The Monographs evaluate cancer hazards, despite the historical presence
of the word risks in the title.
Inclusion of an agent in the Monographs does not imply that it is a carcinogen, only that the
published data have been examined. Equally, the fact that an agent has not yet been evaluated in a
Monograph does not mean that it is not carcinogenic. Similarly, identification of cancer sites with
sufficient evidence or limited evidence in humans should not be viewed as precluding the possibility
that an agent may cause cancer at other sites.
The evaluations of carcinogenic risk are made by international working groups of independent
scientists and are qualitative in nature. No recommendation is given for regulation or legislation.
Anyone who is aware of published data that may alter the evaluation of the carcinogenic risk
of an agent to humans is encouraged to make this information available to the Section of IARC
Monographs, International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon
Cedex 08, France, in order that the agent may be considered for re-evaluation by a future Working
Group.
Although every effort is made to prepare the Monographs as accurately as possible, mistakes may
occur. Readers are requested to communicate any errors to the Section of IARC Monographs, so that
corrections can be reported in future volumes.
1
LIST OF PARTICIPANTS
Members1
Bruce C. Baguley Robert J. Biggar
Auckland Cancer Society Research Centre Queensland University of Technology
University of Auckland Brisbane
Auckland Australia
New Zealand
Esperanza J. Carcache de Blanco

Frederick A. Beland
College of Pharmacy
National Center for Toxicological Research Ohio State University
Food and Drug Administration Columbus, OH
Jefferson, AR USA
USA
Michael L. Cunningham [retired]

Joseph M. Betz (Subgroup chair, Exposure
National Institute of Environmental Health
Data)
Sciences
Office of Dietary Supplements Chapel Hill, NC
National Institutes of Health USA
Rockville, MD
USA
1
Working Group Members and Invited Specialists serve in their individual capacities as scientists and not as
representatives of their government or any organization with which they are affiliated. Affiliations are provided for iden-
tification purposes only. Invited Specialists do not serve as Meeting Chair or Subgroup Chair, draft text that pertains to
the description or interpretation of cancer data, or participate in the evaluations.
Each participant was asked to disclose pertinent research, employment, and financial interests. Current financial
interests and research and employment interests during the past 4 years or anticipated in the future are identified here.
Minor pertinent interests are not listed and include stock valued at no more than US $1000 overall, grants that provide
no more than 5% of the research budget of the experts organization and that do not support the experts research or
position, and consulting or speaking on matters not before a court or government agency that does not exceed 2% of
total professional time or compensation. All grants that support the experts research or position and all consulting
or speaking on behalf of an interested party on matters before a court or government agency are listed as significant
pertinent interests.
3
IARC MONOGRAPHS 108
June K. Dunnick Dirk W. Lachenmeier

Toxicology Branch Ministry of Rural Region and Consumer
National Institute of Environmental Health Protection Baden-Wrttemberg
Sciences Stuttgart
Research Triangle Park, NC Germany
USA
Ruth M. Lunn
Lei Guo National Toxicology Program
National Center for Toxicological Research National Institute of Environmental Health
Food and Drug Administration Sciences
Jefferson, AR Research Triangle Park, NC
USA USA
M. Matilde Marques (Subgroup Chair,

Charles William Jameson (Subgroup Chair, Mechanistic and Other Relevant Data)
Cancer in Experimental Animals) Department of Chemical and Biological
CWJ Consulting, LLC Engineering
Cape Coral, FL Technical University of Lisbon
USA Lisbon
Portugal
Margaret Karagas (Subgroup Chair, Cancer in David L. McCormick

Humans)
IIT Research Institute
Section of Biostatistics and Epidemiology Chicago, IL
Dartmouth Medical School USA
Lebanon, NH
USA
Sonal Singh
Trudy L. Knight Department of Epidemiology
Johns Hopkins School of Medicine and Public
Institute of Occupational Environmental Health
Medicine Baltimore, MD
University of Birmingham USA
Birmingham
United Kingdom
4
Participants
Saranjit Singh [unable to attend] Shufeng Zhou

Department of Pharmaceutical Analysis College of Pharmacy
National Institute of Pharmaceutical University of South Florida
Education and Research Tampa, FL
Punjab USA
India
Invited specialist
Bernard W. Stewart (Overall Chair)
Cancer Control Program Randall S. Stafford 3
South Eastern Sydney Public Health Unit
Program on Prevention Outcomes and
Randwick
Practices
Australia
Stanford Prevention Research Center
Stanford, CA
Chin-Hsiao Tseng 2 USA
College of Medicine
Taiwan University Representatives
Taipei
Taiwan, China None
Hiroyuki Tsuda Observers

Nanotoxicology Project
Nagoya City University David Alexander 4 [unable to attend]
Nagoya Williams Kherkher Hart Boundas, LLP
Japan League City, Texas
USA
Kristine L. Witt [unable to attend]
National Toxicology Program Andrea Bertocco 5
National Institute of Environmental Health Global Products Science & Safety EMEA
Sciences Herbalife Europe Limited
Research Triangle Park, NC Uxbridge
USA England
2
Chin-Hsiao Tseng received non-significant honoraria for serving on advisory boards and presentations from phar-
maceutical companies, including Bristol-Myers-Squibb Company, Eli Lilly and Company, and Takeda Pharmaceutical
Company Limited.
3
Randall S. Stafford provided expert testimony (ceased in 2011) for Mylan Pharmaceuticals, Morgantown, WV, manu-
facturer of generic drugs, including hydrochlorothiazide containing drugs.
4
David Alexander represents clients who claim they have been injured by the drug pioglitazone. He is employed and
sponsored by Williams Kherkher Hart Boundas, LLP, Houston Texas, USA.
5
Andrea Bertocco is employed by Herbalife, where he served as regulatory Manager for Herbalife products containing
Aloe vera, Goldenseal and Ginkgo biloba. He is sponsored by Herbalife Europe Limited, United Kingdom.
5
IARC MONOGRAPHS 108
Paul Dolin 6 Yann Grosse (Responsible Officer, Rapporteur

Cancer in Experimental Animals)
Takeda Global Research & Development Neela Guha (Rapporteur Cancer in Humans)
Centre (Europe) Ltd Kate Z. Guyton
London Batrice Lauby-Secretan (Rapporteur
England Exposure Data)
Ho-Sun Lee
Dana Loomis (Rapporteur Cancer in Humans)
Olaf Kelber 7 Heidi Mattock (Scientific Editor)
World Self-Medication Industry Douglas Puricelli Perin
Steigerwald Arzneimittelwerk GmbH Mnica S. Sierra
Darmstadt Kurt Straif (Head of Programme)
Germany Jiri Zavadil
Egon Koch 8 Administrative Assistance

World Self-Medication Industry
Sandrine Egraz
Dr. Willmar Schwabe GmbH & Co. KG
Brigitte Kajo
Karlsruhe
Michel Javin
Germany
Annick Leroux
Helene Lorenzen-Augros
IARC Secretariat
Melina Arnold
Production Team
Robert Baan (Senior Visiting Scientist)
Elisabeth Elbers
Helen Bailey
Solne Quennehen
Lamia Benbrahim-Tallaa (Rapporteur
Dorothy Russell
Mechanistic and Other Relevant Data)
Vronique Bouvard (Rapporteur Exposure
Data)
Fatiha El Ghissassi (Rapporteur Mechanistic
and Other Relevant Data)
Akram Ghantous
6
Paul Dolin is employed and sponsored by Takeda Pharmaceutical Company Limited, Europe, manufacturing piogli-
tazone-containing and other pharmaceutical drugs. He holds stock of pharmaceutical companies marketing drugs that
are reviewed at this meeting.
7
Olaf Kelber is employed by Steigerwald Arzneimittelwerk GmbH, Darmstadt, Germany, manufacturing herbal medi-
cines. He is sponsored by the World-Self-Medication Industry (WSMI), France.
8
Egon Koch is employed by Dr. Wilmar Schwabe GmbH & Co KG, Karlsruhe, Germany. He provides expert testimony
with respect to the commercialization of Ginkgo biloba extracts. He is sponsored by the World-Self-Medication Indus-
try (WSMI), France.
6
PREAMBLE
The Preamble to the IARC Monographs describes the objective and scope of the programme,
the scientific principles and procedures used in developing a Monograph, the types of
evidence considered and the scientific criteria that guide the evaluations. The Preamble
should be consulted when reading a Monograph or list of evaluations.
A. GENERAL PRINCIPLES AND risk of chemicals to man, which became the ini-
PROCEDURES tial title of the series.
In the succeeding years, the scope of the pro-
gramme broadened as Monographs were devel-
1. Background oped for groups of related chemicals, complex
Soon after IARC was established in 1965, it mixtures, occupational exposures, physical and
received frequent requests for advice on the car- biological agents and lifestyle factors. In 1988,
cinogenic risk of chemicals, including requests the phrase of chemicals was dropped from
for lists of known and suspected human carcino- the title, which assumed its present form, IARC
gens. It was clear that it would not be a simple Monographs on the Evaluation of Carcinogenic
task to summarize adequately the complexity of Risks to Humans.
the information that was available, and IARC Through the Monographs programme, IARC
began to consider means of obtaining interna- seeks to identify the causes of human cancer. This
tional expert opinion on this topic. In 1970, the is the first step in cancer prevention, which is
IARC Advisory Committee on Environmental needed as much today as when IARC was estab-
Carcinogenesis recommended ...that a com- lished. The global burden of cancer is high and
pendium on carcinogenic chemicals be pre- continues to increase: the annual number of new
pared by experts. The biological activity and cases was estimated at 10.1 million in 2000 and
evaluation of practical importance to public is expected to reach 15 million by 2020 (Stewart
health should be referenced and documented. & Kleihues, 2003). With current trends in demo-
The IARC Governing Council adopted a resolu- graphics and exposure, the cancer burden has
tion concerning the role of IARC in providing been shifting from high-resource countries to
government authorities with expert, independ- low- and medium-resource countries. As a result
ent, scientific opinion on environmental carcino- of Monographs evaluations, national health agen-
genesis. As one means to that end, the Governing cies have been able, on scientific grounds, to take
Council recommended that IARC should prepare measures to reduce human exposure to carcino-
monographs on the evaluation of carcinogenic gens in the workplace and in the environment.
7
IARC MONOGRAPHS 108
The criteria established in 1971 to evaluate causation of, and susceptibility to, malignant
carcinogenic risks to humans were adopted by the disease become more fully understood.
Working Groups whose deliberations resulted in A cancer hazard is an agent that is capable
the first 16 volumes of the Monographs series. of causing cancer under some circumstances,
Those criteria were subsequently updated by fur- while a cancer risk is an estimate of the carci-
ther ad hoc Advisory Groups (IARC, 1977, 1978, nogenic effects expected from exposure to a can-
1979, 1982, 1983, 1987, 1988, 1991; Vainio et al., cer hazard. The Monographs are an exercise in
1992; IARC, 2005, 2006). evaluating cancer hazards, despite the historical
The Preamble is primarily a statement of sci- presence of the word risks in the title. The dis-
entific principles, rather than a specification of tinction between hazard and risk is important,
working procedures. The procedures through and the Monographs identify cancer hazards
which a Working Group implements these prin- even when risks are very low at current exposure
ciples are not specified in detail. They usually levels, because new uses or unforeseen exposures
involve operations that have been established could engender risks that are significantly higher.
as being effective during previous Monograph In the Monographs, an agent is termed car-
meetings but remain, predominantly, the pre- cinogenic if it is capable of increasing the inci-
rogative of each individual Working Group. dence of malignant neoplasms, reducing their
latency, or increasing their severity or multiplic-
ity. The induction of benign neoplasms may in
2. Objective and scope some circumstances (see Part B, Section 3a) con-
The objective of the programme is to pre- tribute to the judgement that the agent is carci-
pare, with the help of international Working nogenic. The terms neoplasm and tumour are
Groups of experts, and to publish in the form of used interchangeably.
Monographs, critical reviews and evaluations of The Preamble continues the previous usage
evidence on the carcinogenicity of a wide range of the phrase strength of evidence as a matter
of human exposures. The Monographs repre- of historical continuity, although it should be
sent the first step in carcinogen risk assessment, understood that Monographs evaluations con-
which involves examination of all relevant infor- sider studies that support a finding of a cancer
mation to assess the strength of the available evi- hazard as well as studies that do not.
dence that an agent could alter the age-specific Some epidemiological and experimental
incidence of cancer in humans. The Monographs studies indicate that different agents may act at
may also indicate where additional research different stages in the carcinogenic process, and
efforts are needed, specifically when data imme- several different mechanisms may be involved.
diately relevant to an evaluation are not available. The aim of the Monographs has been, from their
In this Preamble, the term agent refers to inception, to evaluate evidence of carcinogenic-
any entity or circumstance that is subject to ity at any stage in the carcinogenesis process,
evaluation in a Monograph. As the scope of the independently of the underlying mechanisms.
programme has broadened, categories of agents Information on mechanisms may, however, be
now include specific chemicals, groups of related used in making the overall evaluation (IARC,
chemicals, complex mixtures, occupational or 1991; Vainio et al., 1992; IARC, 2005, 2006; see
environmental exposures, cultural or behav- also Part B, Sections 4 and 6). As mechanisms
ioural practices, biological organisms and physi- of carcinogenesis are elucidated, IARC convenes
cal agents. This list of categories may expand as international scientific conferences to determine
whether a broad-based consensus has emerged
8
Preamble
on how specific mechanistic data can be used exposure and (b) there is some evidence or sus-
in an evaluation of human carcinogenicity. The picion of carcinogenicity. Mixed exposures may
results of such conferences are reported in IARC occur in occupational and environmental set-
Scientific Publications, which, as long as they still tings and as a result of individual and cultural
reflect the current state of scientific knowledge, habits (such as tobacco smoking and dietary
may guide subsequent Working Groups. practices). Chemical analogues and compounds
Although the Monographs have emphasized with biological or physical characteristics simi-
hazard identification, important issues may also lar to those of suspected carcinogens may also
involve doseresponse assessment. In many be considered, even in the absence of data on a
cases, the same epidemiological and experimen- possible carcinogenic effect in humans or experi-
tal studies used to evaluate a cancer hazard can mental animals.
also be used to estimate a doseresponse relation- The scientific literature is surveyed for pub-
ship. A Monograph may undertake to estimate lished data relevant to an assessment of carci-
doseresponse relationships within the range nogenicity. Ad hoc Advisory Groups convened
of the available epidemiological data, or it may by IARC in 1984, 1989, 1991, 1993, 1998 and
compare the doseresponse information from 2003 made recommendations as to which
experimental and epidemiological studies. In agents should be evaluated in the Monographs
some cases, a subsequent publication may be pre- series. Recent recommendations are avail-
pared by a separate Working Group with exper- able on the Monographs programme web site
tise in quantitative doseresponse assessment. (http://monographs.iarc.fr). IARC may schedule
The Monographs are used by national and other agents for review as it becomes aware of
international authorities to make risk assess- new scientific information or as national health
ments, formulate decisions concerning preventive agencies identify an urgent public health need
measures, provide effective cancer control pro- related to cancer.
grammes and decide among alternative options As significant new data become available
for public health decisions. The evaluations of on an agent for which a Monograph exists, a re-
IARC Working Groups are scientific, qualita- evaluation may be made at a subsequent meeting,
tive judgements on the evidence for or against and a new Monograph published. In some cases it
carcinogenicity provided by the available data. may be appropriate to review only the data pub-
These evaluations represent only one part of the lished since a prior evaluation. This can be useful
body of information on which public health deci- for updating a database, reviewing new data to
sions may be based. Public health options vary resolve a previously open question or identifying
from one situation to another and from country new tumour sites associated with a carcinogenic
to country and relate to many factors, including agent. Major changes in an evaluation (e.g. a new
different socioeconomic and national priorities. classification in Group 1 or a determination that a
Therefore, no recommendation is given with mechanism does not operate in humans, see Part
regard to regulation or legislation, which are B, Section 6) are more appropriately addressed by
the responsibility of individual governments or a full review.
other international organizations.
4. Data for the Monographs
3. Selection of agents for review
Each Monograph reviews all pertinent epi-
Agents are selected for review on the basis of demiological studies and cancer bioassays in
two main criteria: (a) there is evidence of human experimental animals. Those judged inadequate
9
IARC MONOGRAPHS 108
or irrelevant to the evaluation may be cited but (a) The Working Group
not summarized. If a group of similar studies is
not reviewed, the reasons are indicated. The Working Group is responsible for the crit-
Mechanistic and other relevant data are also ical reviews and evaluations that are developed
reviewed. A Monograph does not necessarily during the meeting. The tasks of Working Group
cite all the mechanistic literature concerning Members are: (i) to ascertain that all appropriate
the agent being evaluated (see Part B, Section data have been collected; (ii) to select the data rel-
4). Only those data considered by the Working evant for the evaluation on the basis of scientific
Group to be relevant to making the evaluation merit; (iii) to prepare accurate summaries of the
are included. data to enable the reader to follow the reasoning
With regard to epidemiological studies, can- of the Working Group; (iv) to evaluate the results
cer bioassays, and mechanistic and other relevant of epidemiological and experimental studies on
data, only reports that have been published or cancer; (v) to evaluate data relevant to the under-
accepted for publication in the openly available standing of mechanisms of carcinogenesis; and
scientific literature are reviewed. The same publi- (vi) to make an overall evaluation of the carci-
cation requirement applies to studies originating nogenicity of the exposure to humans. Working
from IARC, including meta-analyses or pooled Group Members generally have published sig-
analyses commissioned by IARC in advance of a nificant research related to the carcinogenicity of
meeting (see Part B, Section 2c). Data from gov- the agents being reviewed, and IARC uses litera-
ernment agency reports that are publicly avail- ture searches to identify most experts. Working
able are also considered. Exceptionally, doctoral Group Members are selected on the basis of (a)
theses and other material that are in their final knowledge and experience and (b) absence of real
form and publicly available may be reviewed. or apparent conflicts of interests. Consideration
Exposure data and other information on an is also given to demographic diversity and bal-
agent under consideration are also reviewed. In ance of scientific findings and views.
the sections on chemical and physical proper-
ties, on analysis, on production and use and on (b) Invited Specialists
occurrence, published and unpublished sources Invited Specialists are experts who also have
of information may be considered. critical knowledge and experience but have
Inclusion of a study does not imply accept- a real or apparent conflict of interests. These
ance of the adequacy of the study design or of experts are invited when necessary to assist in
the analysis and interpretation of the results, and the Working Group by contributing their unique
limitations are clearly outlined in square brack- knowledge and experience during subgroup and
ets at the end of each study description (see Part plenary discussions. They may also contribute
B). The reasons for not giving further considera- text on non-influential issues in the section on
tion to an individual study also are indicated in exposure, such as a general description of data
the square brackets. on production and use (see Part B, Section 1).
Invited Specialists do not serve as meeting chair
5. Meeting participants or subgroup chair, draft text that pertains to the
description or interpretation of cancer data, or
Five categories of participant can be present participate in the evaluations.
at Monograph meetings.
10
Preamble
(c) Representatives of national and whether there is a conflict that warrants some
international health agencies limitation on participation. The declarations are
updated and reviewed again at the opening of
Representatives of national and interna- the meeting. Interests related to the subject of
tional health agencies often attend meetings the meeting are disclosed to the meeting par-
because their agencies sponsor the programme ticipants and in the published volume (Cogliano
or are interested in the subject of a meeting. et al., 2004).
Representatives do not serve as meeting chair or The names and principal affiliations of par-
subgroup chair, draft any part of a Monograph, ticipants are available on the Monographs pro-
or participate in the evaluations. gramme web site (http://monographs.iarc.fr)
approximately two months before each meeting.
(d) Observers with relevant scientific
It is not acceptable for Observers or third parties
credentials to contact other participants before a meeting or
Observers with relevant scientific credentials to lobby them at any time. Meeting participants
may be admitted to a meeting by IARC in limited are asked to report all such contacts to IARC
numbers. Attention will be given to achieving a (Cogliano et al., 2005).
balance of Observers from constituencies with All participants are listed, with their princi-
differing perspectives. They are invited to observe pal affiliations, at the beginning of each volume.
the meeting and should not attempt to influence Each participant who is a Member of a Working
it. Observers do not serve as meeting chair or Group serves as an individual scientist and not as
subgroup chair, draft any part of a Monograph, a representative of any organization, government
or participate in the evaluations. At the meeting, or industry.
the meeting chair and subgroup chairs may grant
Observers an opportunity to speak, generally 6. Working procedures
after they have observed a discussion. Observers
agree to respect the Guidelines for Observers A separate Working Group is responsible for
at IARC Monographs meetings (available at developing each volume of Monographs. A vol-
http://monographs.iarc.fr). ume contains one or more Monographs, which
can cover either a single agent or several related
(e) The IARC Secretariat agents. Approximately one year in advance of the
The IARC Secretariat consists of scientists meeting of a Working Group, the agents to be
who are designated by IARC and who have rel- reviewed are announced on the Monographs pro-
evant expertise. They serve as rapporteurs and gramme web site (http://monographs.iarc.fr) and
participate in all discussions. When requested by participants are selected by IARC staff in consul-
the meeting chair or subgroup chair, they may tation with other experts. Subsequently, relevant
also draft text or prepare tables and analyses. biological and epidemiological data are collected
Before an invitation is extended, each poten- by IARC from recognized sources of information
tial participant, including the IARC Secretariat, on carcinogenesis, including data storage and
completes the WHO Declaration of Interests to retrieval systems such as PubMed. Meeting par-
report financial interests, employment and con- ticipants who are asked to prepare preliminary
sulting, and individual and institutional research working papers for specific sections are expected
support related to the subject of the meeting. to supplement the IARC literature searches with
IARC assesses these interests to determine their own searches.
11
IARC MONOGRAPHS 108
Industrial associations, labour unions and not necessarily unanimity. The chair may elect
other knowledgeable organizations may be to poll Working Group Members to determine
asked to provide input to the sections on produc- the diversity of scientific opinion on issues where
tion and use, although this involvement is not consensus is not readily apparent.
required as a general rule. Information on pro- After the meeting, the master copy is verified
duction and trade is obtained from governmen- by consulting the original literature, edited and
tal, trade and market research publications and, prepared for publication. The aim is to publish
in some cases, by direct contact with industries. the volume within six months of the Working
Separate production data on some agents may Group meeting. A summary of the outcome is
not be available for a variety of reasons (e.g. not available on the Monographs programme web
collected or made public in all producing coun- site soon after the meeting.
tries, production is small). Information on uses
may be obtained from published sources but is
often complemented by direct contact with man- B. SCIENTIFIC REVIEW AND
ufacturers. Efforts are made to supplement this EVALUATION
information with data from other national and
international sources. The available studies are summarized by the
Six months before the meeting, the mate- Working Group, with particular regard to the
rial obtained is sent to meeting participants to qualitative aspects discussed below. In general,
prepare preliminary working papers. The work- numerical findings are indicated as they appear
ing papers are compiled by IARC staff and sent, in the original report; units are converted when
before the meeting, to Working Group Members necessary for easier comparison. The Working
and Invited Specialists for review. Group may conduct additional analyses of the
The Working Group meets at IARC for seven published data and use them in their assessment
to eight days to discuss and finalize the texts of the evidence; the results of such supplemen-
and to formulate the evaluations. The objectives tary analyses are given in square brackets. When
of the meeting are peer review and consensus. an important aspect of a study that directly
During the first few days, four subgroups (cov- impinges on its interpretation should be brought
ering exposure data, cancer in humans, cancer to the attention of the reader, a Working Group
in experimental animals, and mechanistic and comment is given in square brackets.
other relevant data) review the working papers, The scope of the IARC Monographs pro-
develop a joint subgroup draft and write sum- gramme has expanded beyond chemicals to
maries. Care is taken to ensure that each study include complex mixtures, occupational expo-
summary is written or reviewed by someone sures, physical and biological agents, lifestyle
not associated with the study being considered. factors and other potentially carcinogenic expo-
During the last few days, the Working Group sures. Over time, the structure of a Monograph
meets in plenary session to review the subgroup has evolved to include the following sections:
drafts and develop the evaluations. As a result, Exposure data
the entire volume is the joint product of the Studies of cancer in humans
Working Group, and there are no individually Studies of cancer in experimental animals
authored sections. Mechanistic and other relevant data
IARC Working Groups strive to achieve a Summary
consensus evaluation. Consensus reflects broad Evaluation and rationale
agreement among Working Group Members, but
12
Preamble
In addition, a section of General Remarks at response and clinical disease other than cancer
the front of the volume discusses the reasons the are also presented.
agents were scheduled for evaluation and some For physical agents that are forms of radia-
key issues the Working Group encountered dur- tion, energy and range of the radiation are
ing the meeting. included. For foreign bodies, fibres and respir-
This part of the Preamble discusses the types able particles, size range and relative dimensions
of evidence considered and summarized in each are indicated.
section of a Monograph, followed by the scientific For agents such as mixtures, drugs or lifestyle
criteria that guide the evaluations. factors, a description of the agent, including its
composition, is given.
Whenever appropriate, other information,
1. Exposure data such as historical perspectives or the description
Each Monograph includes general informa- of an industry or habit, may be included.
tion on the agent: this information may vary sub-
stantially between agents and must be adapted (b) Analysis and detection
accordingly. Also included is information on
An overview of methods of analysis and
production and use (when appropriate), meth-
detection of the agent is presented, including
ods of analysis and detection, occurrence, and
their sensitivity, specificity and reproducibility.
sources and routes of human occupational and
Methods widely used for regulatory purposes
environmental exposures. Depending on the
are emphasized. Methods for monitoring human
agent, regulations and guidelines for use may be
exposure are also given. No critical evaluation
presented.
or recommendation of any method is meant or
implied.
(a) General information on the agent
For chemical agents, sections on chemical (c) Production and use
and physical data are included: the Chemical
The dates of first synthesis and of first com-
Abstracts Service Registry Number, the latest pri-
mercial production of a chemical, mixture or
mary name and the IUPAC systematic name are
other agent are provided when available; for
recorded; other synonyms are given, but the list
agents that do not occur naturally, this informa-
is not necessarily comprehensive. Information
tion may allow a reasonable estimate to be made
on chemical and physical properties that are rel-
of the date before which no human exposure to
evant to identification, occurrence and biologi-
the agent could have occurred. The dates of first
cal activity is included. A description of technical
reported occurrence of an exposure are also pro-
products of chemicals includes trade names, rel-
vided when available. In addition, methods of
evant specifications and available information
synthesis used in past and present commercial
on composition and impurities. Some of the
production and different methods of production,
trade names given may be those of mixtures in
which may give rise to different impurities, are
which the agent being evaluated is only one of
described.
the ingredients.
The countries where companies report pro-
For biological agents, taxonomy, struc-
duction of the agent, and the number of compa-
ture and biology are described, and the degree
nies in each country, are identified. Available data
of variability is indicated. Mode of replication,
on production, international trade and uses are
life cycle, target cells, persistence, latency, host
13
IARC MONOGRAPHS 108
obtained for representative regions. It should not, (e) Regulations and guidelines
however, be inferred that those areas or nations
are necessarily the sole or major sources or users Statements concerning regulations and
of the agent. Some identified uses may not be guidelines (e.g. occupational exposure limits,
current or major applications, and the coverage maximal levels permitted in foods and water,
is not necessarily comprehensive. In the case of pesticide registrations) are included, but they
drugs, mention of their therapeutic uses does not may not reflect the most recent situation, since
necessarily represent current practice nor does it such limits are continuously reviewed and modi-
imply judgement as to their therapeutic efficacy. fied. The absence of information on regulatory
status for a country should not be taken to imply
that that country does not have regulations with
(d) Occurrence and exposure regard to the exposure. For biological agents, leg-
Information on the occurrence of an agent in islation and control, including vaccination and
the environment is obtained from data derived therapy, are described.
from the monitoring and surveillance of levels
in occupational environments, air, water, soil,
2. Studies of cancer in humans
plants, foods and animal and human tissues.
When available, data on the generation, per- This section includes all pertinent epidemio-
sistence and bioaccumulation of the agent are logical studies (see Part A, Section 4). Studies of
also included. Such data may be available from biomarkers are included when they are relevant
national databases. to an evaluation of carcinogenicity to humans.
Data that indicate the extent of past and pre-
sent human exposure, the sources of exposure, (a) Types of study considered
the people most likely to be exposed and the fac-
tors that contribute to the exposure are reported. Several types of epidemiological study con-
Information is presented on the range of human tribute to the assessment of carcinogenicity in
exposure, including occupational and environ- humans cohort studies, casecontrol studies,
mental exposures. This includes relevant findings correlation (or ecological) studies and interven-
from both developed and developing countries. tion studies. Rarely, results from randomized tri-
Some of these data are not distributed widely and als may be available. Case reports and case series
may be available from government reports and of cancer in humans may also be reviewed.
other sources. In the case of mixtures, indus- Cohort and casecontrol studies relate indi-
tries, occupations or processes, information is vidual exposures under study to the occurrence of
given about all agents known to be present. For cancer in individuals and provide an estimate of
processes, industries and occupations, a histori- effect (such as relative risk) as the main measure
cal description is also given, noting variations in of association. Intervention studies may provide
chemical composition, physical properties and strong evidence for making causal inferences, as
levels of occupational exposure with date and exemplified by cessation of smoking and the sub-
place. For biological agents, the epidemiology of sequent decrease in risk for lung cancer.
infection is described. In correlation studies, the units of inves-
tigation are usually whole populations (e.g. in
particular geographical areas or at particular
times), and cancer frequency is related to a sum-
mary measure of the exposure of the population
14
Preamble
to the agent under study. In correlation studies, agent and disease. Confounding is a form of bias
individual exposure is not documented, which that occurs when the relationship with disease is
renders this kind of study more prone to con- made to appear stronger or weaker than it truly is
founding. In some circumstances, however, cor- as a result of an association between the apparent
relation studies may be more informative than causal factor and another factor that is associated
analytical study designs (see, for example, the with either an increase or decrease in the inci-
Monograph on arsenic in drinking-water; IARC, dence of the disease. The role of chance is related
2004). to biological variability and the influence of sam-
In some instances, case reports and case series ple size on the precision of estimates of effect.
have provided important information about the In evaluating the extent to which these fac-
carcinogenicity of an agent. These types of study tors have been minimized in an individual study,
generally arise from a suspicion, based on clinical consideration is given to several aspects of design
experience, that the concurrence of two events and analysis as described in the report of the
that is, a particular exposure and occurrence of study. For example, when suspicion of carcino-
a cancer has happened rather more frequently genicity arises largely from a single small study,
than would be expected by chance. Case reports careful consideration is given when interpreting
and case series usually lack complete ascertain- subsequent studies that included these data in an
ment of cases in any population, definition or enlarged population. Most of these considera-
enumeration of the population at risk and esti- tions apply equally to casecontrol, cohort and
mation of the expected number of cases in the correlation studies. Lack of clarity of any of these
absence of exposure. aspects in the reporting of a study can decrease
The uncertainties that surround the inter- its credibility and the weight given to it in the
pretation of case reports, case series and corre- final evaluation of the exposure.
lation studies make them inadequate, except in First, the study population, disease (or dis-
rare instances, to form the sole basis for inferring eases) and exposure should have been well
a causal relationship. When taken together with defined by the authors. Cases of disease in the
casecontrol and cohort studies, however, these study population should have been identified in
types of study may add materially to the judge- a way that was independent of the exposure of
ment that a causal relationship exists. interest, and exposure should have been assessed
Epidemiological studies of benign neo- in a way that was not related to disease status.
plasms, presumed preneoplastic lesions and Second, the authors should have taken into
other end-points thought to be relevant to cancer account in the study design and analysis
are also reviewed. They may, in some instances, other variables that can influence the risk of dis-
strengthen inferences drawn from studies of ease and may have been related to the exposure
cancer itself. of interest. Potential confounding by such vari-
ables should have been dealt with either in the
(b) Quality of studies considered design of the study, such as by matching, or in
the analysis, by statistical adjustment. In cohort
It is necessary to take into account the pos- studies, comparisons with local rates of disease
sible roles of bias, confounding and chance in may or may not be more appropriate than those
the interpretation of epidemiological studies. with national rates. Internal comparisons of fre-
Bias is the effect of factors in study design or quency of disease among individuals at different
execution that lead erroneously to a stronger or levels of exposure are also desirable in cohort
weaker association than in fact exists between an studies, since they minimize the potential for
15
IARC MONOGRAPHS 108
confounding related to the difference in risk fac- that may explain heterogeneity among studies in
tors between an external reference group and the more detail. A disadvantage of combined analy-
study population. ses is the possible lack of compatibility of data
Third, the authors should have reported the from various studies due to differences in sub-
basic data on which the conclusions are founded, ject recruitment, procedures of data collection,
even if sophisticated statistical analyses were methods of measurement and effects of unmeas-
employed. At the very least, they should have ured co-variates that may differ among studies.
given the numbers of exposed and unexposed Despite these limitations, well conducted com-
cases and controls in a casecontrol study and bined analyses may provide a firmer basis than
the numbers of cases observed and expected in individual studies for drawing conclusions about
a cohort study. Further tabulations by time since the potential carcinogenicity of agents.
exposure began and other temporal factors are IARC may commission a meta-analysis or
also important. In a cohort study, data on all pooled analysis that is pertinent to a particular
cancer sites and all causes of death should have Monograph (see Part A, Section 4). Additionally,
been given, to reveal the possibility of reporting as a means of gaining insight from the results of
bias. In a casecontrol study, the effects of inves- multiple individual studies, ad hoc calculations
tigated factors other than the exposure of interest that combine data from different studies may
should have been reported. be conducted by the Working Group during
Finally, the statistical methods used to obtain the course of a Monograph meeting. The results
estimates of relative risk, absolute rates of can- of such original calculations, which would be
cer, confidence intervals and significance tests, specified in the text by presentation in square
and to adjust for confounding should have been brackets, might involve updates of previously
clearly stated by the authors. These methods have conducted analyses that incorporate the results
been reviewed for casecontrol studies (Breslow of more recent studies or de-novo analyses.
& Day, 1980) and for cohort studies (Breslow & Irrespective of the source of data for the meta-
Day, 1987). analyses and pooled analyses, it is important that
the same criteria for data quality be applied as
(c) Meta-analyses and pooled analyses those that would be applied to individual studies
and to ensure also that sources of heterogeneity
Independent epidemiological studies of the between studies be taken into account.
same agent may lead to results that are difficult
to interpret. Combined analyses of data from
(d) Temporal effects
multiple studies are a means of resolving this
ambiguity, and well conducted analyses can be Detailed analyses of both relative and abso-
considered. There are two types of combined lute risks in relation to temporal variables, such
analysis. The first involves combining summary as age at first exposure, time since first exposure,
statistics such as relative risks from individual duration of exposure, cumulative exposure, peak
studies (meta-analysis) and the second involves a exposure (when appropriate) and time since
pooled analysis of the raw data from the individ- cessation of exposure, are reviewed and sum-
ual studies (pooled analysis) (Greenland, 1998). marized when available. Analyses of temporal
The advantages of combined analyses are relationships may be useful in making causal
increased precision due to increased sample size inferences. In addition, such analyses may sug-
and the opportunity to explore potential con- gest whether a carcinogen acts early or late in the
founders, interactions and modifying effects process of carcinogenesis, although, at best, they
16
Preamble
allow only indirect inferences about mechanisms (f) Criteria for causality
of carcinogenesis.
After the quality of individual epidemiologi-
cal studies of cancer has been summarized and
(e) Use of biomarkers in epidemiological assessed, a judgement is made concerning the
studies strength of evidence that the agent in question
Biomarkers indicate molecular, cellular or is carcinogenic to humans. In making its judge-
other biological changes and are increasingly ment, the Working Group considers several crite-
used in epidemiological studies for various pur- ria for causality (Hill, 1965). A strong association
poses (IARC, 1991; Vainio et al., 1992; Toniolo (e.g. a large relative risk) is more likely to indicate
et al., 1997; Vineis et al., 1999; Buffler et al., 2004). causality than a weak association, although it is
These may include evidence of exposure, of early recognized that estimates of effect of small mag-
effects, of cellular, tissue or organism responses, nitude do not imply lack of causality and may be
of individual susceptibility or host responses, important if the disease or exposure is common.
and inference of a mechanism (see Part B, Section Associations that are replicated in several studies
4b). This is a rapidly evolving field that encom- of the same design or that use different epidemi-
passes developments in genomics, epigenomics ological approaches or under different circum-
and other emerging technologies. stances of exposure are more likely to represent
Molecular epidemiological data that identify a causal relationship than isolated observations
associations between genetic polymorphisms from single studies. If there are inconsistent
and interindividual differences in susceptibility results among investigations, possible reasons
to the agent(s) being evaluated may contribute are sought (such as differences in exposure), and
to the identification of carcinogenic hazards to results of studies that are judged to be of high
humans. If the polymorphism has been demon- quality are given more weight than those of stud-
strated experimentally to modify the functional ies that are judged to be methodologically less
activity of the gene product in a manner that is sound.
consistent with increased susceptibility, these If the risk increases with the exposure, this is
data may be useful in making causal inferences. considered to be a strong indication of causality,
Similarly, molecular epidemiological studies that although the absence of a graded response is not
measure cell functions, enzymes or metabolites necessarily evidence against a causal relation-
that are thought to be the basis of susceptibil- ship. The demonstration of a decline in risk after
ity may provide evidence that reinforces biologi- cessation of or reduction in exposure in indi-
cal plausibility. It should be noted, however, that viduals or in whole populations also supports a
when data on genetic susceptibility originate causal interpretation of the findings.
from multiple comparisons that arise from sub- Several scenarios may increase confidence in
group analyses, this can generate false-positive a causal relationship. On the one hand, an agent
results and inconsistencies across studies, and may be specific in causing tumours at one site or
such data therefore require careful evaluation. of one morphological type. On the other, carci-
If the known phenotype of a genetic polymor- nogenicity may be evident through the causation
phism can explain the carcinogenic mechanism of multiple tumour types. Temporality, precision
of the agent being evaluated, data on this pheno- of estimates of effect, biological plausibility and
type may be useful in making causal inferences. coherence of the overall database are consid-
ered. Data on biomarkers may be employed in
17
IARC MONOGRAPHS 108
an assessment of the biological plausibility of epi- 3. Studies of cancer in experimental

demiological observations. animals
Although rarely available, results from rand-
omized trials that show different rates of cancer All known human carcinogens that have been
among exposed and unexposed individuals pro- studied adequately for carcinogenicity in experi-
vide particularly strong evidence for causality. mental animals have produced positive results
When several epidemiological studies show in one or more animal species (Wilbourn et al.,
little or no indication of an association between 1986; Tomatis et al., 1989). For several agents
an exposure and cancer, a judgement may be made (e.g. aflatoxins, diethylstilbestrol, solar radiation,
that, in the aggregate, they show evidence of lack vinyl chloride), carcinogenicity in experimen-
of carcinogenicity. Such a judgement requires tal animals was established or highly suspected
first that the studies meet, to a sufficient degree, before epidemiological studies confirmed their
the standards of design and analysis described carcinogenicity in humans (Vainio et al., 1995).
above. Specifically, the possibility that bias, con- Although this association cannot establish that
founding or misclassification of exposure or out- all agents that cause cancer in experimental ani-
come could explain the observed results should mals also cause cancer in humans, it is biologically
be considered and excluded with reasonable cer- plausible that agents for which there is sufficient
tainty. In addition, all studies that are judged to evidence of carcinogenicity in experimental ani-
be methodologically sound should (a) be con- mals (see Part B, Section 6b) also present a car-
sistent with an estimate of effect of unity for any cinogenic hazard to humans. Accordingly, in
observed level of exposure, (b) when considered the absence of additional scientific information,
together, provide a pooled estimate of relative these agents are considered to pose a carcinogenic
risk that is at or near to unity, and (c) have a nar- hazard to humans. Examples of additional scien-
row confidence interval, due to sufficient popula- tific information are data that demonstrate that
tion size. Moreover, no individual study nor the a given agent causes cancer in animals through
pooled results of all the studies should show any a species-specific mechanism that does not oper-
consistent tendency that the relative risk of can- ate in humans or data that demonstrate that the
cer increases with increasing level of exposure. mechanism in experimental animals also oper-
It is important to note that evidence of lack of ates in humans (see Part B, Section 6).
carcinogenicity obtained from several epidemio- Consideration is given to all available long-
logical studies can apply only to the type(s) of term studies of cancer in experimental animals
cancer studied, to the dose levels reported, and to with the agent under review (see Part A, Section
the intervals between first exposure and disease 4). In all experimental settings, the nature and
onset observed in these studies. Experience with extent of impurities or contaminants present in
human cancer indicates that the period from first the agent being evaluated are given when avail-
exposure to the development of clinical cancer is able. Animal species, strain (including genetic
sometimes longer than 20 years; latent periods background where applicable), sex, numbers per
substantially shorter than 30 years cannot pro- group, age at start of treatment, route of expo-
vide evidence for lack of carcinogenicity. sure, dose levels, duration of exposure, survival
and information on tumours (incidence, latency,
severity or multiplicity of neoplasms or prene-
oplastic lesions) are reported. Those studies in
experimental animals that are judged to be irrel-
evant to the evaluation or judged to be inadequate
18
Preamble
(e.g. too short a duration, too few animals, poor (a) Qualitative aspects
survival; see below) may be omitted. Guidelines
for conducting long-term carcinogenicity exper- An assessment of carcinogenicity involves
iments have been published (e.g. OECD, 2002). several considerations of qualitative impor-
Other studies considered may include: exper- tance, including (i) the experimental conditions
iments in which the agent was administered in under which the test was performed, including
the presence of factors that modify carcinogenic route, schedule and duration of exposure, spe-
effects (e.g. initiationpromotion studies, co- cies, strain (including genetic background where
carcinogenicity studies and studies in geneti- applicable), sex, age and duration of follow-up;
cally modified animals); studies in which the (ii) the consistency of the results, for example,
end-point was not cancer but a defined precan- across species and target organ(s); (iii) the spec-
cerous lesion; experiments on the carcinogenic- trum of neoplastic response, from preneoplastic
ity of known metabolites and derivatives; and lesions and benign tumours to malignant neo-
studies of cancer in non-laboratory animals (e.g. plasms; and (iv) the possible role of modifying
livestock and companion animals) exposed to factors.
the agent. Considerations of importance in the inter-
For studies of mixtures, consideration is pretation and evaluation of a particular study
given to the possibility that changes in the phys- include: (i) how clearly the agent was defined and,
icochemical properties of the individual sub- in the case of mixtures, how adequately the sam-
stances may occur during collection, storage, ple characterization was reported; (ii) whether
extraction, concentration and delivery. Another the dose was monitored adequately, particu-
consideration is that chemical and toxicological larly in inhalation experiments; (iii) whether the
interactions of components in a mixture may doses, duration of treatment and route of expo-
alter doseresponse relationships. The relevance sure were appropriate; (iv) whether the survival
to human exposure of the test mixture adminis- of treated animals was similar to that of con-
tered in the animal experiment is also assessed. trols; (v) whether there were adequate numbers
This may involve consideration of the following of animals per group; (vi) whether both male and
aspects of the mixture tested: (i) physical and female animals were used; (vii) whether animals
chemical characteristics, (ii) identified constitu- were allocated randomly to groups; (viii) whether
ents that may indicate the presence of a class of the duration of observation was adequate; and
substances and (iii) the results of genetic toxicity (ix) whether the data were reported and analysed
and related tests. adequately.
The relevance of results obtained with an When benign tumours (a) occur together
agent that is analogous (e.g. similar in structure with and originate from the same cell type as
or of a similar virus genus) to that being evalu- malignant tumours in an organ or tissue in a
ated is also considered. Such results may provide particular study and (b) appear to represent a
biological and mechanistic information that is stage in the progression to malignancy, they are
relevant to the understanding of the process of usually combined in the assessment of tumour
carcinogenesis in humans and may strengthen incidence (Huff et al., 1989). The occurrence of
the biological plausibility that the agent being lesions presumed to be preneoplastic may in cer-
evaluated is carcinogenic to humans (see Part B, tain instances aid in assessing the biological plau-
Section 2f). sibility of any neoplastic response observed. If an
agent induces only benign neoplasms that appear
to be end-points that do not readily undergo
19
IARC MONOGRAPHS 108
transition to malignancy, the agent should nev- Gart et al., 1986; Portier & Bailer, 1989; Bieler &
ertheless be suspected of being carcinogenic and Williams, 1993). The choice of the most appro-
requires further investigation. priate statistical method requires consideration
of whether or not there are differences in sur-
(b) Quantitative aspects vival among the treatment groups; for example,
The probability that tumours will occur may reduced survival because of non-tumour-related
depend on the species, sex, strain, genetic back- mortality can preclude the occurrence of
ground and age of the animal, and on the dose, tumours later in life. When detailed informa-
route, timing and duration of the exposure. tion on survival is not available, comparisons
Evidence of an increased incidence of neoplasms of the proportions of tumour-bearing animals
with increasing levels of exposure strengthens among the effective number of animals (alive at
the inference of a causal association between the the time the first tumour was discovered) can
exposure and the development of neoplasms. be useful when significant differences in sur-
The form of the doseresponse relation- vival occur before tumours appear. The lethal-
ship can vary widely, depending on the parity of the tumour also requires consideration: for
ticular agent under study and the target organ. rapidly fatal tumours, the time of death provides
Mechanisms such as induction of DNA dam- an indication of the time of tumour onset and
age or inhibition of repair, altered cell division can be assessed using life-table methods; non-
and cell death rates and changes in intercellular fatal or incidental tumours that do not affect
communication are important determinants of survival can be assessed using methods such as
doseresponse relationships for some carcino- the Mantel-Haenzel test for changes in tumour
gens. Since many chemicals require metabolic prevalence. Because tumour lethality is often dif-
activation before being converted to their reac- ficult to determine, methods such as the Poly-K
tive intermediates, both metabolic and toxicoki- test that do not require such information can
netic aspects are important in determining the also be used. When results are available on the
doseresponse pattern. Saturation of steps such number and size of tumours seen in experimen-
as absorption, activation, inactivation and elim- tal animals (e.g. papillomas on mouse skin, liver
ination may produce nonlinearity in the dose tumours observed through nuclear magnetic
response relationship (Hoel et al., 1983; Gart resonance tomography), other more complicated
et al., 1986), as could saturation of processes such statistical procedures may be needed (Sherman
as DNA repair. The doseresponse relationship et al., 1994; Dunson et al., 2003).
can also be affected by differences in survival Formal statistical methods have been devel-
among the treatment groups. oped to incorporate historical control data into
the analysis of data from a given experiment.
These methods assign an appropriate weight to
(c) Statistical analyses
historical and concurrent controls on the basis
Factors considered include the adequacy of of the extent of between-study and within-study
the information given for each treatment group: variability: less weight is given to historical con-
(i) number of animals studied and number exam- trols when they show a high degree of variability,
ined histologically, (ii) number of animals with a and greater weight when they show little varia-
given tumour type and (iii) length of survival. bility. It is generally not appropriate to discount
The statistical methods used should be clearly a tumour response that is significantly increased
stated and should be the generally accepted tech- compared with concurrent controls by arguing
niques refined for this purpose (Peto et al., 1980; that it falls within the range of historical controls,
20
Preamble
particularly when historical controls show high one subsection. For example, a mutation in a
between-study variability and are, thus, of little gene that codes for an enzyme that metabolizes
relevance to the current experiment. In analys- the agent under study could be discussed in the
ing results for uncommon tumours, however, the subsections on toxicokinetics, mechanisms and
analysis may be improved by considering histori- individual susceptibility if it also exists as an
cal control data, particularly when between-study inherited polymorphism.
variability is low. Historical controls should be
selected to resemble the concurrent controls as (a) Toxicokinetic data
closely as possible with respect to species, gen-
der and strain, as well as other factors such as Toxicokinetics refers to the absorption, dis-
basal diet and general laboratory environment, tribution, metabolism and elimination of agents
which may affect tumour-response rates in con- in humans, experimental animals and, where
trol animals (Haseman et al., 1984; Fung et al., relevant, cellular systems. Examples of kinetic
1996; Greim et al., 2003). factors that may affect doseresponse relation-
Although meta-analyses and combined anal- ships include uptake, deposition, biopersis-
yses are conducted less frequently for animal tence and half-life in tissues, protein binding,
experiments than for epidemiological studies metabolic activation and detoxification. Studies
due to differences in animal strains, they can be that indicate the metabolic fate of the agent in
useful aids in interpreting animal data when the humans and in experimental animals are sum-
experimental protocols are sufficiently similar. marized briefly, and comparisons of data from
humans and animals are made when possible.
Comparative information on the relationship
4. Mechanistic and other relevant between exposure and the dose that reaches the
data target site may be important for the extrapola-
tion of hazards between species and in clarifying
Mechanistic and other relevant data may pro- the role of in-vitro findings.
vide evidence of carcinogenicity and also help in
assessing the relevance and importance of find- (b) Data on mechanisms of carcinogenesis
ings of cancer in animals and in humans. The
nature of the mechanistic and other relevant data To provide focus, the Working Group
depends on the biological activity of the agent attempts to identify the possible mechanisms by
being considered. The Working Group considers which the agent may increase the risk of cancer.
representative studies to give a concise descrip- For each possible mechanism, a representative
tion of the relevant data and issues that they con- selection of key data from humans and experi-
sider to be important; thus, not every available mental systems is summarized. Attention is
study is cited. Relevant topics may include toxi- given to gaps in the data and to data that suggests
cokinetics, mechanisms of carcinogenesis, sus- that more than one mechanism may be operat-
ceptible individuals, populations and life-stages, ing. The relevance of the mechanism to humans
other relevant data and other adverse effects. is discussed, in particular, when mechanistic
When data on biomarkers are informative about data are derived from experimental model sys-
the mechanisms of carcinogenesis, they are tems. Changes in the affected organs, tissues or
included in this section. cells can be divided into three non-exclusive lev-
These topics are not mutually exclusive; thus, els as described below.
the same studies may be discussed in more than
21
IARC MONOGRAPHS 108
(i) Changes in physiology described for every possible level and mechanism
Physiological changes refer to exposure- discussed above.
related modifications to the physiology and/or Genotoxicity data are discussed here to illus-
response of cells, tissues and organs. Examples trate the key issues involved in the evaluation of
of potentially adverse physiological changes mechanistic data.
include mitogenesis, compensatory cell division, Tests for genetic and related effects are
escape from apoptosis and/or senescence, pres- described in view of the relevance of gene muta-
ence of inflammation, hyperplasia, metaplasia tion and chromosomal aberration/aneuploidy
and/or preneoplasia, angiogenesis, alterations in to carcinogenesis (Vainio et al., 1992; McGregor
cellular adhesion, changes in steroidal hormones et al., 1999). The adequacy of the reporting of
and changes in immune surveillance. sample characterization is considered and, when
necessary, commented upon; with regard to
(ii) Functional changes at the cellular level complex mixtures, such comments are similar
to those described for animal carcinogenicity
Functional changes refer to exposure-related
tests. The available data are interpreted critically
alterations in the signalling pathways used by
according to the end-points detected, which
cells to manage critical processes that are related
may include DNA damage, gene mutation, sister
to increased risk for cancer. Examples of func-
chromatid exchange, micronucleus formation,
tional changes include modified activities of
chromosomal aberrations and aneuploidy. The
enzymes involved in the metabolism of xenobi-
concentrations employed are given, and men-
otics, alterations in the expression of key genes
tion is made of whether the use of an exogenous
that regulate DNA repair, alterations in cyclin-
metabolic system in vitro affected the test result.
dependent kinases that govern cell cycle progres-
These data are listed in tabular form by phyloge-
sion, changes in the patterns of post-translational
netic classification.
modifications of proteins, changes in regula-
Positive results in tests using prokary-
tory factors that alter apoptotic rates, changes
otes, lower eukaryotes, insects, plants and cul-
in the secretion of factors related to the stimula-
tured mammalian cells suggest that genetic and
tion of DNA replication and transcription and
related effects could occur in mammals. Results
changes in gapjunction-mediated intercellular
from such tests may also give information on
communication.
the types of genetic effect produced and on the
(iii) Changes at the molecular level involvement of metabolic activation. Some end-
points described are clearly genetic in nature
Molecular changes refer to exposure-related
(e.g. gene mutations), while others are associated
changes in key cellular structures at the molec-
with genetic effects (e.g. unscheduled DNA syn-
ular level, including, in particular, genotoxicity.
thesis). In-vitro tests for tumour promotion, cell
Examples of molecular changes include forma-
transformation and gapjunction intercellular
tion of DNA adducts and DNA strand breaks,
communication may be sensitive to changes that
mutations in genes, chromosomal aberrations,
are not necessarily the result of genetic altera-
aneuploidy and changes in DNA methylation
tions but that may have specific relevance to the
patterns. Greater emphasis is given to irrevers-
process of carcinogenesis. Critical appraisals
ible effects.
of these tests have been published (Montesano
The use of mechanistic data in the identifica-
et al., 1986; McGregor et al., 1999).
tion of a carcinogenic hazard is specific to the
Genetic or other activity manifest in humans
mechanism being addressed and is not readily
and experimental mammals is regarded to be of
22
Preamble
greater relevance than that in other organisms. surgical implants of various kinds, and poorly
The demonstration that an agent can induce soluble fibres, dusts and particles of various
gene and chromosomal mutations in mammals sizes, the pathogenic effects of which are a result
in vivo indicates that it may have carcinogenic of their physical presence in tissues or body
activity. Negative results in tests for mutagenicity cavities. Other relevant data for such materials
in selected tissues from animals treated in vivo may include characterization of cellular, tissue
provide less weight, partly because they do not and physiological reactions to these materi-
exclude the possibility of an effect in tissues other als and descriptions of pathological conditions
than those examined. Moreover, negative results other than neoplasia with which they may be
in short-term tests with genetic end-points can- associated.
not be considered to provide evidence that rules
out the carcinogenicity of agents that act through (c) Other data relevant to mechanisms
other mechanisms (e.g. receptor-mediated
effects, cellular toxicity with regenerative cell A description is provided of any structure
division, peroxisome proliferation) (Vainio et al., activity relationships that may be relevant to an
1992). Factors that may give misleading results evaluation of the carcinogenicity of an agent, the
in short-term tests have been discussed in detail toxicological implications of the physical and
elsewhere (Montesano et al., 1986; McGregor chemical properties, and any other data relevant
et al., 1999). to the evaluation that are not included elsewhere.
When there is evidence that an agent acts by High-output data, such as those derived from
a specific mechanism that does not involve gen- gene expression microarrays, and high-through-
otoxicity (e.g. hormonal dysregulation, immune put data, such as those that result from testing
suppression, and formation of calculi and other hundreds of agents for a single end-point, pose a
deposits that cause chronic irritation), that evi- unique problem for the use of mechanistic data
dence is presented and reviewed critically in the in the evaluation of a carcinogenic hazard. In
context of rigorous criteria for the operation of the case of high-output data, there is the possi-
that mechanism in carcinogenesis (e.g. Capen bility to overinterpret changes in individual end-
et al., 1999). points (e.g. changes in expression in one gene)
For biological agents such as viruses, bacteria without considering the consistency of that find-
and parasites, other data relevant to carcinogenic- ing in the broader context of the other end-points
ity may include descriptions of the pathology of (e.g. other genes with linked transcriptional con-
infection, integration and expression of viruses, trol). High-output data can be used in assessing
and genetic alterations seen in human tumours. mechanisms, but all end-points measured in a
Other observations that might comprise cellu- single experiment need to be considered in the
lar and tissue responses to infection, immune proper context. For high-throughput data, where
response and the presence of tumour markers the number of observations far exceeds the num-
are also considered. ber of end-points measured, their utility for iden-
For physical agents that are forms of radia- tifying common mechanisms across multiple
tion, other data relevant to carcinogenicity may agents is enhanced. These data can be used to
include descriptions of damaging effects at the identify mechanisms that not only seem plausi-
physiological, cellular and molecular level, as ble, but also have a consistent pattern of carci-
for chemical agents, and descriptions of how nogenic response across entire classes of related
these effects occur. Physical agents may also be compounds.
considered to comprise foreign bodies, such as
23
IARC MONOGRAPHS 108
(d) Susceptibility data found on the Monographs programme web site

(http://monographs.iarc.fr).
Individuals, populations and life-stages may
have greater or lesser susceptibility to an agent, (a) Exposure data
based on toxicokinetics, mechanisms of carcino-
genesis and other factors. Examples of host and Data are summarized, as appropriate, on the
genetic factors that affect individual susceptibil- basis of elements such as production, use, occur-
ity include sex, genetic polymorphisms of genes rence and exposure levels in the workplace and
involved in the metabolism of the agent under environment and measurements in human tis-
evaluation, differences in metabolic capacity due sues and body fluids. Quantitative data and time
to life-stage or the presence of disease, differ- trends are given to compare exposures in dif-
ences in DNA repair capacity, competition for ferent occupations and environmental settings.
or alteration of metabolic capacity by medica- Exposure to biological agents is described in
tions or other chemical exposures, pre-existing terms of transmission, prevalence and persis-
hormonal imbalance that is exacerbated by a tence of infection.
chemical exposure, a suppressed immune sys- (b) Cancer in humans
tem, periods of higher-than-usual tissue growth
or regeneration and genetic polymorphisms that Results of epidemiological studies pertinent
lead to differences in behaviour (e.g. addiction). to an assessment of human carcinogenicity are
Such data can substantially increase the strength summarized. When relevant, case reports and
of the evidence from epidemiological data and correlation studies are also summarized. The tar-
enhance the linkage of in-vivo and in-vitro labo- get organ(s) or tissue(s) in which an increase in
ratory studies to humans. cancer was observed is identified. Doseresponse
and other quantitative data may be summarized
when available.
(e) Data on other adverse effects
Data on acute, subchronic and chronic (c) Cancer in experimental animals
adverse effects relevant to the cancer evaluation
are summarized. Adverse effects that confirm Data relevant to an evaluation of carcino-
distribution and biological effects at the sites of genicity in animals are summarized. For each
tumour development, or alterations in physiol- animal species, study design and route of admin-
ogy that could lead to tumour development, are istration, it is stated whether an increased inci-
emphasized. Effects on reproduction, embryonic dence, reduced latency, or increased severity
and fetal survival and development are summa- or multiplicity of neoplasms or preneoplastic
rized briefly. The adequacy of epidemiological lesions were observed, and the tumour sites are
studies of reproductive outcome and genetic and indicated. If the agent produced tumours after
related effects in humans is judged by the same prenatal exposure or in single-dose experiments,
criteria as those applied to epidemiological stud- this is also mentioned. Negative findings, inverse
ies of cancer, but fewer details are given. relationships, doseresponse and other quantita-
tive data are also summarized.
5. Summary (d) Mechanistic and other relevant data

This section is a summary of data presented Data relevant to the toxicokinetics (absorp-
in the preceding sections. Summaries can be tion, distribution, metabolism, elimination) and
24
Preamble
the possible mechanism(s) of carcinogenesis (e.g. relationship has been established between expo-
genetic toxicity, epigenetic effects) are summa- sure to the agent and human cancer. That is, a
rized. In addition, information on susceptible positive relationship has been observed between
individuals, populations and life-stages is sum- the exposure and cancer in studies in which
marized. This section also reports on other toxic chance, bias and confounding could be ruled
effects, including reproductive and developmen- out with reasonable confidence. A statement that
tal effects, as well as additional relevant data that there is sufficient evidence is followed by a sepa-
are considered to be important. rate sentence that identifies the target organ(s) or
tissue(s) where an increased risk of cancer was
observed in humans. Identification of a specific
6. Evaluation and rationale target organ or tissue does not preclude the pos-
Evaluations of the strength of the evidence for sibility that the agent may cause cancer at other
carcinogenicity arising from human and experi- sites.
mental animal data are made, using standard Limited evidence of carcinogenicity:
terms. The strength of the mechanistic evidence A positive association has been observed
is also characterized. between exposure to the agent and cancer for
It is recognized that the criteria for these which a causal interpretation is considered by
evaluations, described below, cannot encompass the Working Group to be credible, but chance,
all of the factors that may be relevant to an eval- bias or confounding could not be ruled out with
uation of carcinogenicity. In considering all of reasonable confidence.
the relevant scientific data, the Working Group Inadequate evidence of carcinogenicity: The
may assign the agent to a higher or lower cat- available studies are of insufficient quality, con-
egory than a strict interpretation of these criteria sistency or statistical power to permit a conclu-
would indicate. sion regarding the presence or absence of a causal
These categories refer only to the strength of association between exposure and cancer, or no
the evidence that an exposure is carcinogenic data on cancer in humans are available.
and not to the extent of its carcinogenic activ- Evidence suggesting lack of carcinogenicity:
ity (potency). A classification may change as new There are several adequate studies covering the
information becomes available. full range of levels of exposure that humans are
An evaluation of the degree of evidence is lim- known to encounter, which are mutually consist-
ited to the materials tested, as defined physically, ent in not showing a positive association between
chemically or biologically. When the agents eval- exposure to the agent and any studied cancer
uated are considered by the Working Group to be at any observed level of exposure. The results
sufficiently closely related, they may be grouped from these studies alone or combined should
together for the purpose of a single evaluation of have narrow confidence intervals with an upper
the degree of evidence. limit close to the null value (e.g. a relative risk
of 1.0). Bias and confounding should be ruled
(a) Carcinogenicity in humans out with reasonable confidence, and the studies
should have an adequate length of follow-up. A
The evidence relevant to carcinogenicity from conclusion of evidence suggesting lack of carcino-
studies in humans is classified into one of the fol- genicity is inevitably limited to the cancer sites,
lowing categories: conditions and levels of exposure, and length of
Sufficient evidence of carcinogenicity: observation covered by the available studies. In
The Working Group considers that a causal
25
IARC MONOGRAPHS 108
addition, the possibility of a very small risk at the A single study in one species and sex might be
levels of exposure studied can never be excluded. considered to provide sufficient evidence of carci-
In some instances, the above categories may nogenicity when malignant neoplasms occur to
be used to classify the degree of evidence related an unusual degree with regard to incidence, site,
to carcinogenicity in specific organs or tissues. type of tumour or age at onset, or when there are
When the available epidemiological stud- strong findings of tumours at multiple sites.
ies pertain to a mixture, process, occupation or Limited evidence of carcinogenicity:
industry, the Working Group seeks to identify The data suggest a carcinogenic effect but are
the specific agent considered most likely to be limited for making a definitive evaluation
responsible for any excess risk. The evaluation because, e.g. (a) the evidence of carcinogenicity
is focused as narrowly as the available data on is restricted to a single experiment; (b) there are
exposure and other aspects permit. unresolved questions regarding the adequacy of
the design, conduct or interpretation of the stud-
(b) Carcinogenicity in experimental ies; (c) the agent increases the incidence only of
animals benign neoplasms or lesions of uncertain neo-
plastic potential; or (d) the evidence of carcino-
Carcinogenicity in experimental animals can genicity is restricted to studies that demonstrate
be evaluated using conventional bioassays, bioas- only promoting activity in a narrow range of tis-
says that employ genetically modified animals, sues or organs.
and other in-vivo bioassays that focus on one or Inadequate evidence of carcinogenicity:
more of the critical stages of carcinogenesis. In The studies cannot be interpreted as showing
the absence of data from conventional long-term either the presence or absence of a carcinogenic
bioassays or from assays with neoplasia as the effect because of major qualitative or quantitative
end-point, consistently positive results in several limitations, or no data on cancer in experimental
models that address several stages in the multi- animals are available.
stage process of carcinogenesis should be con- Evidence suggesting lack of carcinogenicity:
sidered in evaluating the degree of evidence of Adequate studies involving at least two species
carcinogenicity in experimental animals. are available which show that, within the limits
The evidence relevant to carcinogenicity in of the tests used, the agent is not carcinogenic.
experimental animals is classified into one of the A conclusion of evidence suggesting lack of car-
following categories: cinogenicity is inevitably limited to the species,
Sufficient evidence of carcinogenicity: The tumour sites, age at exposure, and conditions
Working Group considers that a causal relation- and levels of exposure studied.
ship has been established between the agent and
an increased incidence of malignant neoplasms (c) Mechanistic and other relevant data
or of an appropriate combination of benign and
malignant neoplasms in (a) two or more species Mechanistic and other evidence judged to
of animals or (b) two or more independent stud- be relevant to an evaluation of carcinogenicity
ies in one species carried out at different times and of sufficient importance to affect the over-
or in different laboratories or under different all evaluation is highlighted. This may include
protocols. An increased incidence of tumours in data on preneoplastic lesions, tumour pathol-
both sexes of a single species in a well conducted ogy, genetic and related effects, structureactiv-
study, ideally conducted under Good Laboratory ity relationships, metabolism and toxicokinetics,
Practices, can also provide sufficient evidence.
26
Preamble
physicochemical parameters and analogous bio- have been focused on investigating a favoured
logical agents. mechanism.
The strength of the evidence that any carcino- For complex exposures, including occupa-
genic effect observed is due to a particular mech- tional and industrial exposures, the chemical
anism is evaluated, using terms such as weak, composition and the potential contribution of
moderate or strong. The Working Group then carcinogens known to be present are considered
assesses whether that particular mechanism is by the Working Group in its overall evaluation
likely to be operative in humans. The strongest of human carcinogenicity. The Working Group
indications that a particular mechanism oper- also determines the extent to which the materi-
ates in humans derive from data on humans als tested in experimental systems are related to
or biological specimens obtained from exposed those to which humans are exposed.
humans. The data may be considered to be espe-
cially relevant if they show that the agent in ques- (d) Overall evaluation
tion has caused changes in exposed humans that
are on the causal pathway to carcinogenesis. Finally, the body of evidence is considered as
Such data may, however, never become available, a whole, to reach an overall evaluation of the car-
because it is at least conceivable that certain com- cinogenicity of the agent to humans.
pounds may be kept from human use solely on An evaluation may be made for a group of
the basis of evidence of their toxicity and/or car- agents that have been evaluated by the Working
cinogenicity in experimental systems. Group. In addition, when supporting data indi-
The conclusion that a mechanism operates in cate that other related agents, for which there is
experimental animals is strengthened by find- no direct evidence of their capacity to induce
ings of consistent results in different experimen- cancer in humans or in animals, may also be
tal systems, by the demonstration of biological carcinogenic, a statement describing the ration-
plausibility and by coherence of the overall data- ale for this conclusion is added to the evaluation
base. Strong support can be obtained from stud- narrative; an additional evaluation may be made
ies that challenge the hypothesized mechanism for this broader group of agents if the strength of
experimentally, by demonstrating that the sup- the evidence warrants it.
pression of key mechanistic processes leads to The agent is described according to the word-
the suppression of tumour development. The ing of one of the following categories, and the
Working Group considers whether multiple designated group is given. The categorization of
mechanisms might contribute to tumour devel- an agent is a matter of scientific judgement that
opment, whether different mechanisms might reflects the strength of the evidence derived from
operate in different dose ranges, whether sepa- studies in humans and in experimental animals
rate mechanisms might operate in humans and and from mechanistic and other relevant data.
experimental animals and whether a unique Group 1: The agent is carcinogenic to
mechanism might operate in a susceptible group. humans.
The possible contribution of alternative mecha- This category is used when there is suffi-
nisms must be considered before concluding cient evidence of carcinogenicity in humans.
that tumours observed in experimental animals Exceptionally, an agent may be placed in this
are not relevant to humans. An uneven level of category when evidence of carcinogenicity in
experimental support for different mechanisms humans is less than sufficient but there is suffi-
may reflect that disproportionate resources cient evidence of carcinogenicity in experimental
27
IARC MONOGRAPHS 108
animals and strong evidence in exposed humans Group 2B: The agent is possibly carcinogenic
that the agent acts through a relevant mechanism to humans.
of carcinogenicity. This category is used for agents for which
Group 2. there is limited evidence of carcinogenicity in
This category includes agents for which, at humans and less than sufficient evidence of car-
one extreme, the degree of evidence of carcino- cinogenicity in experimental animals. It may
genicity in humans is almost sufficient, as well as also be used when there is inadequate evidence
those for which, at the other extreme, there are of carcinogenicity in humans but there is suffi-
no human data but for which there is evidence of cient evidence of carcinogenicity in experimental
carcinogenicity in experimental animals. Agents animals. In some instances, an agent for which
are assigned to either Group 2A (probably car- there is inadequate evidence of carcinogenicity in
cinogenic to humans) or Group 2B (possibly humans and less than sufficient evidence of car-
carcinogenic to humans) on the basis of epide- cinogenicity in experimental animals together
miological and experimental evidence of carci- with supporting evidence from mechanistic and
nogenicity and mechanistic and other relevant other relevant data may be placed in this group.
data. The terms probably carcinogenic and possi- An agent may be classified in this category solely
bly carcinogenic have no quantitative significance on the basis of strong evidence from mechanistic
and are used simply as descriptors of different and other relevant data.
levels of evidence of human carcinogenicity, with Group 3: The agent is not classifiable as to its
probably carcinogenic signifying a higher level of carcinogenicity to humans.
evidence than possibly carcinogenic. This category is used most commonly for
Group 2A: The agent is probably agents for which the evidence of carcinogenicity
carcinogenic to humans. is inadequate in humans and inadequate or lim-
This category is used when there is limited ited in experimental animals.
evidence of carcinogenicity in humans and suffi- Exceptionally, agents for which the evidence
cient evidence of carcinogenicity in experimental of carcinogenicity is inadequate in humans but
animals. In some cases, an agent may be classi- sufficient in experimental animals may be placed
fied in this category when there is inadequate evi- in this category when there is strong evidence
dence of carcinogenicity in humans and sufficient that the mechanism of carcinogenicity in experi-
evidence of carcinogenicity in experimental ani- mental animals does not operate in humans.
mals and strong evidence that the carcinogenesis Agents that do not fall into any other group
is mediated by a mechanism that also operates are also placed in this category.
in humans. Exceptionally, an agent may be clas- An evaluation in Group 3 is not a determi-
sified in this category solely on the basis of lim- nation of non-carcinogenicity or overall safety.
ited evidence of carcinogenicity in humans. An It often means that further research is needed,
agent may be assigned to this category if it clearly especially when exposures are widespread or
belongs, based on mechanistic considerations, to the cancer data are consistent with differing
a class of agents for which one or more members interpretations.
have been classified in Group 1 or Group 2A. Group 4: The agent is probably not
carcinogenic to humans.
This category is used for agents for which
there is evidence suggesting lack of carcinogenicity
28
Preamble
in humans and in experimental animals. In 2001. Workshop report. IARC Sci Publ, 157: 127.
some instances, agents for which there is inad- PMID:15055286
Capen CC, Dybing E, Rice JM, Wilbourn JD (1999).
equate evidence of carcinogenicity in humans Species Differences in Thyroid, Kidney and Urinary
but evidence suggesting lack of carcinogenicity in Bladder Carcinogenesis. Proceedings of a consensus
experimental animals, consistently and strongly conference. Lyon, France, 37 November 1997. IARC
Sci Publ, 147: 1225.
supported by a broad range of mechanistic and Cogliano V, Baan R, Straif K etal. (2005). Transparency in
other relevant data, may be classified in this IARC Monographs. Lancet Oncol, 6: 747. doi:10.1016/
group. S1470-2045(05)70380-6
Cogliano VJ, Baan RA, Straif K et al. (2004). The sci-
ence and practice of carcinogen identification and
(e) Rationale evaluation. Environ Health Perspect, 112: 12691274.
doi:10.1289/ehp.6950 PMID:15345338
The reasoning that the Working Group used Dunson DB, Chen Z, Harry J (2003). A Bayesian approach
to reach its evaluation is presented and discussed. for joint modeling of cluster size and subunit-specific
This section integrates the major findings from outcomes. Biometrics, 59: 521530. doi:10.1111/1541-
0420.00062 PMID:14601753
studies of cancer in humans, studies of cancer Fung KY, Krewski D, Smythe RT (1996). A comparison
in experimental animals, and mechanistic and of tests for trend with historical controls in carcinogen
other relevant data. It includes concise state- bioassay. Can J Stat, 24: 431454. doi:10.2307/3315326
ments of the principal line(s) of argument that Gart JJ, Krewski D, Lee PN etal. (1986). Statistical meth-
ods in cancer research. Volume IIIThe design and
emerged, the conclusions of the Working Group analysis of long-term animal experiments. IARC Sci
on the strength of the evidence for each group of Publ, 79: 1219. PMID:3301661
studies, citations to indicate which studies were Greenland S (1998). Meta-analysis. In: Modern
pivotal to these conclusions, and an explanation Epidemiology. Rothman KJ, Greenland S, editors.
Philadelphia: Lippincott Williams & Wilkins, pp.
of the reasoning of the Working Group in weigh- 643673
ing data and making evaluations. When there Greim H, Gelbke H-P, Reuter U et al. (2003).
are significant differences of scientific interpre- Evaluation of historical control data in carcino-
genicity studies. Hum Exp Toxicol, 22: 541549.
tation among Working Group Members, a brief doi:10.1191/0960327103ht394oa PMID:14655720
summary of the alternative interpretations is Haseman JK, Huff J, Boorman GA (1984). Use of historical
provided, together with their scientific rationale control data in carcinogenicity studies in rodents. Toxicol
and an indication of the relative degree of sup- Pathol, 12: 126135. doi:10.1177/019262338401200203
PMID:11478313
port for each alternative. Hill AB (1965). The environment and disease: Association
or causation? Proc R Soc Med, 58: 295300.
PMID:14283879
Hoel DG, Kaplan NL, Anderson MW (1983). Implication
References of nonlinear kinetics on risk estimation in carcino-
genesis. Science, 219: 10321037. doi:10.1126/sci-
Bieler GS & Williams RL (1993). Ratio estimates, ence.6823565 PMID:6823565
the delta method, and quantal response tests for Huff JE, Eustis SL, Haseman JK (1989). Occurrence and
increased carcinogenicity. Biometrics, 49: 793801. relevance of chemically induced benign neoplasms in
doi:10.2307/2532200 PMID:8241374 long-term carcinogenicity studies. Cancer Metastasis
Breslow NE & Day NE (1980). Statistical methods in can- Rev, 8: 122. doi:10.1007/BF00047055 PMID:2667783
cer research. Volume I - The analysis of case-control IARC (1977). IARC Monographs Programme on the
studies. IARC Sci Publ, 32: 5338. PMID:7216345 Evaluation of the Carcinogenic Risk of Chemicals to
Breslow NE & Day NE (1987). Statistical methods in cancer Humans. Preamble (IARC Intern Tech Rep No. 77/002)
research. Volume IIThe design and analysis of cohort IARC (1978). Chemicals with Sufficient Evidence of
studies. IARC Sci Publ, 82: 1406. PMID:3329634 Carcinogenicity in Experimental Animals IARC
Buffler P, Rice J, Baan R et al. (2004). Workshop on Monographs Volumes 117 (IARC Intern Tech Rep No.
Mechanisms of Carcinogenesis: Contributions of 78/003)
Molecular Epidemiology. Lyon, 1417 November
29
IARC MONOGRAPHS 108
IARC (1979). Criteria to Select Chemicals for IARC Tomatis L, Aitio A, Wilbourn J, Shuker L (1989). Human
Monographs (IARC Intern Tech Rep No. 79/003) carcinogens so far identified. Jpn J Cancer Res, 80: 795
IARC (1982). Chemicals, industrial processes and 807. PMID:2513295
industries associated with cancer in humans (IARC Toniolo P, Boffetta P, Shuker DEG et al., editors (1997).
Monographs, Volumes 1 to 29). IARC Monogr Eval Proceedings of the workshop on application of bio-
Carcinog Risk Chem Hum Suppl, 4: 1292. markers to cancer epidemiology. Lyon, France, 2023
IARC (1983). Approaches to Classifying Chemical February 1996. IARC Sci Publ, 142: 1318.
Carcinogens According to Mechanism of Action (IARC Vainio H, Magee P, McGregor D, McMichael A, editors
Intern Tech Rep No. 83/001) (1992). Mechanisms of carcinogenesis in risk identi-
IARC (1987). Overall evaluations of carcinogenicity: an fication. IARC Working Group Meeting. Lyon, 1118
updating of IARC Monographs volumes 1 to 42. IARC June 1991. IARC Sci Publ, 116: 1608.
Monogr Eval Carcinog Risks Hum Suppl, 7: 1440. Vainio H, Wilbourn JD, Sasco AJ et al. (1995).
PMID:3482203 [Identification of human carcinogenic risks in IARC
IARC (1988). Report of an IARC Working Group to Review monographs.] Bull Cancer, 82: 339348. PMID:7626841
the Approaches and Processes Used to Evaluate the Vineis P, Malats N, Lang M etal., editors (1999). Metabolic
Carcinogenicity of Mixtures and Groups of Chemicals Polymorphisms and Susceptibility to Cancer. IARC Sci
(IARC Intern Tech Rep No. 88/002) Publ, 148: 1510. PMID:10493243
IARC (1991). A Consensus Report of an IARC Monographs Wilbourn J, Haroun L, Heseltine E etal. (1986). Response
Working Group on the Use of Mechanisms of of experimental animals to human carcinogens: an
Carcinogenesis in Risk Identification (IARC Intern analysis based upon the IARC Monographs pro-
Tech Rep No. 91/002) gramme. Carcinogenesis, 7: 18531863. doi:10.1093/
IARC (2005). Report of the Advisory Group to Recommend carcin/7.11.1853 PMID:3769134
Updates to the Preamble to the IARC Monographs
(IARC Intern Rep No. 05/001)
IARC (2006). Report of the Advisory Group to Review the
Amended Preamble to the IARC Monographs (IARC
Intern Rep No. 06/001)
IARC (2004). Some drinking-water disinfectants and
contaminants, including arsenic. IARC Monogr Eval
Carcinog Risks Hum, 84: 1477. PMID:15645577
McGregor DB, Rice JM, Venitt S, editors (1999). The use
of short- and medium-term tests for carcinogens and
data on genetic effects in carcinogenic hazard evalua-
tion. Consensus report. IARC Sci Publ, 146: 1536.
Montesano R, Bartsch H, Vainio H etal., editors (1986).
Long-term and short-term assays for carcinogenesis
a critical appraisal. IARC Sci Publ, 83: 1564.
OECD (2002). Guidance Notes for Analysis and Evaluation
of Chronic Toxicity and Carcinogenicity Studies (Series
on Testing and Assessment No. 35), Paris: OECD
Peto R, Pike MC, Day NE et al. (1980). Guidelines for
simple, sensitive significance tests for carcinogenic
effects in long-term animal experiments. IARC Monogr
Eval Carcinog Risk Chem Hum Suppl, 2: 311426.
PMID:6935185
Portier CJ & Bailer AJ (1989). Testing for increased
carcinogenicity using a survival-adjusted quan-
tal response test. Fundam Appl Toxicol, 12: 731737.
doi:10.1016/0272-0590(89)90004-3 PMID:2744275
Sherman CD, Portier CJ, Kopp-Schneider A (1994).
Multistage models of carcinogenesis: an approximation
for the size and number distribution of late-stage clones.
Risk Anal, 14: 10391048. doi:10.1111/j.1539-6924.1994.
tb00074.x PMID:7846311
Stewart BW, Kleihues P, editors (2003). World Cancer
Report, Lyon: IARC
30
GENERAL REMARKS
1. Introduction specific therapeutic usage, a major aspect of the

use of drugs worldwide involves herbal products.
This one-hundred-and-eighth volume of the Estimates from WHO indicate that 80% of the
IARC Monographs includes evaluations of the worlds population has used herbal products as
carcinogenic hazard to humans of exposure to 14 medicines. Use of the term herbal medicine
herbal products or pharmaceutical drugs. None is arbitrary in many contexts. In particular, a
of these, except hydrochlorothiazide, have been wide variety of pharmaceutical drugs, that is,
previously evaluated by the Working Group. agents recognized as having a particular phar-
Hydrochlorothiazide a pharmaceutical macological mode of action and associated
drug was considered in 1989 by an IARC clinical benefit, are derived from plants or other
Monographs Working Group (IARC, 1990), and natural sources. This category of agent is likewise
was evaluated as not classifiable as to its carcino- represented in previous IARC Monographs; and
genicity to humans (Group 3) based on inadequate one aristolochic acid has been classified as a
evidence for carcinogenicity in humans and in Group 1 agent (IARC, 2012).
experimental animals. Pharmaceutical drugs are subject to strict
A summary of the findings of this meeting regulation in most countries, and their avail-
appears in The Lancet Oncology (Grosse et al., ability is highly restricted. This may not be the
2013). case with materials used in the preparation of
Among the agents that are known to cause herbal medicines. The therapeutic benefit of
cancer in humans specifically, there are several such herbal products may have been recognized
pharmaceuticals and other drugs. Volume 100A in certain communities for centuries. Moreover,
of the IARC Monographs (IARC, 2012) reviewed herbal products are available in several regula-
pharmaceuticals that in previous evaluations tory paradigms, ranging from foods and dietary
had been categorized as carcinogenic to humans supplements to cosmetics and over-the-counter
(Group 1), primarily on the basis of epidemiolog- (non-prescription) and prescription drugs.
ical evidence for causation. In respect of specific Worldwide, this means that product quality and
chemical carcinogens, the number of agents clas- composition may vary from country to country
sified as carcinogenic to humans that are thera and within countries, even when different prod-
peutic drugs is second only to the number of ucts bear the same name. In addition, the use of
agents that have been identified in the context of particular herbal products may vary markedly
occupational exposures. between countries and between communities
Apart from pharmaceutical drugs that are within a country.
industrially produced agents identified with a
31
IARC MONOGRAPHS 108
2. Exposure to herbal products Monographs are specified with reference to indi-

and pharmaceuticals vidual components known to occur in particular
plants, as is the case for pulegone and digoxin.
Certain previous IARC Monographs evaluations
Herbal products are complex mixtures
are immediately relevant to the present evalua-
that originate from biological sources. Unlike
tions to the extent that they involve components
single-entity pharmaceuticals, plants contain
(e.g. quercetin for Ginkgo biloba, anthraquinones
thousands of primary and secondary metabolic
for Aloe vera) or metabolites (e.g. phenobarbital
constituents. In addition, raw materials are inher-
for primidone) of agents considered in the present
ently variable because their chemical compo-
volume.
sition depends on factors such as geographical
Over the past several decades, there has been
origin, weather, harvesting practices, while the
a revolution in the production, sale, and use of
chemical composition of the finished herbal
herbal products. In the 1970s, botanicals were
products may not match that of the parent plants,
largely sifted, cut, or powdered plant material
and products frequently contain multiple botan-
in the form of a tablet, capsule, tea, or tincture.
ical ingredients.
More recently, herbal products are often derived
Discussions of exposure to natural products
from intensely processed, carefully controlled
can be complicated by several factors. The first
organic extracts of plant material that have been
is the market category in which the product
spray-dried onto a solid carrier or diluent and
falls. Herbal products can be sold as conven-
then formed into a hard or soft capsule or tablet.
tional foods or food additives (e.g. flavouring
The goal of many such processes is to create
or colouring agents), as dietary supplements,
standardized extracts adjusted to contain
as cosmetic ingredients, or as herbal medicines
consistent amounts of selected compounds
(various national regulatory schemes may clas-
of interest. Unfortunately, most standardized
sify these as natural health products, therapeutic
extracts focus on one or a handful of the thou-
goods, phytomedicines, herbal medicinal prod-
sands of constituents of the whole plant, so that
ucts, traditional medicines, or conventional
even standardized extracts that are created using
drugs). There may also be use of self-collected
different processing techniques (e.g. different
plants that are not marketed products.
solvents, different ratios of plant to solvent) may
Herbal medicine preparations are herbal
achieve the desired levels of the desired chemical
products and consequently constitute complex
constituents while being otherwise chemically
mixtures. The biological impact, and specifi-
dissimilar. Attempts to compare herbal products
cally the carcinogenicity of complex mixtures,
by viewing the entire phytochemical fingerprint
may be addressed by consideration of informa-
are beginning to appear, but these techniques
tion concerning the mixture, and its variability
have not yet had time to have an impact on the
in different contexts, and also by consideration
market or the publicly available scientific litera-
of information concerning biologically active
ture (van Beek & Montoro, 2009).
components within such mixtures. Information
There are several advantages to using such
relevant to possible carcinogenicity may be most highly processed raw materials. These include
adequately addressed with reference either to the the ability to produce dosage forms that are more
mixture or to the active component(s). Therefore, uniform in their composition, and the ability to
some Monographs in the present volume are spec- preferentially concentrate the desirable constitu-
ified with reference to the plant itself, i.e. Aloe ents of a plant while leaving behind undesirable
vera, Ginkgo biloba, goldenseal, or kava. Other constituents. Because products are frequently
32
General remarks
referred to generically by the name of the plant tested from among dozens or hundreds of prod-
in marketing and consumer-use surveys, it is ucts with similar or identical names but widely
difficult to differentiate between exposure to divergent compositions remains a major obstacle.
the crude plant material or to unique, highly As with most herbal products, there may be
processed proprietary extracts that differ signifi- some controversy surrounding generalizability
cantly from both the plant source and from other of conclusions for a commercial entity, because
proprietary products. In countries where there is commercial products are very diverse in terms
pre-market review and product licensing, prod- of processing, composition, and intended use.
ucts must often conform to published composi- Attempts to identify the predominant form of
tional standards, such as those in the United States an herbal product in the market place are pure
Pharmacopoeia or the European Pharmacopoeia; conjecture in the absence of data. This is a recur-
and similarly named products marketed in this ring theme for all discussions on herbal products.
regulatory environment are likely to be relatively The ability of the Working Group to gauge
similar to each other in composition, but may be the extent of global exposure to herbal products
very dissimilar from products that do not meet was very limited, since the quality and quantity
such standards. of data available were inconsistent across coun-
In addition to the broad variability in compo- tries. Having better information on patterns of
sition of herbal products that are available to use and on product composition would provide
consumers, problems in interpreting published a means to prioritize the herbal products consid-
scientific studies of herbal products have been ered in this volume for such activities as policy
reported. Wolsko et al. (2005) performed a formulation or further research needs.
systematic review of the Materials and Methods While the available information on expo-
sections of 81 published studies on herbal prod- sure to pharmaceuticals was more abundant and
ucts. They noted that only 12 (15%) of the studies accessible than that on herbal products, limita-
reported any kind of quantitative chemical anal- tions remain. For the most part, information on
ysis of the study material, and that only 8 (10%) prescribing patterns outside the USA was not
of those reporting analysis reported results of available to the Working Group. In addition,
the analysis. In addition, only 40 of the studies published studies indicated that patterns of adher-
(49%) provided the Latin binomial name of the ence and persistence are suboptimal for medica-
study material, only 8 (10%) identified the part tions used to manage or treat chronic conditions.
of the plant used, and only 23 (28%) described And while prescribing patterns are available for
the extraction/processing method used to create
some drugs, such data do not exist for over-the-
the product. A larger review by Gagnier et al.
counter drugs; consequently, exposure estimates
(2011) reported similar findings. To prevent such
must be made using means similar to those used
problems in future studies, Swanson (2002) and
to estimate exposure to herbal products (e.g. Aloe
Gagnier et al. (2006) have published guidelines
vera, for which over-the-counter use is difficult
for the reporting of studies on natural products.
to quantify), namely sales data and consumer use
While some organizations that conduct safety
surveys. Although not widely available or widely
studies adhere to or surpass the above guide-
accessible, such data for over-the-counter drugs
lines when selecting test articles or designing
is more informative than for herbal products sold
studies, it would be useful if these guidelines
as foods or dietary supplements because drug
became standard practice, as the reproducibility
products with similar names are required to be
and reliability of safety studies would be greatly
similar in composition.
enhanced. Unfortunately, while such recom-
mendations are useful, selecting the article to be
33
IARC MONOGRAPHS 108
As indicated above, exposure can generally adopted within a particular Monograph. In some
be much more accurately measured for pharma- instances, studies may generate epidemiological
ceuticals than for other agents, and therapeutic data that refer to the use of particular classes of
doses used in humans are often closer to those drug, rather than particular individual drugs.
tested in experimental animals. Nonetheless, Interpretation of such data to infer effects attrib-
characterizing the true nature of exposure to utable to particular drugs, such as pioglitazone,
drugs in relation to carcinogenicity is compli- rosiglitazone and hydrochlorothiazide, and
cated by the variability in adherence to drugs the small number of available epidemiological
and their varying patterns of use intermittent studies, may render this task difficult or almost
versus continuous. impossible.
Exposure to herbal products or pharma- Historically, in IARC Monographs evalua-
ceutical drugs may occur as a consequence of tions for which relevant epidemiological data
occupational exposure of people involved in were available, determination of causality on the
production or manufacture of these agents. basis of associations reported in epidemiological
Exposure may also occur as a result of water studies has always been recognized as both chal-
pollution by these agents. Generally, levels of lenging and of critical importance. In general
occupational or environmental exposure are terms, this subject is addressed in the Preamble
much lower than levels of exposure experienced to the IARC Monographs, and the matters raised
by people using the respective herbal products or in that context are fundamental to all such epide-
drugs. Almost no information was available to miological data. In the specific case of epidemio
the Working Group concerning occupational or logical findings in relation to pharmaceutical
environmental circumstances of exposure to the drugs, it is self-evident that the exposed indi-
agents evaluated in this volume. viduals are not a representative sample of the
community, but rather are individuals identified
by a diagnosis in consequence of which they have
3. Epidemiological studies of received the drug in question. At one extreme,
populations using drugs increased risk of cancer in such individuals may
be caused by the drug they have received. At the
other extreme, an association between increased
The Monograph on digoxin exemplifies the
risk of cancer and use of a particular drug may
complications in drug nomenclature that may
be totally independent of causality and arise for
arise due to differences in professional prac-
several reasons: because patients with a particular
tice and disciplines (e.g. manufacturer, medical
disease are at greater risk of malignancy; because
professional). As explained therein, the term
patients with a particular disease are more liable
digitalis as used with reference to chemical
than the community in general to be exposed to
specifications may refer to a plant extract, while
an independent factor that causes or is correlated
the same word in a medical therapeutic context
with increased risk of cancer; or because the
may refer to a particular category of agents (e.g.
symptoms of an undiagnosed cancer may also
digoxin, digitoxin). Such incongruities not only
prompt the use of a drug, which can subsequently
contribute to potential misunderstanding of
be suspected as its cause. An additional problem
data; in an immediate sense, they may compli-
is that patients commonly receive more than one
cate adoption of a particular term as the appro-
drug, and determination of the carcinogenicity
priate identification of the subject of a Monograph
of any single drug may be difficult.
and/or the subject of evaluation statements
34
General remarks
4. Extrapolating from specific 5. Considerations beyond hazard

scientific findings identification
While historically, multiple studies of carcino- Many (if not most) regulatory decisions
genicity in experimental animals may have been concerning putative carcinogens necessitate
conducted on a single test agent in several inde- consideration not only of perceived hazard, but
pendent laboratories, today the massive expense also of potential benefit. It is crucial, therefore,
involved in rigorous testing for carcinogenicity that regulatory decisions affecting drug availa-
in experimental animals often means that only bility include assessment not only of potential
one or two well conducted studies of carcino- carcinogenicity (and other adverse effects), but
genicity (often in one strain of rat and/or one also of the health benefits derived from their
strain of mouse, and typically involving males usage.
and females) may be available in the peer-re-
viewed literature or from government agency
reports that are publicly available. As indicated References
in the Preamble, such studies, ideally conducted
under good laboratory practice, may be able to Gagnier JJ, Boon H, Rochon P, Moher D, Barnes
establish sufficient evidence of carcinogenicity J, Bombardier C; CONSORT Group(2006).
in experimental animals, depending upon the Recommendations for reporting randomized
nature of results obtained. controlled trials of herbal interventions: Explanation
and elaboration. J Clin Epidemiol, 59(11):113449.
Again, in relation to single studies as doi:10.1016/j.jclinepi.2005.12.020 PMID:17027423
outlined above, when the test agent is a complex Gagnier JJ, Moher D, Boon H, Beyene J, Bombardier C
mixture, exemplified, for example, by an herbal (2011). Randomized controlled trials of herbal inter-
ventions underreport important details of the inter-
product, it may not be possible to assume that vention. J Clin Epidemiol, 64(7):7609. doi:10.1016/j.
the agent being tested is identical to either the jclinepi.2010.10.005 PMID:21208777
material marketed under the same name and/or Grosse Y, Loomis D, Lauby-Secretan B, El Ghissassi F,
material tested in other studies that might other- Bouvard V, Benbrahim-Tallaa L et al. International
Agency for Research on Cancer Monograph Working
wise be understood to indicate possible mecha- Group (2013). Carcinogenicity of some drugs and
nism(s) of carcinogenesis or to exclude particular herbal products. Lancet Oncol, 14(9):8078. doi:10.1016/
mechanism(s) of carcinogenesis. S1470-2045(13)70329-2 PMID:24058961
As indicated in the Preamble, the IARC IARC (1990). Hydrochlorothiazide. IARC Monogr Eval
Carcinog Risks Hum, 50:293305. PMID:2292804
Monographs evaluations are wholly dependent IARC (2012). Pharmaceuticals. Volume 100 A. A review of
on publicly available data that are exemplified by human carcinogens. IARC Monogr Eval Carcinog Risks
published research results in the peer-reviewed Hum, 100:Pt A: 1401. PMID:23189749
Swanson CA (2002). Suggested guidelines for articles
literature. This information comprises only a about botanical dietary supplements. Am J Clin Nutr,
subset of data on pharmaceutical drugs, specif- 75(1):810. PMID:11756054
ically excluding commercial in-confidence van Beek TA, Montoro P (2009). Chemical analysis and
findings of the type provided by industry to quality control of Ginkgo biloba leaves, extracts, and
phytopharmaceuticals. J Chromatogr A, 1216(11):2002
national or multinational regulatory authorities 32. doi:10.1016/j.chroma.2009.01.013 PMID:19195661
in the context of applications to market particular Wolsko PM, Solondz DK, Phillips RS, Schachter SC,
drugs. The initiatives of the European Medicines Eisenberg DM (2005). Lack of herbal supplement
Agency and other organizations to make such characterization in published randomized controlled
trials. Am J Med, 118(10):108793. doi:10.1016/j.
data publicly available are properly noted in this amjmed.2005.01.076 PMID:16194636
context.
35
ALOE VERA
1. Exposure Data A glossary of commonly used terms for Aloe

vera products is provided in Table1.1.
The first record of human use of Aloe vera is in
Sumerian hieroglyphics engraved on clay tablets 1.1 Identification of the agent
during the Mesopotamia civilization circa 2200
BC, in which it is described as a laxative. Use 1.1.1 Botanical data
of aloe in ancient times is also documented in (a) Nomenclature
Egypt, Greece, and China. Aloe vera was culti-
vated on the islands of Barbados and Curacao For details on botanical nomenclature, see
in the Caribbean by Spain and the Netherlands, Newton (2004).
and was sold in various parts of Europe during Chem. Abstr. Serv. Reg. No.: 8001-97-6
the 17th century (Park & Jo, 2006). Commercial Chem. Abstr. Name: Aloe barbadensis
cultivation of Aloe vera in the USA began in the
Botanical name: Aloe vera (L.) Burm. f.
1920s in Florida (Grindlay & Reynolds, 1986).
(synonym, Aloe barbadensis, Aloe humilis
Although Aloe vera originated in the warm, dry
Blanco, Aloe indica Royle, nomen nudum,
climates of Africa, the plant is readily adaptable
Aloe perfoliata var. vera L., Aloe vulgaris
and grows worldwide (Steenkamp & Stewart,
Lam.) (GRIN, 2013).
2007).
Use of Aloe vera gel extracts in health foods Family: Xanthorrhoeaceae
and beverages, and moisturizing cosmetics, Genus: Aloe
began during the 1970s, starting in the USA and Plant part: Leaf
parts of Europe (Park & Jo, 2006). Historically, Common names: Aloe vera; Aloe vera Linn;
Aloe vera was used topically to heal wounds and True aloe; Aloe barbadensis; Barbados
for various skin conditions, and orally as a laxa- aloe; Curaao aloe; Mediterranean aloe;
tive (Steenkamp & Stewart, 2007). The dried latex Ghritakumari; Lu Hui; Luhui, etc.
of other Aloe species, such as Aloe ferox Miller
(Cape aloe or bitter aloe) has also been used as From WHO (1999), Eur Ph (2008), ONeil
a laxative (EMA, 2006). Today, Aloe vera is also et al. (2006), SciFinder (2013), IASC (2013a),
used as a folk or traditional remedy for a variety Boudreau et al. (2013a).
of conditions and is found in some dietary
supplements and food products. Aloe vera gel can
be found in hundreds of skin products, including
lotions and sunblocks (NCCAM, 2012).
37
IARC MONOGRAPHS 108
Table 1.1 Definition of terms commonly used in the Aloe industry
Term Definition
Leaf The part of the Aloe vera plant used in commerce, where processing is begun without stripping off the rind.
Whole leaf Historically used to describe products derived from the entire leaf that were filtered/purified. However, use of
this terminology without adequate additional descriptors is not recommended. This terminology is now seen
on products or in reference to raw material where the entire leaf is used as a starting ingredient to create Aloe
vera juice.
Decolorized A process, usually involving filtration with activated charcoal, that clarifies the liquid aloe mass.
whole leaf
Inner leaf Plant part used to describe the clear, central parenchymatous tissues of the aloe leaf.
Aloe latex Brown, yellow-brown, or occasionally red exudate found between the rind and inner leaf. Also called sap, it
contains several constituents, but most notably anthraquinones.
Anthraquinone An organic compound primarily found in the aloe latex, whose structure serves as the basic building block
for several naturally occurring plant pigments. The substance is commonly used for laxative purposes.
Gel Liquid product typically derived from the inner leaf.
Juice Liquid product derived from Aloe vera leaf [the Working Group noted that the term juice is used arbitrarily
and may either apply to products from the latex or from the gel].
Adapted from IASC (2009)
(b) Description Aloe vera plants contain two major liquid

Aloes are perennial succulents or xero- materials (Fig. 1.2): first, a bitter yellow latex
phytes; they can adapt to habitats with low or located under the strongly cutinized epidermis
erratic water availability, are characterized by of the leaves in the vascular layer and containing
the capacity to store large volumes of water in a high concentration of anthraquinone
their tissue, and are able to use crassulacean acid compounds, which has been used throughout
metabolism, an adaptation to the photosynthetic the centuries as a cathartic and for medicinal
pathway that involves the formation of malic purges; and, second, a clear mucilaginous gel
acid (Boudreau et al., 2013a). Aloe plants, such produced by the thin-walled tubular cells in the
as Aloe vera (Fig.1.1), all have green fleshy leaves inner central zone (parenchyma) that has been
covered by a thick cuticle or rind, under which used since ancient times to treat burns and other
is a thin vascular layer covering an inner clear wounds, where it is thought to increase the rate of
pulp (Boudreau et al., 2013a; Fig.1.2) The leaves healing and reduce the risk of infection (Joseph
are 3050 cm in length and 10 cm in width at the & Raj, 2010). A third liquid may also be obtained
base, pea-green in colour (when young spotted by macerating the whole leaf.
with white), and with bright yellow tubular [Both the scientific and the lay literature
flowers 2535 cm in length arranged in a slender (e.g. on internet sites) are extremely inconsistent
loose spike (WHO, 1999). when referring to products obtained from Aloe
The vascular bundles, located within the leaf vera. The problem starts with the fact that the
pulp, transport (i) water and minerals from the three types of liquids that are obtained from Aloe
roots to the leaves; (ii) synthesized materials to vera leaves are interchangeably referred to as
the roots; and (iii) latex along the margins of Aloe juice, which has caused confusion in the
the leaf for storage (Ni et al., 2004; Fig.1.2). The literature. For disambiguation reasons, the term
number of vascular bundles varies depending on Aloe juice should be restricted if used at all
the size of the leaves and the age of the plant (Ni to the latex material of the pericycle, which is in
et al., 2004). accordance with the pharmacopoeial definitions
38
Aloe vera
Fig.1.1 Aloe vera (L.) Burm. F, plant and flower The main feature of the Aloe vera plant is its
high water content, ranging from 99% to 99.5%,
while the remaining 0.51.0% solid material is
reported to contain over 200 different poten-
tially active compounds, including vitamins,
minerals, enzymes, simple and complex poly-
saccharides, phenolic compounds, and organic
acids (Boudreau et al., 2013a; Rodrguez et al.,
2010).
In compositional studies on the structural
components of leaf portions of the Aloe vera
plant, the rind was found to compose 2030%
and the pulp 7080% of the whole leaf weight.
On a dry-weight basis, the rind and pulp contain
2.7% and 4.2% lipids, and 6.3% and 7.3%
proteins, respectively (Femenia et al., 1999). The
percentages of soluble sugars (11.2% and 16.5%),
primarily as glucose, and the percentages of ash
(13.5% and 15.4%), in particular calcium, were
relatively high in the rind and pulp, respectively.
Non-starch polysaccharides and lignin repre-
sented the bulk of each leaf fraction and were
found to be 62.3% and 57.6% of the dry weight of
the rind and pulp, respectively (Boudreau et al.,
2013a). Acetylated mannan is the primary poly-
From Spohn (2013) saccharide in Aloe vera gel (Ni et al., 2004). Other
Roland Spohn chemical constituents of Aloe vera include lectins
such as aloctins A and B (Kuzuya et al., 2004).
(WHO, 1999; Eur Ph, 2008; JP XVI, 2011); and The physical and chemical constituents of
the inner leaf liquid material should be referred the products derived from Aloe vera plants differ
to as gel (WHO, 1999). Interchangable terms depending on the source (e.g. part of the plant),
found in the literature for the gel are inner the species of the plant, the climate conditions,
pulp, mucilage tissue, mucilaginous gel, muci- seasonal and grower influences (Boudreau et al.,
laginous jelly, inner gel, and leaf parenchyma 2013a), and processing techniques (Waller et al.,
tissue (Hamman, 2008).] 2004).
1.1.2 Chemical constituents and their 1.1.3 Technical and commercial products
properties
Three types of Aloe vera extracts can be
A review of the chemistry of Aloe vera was distinguished gel extract, whole leaf extract,
provided by Reynolds (2004), and a summary of and decolorized whole leaf extract (Boudreau
the chemical constituents of Aloe vera is provided et al., 2013a), and a fourth type of commercial
in Table1.2. material is available as dried latex, which has
39
IARC MONOGRAPHS 108
Fig.1.2 Schematic representation of the Aloe vera plant, showing a cross-section through a leaf
The green rind or cuticle of the
The perimeter of the Aloe vera Aloe vera plant consists of
leaf pulp is interspersed with multiple layers interspersed with
vascular bundles that are chloroplasts.
composed of three types of
tubular structures: the xylem,
the phloem, and the pericyclic
tubules. The pericyclic tubules
transport the Aloe latex.
The Aloe vera inner leaf pulp is

composed of large thin-walled
parenchyma cells that store the
Aloe gel.
Aloe vera Plant
From Boudreau et al. (2013a)
been traditionally used as the laxative (Eur Ph, and, because there is considerably more mannose
2008). present than glucose, the molecules are referred
to as polymannans. These linear chains range
(a) Aloe vera gel extract in size from a few to several thousand mono-
The inner leaf pulp of the Aloe vera plant saccharide molecules. The major polysaccha-
contains large, thin-walled cells that produce gel, ride, acetylated mannan, is composed of one or
the clear, mucilaginous, and aqueous extract of more polymers of various chain lengths with
the inner central area of the leaf pulp (Fig.1.2). molecular weights ranging from 30 to 40 kDa
Aloe vera gel serves as the water and energy or greater, and consisting of repeating units of
storage component of the plant. The mechanical glucose and mannose in a 1:3 ratio (Channe
extrusion of the mucilaginous gel from the inner Gowda et al., 1979; Mandal & Das, 1980; Yaron,
leaf pulp gives a 70% yield with a water content 1993; Femenia et al., 1999; Boudreau et al., 2013a;
of 9999.5% (Femenia et al., 1999). Fig. 1.3). Chemically preserved fresh Aloe vera
Polysaccharides in Aloe vera gel consist of gel stored at room temperature or incubated at
linear chains of glucose and mannose molecules, 40 C for 48 hours exhibited degradation in its
40
Aloe vera
Table 1.2 Summary of chemical constituents of Aloe vera products
Class Compounds
Anthraquinones/ Aloe-emodin, aloetic acid, anthranol, aloin A and B (or collectively known as barbaloin),
anthrones isobarbaloin, emodin, ester of cinnamic acid
Carbohydrates Pure mannan, acetylated mannan, acetylated glucomannan, glucogalactomannan, galactan,
galactogalacturan, arabinogalactan, galactoglucoarabinomannan, pectic substance, xylan, cellulose
Chromones 8-C-Glucosyl-(2-O-cinnamoyl)-7-O-methylaloediol A, 8-C-glucosyl-(S)-aloesol, 8-C-glucosyl-7-
O-methyl-(S)-aloesol, 8-C-glucosyl-7-O-methylaloediol, 8-C-glucosyl-noreugenin, isoaloeresin D,
isorabaichromone, neoaloesin A
Enzymes Alkaline phosphatase, amylase, carboxypeptidase, catalase, cyclooxidase, cyclooxygenase, lipase,
oxidase, phosphoenolpyruvate carboxylase, superoxide dismutase
Minerals Calcium, chlorine, chromium, copper, iron, magnesium, manganese, potassium, phosphorous,
sodium, zinc
Lipids and Arachidonic acid, -linolenic acid, steroids (campestrol, cholesterol, -sitosterol), triglycerides,
miscellaneous organic triterpenoid, gibberillin, lignins, potassium sorbate, salicylic acid, uric acid
compounds
Amino acids Alanine, arginine, aspartic acid, glutamic acid, glycine, histidine, hydroxyproline, isoleucine,
leucine, lysine, methionine, phenylalanine, proline, threonine, tyrosine, valine
Proteins Lectins, lectin-like substance
Saccharides Mannose, glucose, L-rhamnose, aldopentose
Vitamins B1, B2, B6, C, -carotene, choline, folic acid, -tocopherol
Adapted from Hamman (2008)
rheological properties, a decrease in the content largely phenolic in nature, and many are anth-
and composition of polysaccharides, and a raquinone C-glycosides, anthrones, and free
substantial increase in the mannose:glucose anthraquinones (Park et al., 1998). The levels of
ratio, from 2.9 in the fresh gel to 13.4 in the incu- anthraquinone C-glycosides in Aloe vera latex
bated gel (Yaron, 1993). are quite variable; however, they may constitute
up to 30% of the dry weight of the latex (Groom
(b) Aloe vera whole leaf extract & Reynolds, 1987). Aloe vera latex contains
The Aloe vera whole leaf extract (sometimes four major C-glycosyl constituents: aloin A,
referred to as whole leaf Aloe vera juice, Aloe aloin B, aloesin, and aloeresin A (Fig.1.3; Sacc
juice or nondecolorized whole leaf extract), is the et al., 2001). Aloin A, a C-glycosyl anthrone, also
aqueous extract of the whole leaf with lignified referred to as barbaloin, is the major component
fibres removed. The whole leaf extract contains of aloe latex. Aloin A and its epimer, aloin B, also
both the gel from the inner parenchyma leaf pulp referred to as isobarbaloin, have a 9-anthrone
and the latex. The restricted distribution of the skeleton and a -D-glucopyranosyl substit-
bitter latex within the margins of the leaves of uent. Aloesin, also known as aloeresin B, is a
the Aloe vera plant suggests that this thin layer 5-methyl chromone with an 8--D-glucopyra-
is the primary site of secondary metabolites nosyl substituent, and aloeresin A is a 5-methyl
biosynthesis: compounds that do not function chromone with an 8--D-glucopyranosyl-2-
directly in plant growth and development and O-trans-p-coumarol substituent. Several other
serve as a plant defence strategy (Boudreau et al., C-glycosyl-chromones and anthrones have been
2013a). A wide variety of secondary compounds isolated from Aloe vera, including aloe-emodin,
have been isolated from the Aloe vera latex the anthraquinone of barbaloin and isobarbaloin
(Reynolds, 2004). The isolated compounds are (Boudreau et al., 2013a).
41
Fig.1.3 Chemicals present in gel and latex prepared from Aloe vera
42
Aloe vera whole leaf
Aloe vera latex
OH O OH OH O OH
8 9 1
7 2
6 3 OH
OH
5 10 4
IARC MONOGRAPHS 108
H
HO H HO
Aloe vera gel
O O
OAc OH OH OH
A cO OAc OH
A cO
HO HO OH
OH
O O OH OH
A loin A (Barbaloin) A loin B (Isobarbaloin)

O O O O O
O O O O
OAc OH OAc OH
O O
OH OH
A cetylated mannan (A cemannan) HO O HO O

HO HO
O O
OH
OH OH
OH OH
O OH
A loeresin A A loesin (A loeresin B )

O
HO
O O
HO
O O
OH
OH O
OH
A loenin
Ac, acetyl group

From Boudreau et al. (2013a)
Aloe vera
The occurrence in Aloe vera latex of endoge- et al. (2013) was reported to contain combined
nous free anthraquinones and anthrones results Aloin A and Aloin B at <0.1 ppm.
from oxidative processes acting on the glycosides Although Aloe vera gel and the decolor-
rather than from metabolic synthesis (Boudreau ized whole leaf extract are similar in that each
et al., 2013a). In addition, the latex from Aloe vera contain little or no latex anthraquinones, carbon
contains several aromatic compounds, such as adsorption changes the physical and chemical
aldehydes and ketones (Sacc et al., 2001). The properties of the whole leaf extract. Aloe vera
sugar moiety in aloins is D-glucose, and studies decolorized whole leaf extract differs from the
indicate that carbon atom 1 of the D-glucose gel in that it exhibits a degradation in rheological
moiety is linked directly to carbon atom 10 of the properties and a loss of approximately 1923% of
anthracene ring in a -configuration (Fig. 1.3). the complex polysaccharide content (Pelley et al.,
The carboncarbon bond is quite resistant to acid 1998).
and alkaline conditions; however, the intestinal
microflora of humans and animals have been (d) Dried Aloe vera latex (pharmaceutical
shown to cleave the -C-glucosyl bond, although material)
considerable variation in response among animal The dried Aloe vera latex is the solidified
species occurs. Cleavage of the -C-glucosyl liquid originating in the cells of the pericycle and
bond results in the formation of aloe-emodin, adjacent leaf parenchyma, and flowing sponta-
the cathartic principle of the latex, and other free neously from the cut leaf, allowed to dry with
anthraquinones and anthrones (Boudreau et al., or without the aid of heat (WHO, 1999). The
2013a; see Section 4.1.1b). In commercial prod- material is used for medicinal purposes and its
ucts containing whole leaf extract, a rapid dete- composition is specified in several official phar-
rioration of aloin was detected during storage, macopoeias (see Section 1.6).
especially at higher temperatures (Pellizzoni
et al., 2011).
1.2 Analysis
(c) Aloe vera decolorized whole leaf extract
For Aloe vera sold for medicinal purposes,
Activated carbon treatment of the Aloe vera analyses are defined in pharmacopoeial mono-
whole leaf extract is used to remove bitterness graphs (see Section 1.6). Most of the published
and colour caused by the anthraquinone compo- analytical methods (Table 1.3) deal with the
nents of the latex. This results in a product determination of the anthraquinone compounds
termed decolorized whole leaf extract that has in the latex, and fewer and mostly qualitative
quite different properties from the whole leaf methods are available for authentication.
extract. Aloe vera decolorized whole leaf extract To carry out an exhaustive quality control of
is also referred to as whole leaf Aloe vera gel commercial Aloe vera gel products (e.g. for food
(Boudreau et al., 2013a). Dentali (2013) noted or cosmetic uses), the following analyses should
that an industry standard for aloin content of be carried out: (i) investigation of authenticity;
decolorized Aloe vera whole leaf extract is <10 (ii) test for identification of additives (to control
ppm. Sehgal et al. (2013) reported results of the labelling or regulatory limits); and (iii) deter-
toxicological assessment of a commercial decol- mination of the aloin content (Lachenmeier et al.,
orized whole leaf extract that contained approx- 2005; Rodrguez et al., 2010). The investigation
imately Aloin A at 0.9 ppm, Aloin B at 1.3 ppm, of authenticity aims at confirming the amount
and aloe-emodin at 0.2 ppm. A decolorized Aloe of Aloe vera in the preparation; adulteration
vera whole leaf extract assessed for safety by Shao
43
44
Table 1.3 Selected methods of analysis of Aloe vera constituents in various matrices
Sample matrix Analyte/purpose of analysis Sample preparation Assay method Detection Reference
limit
Urine Aloin/detection of laxative abuse Glucuronidase, SPE HPTLC 1020 mg/L Perkins & Livesey (1993)
Urine Aloe-emodin/detection of laxative Glucuronidase, Extraction with HPLC/UV 0.015 mg/L Stolk & Hoogtanders
abuse chloroform/isopropanol 9+1 (1999)
Serum Aloin/pharmacokinetic study Extraction with ethyl acetate TLC 0.033 mg/L Ishii et al. (1987)
Plasma Aloe-emodin/pharmacokinetic study Dichloromethane extraction HPLC/FD 4.5 g/L Zaffaroni et al. (2003)
Aloe leave exudates Aloin/taxonomy Methanolic solution HPLC/UV NA Groom & Reynolds
IARC MONOGRAPHS 108
(1987)
Aloe vera gel 13 phenolic compounds/quality Liquidliquid extraction HPLC/UV NA Kim & Park (2006)
control and standardization
Aloe vera products Acetylated polysaccharides, glucose, None (dissolve in D2O) NMR <0.05 g/L Jiao et al. (2010)
malic acid, lactic acid, and acetic acid/
quality control
Aloe extracts Aloe-emodin/identity confirmation Preparative TLC HPLC/UV and UV: 3 g/L Mandrioli et al. (2011)
and commercial FD FD: 0.8 g/L
formulations
Aloe vera plants Metabolite profiling/metabolomics Extraction with methanol. GC-IT-MS and NA Lee et al. (2012)
Derivatization with MSTFA for UPLC-Q-TOF-
GC-IT-MS analysis MS
Aloe vera leaves Aloin derivatives/regulatory control Ultrasound-assisted extraction in HPLC-DAD, 310 mg/L Azaroual et al. (2012)
methanol HPLC-MS,
UPLC
Pharmaceutical Aloe vera polysaccharides/control for None (dissolve in D2O) NMR 2 g/L Davis & Goux (2009)
formulations adulteration or degradation
Cosmetics Mannose/determination of quantity Extraction with diethyl ether and HPTLC 3% of Aloe vera Geisser & Kratz (2010)
of Aloe in product hydrolysis with sulfuric acid in cosmetic
product
Aloe species Aloin derivatives/authenticity control Methanolic extraction HPLC/UV <0.05 mg/L Okamura et al. (1996)
Aloe exudate Volatiles/flavour characterization for Ethanol 40% for HPLC, HS HPLC, HS-GC/ NA Sacc et al. (2001)
beverage industry sampling for GC MS
Table 1.3 (continued)
Sample matrix Analyte/purpose of analysis Sample preparation Assay method Detection Reference
limit
Aloe vera beverages Profiling for identity, adulteration, None HPTLC, HS- NA Lachenmeier et al. (2005)
dilution SPME-GC/MS
Aloe species 13 Phenolic compounds/seasonal Extraction with ethanol HPLC/UV NA Park et al. (1998)
variation
Commercial aloe High molecular-weight Dilution with water and 0.2 M SEC NA Turner et al. (2004)
products polysaccharides NaCl
Commercial aloe Aloe-emodin, aloin A Extraction with ethyl acetate/ HPLC/MS Aloin-A, Elsohly et al. (2007)
products methanol 9+1 1 g/L; Aloe
emodin,
2.5 g/L
Commercial aloe Aloin A, aloin B Extraction with ethanol/water HPLC 0.06 mg/L Ramrez Durn et al.
products (90+10) (2008)
DAD, diode array detector; FD, fluorescence detection; GC, gas chromatography; HPLC, high-performance liquid chromatography; HPTLC, high-performance thin-layer
chromatography; HS, headspace; IT; ion trap; MS, mass spectrometry; MSTFA, N-methyl-N-(trimethylsilyl)trifluoroacetamide; NA, not applicable; NMR, nuclear magnetic
resonance spectroscopy; Q-TOF, quadrupole-time of flight; SEC, size-exclusion chromatography; SPE, solid-phase extraction; SPME, solid-phase microextraction; TLC, thin-layer
chromatography; UPLC, ultra-performance liquid chromatography; UV, ultraviolet
45
Aloe vera
IARC MONOGRAPHS 108
has been a major concern as a consequence of Agency also found that the therapeutic indica-
the high cost of the raw materials. Common tion as an herbal product for short-term use in
adulterants have included maltodextrin in Aloe cases of occasional constipation is a well estab-
vera gel, powder or water in the liquid prepa- lished use of Aloe vera latex (EMA, 2006).
rations (Pelley et al., 1998). Many authors have For the gel, WHO identified no uses supported
reviewed the considerable available amount of by clinical data. Traditional uses include the
literature for analysis and authenticity control external treatment of minor wounds and inflam-
of Aloe vera. Besides various chromatographic matory skin disorders. The gel may be used in
approaches, nuclear magnetic resonance spec- the treatment of minor skin irritations, including
troscopy appears to be the method of choice burns, bruises, and abrasions (WHO, 1999).
for this purpose (Table 1.3). Common addi- In recent times, the oral consumption of
tives found in Aloe vera gel preparations, which Aloe vera has been promoted as prophylaxis
can be detected by chromatographic methods, and therapy for a variety of unrelated systemic
include preservatives such as benzoic acid and conditions. The scientific literature yields little
sorbic acid, or antioxidants such as ascorbic acid to substantiate claims of usefulness for systemic
(Lachenmeier et al., 2005). Several methods to conditions by the ingestion of Aloe vera (Boudreau
control the gel material for contamination with et al., 2013a).
aloin are available (see review by Rodrguez et al. Aloe vera may be used in veterinary medicine
(2010) and Table1.3). as laxative or in topical applications, e.g. in udder
disinfectants (Leon, 2003).
1.3 Use (b) Food use
1.3.1 Indications Aloe vera extracts may be used in beverages
as bitter flavouring agent (ONeil et al., 2006).
(a) Medicinal use Food products include health and soft drinks,
The Aloe vera plant has been used in folk yoghurts, jams, instant tea granules, candies,
medicine for more than 2000 years, and it alcoholic beverages, and ice cream (Ahlawat &
remains an important component of traditional Khatkar, 2011). Aloe vera may also be used in
medicine in many contemporary cultures, food supplements (Steenkamp & Stewart, 2007).
such as China, India, the Caribbean, and Japan The Dietary Supplements Label Database lists
(Grindlay & Reynolds, 1986). Aloe vera first 43 products that contain Aloe vera as an active
gained popularity in the USA in the 1930s with ingredient in amounts of 0.33 to 750 mg per
reports of successful use of freshly cut leaves in capsule (NLM, 2012). Aloe vera whole leaf extract
treating X-ray burns (Ulbricht et al., 2007). Both (which combines both the gel and latex) and
classes of Aloe vera leaf products, gel and latex, Aloe vera decolorized whole leaf extract (from
are reported to possess a wide range of pharma- which most of the latex components have been
ceutical activities. removed) are popular as dietary supplements for
WHO lists the short-term treatment of occa- various systemic ailments. The anthraquinone
sional constipation as a use for Aloe vera latex components of these products appear to vary
that is supported by clinical data (WHO, 1999). significantly in their content of aloe-emodin and
The well established cathartic properties of anth- aloin A, the major anthraquinone constituent of
raquinone glycosides provide strong evidence in Aloe vera latex. (Elsohly et al., 2007) evaluated
support of the laxative properties of Aloe vera 53 liquid and 30 semisolid and solid aloe-based
(Ulbricht et al., 2007). The European Medicines commercial products. The liquid samples all
46
Aloe vera
contained either aloe-emodin or aloin A at 10 1.3.2 Dosage

ppm, with many having no detectable levels of
either of the two compounds. Unlike liquid prod- For medicinal use as a laxative, the correct
ucts, many solid and semisolid products (11 out individual dose is the smallest amount required
of 30) contained one or both of the compounds, to produce a soft-formed stool. For adults and
aloe-emodin and aloin A, at 10 ppm. children aged more than 10 years, the dose is
40110 mg of the dried latex, corresponding to
(c) Cosmetic use 1030 mg of hydroxyanthraquinones per day, or
100 mg as a single dose in the evening (WHO,
The gel may be used as emollient and mois-
1999). The European Medicines Agency suggests
turizer in cosmetics and personal care prod-
a maximum daily dose of hydroxyanthracene
ucts (ONeil et al., 2006). The gel is used in the
glycosides of 30 mg, and that the correct indi-
cosmetics industry as a hydrating ingredient in
vidual dose is the smallest required to produce
liquids, creams, sun lotions, shaving creams, lip
a comfortable soft-formed motion (EMA, 2006).
balms, healing ointments, and face packs (WHO,
As for other laxatives, there is potential for abuse
1999). Other products containing Aloe vera
of Aloe vera latex (Perkins & Livesey, 1993; Stolk
include after-shave gel, mouthwash, hair tonic,
& Hoogtanders, 1999). It is difficult to estimate
shampoo, and skin-moistening gel (Newton,
rates of laxative abuse, and more so for cases of
2004).
abuse attributable to Aloe vera alone.
Aloe vera may be used in cosmetics for
For medicinal use of Aloe vera gel, 25 to 100 mL
marketing reasons (i.e. to impart a touch of
per day of a 4.5:1 gel concentrate was suggested as
nature to the product) rather than for actual
typical oral dose range in adults (Morgan et al.,
effects, and the content may be normally kept at
2005). The International Aloe Science Council
a low level (Committee of Experts on Cosmetic
recommended a total daily consumption of Aloe
Products, 2008).
vera of 28 fluid ounces (59237 mL) of single-
A study on skin hydration found that a single
strength leaf gel (IASC, 2013b). For topical use,
application of a cosmetic formulation containing
pure Aloe vera gel is often used liberally on the
>0.25% of a commercial freeze-dried Aloe vera
skin. Hydrophilic cream of 0.5% (by weight) of a
gel 200:1 concentrate improved the water content
50% ethanol extract of Aloe vera, three times per
of the stratum corneum (DalBelo et al., 2006).
day for five consecutive days per week has been
However, the concentrations of Aloe vera raw
used for treatment of genital herpes and psoriasis
materials in cosmetics vary widely from 0.1%
vulgaris (Ulbricht et al., 2007).
or less up to 20% (Cosmetic Ingredient Review
Expert Panel, 2007).
Anthraquinone-rich Aloe vera extracts may 1.4 Production, sales, and
function as absorbers of ultraviolet radiation consumption
in suncreens, because anthraquinones absorb
ultraviolet radiation (Committee of Experts on 1.4.1 Production
Cosmetic Products, 2008). Regulatory author-
(a) Production process
ities in Germany have proposed that cosmetic
products for which claims are made regarding Aloe vera grows best in dry chalky soil or in a
Aloe vera should contain at least 5 g of Aloe vera sandy loam (Grindlay & Reynolds, 1986). While
per 100 g of product (Kratz, 2009). the plant needs warm semi-tropical conditions,
overexposure to sun results in stunted plants with
low gel yield. Therefore, Aloe vera is commonly
47
IARC MONOGRAPHS 108
interplanted with other crops, such as fruit trees. Aloe vera whole leaf extract is obtained by
The quality of Aloe vera plant products varies grinding the whole fresh leaves, without removal
considerably due to differences in growing, of the rind. Extraneous material and lignified
harvesting, processing, and storage techniques fibres are then removed by homogenizing and
(Boudreau et al., 2013a), and may also depend on filtering the crude gel or whole leaf extracts
the regulatory regime under which the product is (Yaron, 1993). Since various amounts of latex and
sold (see Section 1.6). rind may be present in the whole leaf extracts,
Mexico, followed by the rest of Latin America, the extracts may appear yellow to yellow-green
China, Thailand, and the USA were described in colour.
as main producing countries (Rodrguez et al., Activated carbon adsorption to produce Aloe
2010). Aloe vera has become an important plant vera decolorized whole leaf extract is the first
crop in Arizona and in the Rio Grande valley of processing step where an extract is intention-
southern Texas (Boudreau et al., 2013a). ally subjected to chemical alteration. Aloe vera
The production processes for Aloe vera prod- decolorized whole leaf has lower rheological
ucts include various steps such as crushing, values than the gel and has a lower content of
grinding or pressing, filtration, decoloriza- complex carbohydrates than either gel or whole
tion, stabilization, heat processing, and may leaf extracts (Pelley et al., 1998).
be followed by addition of preservatives and The processed extracts are difficult to keep
stabilizers. A complete overview of production stable, a problem that may cause differences
was provided by Ahlawat & Khatkar (2011). The in product potency; therefore, the gel or whole
technology for processing of Aloe vera gel was leaf extracts can undergo a stabilization process
reviewed by Ramachandra & Rao (2008). before being bottled. This process may involve
Harvesting of the leaves of the Aloe vera pasteurization, ultraviolet stabilization, chemical
plant is generally performed by hand, with the oxidation with hydrogen peroxide, addition of
leaves cut from the base of the plant (Grindlay & chemical preservatives and additives, or concen-
Reynolds, 1986). Individual leaves are wrapped, tration, and/or drying (Boudreau et al., 2013a).
crated, and transported to processing plants.
Ideally, the leaves are processed within a few (b) Production volume
hours after harvesting, as temperature, light, air, In the cosmetic industry, Aloe vera ingredi-
and humidity can affect the stability of the plant ents hold a prominent position at the top of the
components (Paez et al., 2000). At the processing list showing the relative frequency of use of plant
step, the leaves may be cleaned with water and ingredients within formulations filed with the
a mild chlorine solution (Grindlay & Reynolds, United States Food and Drug Administration
1986). (FDA) (Committee of Experts on Cosmetic
Aloe vera gel from the fillet of the inner leaf Products, 2008).
pulp is obtained either by manual removal of the
outer layers of the leaf with a knife or by machine. 1.4.2 Sales
Either method can be flawed and has the poten-
tial to contaminate the gel with latex (Grindlay According to the 2012 Nutrition Business
& Reynolds, 1986). This process yields crude Journal Annual Report, Aloe vera was 20th
Aloe vera gel. High quality gel appears opaque, among best-selling dietary supplements in the
slightly off-white in colour, and is viscous (Vogler USA. There has been a general upward trend
& Ernst, 1999). in sales from US$ 31 million in 2000 to US$
48
Aloe vera
Fig.1.4 Sales of dietary supplements containing Aloe vera in the USA

80
70
60
Aloe vera sales (million US$)
50
40
30
20
10
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
Compiled by the Working Group from data in Nutrition Business Journal (2010, 2012).
72 million in 2011 (Fig. 1.4; Nutrition Business 1.4.3 Consumption

Journal, 2010, 2012).
In 2006, the industry size for Aloe species Consumers of products specified in Section
raw material was estimated to be about US$ 125 1.3 are exposed to Aloe vera. While the occa-
million worldwide, while the industry for finished sional short-term use of the latex as a laxative
products containing Aloe vera was around US$ may allow exposure to be estimated for use in
110 billion (Ahlawat & Khatkar, 2011). that context, it is unclear whether or not the gel
Global sales of Aloe species products in 2012 products or liquid preparations are used over the
totalled US$ 351 million, according to IMS Health short or long-term.
MIDAS data. Most products were reported as According to a representative survey
derived from Aloe vera (90%). Substantial sales conducted by the National Health and Nutrition
as a dietary supplement were reported in Brazil Examination Survey from 1999 to 2010
(US$ 74 million), Indonesia (US$ 50 million), (NHANES, 2010), the consumption of dietary
India (US$ 34 million), USA (US$ 29 million), supplements containing Aloe vera in the USA
the Russian Federation (US$ 19 million), Japan (prevalence of use in the past 30 days among
(US$ 15 million), and Mexico (US$ 12 million) adults in the USA) was 0.3% in 19992006 and
(IMS Health, 2012). 0.1% in 20072010 [figures calculated by the
Working Group from publicly available data;
due to the small use, the coefficient of variation
is >30%, so that the data for Aloe vera are less
reliable than for other herbs]. In the context of
complementary and alternative medicine, use
49
50
Table 1.4 Regulations for different Aloe vera products
Regulation WHO Monograph on Selected Medicinal Japanese Pharmacopoeia European The International Aloe Science
Plants (1999)a Sixteenth Edition (2011)b Pharmacopoeia 7.0 Council (2013)d
(2008)c
Regulated Aloe Dried juice Gel Dried juice Concentrated and Raw materials for use in products for
product dried juice oral consumption
Content Min. 28% of Min. 4% aloin (dried Min. 28% of Max. 10 ppm (aloin A + B)
hydroxyanthracene material) hydroxyanthracene
derivatives, expressed as derivatives, expressed
aloin as aloin (dried drug)
IARC MONOGRAPHS 108
Identity tests Macroscopic and Colour reactions with TLC; fluorescence Min. 5% acetylated mannan content by
microscopic examinations, sodium tetraborate and with disodium dry weight
solvent solubility; TLC nitric acid; TLC tetraborate; colour Organoleptic standards:
reaction with Aloe solids in single strength juice (1%
bromine water in leaf juice and 0.5% for inner leaf
juice)
Malic acid and glucose must be present
at a minimum
Whole leaf marker (isocitrate). Max.
5% for inner leaf by dry weight.
Moisture Max. 12% for Curacao or Contains 98.5%
Barbados Aloe water
Total ash Max. 2% Max. 2% Max. 2% < 40%
Loss on drying Max. 12% Max. 12%
Foreign substances/ Limits for certain Limits for Two different purity tests Microbiologicals (pathogens, lactic
contaminants microorganisms, absence certain are specified acid, mould, yeast), heavy metals,
of adulterants such as black microorganisms, maltodextrin
catechu, pieces of iron, and pesticides,
stones; limits for certain heavy metals,
pesticides, heavy metals, radioactive
and radioactive residues residues
Extract content Min. 50% (water-soluble Min. 40% (water-soluble
extract), max. 10% (alcohol- extract)
insoluble extract)
a WHO (1999)
b Eur Ph (2008)
c JP XVI (2011)
d IASC (2013b)
max., maximum; min., minimum; TLC, thin-layer chromatography

Aloe vera
of Aloe vera has been reported in 8.513.8% of A published tabulation of acceptable levels
people in predominantly Hispanic populations of natural flavourings by the Flavor and Extract
in the southern USA; according to surveys, it is Manufacturers Association indicates that an
also used frequently by 10.8%, 10.3%, and 7.6% acceptable level of Aloe vera extract is 52000
of adults in Australia, Italy, and Jamaica, respec- ppm. No distinction is given for the part of the
tively (Ngo et al., 2010). plant or type of plant extract used to produce
the extract used as a flavouring additive (Duke
& Beckstrom-Sternberg, 1994).
1.5 Occupational exposure For cosmetic uses, many of the manufac-
No specific studies on occupational exposure turers of Aloe vera gel take care to supply an
were identified. It can be assumed that workers ingredient containing anthraquinones at no
in the production of Aloe vera may be exposed, more than 50 ppm (Committee of Experts on
as well as workers in pharmaceutical, cosmetic, Cosmetic Products, 2008). This maximum level
and food industries that use Aloe vera as an was also demanded in a safety assessment of the
ingredient. cosmetic industry (Cosmetic Ingredient Review
Expert Panel, 2007).
Aloe vera is specified in several official phar-
1.6 Regulations and guidelines macopoeias, and an industry quality standard
Products made with various components of of the International Aloe Science Council is
Aloe vera (aloin, aloe-emodin, and barbaloin) also available (Table 1.4). An American Herbal
were at one time regulated by the FDA as oral Pharmacopoeia on Aloe vera leaf, Aloe vera leaf
over-the-counter (OTC) laxatives (NCCAM, juice, Aloe vera inner leaf juice was provided
2012). In 2002, the FDA promulgated a regula- (AHP, 2012).
tion stating that the stimulant laxative ingre-
dient Aloe vera in over the counter (OTC) drug 2. Cancer in Humans
products is not generally recognized as safe and
effective or is misbranded (FDA, 2002). Because
No data were available to the Working Group.
the companies that manufactured such products
did not provide the necessary safety data, the
FDA required that all OTC Aloe vera laxative 3. Cancer in Experimental Animals
products be removed from the USA market or
reformulated (NCCAM, 2012). [The Working 3.1 Studies of carcinogenicity
Group noted that currently no medicinal OTC
Aloe vera products are available in the USA, Whole leaf extract of Aloe barbadensis Miller
unlike Europe where some medicinal Aloe vera [Aloe vera] was tested for carcinogenicity by oral
products are still available.] administration (drinking-water) in one study in
According to FDA regulations, Aloe vera may mice and one study in rats.
be safely used as a flavouring in foods as defined 3.1.1 Mouse
in 21CFR172.510. The Environmental Protection
Agency (EPA) classified Aloe vera gel as a List 3 In a 2-year study of carcinogenicity, groups
substance (inerts of unknown toxicity), and also of 48 male and 48 female B6C3F1 mice (age, 67
listed Aloe vera gel as an inert ingredient of pesti- weeks) were given drinking-water containing 0
cide products (SciFinder, 2013). (controls), 1.0%, 2.0%, or 3.0% (wt/wt) whole leaf
extract of Aloe barbadensis Miller [Aloe vera]
51
IARC MONOGRAPHS 108
for 104 weeks. The average content of aloin A Groups of 36 male and 36 female Crl:SKH-1
and aloe-emodin of the whole leaf test material (hr-/hr-) hairless mice (age, 8 weeks) received
was 6.40 and 0.071 mg/g respectively. The doses topical applications of control cream or creams
of whole leaf extract were equivalent to average containing: 3% or 6% (w/w) gel; 3% or 6% (w/w)
daily doses of approximately 0, 2.9, 7.0, or 11.8 g/ whole leaf extract; 3% or 6% (w/w) decolorized
kg body weight (bw) in males; and 0, 2.2, 6.3, or whole leaf extract; or 7.46 or 74.6 g/g of aloe-
11.8 g/kg bw in females (Boudreau et al., 2013a). emodin to the dorsal skin region, for 5days per
Survival of exposed groups was similar to that week, for up to 40 weeks. After application of
of controls. There was no significantly increased the cream in the morning, mice were exposed
incidence of any tumour type in male or female to filtered solar simulated light (SSL) at 0 (0.00
mice. Whole leaf extract increased the incidence mJ.CIE/cm2 per day) or 0.6 (13.70 mJ.CIE/cm2
of goblet cell hyperplasia in the intestine of male per day) minimal erythema doses of light (NTP,
and female mice (Table3.1). 2010). The minimal erythema dose is defined
as the minimal amount of radiation that causes
3.1.2 Rat
slight erythema within 24 hours after irradiation
In a 2-year study of carcinogenicity, groups (Table3.2). The mice were killed after a recovery/
of 48 male and 48 female F344/N rats were observation period of 12 weeks.
given drinking-water containing whole leaf At 52 weeks, there was no significant increase
extract of Aloe barbadensis Miller [Aloe vera] at in the incidence of skin neoplasms in any group
0 (controls), 0.5%, 1.0%, or 1.5% (wt/wt) for 104 receiving any of the four creams containing
weeks. The average content of aloin A and aloe- Aloe preparations without exposure to SSL. [The
emodin of the whole leaf test material was 6.40 Working Group noted that the duration of the
and 0.071 mg/g, respectively. The doses of whole experiment, 1year, was too short to consider this
leaf extract were equivalent to average daily arm of the experiment as a full carcinogenicity
doses of approximately 0, 0.2, 0.6, or 1.1 g/kg bw study.]
in males and 0, 0.3, 0.7, or 1.3 g/kg bw in females There was no treatment-related increase in
(Boudreau et al., 2013a). Survival of exposed the incidence of skin neoplasms in any groups
groups was similar to that of controls. Whole leaf receiving any of the four creams containing Aloe
extract caused increased incidences of adenoma preparations followed by SSL when compared
and carcinoma of the large intestine (colon and with the groups receiving control cream followed
caecum) in males and females. Other treat- by SSL. Almost all mice in groups exposed to
ment-related lesions included hyperplasia and/ SSL presented with skin neoplasms due to SSL
or inflammation in the mesenteric lymph node, exposure. [As a result, the primary experimental
forestomach, small intestine, and large intestine end-point was multiplicity of skin tumours.]
in males and females (Table3.1). [The Working There was a significant enhancing effect of
Group noted that large intestine tumours are Aloe gel cream or of aloe-emodin cream on the
rare spontaneous neoplasms in F344/N rats.] photocarcinogenic activity of SSL in female mice,
and there was a significant enhancing effect of the
3.2 Photo-co-carcinogenicity studies cream containing whole leaf extract, or cream
containing decolorized whole leaf extract, on the
Mouse photocarcinogenic activity of SSL in male and
There has been one study reported in which female mice, based on an increase in the multi-
Aloe barbadensis Miller [Aloe vera] test articles plicity of squamous cell papilloma, carcinoma or
were studied by dermal application in mice. carcinoma in situ (combined) (NTP, 2010).
52
Table 3.1 Studies of carcinogenicity in mice and rats given Aloe vera by oral administration
Species, strain Dosing regimen Incidence of tumours Significance Comments

(sex) Animals/group at start
Duration
Reference
Mouse, B6C3F1 Whole leaf extract of Aloe No significantly increased incidence of any tumour type Aloin A content,
(M,F) barbadensis Miller [Aloe vera]: 0 6.40mg/g whole leaf
104wk (control), 1.0%, 2.0%, or 3.0% (w/w) test material
Boudreau et al. in drinking-water (estimated to be 0, Aloin-emodin
(2013a) 2.9, 7.0, or 11.8g/kg bw (M); 0, 2.2, content, 0.071mg/g
6.3, or 11.8g/kg bw (F) whole leaf.
48 M and 48 F/group (age, 67 wk)
Rat, F344/N Whole leaf extract of Aloe All large intestine (colon and caecum) adenoma: # P0.001 Aloin A content,
(M,F) barbadensis Miller [Aloe vera]: 0 M: 0/47 (0%)#, 0/48 (0%), 26/48 (54%)*, 23/48 (48%)* (trend) 6.40mg/g whole leaf.
104wk (control), 0.5%, 1.0%, or 1.5% (w/w) F: 0/48 (0%)#, 0/48 (0%), 6/48 (13%)**, 13/48 (27%)* ## P0.01 Aloin-emodin
Boudreau et al. in drinking-water (estimated to be All large intestine (colon and caecum) carcinoma: (trend) content, 0.071mg/g
(2013a) 0, 0.2, 0.6, or 1.1g whole leaf/kg bw M: 0/47 (0%)#, 0/48 (0%), 10/48 (21%)*, 14/48 (29%)* * P0.001 whole leaf
(M); 0, 0.3, 0.7, 1.3g whole leaf/kg F: 0/48 (0%)##, 0/48 (0%), 3/48 (6%), 4/48 (8%)** ** P<0.05
bw (F) All large intestine (colon and caecum) adenoma or carcinoma *** P0.01
48 M and 48 F/group (age, 67 wk) (combined):
M: 0/47 (0%)#, 0/48 (0%), 28/48 (58%)*, 31/48 (65%)*
F: 0/48 (0%)#, 0/48 (0%), 8/48 (17%)***, 15/48 (31%)*
bw, body weight; F, female; M, male; wk, week; w, weight
53
Aloe vera
54
Table 3.2 Co-carcinogenicity studies in SKH-1 mice given Aloe vera or aloe-emodin by skin application followed by exposure
to simulated solar light
Species, strain Dosing regimen Overall age-adjusted tumour multiplicity Significance Comments
Duration
Reference
Mouse, Aloe vera gel cream at 0%, 3%, or 6% (w/w) + SSL Squamous cell papilloma, squamous cell * P<0.05 No significant increase
Crl:SKH-1 (hr-/ (13.70mJCIE/cm2 per day). Cream applied in the carcinoma in situ, and/or squamous cell (trend) in the incidence of
hr-) hairless morning; SSL in the afternoon. carcinoma of the skin: ** P=0.006 skin neoplasms in
IARC MONOGRAPHS 108
(M,F) 5d per wk for 40wk, followed by 12-wk recovery/ M: 5.8 (4.77.2), 6.8 (5.68.4), 7.1 (5.88.7) *** P<0.05 any group receiving
52wk observation period F: 6.4 (5.37.6)*, 9.2 (7.810.8)**, 8.1 (6.99.6)*** creams containing
NTP (2010) 36 M and 36 F/group (age 8 wk) Aloe vera preparations
without exposure to
SSL
Mouse, Whole leaf Aloe vera cream at 0%, 3%, or 6% (w/w) Squamous cell papilloma, squamous cell * P<0.05
Crl:SKH-1 (hr-/ + SSL (13.70mJCIE/cm2 per day). Cream applied carcinoma in situ, and/or squamous cell (trend)
hr-) hairless in the morning; SSL in the afternoon. carcinoma of the skin: ** P<0.05
(M,F) 5d per wk for 40wk, followed by 12-wk recovery/ M: 5.8* (4.77.2), 6.4 (5.27.9), 8.4** (6.810.3)
52wk observation period F: 6.4* (5.37.6), 8.7** (7.410.3), 7.7 (6.59.1)
NTP (2010) 36 M and 36 F/group (age 8 wk)
Mouse, Decolorized whole leaf Aloe vera cream at 0%, Squamous cell papilloma, squamous cell * P<0.05
Crl:SKH-1 (hr-/ 3%, or 6% (w/w) + SSL (13.70mJCIE/cm2 per carcinoma in situ, and/or squamous cell ** P=0.007
hr-) hairless day). Cream applied in the morning; SSL in the carcinoma of the skin: (trend)
(M,F) afternoon. M: 5.8 (4.67.3), 8.0* (6.59.9), 6.4 (5.2 8.0) *** P=0.002
52wk 5d per wk for 40wk, followed by 12-wk recovery/ F: 6.4** (5.27.7), 10.0*** (8.412.0), 9.3**** **** P=0.007
NTP (2010) observation period (7.811.1)
36 M and 36 F/group (age 8 wk)
Mouse, Aloe-emodin cream at 0, 7.46, or 74.6g/g + SSL Squamous cell papilloma, squamous cell * P<0.05
Crl:SKH-1 (hr-/ (13.70mJCIE/cm2 per day). Cream applied in the carcinoma in situ, and/or squamous cell (trend)
hr-) hairless morning; SSL in the afternoon. carcinoma of the skin: ** P<0.05
(M,F) 5d per wk for 40wk, followed by 12 wk recovery/ M: 5.8 (4.77.2), 6.3 (5.17.8), 7.1 (5.88.7)
52wk observation period F: 6.4* (5.37.7) 7.9 (6.69.4), 8.9** (7.510.6)
NTP (2010) 36 M and 36 F/group (age, 8 wk)
bw, body weight; CIE, Commission Internationale de lEclairage [International Commission on Illumination]; d, day; F, female; M, male; SSL, simulated solar light; wk, week
Aloe vera
4. Mechanistic and Other (a) Components of Aloe vera gel: metabolism

Relevant Data ex vivo
Incubation of acemannan (aloemannan;
In reviewing studies relevant to the possible molecular weight >400kDa) labelled with fluo-
carcinogenicity of Aloe vera, the Working Group resceinyl isothiocyanate (FITC) with a suspen-
noted that attributing appropriate weight to indi- sion of fresh human faeces for 5days gave two
vidual studies was complicated by the consid- metabolites, with molecular weights of 10 and
eration that, despite the terminology used, the 30kDa, in 1% yield, meaning that aloemannan
material tested may not have been identical is catabolized by human intestinal bacteria (Yagi
across various studies and/or may not have been et al., 1999).
identical to the material that was studied in
(b) Components of Aloe vera latex
experimental animals, as described in Section 3
of this Monograph. Orally ingested anthrone C-glycosides (i.e.
aloin A and aloin B) pass intact through the
upper portion of the gastrointestinal tract and
4.1 Absorption, distribution, upon reaching the lower gastrointestinal tract
metabolism, and excretion are cleaved to aloe-emodin-9-anthrone by
human Eubacterium sp. BAR given to germ-free
4.1.1 Humans
rats (Che et al., 1991; Hattori et al., 1993; Akao
There were no reports of studies to determine et al., 1996). The free aglycone is then absorbed,
the absorption, distribution, metabolism, or undergoes oxidation, and is excreted in the
excretion of topically applied Aloe vera gel, whole urine as rhein, as was shown in three volunteers
leaf extract or decolorized whole leaf extract in receiving Aloe vera or barbaloin (Vyth & Kamp,
experimental animals or humans. 1979; Fig.4.1).
Aloe vera whole leaf extract is composed of
gel and latex. Aloe vera gel contains non-starch 4.1.2 Experimental systems
polysaccharides of high molecular weight (the
major one being acemannan) that are composed (a) Components of Aloe vera gel
of sugar moieties linked by -1,4-glycosyl bonds In beagle dogs, the oral administration of
(Fig. 1.3 in Section 1). Aloe vera latex contains radiolabelled acemannan at a dose of 20mg/kg
the anthrone C-glycosides aloin A (barbaloin) body weight (bw) per day for 3months resulted
and aloin B (isobarbaloin) that are linked by in peak blood concentrations at 46hours and
-glycosyl bonds to D-glucopyranose. Other a half-life of >48hours (Fogleman et al., 1992).
C-glycosides found in Aloe vera latex include Male ddY mice were given FITC-labelled
aloesin (aloeresin B) and aloeresin A in which acemannan (aloemannan; molecular weight,
the glycosyl linkage is to the benzo ring of 500kDa) at a dose of 120 mg/kg bw by gavage,
benzopyran-4-one (Boudreau & Beland, 2006). and urinary and faecal excretion was monitored
Aloenin, an O--glucoside, is also a component for 48hours. Of the administered dose, 95% was
of Aloe vera latex (Hirata et al., 1981; Matsuda excreted in the faeces, with > 90% occurring
et al., 2008). within 24hours. Only 0.3% of the material was
found in the urine. In both urine and faeces,
FITC-labelled acemannan was converted to
substances of low molecular weight (< 9 kDa).
55
IARC MONOGRAPHS 108
Fig.4.1 Metabolites of aloin A and aloin B

OH O OH OH O OH
8 9 1
7 2
3
6
OH OH
5 10 4
H H
HO HO
O O
OH OH
OH OH
OH OH
A loin A (Barbaloin) A loin B (Isobarbaloin)
H ydrolysis of the -glycosidic

bond by intestinal bacteria
OH O OH
OH
A loe-emodin-9-anthrone
[O ]
OH O OH
OH
O
A loe-emodin anthraquinone
[O ]
OH O OH
CO O H
Rhein
Aloin A and B are constituents of Aloe vera whole leaf latex.

Compiled by the Working Group.
56
Aloe vera
FITC-labelled acemannan was also adminis- This material was characterized as aloe-emodin,
tered to mice by intravenous injection at a dose rhein, and their conjugates (Lang, 1993).
of 120 mg/kg bw. Of the administered dose, 73% Intracaecal administration of [14C]rhein to
was excreted in the urine, with >60% occurring male Wistar rats resulted in 37% of the radioac-
within 24hours; 13% of the material was found tivity being excreted in the urine and 53% in the
in the faeces. In both urine and faeces, FITC- faeces. The highest tissue concentrations were
labelled acemannan was converted to substances found in the kidney (De Witte & Lemli, 1988).
of low molecular weight (1070 kDa in urine; The oral administration of [14C]-labelled
5kDa in faeces) (Yagi et al., 1999). aloenin to rats resulted in the faecal and urinary
excretion of the aglycone 4-methoxy-6-(2,4-di-
(b) Components of Aloe vera latex hydroxy-6-methylphenyl)-2-pyrone (Hirata
In male Wistar rats, oral administration of et al., 1981).
aloin A (barbaloin; 100 mg/kg bw) resulted in
maximum serum concentrations of aloin A 4.1.3 Alterations in enzymes involved in
[340 ng/mL; ~0.8 M, based upon the molec- metabolism
ular weight of aloin A of 404Da] 1.5hours after
administration, followed by a decrease in concen- Incubation of the human colon carcinoma
tration, with aloin A still detectable 6hours after cell line LS180 with Aloe vera juice resulted in a
dosing (Ishii et al., 1987). significant increase in the expression of CYP1A2,
The ability to cleave anthrone C-glycosides CYP3A4, and multidrug resistance 1 genes
varies among species; free anthrones are detected (Brandin et al., 2007).
in faecal contents from humans and rats, but not Commercial preparations of Aloe vera were
mice or guinea-pigs (Dreessen & Lemli, 1988; tested for their ability to inhibit the activities of
Hattori et al., 1988). CYP3A4 and CYP2D6 in vitro. Inhibition was
In male and female Brown-Norway rats, oral observed with half maximal inhibitory concen-
administration of [14C]aloe-emodin (4.5 mg/kg trations IC50 in the range 843mg/mL, concen-
bw) resulted in maximum blood concentrations trations that were probably sufficiently high as to
[~350ng/mL; ~1.3M, based upon the molecular preclude any significant inhibition in vivo (Djuv
weight of aloe-emodin of 270Da] 2hours after & Nilsen, 2012). Rhein was shown to inhibit
dosing, and a terminal half-life of elimination CYP1A2, CYP2C9, CYP2D6, CYP2E1, and
of approximately 50 hours. Seven metabolites CYP3A activities in rat liver microsomes, with
were detected in the plasma. These were char- Ki in the range of 1074M (Tang et al., 2009).
acterized as aloe-emodin, rhein, an unidentified
aglycone, and conjugates of these aglycones. 4.2 Genetic and related effects
Approximately 20% of the radiolabel was elim-
inated in the urine, primarily as aloe-emodin, 4.2.1 Humans
rhein, and their conjugates. More than 75% of the No data were available to the Working Group.
radiolabel was excreted in the faeces and nearly
all of this was aloe-emodin. At early time-points
(<48hours), a great majority of the radiolabel was
associated with the gastrointestinal tract. At later
time-points (i.e. 96hours), the highest levels of
radioactivity were found in the kidney and liver.
57
IARC MONOGRAPHS 108
4.2.2 Experimental systems leaf extract, Aloe vera gel, acemannan, and aloin
were tested in Salmonella typhimurium reverse
(a) DNA damage mutation assays, Bacillus subtilis rec-assays,
An Aloe vera whole leaf extract induced and/or SOS DNA damage-repair assays. With
single-strand breaks in pUC 9.1 plasmid DNA, the exception of the Bacillus subtilis rec-assay
and this was associated with decreased transfor- with water extracts of Aloe ferox Mill., all gave
mation efficiency of the plasmid (Table4.1; Paes- negative results (Table 4.1; Table 4.2; Brown &
Leme et al., 2005). Dietrich, 1979; Morimoto et al., 1982; Boudreau
Aloe-emodin and/or rhein induced DNA et al., 2013a; Sehgal et al., 2013a, b).
damage in NPC-039 and NPC-076 human naso- Aloe-emodin was mutagenic in reversion
pharyngeal carcinoma cells and SCC-4 human assays with various strains of Salmonella typh-
tongue cancer cells, as measured by comet assays imurium, at the Tk+/ locus in mouse lymphoma
(Table4.2; Lin et al., 2007, 2010; Chen et al., 2010). L5178Y cells, and the Gpt locus in AS52 Chinese
hamster cells (Table 4.2; Brown et al., 1977;
(b) End-points associated with DNA damage Brown & Dietrich, 1979; Westendorf et al., 1990;
In addition to inducing DNA damage (as Heidemann et al., 1996; Mller et al., 1996;
indicated by comet assays), aloe-emodin signif- Mller & Stopper, 1999; Nesslany et al., 2009).
icantly inhibited expression of genes associated Mutation analysis of eight adenomas and
with DNA damage and repair: ataxia telangiec- four carcinomas from the large intestine of F344
tasia mutated (ATM), ataxia telangiectasia and rats given drinking-water containing an Aloe
Rad3-related (ATR), 143-3, breast cancer 1, vera whole leaf extract (Boudreau et al., 2013a,
early onset (BRCA1), and DNA-dependent serine/ b) indicated four point mutations in exons 1 and
threonine protein kinase (DNA-PK) in SCC-4 2 of the Kras gene and four point mutations in
human tongue squamous cancer cells (Chen et al., exon 2 of the Ctnnb1 gene (Table 4.1; Pandiri
2010). Aloe-emodin also induced the formation et al., 2011).
of reactive oxygen species (ROS) in SCC-4 cells,
which was accompanied by S-phase cell-cycle (d) Other genotoxicity end-points
arrest, apoptosis, and several molecular markers Aloe vera inner leaf fillet Qmatrix did not
associated with apoptosis (Chiu et al., 2009). induce chromosomal aberration in Chinese
Rhein induced the formation of ROS in NPC-039 hamster lung cells in vitro; micronuclei were not
human nasopharyngeal carcinoma cells, SCC-4 formed in bone-marrow cells of mice treated
human tongue squamous cancer cells, and A-549 orally in vivo (Williams et al., 2010; Table4.1).
human lung cancer cells, which was accompa- Aloe-emodin induced unscheduled DNA
nied by apoptosis and several molecular markers synthesis in primary hepatocytes from male
associated with apoptosis (Lin et al., 2007; Hsia Wistar rats, micronucleus formation in mouse
et al. 2009; Lai et al., 2009). lymphoma L5178Y cells and TK6 human lymph-
Aloe-emodin induced DNA damage in oblastoid cells, and chromosomal aberrations in
mouse lymphoma L5178 cells, as measured by Chinese hamster ovary cells. Aloe-emodin also
comet assay (Table4.2; Mller et al., 1996). inhibited topoisomerase II, gave positive results
in comet assays in mouse lymphoma L5178Y
(c) Gene mutation cells, SCC-4 human tongue cancer cells, and
Extracts in water, ethanol or methanol of NPC-039 human nasopharyngeal carcinoma
Aloe ferox Mill., stabilized Aloe vera gel, Aloe vera cells, and transformed C3H/M2 mouse cells
whole leaf extract, Aloe vera decolorized whole
58
Table 4.1 Genetic and related effects of Aloe vera preparations
Test system Results Dose Aloe vera preparation Reference

(LED or
Without With HID)
exogenous exogenous
metabolic metabolic
system system
In vitro
Single-strand breaks in pUC 9.1 plasmid + NT 3g/mL Whole leaf extract Paes-Leme et al. (2005)
DNA
Bacillus subtilis rec-assay + NT 6mg/disk Aloe ferox Mill. water extract Morimoto et al. (1982)
Bacillus subtilis rec-assay NT 6mg/disk Aloe ferox Mill. methanol extract Morimoto et al. (1982)
Salmonella typhimurium, TA98, TA100, 10mg/platea Aloe ferox Mill. water extract Morimoto et al. (1982)
reverse mutation
Salmonella typhimurium, TA98, TA100, 10mg/plateb Aloe ferox Mill. methanol extract Morimoto et al. (1982)
reverse mutation
Salmonella typhimurium, TA100, reverse NR Stabilized gel; aloin A and aloin Sehgal et al. (2013a)
mutation B, 10 ppm; in some instances,
material was sterilized by filtration
or autoclaving
Salmonella typhimurium, TA98, TA100, 10mg/plate Qmatrix inner leaf fillet; aloin, Williams et al. (2010)
TA1535, TA1537, reverse mutation <10 ppm
Salmonella typhimurium, TA98, TA100, 6mg/plate Whole leaf extract Boudreau et al. (2013a)
reverse mutation
Salmonella typhimurium, TA98, TA100, 6mg/plate Decolorized whole leaf extract Boudreau et al. (2013a)
reverse mutation
Salmonella typhimurium, TA97, TA98, 10mg/plate Gel Boudreau et al. (2013a)
TA100, TA1535, reverse mutation
Salmonella typhimurium, TA98, TA100, 21initial Decolorized whole leaf extract; Sehgal et al. (2013b)
reverse mutation concentration aloin A & aloin B, ~1ppm. Material
sterilized by filtration.
Escherichia coli, WP2 uvrA/pKM101 6mg/plate Whole leaf extract Boudreau et al. (2013a)
Escherichia coli, WP2 uvrA/pKM101 6mg/plate Decolorized whole leaf extract Boudreau et al. (2013a)
Escherichia coli, WP2 uvrA/pKM101 3 mg/plate Gel Boudreau et al. (2013a)
Escherichia coli, SOS DNA damage repair 10initial Stabilized gel; aloin A and aloin B, Sehgal et al. (2013a)
assay concentration 10 ppm
59
Aloe vera
60
Test system Results Dose Aloe vera preparation Reference

(LED or
Without With HID)
exogenous exogenous
metabolic metabolic
system system
Escherichia coli, SOS DNA damage repair 21initial Decolorized whole leaf extract; Sehgal et al. (2013b)
assay concentration aloin A and aloin B, ~1ppm.
Material sterilized by filtration.
IARC MONOGRAPHS 108
Chromosomal aberrations, Chinese hamster 10mg/plate Qmatrix inner leaf fillet; aloin,<10 Williams et al. (2010)
lung cells ppm
In vivo
Micronucleus formation, male ICR mice, NT 5000mg/kg Qmatrix inner leaf fillet; aloin,<10 Williams et al. (2010)
bone-marrow cells bw, po ppm
Gene mutations in exon 1 and 2 of Kras gene + NT 1% Whole leaf preparation Pandiri et al. (2011)
in large intestine adenomas and carcinomas,
F344 rats
Gene mutations in exon 2 of Ctnnb1 gene in + NT 1% Whole leaf preparation Pandiri et al. (2011)
large intestine adenomas and carcinomas,
F344 rats
+, positive; , negative; LED, lowest effective dose; HID, highest ineffective dose; NR, not reported; NT, not tested; po, per oral
a Toxic in TA98, without exogenous metabolic system, and in TA100, with or without exogenous metabolic system
b Toxic in TA98, without exogenous metabolic system

Table 4.2 Genetic and related effects of constituents of Aloe vera
Constituent Test system Results Dose Reference

(LED or
Without With HID)
exogenous exogenous
metabolic metabolic
system system
Acemannan
Salmonella typhimurium reverse mutation in vitro 800L/plate Fogleman et al. (1992)
Aloin
Salmonella typhimurium, TA1535, TA100, TA1537, TA1538, TA98, TA100FR50, reverse 250g/plate Brown & Dietrich (1979)
mutation in vitro
Aloe-emodin in vitro
Salmonella typhimurium, TA1537, reverse mutation + 100g/plate Brown et al. (1977),
Brown & Dietrich (1979)
Salmonella typhimurium, TA1535, TA100, TA1538, TA98, TA100FR50, reverse mutation 250g/plate Brown et al. (1977),
Brown & Dietrich (1979)
Salmonella typhimurium, TA1537, TA102, TA1538, TA98, TA1978, reverse mutation +a +b 10g/plate Westendorf et al. (1990)
Salmonella typhimurium, TA98, TA100, TA1535, TA1537, TA1538, reverse mutation +c +d 10g/plate Heidemann et al. (1996)
Salmonella typhimurium, TA1537, TA98, TA100, reverse mutation + + 0.62g/plate Nesslany et al. (2009)
Unscheduled DNA synthesis, male Wistar rat primary hepatocytes + NT 25g/mL Westendorf et al. (1990)
Gene mutation, Chinese hamster lung V79 cells, 8-azaguanine resistance (+) NT 10g/mL Westendorf et al. (1990)
Gene mutation, Chinese hamster lung V79 cells, Hprt locus, 6-thioguanine resistance 350g/mL Heidemann et al. (1996)
Gene mutation, Chinese hamster lung V79 cells, Hprt locus, 6-thioguanine resistance 350g/mL Brusick & Mengs (1997)
Gene mutation, mouse lymphoma L5178Y cells, Tk+/ locus, trifluorothymidine resistance + NT 37M Mller et al. (1996)
Gene mutation, mouse lymphoma L5178Y cells, Tk +/ locus, trifluorothymidine resistance +e NT 100M Mller & Stopper (1999)
Gene mutation, AS52 Chinese hamster cells, Gpt locus + NT 28 M Mller et al. (1996)
Micronucleus formation, mouse lymphoma L5178Y cells + NT 37M Mller et al. (1996)
Micronucleus formation, TK6 human lymphoblastoid cells + + 3.12g/mL Nesslany et al. (2009)
Chromosomal aberrations, Chinese hamster ovary cells + + 18.75g/mL Heidemann et al. (1996)
Topoisomerase II inhibition, decatenation of kDNA f + NT 1mM Mller et al. (1996)
Topoisomerase II inhibition, decatenation of kDNA f + NT 741Mg Mller & Stopper (1999)
DNA damage, comet assay, mouse lymphoma L5178Y cells + NT 55M Mller et al. (1996)
DNA fragmentation, comet assay, SCC-4 human tongue cancer cells + NT 100M Chen et al. (2010)
61
Aloe vera
62
Constituent Test system Results Dose Reference

(LED or
Without With HID)
exogenous exogenous
metabolic metabolic
system system
DNA fragmentation, comet assay, NPC-039 & NPC-076 human nasopharyngeal carcinoma + NT 60M Lin et al. (2010)
cells
Cell transformation, C3H/M2 mouse fibroblast cells + NT 3g/mL Westendorf et al. (1990)
IARC MONOGRAPHS 108
Aloe-emodin in vivo
Chromosomal aberrations, Wistar rats, bone-marrow cells NT 200mg/kg Heidemann et al. (1996)
bw, po
Micronucleus formation, polychromatic erythrocytes, NMRI mice, bone-marrow cells NT 1500mg/kg Heidemann et al. (1996)
bw, po
Spot test in F1 offspring of NMRI female mice, X DBA male mice NT 2000mg/kg Heidemann et al. (1996)
bw, poh
Unscheduled DNA synthesis, hepatocytes of Wistar rats treated in vivo NT 1000mg/kg Heidemann et al. (1996)
bw, po
DNA damage, comet assay, male OF1 mice, kidney and colon, treated in vivo + NT 500mg/kg Nesslany et al. (2009)
bw, po
Rhein
Salmonella typhimurium, TA1537, TA102, TA1538, TA98, TA1978, reverse mutation (+)i NR Westendorf et al. (1990)
Unscheduled DNA synthesis, male Wistar rat primary hepatocytes NT NR Westendorf et al. (1990)
Gene mutation, Chinese hamster lung V79 cells, 8-azaguanine resistance NT NR Westendorf et al. (1990)
DNA fragmentation, comet assay, NPC-039 human nasopharyngeal carcinoma cells + NT 180 M Lin et al. (2007)
DNA fragmentation, comet assay, SCC-4 human cancer tongue cells + NT 50 M Chen et al. (2010)
Cell transformation, C3H/M2 mouse cells NT NR Westendorf et al. (1990)
Cell transformation, C3H/M2 mouse fibroblast cells NT 10 g/mL Wlfle et al. (1990)
+, positive; (+), weakly positive; , negative; bw, body weight; HID, highest ineffective dose; LED, lowest effective dose; NR, not reported; NT, not tested; po, oral
a Not positive in TA102
b Not positive in TA102 or TA1978
c Positive only in TA98, TA1537, and TA1538
d Positive only in TA1538
e Mutations associated with loss of heterozygosity.
f kDNA, kinetoplast DNA, a catenated network of mitochondrial DNA rings isolated from Crithidia fasciculate.
g IC =741272 M, represents the concentration at which the catalytic activity of the topoisomerase II was reduced to 50%.
50
h The mother mice were treated on day 9 of the pregnancy.
i In TA1537 only.
Aloe vera
(Table4.2; Heidemann et al., 1996; Mller et al., Male and female Sprague-Dawley rats were
1996; Mller & Stopper, 1999; Lin et al., 2010). treated by gavage with an Aloe vera preparation
Rhein gave positive results in comet assays designated Qmatrix at a dose of 0, 500, 1000, or
in SCC-4 human tongue cancer cells and 2000mg/kg bw per day for 90 days. This material
NPC-039 human nasopharyngeal carcinoma was prepared from Aloe vera inner leaf fillets and
cells (Table4.2; Lin et al., 2007; Chen et al., 2010). was further treated to reduce the aloin content
to <10ppm. At necroscopy, there were no gross
4.3 Toxic effects and other or histopathological alterations that could be
ascribed to treatment with Aloe vera (Williams
mechanistic data
et al., 2010).
[The Working Group noted that the following Short-term studies were conducted in which
studies would have been strengthened by the charcoal-treated Aloe vera gel (designated Aloe
inclusion of positive controls, and Aloe vera juice) was fed to male and female B6C3F1 mice
whole leaf was not included for comparison.] at a dose of 350540 mg per day for 13 weeks.
Short-term studies were conducted in which At necropsy, no significant differences were
technical-grade acemannan (acemannan, 78%) observed between mice fed Aloe juice and the
was fed to male and female Sprague-Dawley control mice. Likewise, histopathological exami-
rats at doses up to 2000 mg/kg bw per day for nation of the livers revealed no treatment-related
6months, and to male and female beagle dogs at lesions (Sehgal et al., 2013a).
doses up to 1500mg/kg bw per day for 90days. Male and female Wistar Hannover rats fed
There were no significant gross or microscopic whole leaf powder from Aloe arborescens Miller
lesions in either species that could be associated var. natalensis Berger at a dietary concentra-
with the dietary administration of acemannan tion of 0%, 0.16%, 0.8%, or 4% [corresponding
(Fogleman et al., 1992). to doses of whole leaf powder of 0, 99, 486, and
Short-term studies were conducted in which 2447 mg per kg bw per day] for 1 or 2 years
Aloe vera gel (designated process A aloe) showed severe sinus dilatation and yellowish
or charcoal-treated Aloe vera gel (designated pigmentation of the ileocaecal lymph nodes, as
process B aloe) were fed to male F344 rats at well as yellow-brown pigmentation of the renal
dietary concentrations of 1% or 10% for up to tubules. Rats from the 2-year study also showed
7 months [corresponding to doses of Aloe vera pigmentation, epithelial thickening, and atypical
gel of ~0.33 and 3.3g per kg bw per day]. Gross hyperplasia of the large intestine (Matsuda et al.,
and microscopic histopathological analyses indi- 2008; Yokohira et al., 2009).
cated that there were no significant differences Male and female F344Du rats given drink-
between the control rats and rats fed either of the ing-water containing Aloe vera decolorized
Aloe vera preparations (Herlihy et al., 1998). whole leaf juice at 0%, 0.5%, 1%, or 2% [corre-
Male F344 rats were given drinking-water sponding to doses of decolorized whole leaf juice
containing Aloe vera gel at a concentration of of 0, 565, 1201, and 2382mg per kg bw per day]
1% [~0.2 g/kg bw per day], or charcoal-treated for 3months. No alterations were detected upon
Aloe vera gel at 1% [~0.2 g/kg bw per day], or histopathological examination of the caecum
charcoal-treated Aloe vera whole leaf at 0.02% and colon (Shao et al., 2013).
[amount consumed could not be determined] for Male and female B6C3F1 mice were given
30 months. None of the treatments caused any an Aloe vera decolorized whole leaf extract to
obvious histopathological changes (Ikeno et al., which had been added a high-molecular-weight
2002). Aloe vera polysaccharide designated Aloesorb,
63
IARC MONOGRAPHS 108
by gavage, twice in 24 hours. The material was 4.4 Other mechanistic data
administered at 1% of the mouse body weight
[the absolute amount of the decolorized whole Gastrointestinal transit times were meas-
leaf extract administered could not be deter- ured in B6C3F1 mice and F344/N rats given
mined] and the mice were monitored for up to drinking-water containing Aloe vera whole leaf
14days after treatment. No adverse effects were extract (aloin A, 14.115.9mg per g of extract),
detected. The same Aloe preparation was also Aloe vera decolorized whole leaf extract (aloin
mixed in the diet at a concentration of 100g/kg A, 0.060.2 mg per g of extract), or Aloe vera
and fed to male and female F344 rats for 3months gel (aloin A, 1.11.4mg per g of gel) for 14 days.
[total amount administered, 10g of decolorized Without treatment, B6C3F1 mice have shorter
whole leaf extract per g bw]. Histopathological transit times than F344/N rats. Aloe vera whole
examination of the caecum, colon, and rectum leaf extract decreased transit times in the rats, but
indicated no significant alterations (Sehgal et al., not in mice. A decrease in gastrointestinal transit
2013b). times was also observed in a 13-week study in
Male and female F344/N rats exposed to F344/N rats, but not B6C3F1 mice, receiving Aloe
drinking-water containing Aloe vera whole leaf vera whole leaf extract (Boudreau et al., 2013a, b).
extract at 1%, 2%, or 3% [corresponding to whole Molecular pathways shown to be important
leaf extract at approximately 1.2, 2.4, and 3.6g per in human colorectal carcinogenesis, including
kg bw per day] for 13weeks, or Aloe vera whole mitogen-activated protein kinases MAPK, WNT,
leaf extract at 0.5%, 1.0%, or 1.5% [corresponding and transforming growth factor TGF- signal-
to whole leaf extract at 0.25, 0.65, and 1.2g per ling pathways, were also altered in adenomas
kg bw per day] for 2 years, had dose-related and carcinomas of the large intestine in F344
increases in the incidence and severity of goblet rats given drinking-water containing Aloe vera
cell and lymph node hyperplasia in the large whole leaf preparations at 1% and 1.5% (Pandiri
intestine. Goblet cell hyperplasia also occurred et al., 2011).
in the large intestine of B6C3F1 mice given Aloe vera whole leaf preparations were incu-
drinking-water containing Aloe vera whole leaf bated with pure and mixed human gut-bacteria
extract at 1%, 2%, or 3% [corresponding to whole cultures. The Aloe vera preparations possessed
leaf extract at 2.55, 6.65, and 11.8g per kg bw per bacteriogenic activity and altered the production
day] for 13weeks, but with lesser severity than of acetic acid, butyric acid, and propionic acid
that observed in rats (Boudreau et al., 2013a, b). (Pogribna et al., 2008).
Aloe vera Liliaceae was extracted with 95%
ethanol, the solvent was evaporated, and the 4.5 Susceptibility
residue was administered in the drinking-water
at a dose of 100mg/kg bw to male Swiss albino No data were available to the Working Group.
mice for 3 months. Of the treated mice, 30%
(6/20) died compared with 10% (2/20) of the
control mice, a difference that was not statisti-
4.6 Mechanistic considerations
cally significant. Mice treated with the Aloe vera Upon oral ingestion, Aloe vera components
preparation had an increased incidence (P<0.01) pass through the upper portion of the gastroin-
of sperm abnormalities, including megacephaly, testinal tract; upon reaching the lower gastroin-
flat head, swollen acrosome, and rotated head testinal tract, the anthrone C-glycosides aloin
(Shah et al., 1989). [The identity of the material A and aloin B are converted by the intestinal
being tested was not certain.] microflora to aloe-emodin-9-anthrone, which
64
Aloe vera
undergoes sequential oxidation to aloe-emodin Although the carcinogenicity of Aloe vera

and rhein (Fig.4.1). Likewise, intestinal microflora appears to be dependent upon the presence of
metabolize acemannan to smaller compounds by the anthraquinone fraction, in particular aloe-
cleavage of the -14 linkages. emodin, the mechanism by which this fraction
The oral administration of Aloe vera whole causes intestinal tumours is presently unknown.
leaf preparation induces hyperplasia of the large The administration of Aloe vera whole leaf prepa-
intestine in mice and rats, and adenomas and rations to rats induced non-neoplastic lesions
carcinomas of the large intestine in rats. Rats (i.e. goblet cell hyperplasia) that are not observed
dosed orally with acemannan for up to 6months with other Aloe vera preparations (e.g. Aloe vera
did not display any treatment-related patholog- gel, decolorized gel, or decolorized whole leaf).
ical changes. Likewise, rats given Aloe vera gel These non-neoplastic changes may result in the
orally, and mice and rats given charcoal-treated formation of ROS that could promote neoplastic
Aloe vera gel did not show any treatment-related progression. Aloe-emodin has been shown to
non-neoplastic or neoplastic lesions. generate ROS after incubation with human cells
Aloe vera preparations, acemannan, and in vitro, probably as a result of its anthraquinone
aloin A do not display genotoxic activity in structure. Aloe-emodin also contains a benzylic
bacterial assays for mutagenesis and/or other hydroxy moiety that has the potential to undergo
assays for genotoxicity. In contrast, aloe-emodin esterification (e.g. sulfation) to a reactive elec-
is mutagenic in Salmonella typhimurium trophile that could bind to DNA. This pathway
reversion assays, induces unscheduled DNA does not appear to have been investigated. The
synthesis, gene mutations, micronucleus forma- incubation of Aloe vera whole leaf preparations
tion, and chromosomal aberrations, inhibits with human intestinal microflora results in an
topoisomerase II, and gives positive results increased production of short-chain fatty acids.
in comet assays. These data suggest that the The impact of this increased production is pres-
neoplastic response observed with Aloe vera is a ently not clear. Although the mechanism by
consequence of the conversion of the anthrone which Aloe vera whole leaf preparations induce
C-glycosides to aloe-emodin, which by itself or intestinal neoplasms in rats is not fully under-
in combination with other Aloe vera components stood, it is clear that the molecular pathways
is responsible for the development of adenomas observed in the intestinal neoplasms induced in
and carcinomas in the large intestine. rats by Aloe vera whole leaf preparations are also
In the 2-year bioassays with Aloe vera whole observed in human colorectal cancers.
leaf preparations, mice were exposed to nearly 10
times more test material (on a per kg bw basis)
than rats. In spite of this higher exposure, the 5. Summary of Data Reported
mice did not develop adenoma or carcinoma of
the large intestine, which may be due to the fact 5.1 Exposure data
that the intestinal bacteria in mice are less efficient
than the intestinal bacteria of rats in converting Aloe vera, also known as Aloe barbadensis,
aloin A and aloin B to aloe-emodin anthrone and is a perennial succulent plant with green fleshy
subequently aloe-emodin. In addition, mice have leaves. The leaves contain two types of liquids: a
shorter gastrointestinal tracts and faster gastro- yellow bitter latex under the skin, and a viscous
intestinal transit times than rats, which could gel in the inner section. Commercial products are
contribute to the lack of a tumour response in made from processed leaves. Four major types
mice. of processed products were identified: whole leaf
65
IARC MONOGRAPHS 108
extract; decolorized whole leaf extract; inner-leaf enhancing effect of Aloe vera gel cream or aloe-
gel; and dried bitter latex. Decolorization removes emodin cream on the photocarcinogenic activity
pigments and anthraquinones from the whole of simulated sunlight in female mice based on
leaf extract. The dried latex has medicinal uses an increase in the multiplicity of squamous
as a laxative. The other forms are used in foods, cell papilloma, carcinoma or carcinoma in situ
dietary supplements, beverages, and cosmetic (combined). There was a significant enhancing
products. Exposure data, where they exist, do not effect of the whole leaf extract cream or decol-
identify the nature of products containing Aloe orized whole leaf extract cream on the photocar-
vera used by consumers. cinogenic activity of simulated sunlight in both
male and female mice, based on an increase in
the multiplicity of squamous cell papilloma,
5.2 Human carcinogenicity data carcinoma or carcinoma in situ (combined).
5.4 Mechanistic and other relevant
5.3 Animal carcinogenicity data data
Whole leaf extract of Aloe vera was tested for The C-glycosides aloin A and aloin B, which
carcinogenicity after oral administration in one are components of Aloe vera latex, are converted
2-year study in mice, and one 2-year study in rats. to aloe-emodin-9-anthrone by bacteria present
In male and female rats, drinking-water in the gastrointestinal tract of rats and humans.
containing whole leaf extract of Aloe vera caused Aloe-emodin-9-anthrone undergoes sequen-
significantly increased incidences of adenoma of tial oxidation to aloe-emodin and rhein.
the large intestine (colon and caecum) and carci- Preparations of Aloe vera, acemannan, and aloin
noma of the large intestine (colon and caecum), A, do not display genotoxic activity in assays
tumours rarely developed spontaneously in rats. for mutagenesis in bacteria and/or other assays
In the 2-year study in mice, there was no for genotoxicity. In contrast, aloe-emodin has
significantly increased incidence of any type genotoxic activity. These data suggest that the
of tumours in males or females given drink- neoplastic response observed with Aloe vera is a
ing-water containing whole leaf extract of Aloe consequence of the conversion of the anthrone
vera. C-glycosides to aloe-emodin, which by itself or
In a study of photo-co-carcinogenesis with in combination with other components of Aloe
simulated sunlight, four articles were studied by vera is responsible for the adenomas and carci-
skin application in hairless mice: three test arti- nomas in the large intestine of rats.
cles containing Aloe vera that included gel, whole
leaf extract, and decolorized whole leaf extract;
and an aloe-emodin preparation. Almost all mice 6. Evaluation
exposed to simulated sunlight developed skin
neoplasms. No increase in the incidence of skin
neoplasms was observed in the groups receiving
6.1 Cancer in humans
any of the four test articles applied as a cream There is inadequate evidence in humans for
followed by simulated sunlight when compared the carcinogenicity of Aloe vera.
with the group receiving control cream followed
by simulated sunlight. There was a significant
66
Aloe vera
6.2 Cancer in experimental animals Phytother Res, 21(3):23944. doi:10.1002/ptr.2057

PMID:17163579
There is sufficient evidence in experimental Brown JP & Dietrich PS (1979). Mutagenicity of anthraqui-
none and benzanthrone derivatives in the Salmonella/
animals for the carcinogenicity of whole leaf microsome test: activation of anthraquinone glyco-
extract of Aloe vera. sides by enzymic extracts of rat cecal bacteria. Mutat
Res, 66(1):924. doi:10.1016/0165-1218(79)90003-X
PMID:370585
6.3 Overall evaluation Brown JP, Dietrich PS, Brown RJ (1977). Frameshift
mutagenicity of certain naturally occurring phenolic
Whole leaf extract of Aloe vera is possibly compounds in the Salmonella/microsome test: acti-
vation of anthraquinone and flavonol glycosides by gut
carcinogenic to humans (Group 2B). bacterial enzymes [proceedings]. Biochem Soc Trans,
5(5):148992. PMID:336434
Brusick D & Mengs U (1997). Assessment of the
References genotoxic risk from laxative senna products.
Environ Mol Mutagen, 29(1):19. doi:10.1002/
(SICI)1098-2280(1997)29:1<1::AID-EM1>3.0.CO;2-J
Ahlawat KS & Khatkar BS (2011). Processing, food appli- PMID:9020301
cations and safety of aloe vera products: a review. J Channe Gowda D, Neelisiddaiah B, Anjaneyalu YV
Food Sci Technol, 48(5):52533. doi:10.1007/s13197-011- (1979). Structural studies of polysaccharides from
0229-z PMID:23572784 Aloe vera. Carbohydr Res, 72:2015. doi:10.1016/
AHP; American Herbal Pharmacopoeia (2012). Aloe Vera S0008-6215(00)83936-1
Leaf, Aloe Vera Leaf Juice, Aloe Vera Inner Leaf Juice. Che Q-M, Akao T, Hattori M, Kobashi K, Namba T (1991).
Scotts Valley (CA): American Herbal Pharmacopoeia. Isolation of a human intestinal bacterium capable of
Akao T, Che Q-M, Kobashi K, Hattori M, Namba T transforming barbaloin to aloe-emodin anthrone.
(1996). A purgative action of barbaloin is induced Planta Med, 57(1):159. doi:10.1055/s-2006-960007
by Eubacterium sp. strain BAR, a human intestinal PMID:2062951
anaerobe, capable of transforming barbaloin to aloe- Chen Y-Y, Chiang S-Y, Lin J-G, Yang JS, Ma YS, Liao CL
emodin anthrone. Biol Pharm Bull, 19(1):1368. etal. (2010). Emodin, aloe-emodin and rhein induced
doi:10.1248/bpb.19.136 PMID:8820926 DNA damage and inhibited DNA repair gene expres-
Azaroual L, Liazid A, Barbero GF, Brigui J, Palma M, sion in SCC-4 human tongue cancer cells. Anticancer
Barroso CG (2012). Improved chromatographic Res, 30(3):94551. PMID:20393018
methods for determination of bioactive compounds Chiu T-H, Lai W-W, Hsia T-C, Yang JS, Lai TY, Wu
from Aloe vera leaves. ISRN Chromatography, 2012:17. PP et al. (2009). Aloe-emodin induces cell death
doi:10.5402/2012/609095 through S-phase arrest and caspase-dependent path-
Boudreau MD & Beland FA (2006). An evaluation of ways in human tongue squamous cancer SCC-4 cells.
the biological and toxicological properties of Aloe Anticancer Res, 29(11):450311. PMID:20032398
barbadensis (miller), Aloe vera. J Environ Sci Health Committee of Experts on Cosmetic Products (2008).
C Environ Carcinog Ecotoxicol Rev, 24(1):10354. Aloe extracts with anthraquinones. Active ingredients
doi:10.1080/10590500600614303 PMID:16690538 used in cosmetics: safety survey. Strasbourg, France:
Boudreau MD, Beland FA, Nichols JA, Pogribna M Council of Europe Publishing; pp. 927.
(2013a). Toxicology and carcinogenesis studies of a Cosmetic Ingredient Review Expert Panel (2007). Final
noncolorized whole leaf extract of Aloe barbadensis report on the safety assessment of AloeAndongensis
Miller (Aloe vera) in F344/N rats and B6C3F1 mice Extract, Aloe Andongensis Leaf Juice,aloe Arborescens
(drinking water study). Natl Toxicol Program Tech Rep Leaf Extract, Aloe Arborescens Leaf Juice, Aloe
Ser, 577(577):1266. PMID:24042237 Arborescens Leaf Protoplasts, Aloe Barbadensis Flower
Boudreau MD, Mellick PW, Olson GR, Felton RP, Thorn Extract, Aloe Barbadensis Leaf, Aloe Barbadensis Leaf
BT, Beland FA (2013b). Clear evidence of carcinogenic Extract, Aloe Barbadensis Leaf Juice,aloe Barbadensis
activity by a whole-leaf extract of Aloe barbadensis Leaf Polysaccharides, Aloe Barbadensis Leaf Water,
miller (aloe vera) in F344/N rats. Toxicol Sci, 131(1):26 Aloe Ferox Leaf Extract, Aloe Ferox Leaf Juice, and Aloe
39. doi:10.1093/toxsci/kfs275 PMID:22968693 Ferox Leaf Juice Extract. Int J Toxicol, 26(1):Suppl 2:
Brandin H, Viitanen E, Myrberg O, Arvidsson A-K 150. doi:10.1080/10915810701351186 PMID:17613130
(2007). Effects of herbal medicinal products and food DalBelo SE, Gaspar LR, Maia Campos PM (2006).
supplements on induction of CYP1A2, CYP3A4 and Moisturizing effect of cosmetic formulations containing
MDR1 in the human colon carcinoma cell line LS180. Aloe vera extract in different concentrations assessed
67
IARC MONOGRAPHS 108
by skin bioengineering techniques. Skin Res Technol, GRIN; Germplasm Resources Information Network
12(4):2416. doi:10.1111/j.0909-752X.2006.00155.x (2013). National Germplasm Resources Laboratory,
PMID:17026654 Beltsville (Md): USDA. Available from: http://www.
Davis B & Goux WJ (2009). Single-laboratory validation ars-grin.gov/cgi-bin/npgs/html/genform.pl, accessed
of an NMR method for the determination of aloe vera 5 June 2013
polysaccharide in pharmaceutical formulations. J Grindlay D & Reynolds T (1986). The Aloe vera phenom-
AOAC Int, 92(6):160716. PMID:20166576 enon: a review of the properties and modern uses of the
De Witte P & Lemli J (1988). Excretion and distribution leaf parenchyma gel. J Ethnopharmacol, 16(2-3):11751.
of [14C]rhein and [14C]rhein anthrone in rat. J Pharm doi:10.1016/0378-8741(86)90085-1 PMID:3528673
Pharmacol, 40(9):6525. doi:10.1111/j.2042-7158.1988. Groom QJ & Reynolds T (1987). Barbaloin in aloe species.
tb05330.x PMID:2907037 Planta Med, 53(4):3458. doi:10.1055/s-2006-962735
Dentali S (2013). Nondecolorized essential qualifier for PMID:17269040
NTP aloe vera study material. Toxicol Sci, 133(2):342 Hamman JH (2008). Composition and applications of Aloe
doi:10.1093/toxsci/kft072 PMID:23492813 vera leaf gel. Molecules, 13(8):1599616. doi:10.3390/
Djuv A & Nilsen OG (2012). Aloe vera juice: IC and molecules13081599 PMID:18794775
dual mechanistic inhibition of CYP3A4 and CYP2D6. Hattori M, Akao T, Kobashi K, Namba T (1993).
Phytother Res, 26(3):44551. PMID:21842479 Cleavages of the O- and C-glucosyl bonds of anthrone
Dreessen M & Lemli J (1988). Studies in the field of drugs and 10,10-bianthrone derivatives by human intes-
containing anthraquinone derivatives. XXXVI. The tinal bacteria. Pharmacology, 47:Suppl 1: 12533.
metabolism of cascarosides by intestinal bacteria. doi:10.1159/000139851 PMID:8234419
Pharm Acta Helv, 63(9-10):2879. PMID:3237734 Hattori M, Kanda T, Shu Y-Z, Akao T, Kobashi K, Namba T
Duke JA & Beckstrom-Sternberg SM (1994). Acceptable (1988). Metabolism of barbaloin by intestinal bacteria.
Levels of Flavoring Ingredients? Developmental Food Chem Pharm Bull (Tokyo), 36(11):44626. doi:10.1248/
Science, 34:741757. cpb.36.4462 PMID:3246014
Elsohly MA, Gul W, Avula B, Khan IA (2007). Heidemann A, Vlkner W, Mengs U (1996). Genotoxicity
Determination of the anthraquinones aloe-emodin of aloeemodin in vitro and in vivo. Mutat Res,
and aloin-A by liquid chromatography with mass 367(3):12333. doi:10.1016/0165-1218(95)00084-4
spectrometric and diode array detection. J AOAC Int, PMID:8600368
90(1):2842. PMID:17373434 Herlihy JT, Bertrand HA, Kim JD, Ikeno Y, Yu BP
EMA (2006). Community herbal monograph on Aloe (1998). Effects of Aloe vera ingestion in the rat I.
barbadensis MILLER and on Aloe (various species, Growth, food and fluid intake and serum chemistry.
mainly Aloe ferox MILLER and its hybrids). London, Phytother Res, 12(3):1838. doi:10.1002/(SICI)1099-
UK: European Medicines Agency. 1573(199805)12:3<183::AID-PTR219>3.0.CO;2-Q
Eur Ph; European Pharmacopoeia (2008). European Hirata T, Sakano S, Suga T (1981). Biotransformation of
pharmacopoeia. 7.0. Strasbourg, France: European aloenin, a bitter glucoside constituent of Aloe arbo-
Directorate for the Quality of Medicines & HealthCare. rescens, by rats. Experientia, 37(12):12523. doi:10.1007/
FDA; Food and Drug Administration (2002). Status BF01948341 PMID:7035211
of certain additional over-the-counter drug cate- Hsia T-C, Yang J-S, Chen G-W, Chiu TH, Lu HF, Yang
gory II and III active ingredients. Fed Regist, MD etal. (2009). The roles of endoplasmic reticulum
67(90):311257. Availabe from: http://www.fda.gov/ stress and Ca2+ on rhein-induced apoptosis in A-549
downloads/Drugs/DevelopmentApprovalProcess/ human lung cancer cells. Anticancer Res, 29(1):30918.
DevelopmentResources/Over-the- PMID:19331167
CounterOTCDrugs/StatusofOTCRu lema k ings/ IASC (2009). IASC Labeling Guidance. Silver Spring
ucm094018.pdf. PMID:12001972 (MD): International Aloe Science Council. Available
Femenia A, Sanchez E, Simal S, Rossell C (1999). from: http://www.iasc.org/pdfs/10_0405_IASC_
Compositional features of polysacchardies from Labeling_Guidance_Definitions.pdf.
Aloe vera (Aloe barbadensis Miller) plant tissues. IASC (2013a). The nomenclature of Aloe vera. Silver
Carbohydr Polym, 39(2):10917. doi:10.1016/ Spring (MD): International Aloe Science Council.
S0144-8617(98)00163-5 Available from: http://www.iasc.org/Aloe_Taxonomy.
Fogleman RW, Shellenberger TE, Balmer MF, Carpenter pdf, accessed 3 June 2014.
RH, McAnalley BH (1992). Subchronic oral admin- IASC (2013b). Aloe vera quality standard. Silver
istration of acemannan in the rat and dog. Vet Hum Spring (MD), USA: International Aloe Science
Toxicol, 34(2):1447. PMID:1509675 Council. Available from: http://iasc.org/pdfs/
Geisser J & Kratz E (2010). Determination of Aloe vera AloeVeraQualityStandard.pdf, accessed 3 June 2014.
gel in cosmetics. CBS Camag Bibliography Service, Ikeno Y, Hubbard GB, Lee S, Yu BP, Herlihy JT (2002).
104:1315. The influence of long-term Aloe vera ingestion on
68
Aloe vera
age-related disease in male Fischer 344 rats. Phytother Lin M-L, Lu Y-C, Chung J-G, Li YC, Wang SG, N G SH etal.
Res, 16(8):7128. doi:10.1002/ptr.1022 PMID:12458471 (2010). Aloe-emodin induces apoptosis of human naso-
IMS Health (2012). Multinational Integrated Data Analysis pharyngeal carcinoma cells via caspase-8-mediated
(MIDAS). Worldwide Sales of Selected Products. activation of the mitochondrial death pathway. Cancer
Plymouth Meeting, Pennsylvania: IMS Health Lett, 291(1):4658. doi:10.1016/j.canlet.2009.09.016
Ishii Y, Tanizawa H, Takino Y (1987). Determination PMID:19942342
of barbaloin in rat serum. Chem Pharm Bull (Tokyo), Mandal G & Das A (1980). Structure of the D-galactan
35(11):46424. doi:10.1248/cpb.35.4642 PMID:3442898 isolated from Aloe barbadensis Miller. Carbohydr Res,
Jiao P, Jia Q, Randel G, Diehl B, Weaver S, Milligan G 86(2):24757. doi:10.1016/S0008-6215(00)85902-9
(2010). Quantitative 1H-NMR spectrometry method Mandrioli R, Mercolini L, Ferranti A, Fanali S, Raggi
for quality control of Aloe vera products. J AOAC Int, MA (2011). Determination of aloe emodin in Aloe
93(3):8428. PMID:20629385 vera extracts and commercial formulations by
Joseph B & Raj SJ (2010). Pharmacognostic and phyto- HPLC with tandem UV absorption and fluorescence
chemical properties of Aloe vera Linnan overview. detection. Food Chem, 126(1):38793. doi:10.1016/j.
International Journal of Pharmaceutical Sciences foodchem.2010.10.112
Review and Research, 4:106110. Matsuda Y, Yokohira M, Suzuki S, Hosokawa K,
JP XVI, The Japanese Pharmacopoeia (2011). The Japanese Yamakawa K, Zeng Y etal. (2008). One-year chronic
Pharmacopoeia. 16th ed. English Version, Tokyo, Japan: toxicity study of Aloe arborescens Miller var. natalensis
Ministry of Health, Labour and Welfare. Berger in Wistar Hannover rats. A pilot study. Food
Kim KH, Park JH (2006). Quality control and standard- Chem Toxicol, 46(2):7339. doi:10.1016/j.fct.2007.09.107
ization of Aloe products. In: Park YI, Lee SK, editors. PMID:17981383
New Perspectives on Aloe. New York (NY), USA: Morgan M, Bone K, Mills S et al. (2005). Aloe. Safety
Springer Science and Business Media; pp. 169189 monograph. In: Mills S, Bone K, editors. The essential
Kratz E (2009). Aloe vera. Evidence of active content. guide to herbal safety. St. Louis (MO), USA: Elsevier
Cossma, 10:1213. Churchill Livingstone; pp. 233240
Kuzuya H, Shimpo K, Beppu H (2004). Aloe lectins and Morimoto I, Watanabe F, Osawa T, Okitsu T, Kada T (1982).
their activities. In: Reynolds T, editor. Aloes: The genus Mutagenicity screening of crude drugs with Bacillus
Aloe. 38th ed. Boca Raton (FL), USA: CRC Press; pp. subtilis rec-assay and Salmonella/microsome rever-
88110 sion assay. Mutat Res, 97(2):81102. doi:10.1016/0165-
Lachenmeier K, Kuepper U, Musshoff F et al. (2005). 1161(82)90007-3 PMID:6804865
Quality control of Aloe vera beverages. Electronic Mller SO & Stopper H (1999). Characterization of the
Journal of Environmental, Agricultural and Food genotoxicity of anthraquinones in mammalian cells.
Chemistry, 4:10331042. Biochim Biophys Acta, 1428(2-3):40614. doi:10.1016/
Lai W-W, Yang J-S, Lai K-C, Kuo CL, Hsu CK, Wang S0304-4165(99)00064-1 PMID:10434060
CK et al. (2009). Rhein induced apoptosis through Mller SO, Eckert I, Lutz WK, Stopper H (1996).
the endoplasmic reticulum stress, caspase- and mito- Genotoxicity of the laxative drug components emodin,
chondria-dependent pathways in SCC-4 human aloe-emodin and danthron in mammalian cells: topoi-
tongue squamous cancer cells. In Vivo, 23(2):30916. somerase II mediated? Mutat Res, 371(3-4):16573.
PMID:19414420 doi:10.1016/S0165-1218(96)90105-6 PMID:9008718
Lang W (1993). Pharmacokinetic-metabolic studies with NCCAM (2012). Aloe vera. Bethesda (MD): National
14C-aloe emodin after oral administration to male Center for Complementary and Alternative Medicine.
and female rats. Pharmacology, 47:Suppl 1: 1109. Available from: http://nccam.nih.gov/health/aloevera,
doi:10.1159/000139849 PMID:8234417 accessed 5 June 2014.
Lee S, Do SG, Kim SY, Kim J, Jin Y, Lee CH (2012). Nesslany F, Simar-Meintires S, Ficheux H, Marzin D
Mass spectrometry-based metabolite profiling and (2009). Aloe-emodin-induced DNA fragmentation in
antioxidant activity of Aloe vera ( Aloe barbadensis the mouse in vivo comet assay. Mutat Res, 678(1):139.
Miller) in different growth stages. J Agric Food Chem, doi:10.1016/j.mrgentox.2009.06.004 PMID:19559101
60(45):112228. doi:10.1021/jf3026309 PMID:23050594 Newton LE (2004). Aloes in habitat. In: Reynolds T, editor.
Leon L (2003). The medicinal plant Aloe. Ganzheitliche Aloes: the genus Aloe. Boca Raton (FL), USA: CRC
Tiermedizin, 17:138143. Press; pp. 314
Lin M-L, Chen S-S, Lu Y-C, Liang RY, Ho YT, Yang CY Ngo MQ, Nguyen NN, Shah SA (2010). Oral aloe vera for
etal. (2007). Rhein induces apoptosis through induction treatment of diabetes mellitus and dyslipidemia. Am J
of endoplasmic reticulum stress and Ca2+-dependent Health Syst Pharm, 67(21):180411, 1806, 1808 passim.
mitochondrial death pathway in human nasopharyn- doi:10.2146/ajhp100182 PMID:20966143
geal carcinoma cells. Anticancer Res, 27:5A: 331322. NHANES (2010). National Health and Nutrition
PMID:17970076 Examination Survey (19992010). Atlanta (GA):
69
IARC MONOGRAPHS 108
Centers for Disease Control and Prevention. Available Phytochem Anal, 9(4):18691. doi:10.1002/(SICI)1099-
from: http://www.cdc.gov/nchs/nhanes.htm; accessed 1565(199807/08)9:4<186::AID-PCA406>3.0.CO;2-#
5 June 2014. Park YI, Jo TH (2006). Perspectives of industrial appli-
Ni Y, Turner D, Yates KM, Tizard I (2004). Isolation and cation of Aloe vera. In: Park YI, Lee SK, editors. New
characterization of structural components of Aloe vera Perspectives on Aloe. New York (NY), USA: Springer
L. leaf pulp. Int Immunopharmacol, 4(14):174555. Science+Business Media; pp. 191200
doi:10.1016/j.intimp.2004.07.006 PMID:15531291 Pelley RP, Martini WJ, Liu DQ etal. (1998). Multiparameter
NLM (2012). Products that contain active ingredient - analysis of commercial Aloe vera materials and
Aloe vera. Dietary supplements labels database. United comparison to Aloe barbadensis Miller extracts.
States National Library of Medicine. Available from: Subtropical Plant Science, 50:114.
http://www.dsld.nlm.nih.gov/dsld/; accessed 5 June Pellizzoni M, Molinari GP, Lucini L (2011). Stability of the
2014. main Aloe fractions and Aloe-based commercial prod-
NTP; National Toxicology Program (2010). ucts under different storage conditions Agrochimica,
Photocarcinogenesis study of aloe vera [CAS NO. 5:288296.
481721(Aloe-emodin)] in SKH-1 mice (simulated Perkins SL & Livesey JF (1993). A rapid high-performance
solar light and topical application study). Natl Toxicol thin-layer chromatographic urine screen for laxative
Program Tech Rep Ser, 553(553):733, 3597, 99103 abuse. Clin Biochem, 26(3):17981. doi:10.1016/0009-
passim. PMID:21031007 9120(93)90023-Y PMID:8330386
Nutrition Business Journal (2010). NBJs Supplement Pogribna M, Freeman JP, Paine D, Boudreau MD
Business Report. An analysis of markets, trends, (2008). Effect of Aloe vera whole leaf extract on short
competition and strategy in the U.S. dietary supple- chain fatty acids production by Bacteroides fragilis,
ment industry. New York (NY), USA: Penton Media, Bifidobacterium infantis and Eubacterium limosum.
Inc. Lett Appl Microbiol, 46(5):57580. doi:10.1111/j.1472-
Nutrition Business Journal (2012). NBJs Supplement 765X.2008.02346.x PMID:18363656
Business Report. An analysis of markets, trends, Ramachandra CT & Rao PS (2008). Processing of Aloe
competition and strategy in the U.S. dietary supple- vera leaf gel: a review. Am J Agric Biol Sci, 3(2):50210.
ment industry. . New York (NY), USA: Penton Media, doi:10.3844/ajabssp.2008.502.510
Inc. Available from: http://newhope360.com/2012- Ramrez Durn R, Ceniceros Almaguer L, Cavazos Rocha
supplement-business-report; accessed 5 June 2014. NC, Silva Flores PG, De Torres NW (2008). Comparison
ONeil MJ, Heckelman PE, Koch CB et al. (2006). The of high-performance liquid chromatographic and thin-
Merck Index - An Encyclopedia of Chemicals, Drugs, layer chromatographic methods for determination of
and Biologicals. 14th Ed. Version 14.6. Whitehouse aloin in herbal products containing Aloe vera. J AOAC
Station (NJ), USA: Merck & Co., Inc. Int, 91(6):126570. PMID:19202785
Okamura N, Asai M, Hine N, Yagi A (1996). High- Reynolds T (2004). Aloe chemistry. In: Reynolds T, editor.
performance liquid chromatographic determination of Aloes: The genus Aloe. 38th Ed. Boca Raton (FL), USA:
phenolic compounds in Aloe species. J Chromatogr A, CRC Press; pp. 3974
746(2):22531. doi:10.1016/0021-9673(96)00342-1 Rodrguez ER, Martn JD, Romero CD (2010). Aloe vera
Paes-Leme AA, Motta ES, De Mattos JCP, Dantas FJ, as a functional ingredient in foods. Crit Rev Food Sci
Bezerra RJ, Caldeira-de-Araujo A (2005). Assessment Nutr, 50(4):30526. doi:10.1080/10408390802544454
of Aloe vera (L.) genotoxic potential on Escherichia coli PMID:20301017
and plasmid DNA. J Ethnopharmacol, 102(2):197201. Sacc D, Bogoni P, Procida G (2001). Aloe exudate:
doi:10.1016/j.jep.2005.06.013 PMID:16054315 characterization by reversed phase HPLC and head-
Paez A, Michael Gebre G, Gonzalez ME, Tschaplinski space GC-MS. J Agric Food Chem, 49(10):452630.
TJ (2000). Growth, soluble carbohydrates, and aloin doi:10.1021/jf010179c PMID:11599983
concentration of Aloe vera plants exposed to three irra- SciFinder (2013). CAS Registry Number 8001976.
diance levels. Environ Exp Bot, 44(2):1339. doi:10.1016/ Columbus (OH). USA: Chemical Abstracts Service,
S0098-8472(00)00062-9 PMID:10996366 American Chemical Society. Accessed: 15 January 2013
Pandiri AR, Sills RC, Hoenerhoff MJ, Peddada SD, Ton Sehgal I, Winters WD, Scott M, David A, Gillis G, Stoufflet
TV, Hong HH etal. (2011). Aloe vera non-decolorized T etal. (2013b). Toxicologic assessment of a commer-
whole leaf extract-induced large intestinal tumors in cial decolorized whole leaf aloe vera juice, lily of the
F344 rats share similar molecular pathways with human desert filtered whole leaf juice with aloesorb. J Toxicol,
sporadic colorectal tumors. Toxicol Pathol, 39(7):1065 Epub802453 doi:10.1155/2013/802453 PMID:23554812
74. doi:10.1177/0192623311422081 PMID:21937742 Sehgal I, Winters WD, Scott M, Kousoulas K (2013). An
Park MK, Park JH, Kim NY, Shin YG, Choi YS, Lee JG in vitro and in vivo toxicologic evaluation of a stabi-
etal. (1998). Analysis of 13 phenolic compounds in Aloe lized aloe vera gel supplement drink in mice. Food
species by high performance liquid chromatography.
70
Aloe vera
Chem Toxicol, 55:36370. doi:10.1016/j.fct.2013.01.012 Westendorf J, Marquardt H, Poginsky B, Dominiak

PMID:23376510 M, Schmidt J, Marquardt H (1990). Genotoxicity of
Sehgal I, Winters WD, Scott M, Kousoulas K (2013a). An naturally occurring hydroxyanthraquinones. Mutat
in vitro and in vivo toxicologic evaluation of a stabi- Res, 240(1):112. doi:10.1016/0165-1218(90)90002-J
lized aloe vera gel supplement drink in mice. Food PMID:2294411
Chem Toxicol, 55:36370. doi:10.1016/j.fct.2013.01.012 WHO (1999). Aloe and Aloe vera gel. WHO Monographs
PMID:23376510 on selected medicinal plants. Vol. 1. Geneva,
Shah AH, Qureshi S, Tariq M, Ageel M (1989). Toxicity Switzerland: World Health Organization; pp. 3349
studies on six plants used in the traditional Arab system (available from http://apps.who.int/medicinedocs/
of medicine. Phytother Res, 3(1):259. doi:10.1002/ en/d/Js2200e/5.html)
ptr.2650030107 Williams LD, Burdock GA, Shin E, Kim S, Jo TH, Jones
Shao A, Broadmeadow A, Goddard G, Bejar E, Frankos KN etal. (2010). Safety studies conducted on a propri-
V (2013). Safety of purified decolorized (low anth- etary high-purity aloe vera inner leaf fillet prepara-
raquinone) whole leaf Aloe vera (L) Burm. f. juice tion, Qmatrix. Regul Toxicol Pharmacol, 57(1):908.
in a 3-month drinking water toxicity study in F344 doi:10.1016/j.yrtph.2010.01.002 PMID:20096744
rats. Food Chem Toxicol, 57:2131. doi:10.1016/j. Wlfle D, Schmutte C, Westendorf J, Marquardt H
fct.2013.03.002 PMID:23500775 (1990). Hydroxyanthraquinones as tumor promoters:
Spohn R (2013). Water colour of Aloe vera. Available from: enhancement of malignant transformation of
http://www.spohns.de/heilpf lanzenillus/aloevera. C3H mouse fibroblasts and growth stimulation of
html; accessed 28 May 2014. primary rat hepatocytes. Cancer Res, 50(20):65404.
Steenkamp V & Stewart MJ (2007). Medicinal appli- PMID:2208114
cations and toxicological activities of Aloe prod- Yagi A, Nakamori J, Yamada T, Iwase H, Tanaka T, Kaneo
ucts. Pharmaceutical Biology, 45(5):41120. Y et al. (1999). In vivo metabolism of aloemannan.
doi:10.1080/13880200701215307 Planta Med, 65(5):41720. doi:10.1055/s-1999-14018
Stolk LM & Hoogtanders K (1999). Detection of laxa- PMID:10418327
tive abuse by urine analysis with HPLC and diode Yaron A (1993). Characterization of Aloe vera gel before
array detection. Pharm World Sci, 21(1):403. and after autodegradation, and stabilization of the
doi:10.1023/A:1008647424809 PMID:10214668 natural fresh gel. Phytother Res, 7(7):11 doi:10.1002/
Tang JC, Yang H, Song XY, Song XH, Yan SL, Shao JQ ptr.2650070706
etal. (2009). Inhibition of cytochrome P450 enzymes Yokohira M, Matsuda Y, Suzuki S, Hosokawa K,
by rhein in rat liver microsomes. Phytother Res, Yamakawa K, Hashimoto N et al. (2009). Equivocal
23(2):15964. doi:10.1002/ptr.2572 PMID:18814214 colonic carcinogenicity of Aloe arborescens Miller
Turner CE, Williamson DA, Stroud PA, Talley DJ (2004). var. natalensis berger at high-dose level in a Wistar
Evaluation and comparison of commercially available Hannover rat 2-y study. J Food Sci, 74(2):T2430.
Aloe vera L. products using size exclusion chroma- doi:10.1111/j.1750-3841.2009.01070.x PMID:19323775
tography with refractive index and multi-angle laser Zaffaroni M, Mucignat C, Pecere T, Zagotto G, Frapolli
light scattering detection. Int Immunopharmacol, R, DIncalci M etal. (2003). High-performance liquid
4(14):172737. doi:10.1016/j.intimp.2004.07.004 chromatographic assay for the determination of Aloe
PMID:15531289 Emodin in mouse plasma. J Chromatogr B Analyt
Ulbricht C, Armstrong J, Basch E, Basch S, Bent S, Technol Biomed Life Sci, 796(1):1139. doi:10.1016/j.
Dacey C et al. (2007). An evidence-based systematic jchromb.2003.08.012 PMID:14552822
review of Aloe vera by the natural standard research
collaboration. J Herb Pharmacother, 7(3-4):279323.
doi:10.1080/15228940802153339 PMID:18928148
Vogler BK & Ernst E (1999). Aloe vera: a systematic
review of its clinical effectiveness. Br J Gen Pract,
49(447):8238. PMID:10885091
Vyth A & Kamp PE (1979). Detection of anthraquinone
laxatives in the urine. Pharm Weekbl, 1(1):4569.
doi:10.1007/BF02293250
Waller TA, Pelley RP, Strickland FM (2004). Industrial
processing and quality control of Aloe barbadensis
(Aloe vera) gel. In: Reynolds T, editor. Aloes: The genus
Aloe. 38th Ed. Boca Raton (FL), USA: CRC Press; pp.
139205
71
GOLDENSEAL
1. Exposure Data Common names: Hydrastis; Golden seal;

Yellow Indian plant; Yellow seal
Goldenseal (Hydrastis canadensis L.) was
traditionally used by native Americans as a medic- (b) Description
inal remedy and as a colouring agent (Sinclair & Goldenseal is a perennial plant, which grows
Catling, 2000; McKenna & Plotnikoff, 2010). In naturally in eastern USA and Canada (AHP,
his Collections for an Essay Towards a Materia 2001; Zieger & Tice, 1997). Goldenseal has one
Medica of the United States, the American bota- long-trunked basal leaves, a single stem, and two
nist Benjamin Smith Barton first mentioned the smaller leaves attached to the flowering stem.
medicinal use of H. canadensis by the Cherokee Usually, the plant has two rounded, lobed, and
(Hobbs, 1990). Today, the main application of H. double-toothed leaves on a forked branch, with
canadensis is for the prevention and treatment of one being larger than the other. The plant has a
skin disorders, dyspepsia, gastritis, peptic ulcer, knotted yellow rhizome and a solitary terminal
colitis, anorexia, menorrhagia, dysmenorrhoea, flower with three white sepals and many green-
sinusitis, mucosal inflammation, and other ish-white stamens in clusters, while the fruit
inflammatory conditions or infectious diseases is small and red raspberry-like (Zieger & Tice,
(BHMA, 1983; BHMA, 1992; NTP, 2010; Sun 1997; AHP, 2001). The plant grows up to about
et al., 2009; McKenna & Plotnikoff, 2010). 1 foot in height [30.5 cm] (Palmer, 1975; Duke,
2002).
1.1 Identification of the agent and
its major constituents 1.1.2 Chemical constituents and their
properties
1.1.1 Botanical data
The major constituents of goldenseal root
(a) Nomenclature are isoquinoline alkaloids such as hydrastine
(1.54%), berberine (2.5%), canadine (0.5%),
Chem. Abstr. Serv. Reg. No.: 84603-60-1
and other alkaloids (see Fig.1.2; BHMA, 1992).
Chem. Abstr. Name: Golden seal root Berberine is usually found in the roots of golden-
Botanical name: Hydrastis canadensis L. seal as a sulfate with a yield of 500060000ppm
(Fig.1.1) (HSDB, 1997). Hydrastine is also found in gold-
Family: Ranunculaceae enseal in concentrations of 1500040000ppm
Genus: Hydrastis (NTP, 2010). Junio et al. (2011) have also iden-
tified sideroxylin, 8-desmethyl-sideroxylin, and
Plant part: Root
73
IARC MONOGRAPHS 108
Fig.1.1 Goldenseal (Hydrastis canadensis L.), (b) Hydrastine

showing leaves, root, flower and seed
IUPAC name: 1(3H)-Isobenzofuranone,
6,7-dimethoxy-3-(5,6,7,8-tetrahydro-6-me-
thyl-1,3-dioxolo[4,5-g]isoquinolin-5-yl)-
Description: Solid with a melting point of
132 C and highly soluble in acetone and
benzene; insoluble in water (ONeil, 2013).
(c) Canadine

IUPAC name: 6H-benzo[g]-1,3-
b e n z o d i ox o l o[5 , 6 - ]q u i n o l i z i n e ,
5,8,13,13-tetrahydro-9,10-dimethoxy-(13S)
Description: Yellow pale solid with a melting
point of 172 C and soluble in methanol
(ONeil, 2013).
(d) Others
Other goldenseal components of the flavo-
From Koehlers Medicinal-Plants (1887) noid class include:
sideroxylin (CAS No., 3122-87-0; IUPAC
6-desmethyl-sideroxylin. Chemical Abstract
name, 4,5-dihydroxy-7-methoxy-6,8-di-
Registry (CAS) number, International Union of
methylflavone)
Pure and Applied Chemistry (IUPAC) names
and some physical properties of goldenseal 8-desmethyl-sideroxylin (CAS No., 80621-
major alkaloids are presented below (American 54-1; IUPAC name, 4H-1-benzopyran-4-one,
Chemical Society, 2014; Zieger & Tice, 1997). 5-hydroxy-2-(4-hydroxyphenyl)-7-methoxy-
6-methyl-)
(a) Berberine 6-desmethyl-sideroxylin (CAS No., 1194721-
03-3; IUPAC name, 4H-1-benzopyran-4-one,
5-hydroxy-2-(4-hydroxyphenyl)-7-methoxy-
IUPAC name: 7,8,13,13-Tetradehy- 8-methyl-).
dro-9,10-dimethoxy-2,3-(methylenedioxy)
berbinium For structural and molecular formulae and
relative molecular mass, see Huang & Johnston
Description: Yellow solid with a melting point
(1990), Junio et al. (2011), Li et al. (2011), and
of 145 C and slowly soluble in water (ONeil,
Fig.1.2.
2013).
74
Goldenseal
1.1.3 Technical and commercial products 1.2 Analysis

Official technical products are goldenseal Botanical identity and composition are
root, powdered goldenseal root, and powdered confirmed by macroscopic and microscopic
goldenseal root extract (NTP, 2010; USP, 2013). examinations of rhizome and root, as well as by
Other minor technical products may include thin-layer chromatography (TLC) (USP, 2013).
powdered goldenseal leaf and its derived extracts, TLC identification tests for the major golden-
as well as fluid extracts (Mikkelsen & Ash, 1988; seal alkaloids are reported in the United States
Oldfield, 2005; NTP, 2010). Pharmacopeia: a dry extract from roots and
Powdered goldenseal root and leaf products rhizomes contains at least 2% hydrastine and
are available as capsules and teas in combination 2.5% berberine calculated on a dried basis; a
with other herbs, in some over-the-counter (OTC) standardized developing solvent system includes
herbal supplements (Mikkelsen & Ash, 1988; ethyl acetate, butyl alcohol, formic acid, and water
Zieger & Tice, 1997; AHP, 2001). Goldenseal is (5:3:1:1); chromatogram analysis with ultraviolet
also found in eardrops, feminine cleansing prod- (UV) light at 365 nm demonstrates a lemon-
ucts, cold/flu remedies, allergy relief products, yellow fluorescence for berberine and a blue-
laxative products, and aids to digestion (Zieger & white fluorescence for hydrastine (USP, 2013).
Tice, 1997; AHP, 2001). Chemical derivatives of Content of the most active alkaloids is deter-
purified major components of goldenseal, such as mined by high-performance liquid chromatog-
berberine hydrochloride and berberine bisulfate, raphy (HPLC) using a mobile phase composed
are found in some commercial eyewash formula- of 0.1 M monobasic potassium phosphate and
tions (Zeiger & Tice, 1997). Hydrastine, another acetonitrile (60:40) (USP, 2013). An official
derivative, is commercially available in the form HPLC/UV method for hydrastine and berberine
of ()-hydrastine and used as an ingredient in in goldenseal has been published (Brown &
some decongestant nose sprays and feminine Roman, 2008), but other HPLC methods have
hygiene products (Zeiger & Tice, 1997; Huang & also been developed (Abourashed & Khan, 2001).
Johnston, 1990). Additional analyses report on the use of gold-
As a consequence of the high cost of genuine enseal and illicit drugs. Goldenseal may prevent
goldenseal, some commercial products contain the detection of illicit drugs (such as tetrahydro-
little or no goldenseal plant material (Govindan canabinol and barbiturates) in urine by inducing
& Govindan, 2000). Coptis chinensis has been their rapid elimination (Mikkelsen & Ash, 1988;
sold in place of Chinese goldenseal and has Hamon, 1990; Schwarzhoff & Cody, 1993).
been found as an adulterant of goldenseal
(Brown & Roman, 2008). Moreover, other species
that contain berberine, such as Coptis japonica, 1.3 Use
Berberis aquifolium, Berberis spp., Rumex spp.,
1.3.1 Indications and applications
and Xanthorhiza simplicissima, have been used
to adulterate goldenseal (Brown & Roman, 2008; (a) Medicinal use
Pengelly, 2012). Native Americans used goldenseal to treat
common conditions such as wounds, ulcers,
digestive disorders, cancer, and skin and eye
ailments (Hamon, 1990; Hobbs, 1990). Over the
years, goldenseal has been used to treat a variety
of digestive and haemorrhagic disorders. When
75
IARC MONOGRAPHS 108
Fig.1.2 Selected alkaloids and flavonoids from goldenseal

O O O
O O
N
O CH 3
H H H
O
N N
H 3 CO + H 3 CO H 3C O
O
O CH 3 OCH3 O CH 3
B erberine H ydrastine Canadine

C 20 H 18 N O 4 C 21 H 21 N O 6 C 20 H 21 N O 4
R M M = 336 R M M = 383 RM M = 339
OH OH OH
CH 3 H CH3
H 3C O O H 3 CO O H 3 CO O
H 3C H 3C H
OH O OH O OH O
Sideroxylin 8-D esmethyl-sideroxylin 6-D esmethyl-sideroxylin

C 18 H 16 O 5 C 17 H 14 O 5 C 17 H 14 O 5
R M M = 312.32 R M M = 298.29 RM M = 298.29
RMM, relative molecular mass

Compiled by the Working Group
applied topically, it is thought to possess slight Berberine, a major goldenseal alkaloid, has
antiseptic, astringent, and haemostatic qualities been used as a bitter tonic, diaphoretic, and
(NTP, 2010). antipyretic (Kulkarni et al., 1972), for the treat-
Some OTC dietary supplements containing ment of skin diseases, liver diseases, eye infec-
goldenseal are used as to treat menstrual disor- tions, diarrhoea, cholera, giardiasis, amoebiasis,
ders, minor sciatica, rheumatic and muscular and dermal leishmaniasis (Choudhry et al.,
pain (Hamon, 1990), allergy symptoms, cold 1972; Kulkarni et al., 1972; Martin et al., 1978;
and flu symptoms, motion sickness and nausea, Sabir et al., 1978; Khin-Maung-U et al., 1985;
earaches and ear infections, and chronic diar- Vennerstrom et al., 1990; Chi et al., 1994; Mller
rhoea from protozoal, fungal, and bacterial et al., 1995). Berberine appears to control psori-
infections (Zieger & Tice, 1997; AHP, 2001; asis through its ability to inhibit hyperprolifer-
Hwang et al., 2003). Goldenseal has also been ation (Mller et al., 1995). The evidence on the
used in combination with dietary vitamins and efficacy of berberine in treating peptic ulcers and
minerals in an attempt to treat symptoms of hyperacidity, and malaria, is conflicting (Sabir
AIDS (Zieger & Tice, 1997). It is also claimed to et al., 1978; Vennerstrom & Klayman, 1988).
have the ability to cleanse the body from mucus, Some other pharmacological properties that
toxins, and waste (Zieger & Tice, 1997). have been identified for berberine include anti-
platelet, anticerebral ischaemia, vasodilation, and
76
Goldenseal
anti-arrythmia (Peng et al., 1997). Berberine is 1.4 Production, sales, and

thought to increase ileal contractility and acetyl- consumption
choline retention by cholinesterase activity and
is believed to be the active ingredient of Coptis 1.4.1 Production
rhizome, used to treat amnesia (Peng et al., 1997).
A clinical trial using berberine suggested No data on production processes or volumes
that it is effective in improving cardiac perfor- were available to the Working Group.
mance in patients with heart failure (Marin-
Neto et al., 1988). Berberine appears to exercise a 1.4.2 Sales
direct depressive effect on myocardial, vascular, According to data from 2012 IMS Health
and smooth musculature (Sabir & Bhide, 1971; MIDAS, worldwide sales of goldenseal root
Creasey, 1977, 1979). Berberine may also have (Hydrastis canadensis) as a dietary supplement in
anticholinesterolase activity (Sabir & Bhide, pharmaceutical outlets totalled US$ 25 million.
1971). It is also suggested that berberine exerts Appreciable sales occurred in Germany (US$ 8
anticancer activities both in vitro and in vivo million), France (US$ 5 million), and the United
through different mechanisms (Sun et al., 2009). States (US$ 5 million) (IMS Health, 2012). Other
Hydrastine, another major goldenseal alka- countries known to sell products containing
loid, is claimed to be an abortifacient, antibiotic, goldenseal include Canada.
antiuterotic, antivaginitic, bactericide, central According to the 2012 Nutrition Business
nervous system depressant, choleretic, convul- Journal Report, goldenseal was the 37th best-
sant, haemostat, hypertensive, hypotensive, selling herb in the USA in 2011. Following a
pesticide, sedative, uterotonic, and vasocon- decline from US$ 40 million in 2003, sales have
strictor (NTP, 2010). remained constant (NBJ, 2012a, b). According
to data from the United States National Health
(b) Medical research
and Nutrition Examination Survey, there has
In medical research, berberine is used as a been a decline in the prevalence of goldenseal
fluorescent stain for cells, chromosomes, and use as follows: 19992002 (0.6%), 20032006
energized mitochondria (Borodina et al., 1979; (0.3%), and 20072010 (0.2%), with similar data
Ridler & Jennings, 1983; Mike & Dadk, 1983; for males and females. These data showed a large
Mike & Yaguzhinskij, 1985; Kim et al., 1990). coefficient of variation and caution should be
used in interpretation (CDC, 2013).
1.3.2 Dosage
As a dietary supplement, goldenseal can be 1.4.3 Consumption
given at a wide range of doses: decoction of dried Goldenseal is consumed orally as a tea and in
roots, 0.510 g three times per day; alcoholic capsules, applied dermally as a skin lotion, applied
tincture, 24 mL three times per day; and fluid to the eye as eyewash, to the ear as eardrops, and
extract, 0.310 mL three times per day (Newall applied as a vaginal douche (Zieger & Tice, 1997;
et al., 1996; Zieger & Tice, 1997; AHP, 2001; Hamon, 1990; AHP, 2001; NTP, 2010). Exposure
McKenna & Plotnikoff, 2010). OTC preparations to hydrastine also occurs when used in decon-
of goldenseal are available in doses of 100 mg up gestant nose sprays and feminine hygiene prod-
to 470 mg (Zieger & Tice, 1997; AHP, 2001). ucts (Zeiger & Tice, 1997).
77
IARC MONOGRAPHS 108
1.5 Occupational exposure 3. Cancer in Experimental Animals

See Table3.1
Workers on goldenseal plantations and in gold-
Goldenseal (H. canadensis) root powder was
enseal processing plants are probably exposed.
tested for carcinogenicity by oral administration
in one study in mice and one study in rats.
1.6 Regulations and guidelines
According to the 1994 Dietary Supplement 3.1 Mouse
Health and Education Act (DSHEA) in the
In a 2-year study of carcinogenicity, groups
USA, goldenseal is considered a dietary supple-
of 50 male and 50 female B6C3F1 mice (age,
ment under the general umbrella of foods
56 weeks) were fed ad libitum with diet
(FDA, 1994). In the USA, dietary supplements
containing well-characterized goldenseal root
put on the market before 15 October 1994 do
powder (H. canadensis) shown to contain all
not require proof of safety; however, the label-
the major alkaloids characteristic of goldenseal
ling recommendations for dietary supplements
(berberine, hydrastine and canadine) at a concen-
include warnings, dosage recommendations, and
tration of 0 (control), 3000, 9000, or 25000ppm.
substantiated structure or function claims. The
Concentrations were equivalent to average daily
product label must declare prominently that the
doses of goldenseal root powder of approximately
claims have not been evaluated by the Food and
0, 375, 1120, or 3275mg/kg body weight (bw) for
Drug Administration, and bear the statement
males and 0, 330, 1000, or 2875 mg/kg bw for
This product is not intended to diagnose, treat,
females (NTP, 2010). Survival of the females at
cure, or prevent any disease (Croom & Walker,
9000 ppm was lower than that of controls. At
1995).
105106 weeks, goldenseal root powder caused
The Natural Medicines Comprehensive
a significant positive trend in the incidence of
Database ranks goldenseal as a possibly safe
hepatoblastoma, and of hepatocellular adenoma,
dietary supplement when used at a single dose,
in males. There were no significant increases in
and at recommended oral dosages for short-term
the incidence of tumours in female mice. The
administration (NCCAM, 2012; NMCD, 2013).
incidence of liver eosinophilic foci was signifi-
In Canada, goldenseal is a Natural Health
cantly increased in males at the intermediate and
Product (a form of OTC drug) and requires a
highest doses, and in all exposed females.
product license (pre-market authorization) to be
sold (Goldenseal Buccal, 2010). Goldenseal (H.
canadensis) is also regulated as an endangered 3.2 Rat
species under Appendix II of the Convention on
International Trade in Endangered Species of In a 2-year study of carcinogenicity, groups
Wild Fauna and Flora (CITES, 2006). of 50 male and 50 female F344/N rats (age,
56 weeks) were fed ad libitum with diet
containing a well-characterized goldenseal root
powder (H. canadensis) shown to contain all the
2. Cancer in Humans
major alkaloids (berberine, hydrastine and cana-
dine) characteristic of goldenseal at a concentra-
No data were available to the Working Group. tion of 0 (control), 3000, 9000, or 25 000 ppm.
Concentrations were equivalent to average daily
78
Table 3.1 Studies of carcinogenicity with goldenseal root powder in mice and rats
Species, Dosing regimen Incidence of tumours Significance Comments

strain (sex) Animals/group at start
Duration
Reference
Mouse, Goldenseal root powder (H. Hepatoblastoma: *P=0.030 Major alkaloids content:
B6C3F1 canadensis): 0 (control), 3000, 9000, or 1/50 (2%)*, 2/50 (4%), 1/50 (25), 6/50 (12%) (M)a (trend test) berberine, 3.89%; hydrastine,
(M,F) 25000 ppm in feed, corresponding to Hepatocellular adenoma: **P=0.016 2.8%; and canadine, 0.17%
105106 wk 0, 375, 1120, or 3275 mg/kg bw (M); 0, 22/50 (44%)**, 16/50 (32%), 23/50 (46%), 29/50 (58%)
NTP (2010) 330, 1000, or 2875 mg/kg bw (F) (M)
50 M and 50 F/group (age, 56-wk) 3/50 (6%), 6/50 (12%), 7/50 (14%), 7/50 (14%) (F)
Hepatocellular carcinoma:
8/50 (16%), 14/50 (28%), 15/50 (30%), 12/50 (24%) (M)
Rat, F344/N Goldenseal root powder (H. Hepatocellular adenoma: *P<0.001 Major alkaloids content:
(M,F) canadensis): 0 (control), 3000, 9000, 1/50 (2%)*, 1/50 (2%), 2/50 (4%), 10/50 (20%)** (M) (trend test) berberine, 3.89%; hydrastine,
105106 wk or 25000 ppm in feed: 0, 135, 400, 0/50 (0%)*, 0/50 (05), 1/50 (2%), 8/50 (16%)** (F)b **P0.004 2.8%; and canadine, 0.17%
NTP (2010) or 1175 mg/kg bw (M); 0, 150, 470, Hepatocellular carcinoma: ***P<0.001
1340 mg/kg bw (F) 0/50 (0%), 0/50 (0%), 0/50 (0%), 1/50 (2%) (M)c
50 M and 50 F/group (age, 56 wk) Hepatocellular adenoma or carcinoma (combined):
1/50 (2%)*, 1/50 (2%), 2/50 (4%), 11/50 (22%)*** (M)d
a Historical incidence of hepatoblastoma (includes multiple) (meanstandard deviation) for feed studies in untreated control male mice: 1/250 (0.4%0.9%); range, 02%; all routes:
48/1447 (3.3%6.4%); range, 034%.
b Historical incidence of hepatocellular adenoma (meanstandard deviation) for feed studies in untreated control female rats: 4/250 (1.6%2.2%); range, 04%; all routes: 16/1350
(1.2%2.6%); range, 012%

c Historical incidence of hepatocellular carcinoma in untreated control male rats: 1/1250 (all routes)
d Historical incidence of hepatocellular adenoma or carcinoma (meanstandard deviation) for feed studies in untreated control male rats: 7/300 (2.3%2.3%); range, 06%; all routes:
22/1399 (1.6%1.7%); range, 06%

bw, body weight; F, female; M, male; wk, week
79
Goldenseal
IARC MONOGRAPHS 108
doses of goldenseal root powder of approximately berberine were 1.11.2 ng/mL, 3.03.3 hours
0, 135, 400, or 1175 mg/kg bw to males and 0, and 0.150.09 ngh/mLkg, respectively. Phase
150, 470, or 1340 mg/kg bw for females (NTP, I metabolites of hydrastine underwent extensive
2010). Survival of the females at 9000ppm was glucuronidation, but not sulfation, while the
greater than that of controls. At 105106weeks, phase I metabolites of berberine were primarily
goldenseal root powder caused increased inci- sulfated (Gupta et al., 2010). The area under the
dences of hepatocellular adenoma in males and curve of hydrastine in plasma is significantly
females at the highest dose. One male rat at the higher than that of berberine, suggesting that the
highest dose also developed a rare hepatocel- oral bioavailability of hydrastine is also higher.
lular carcinoma. There was a treatment-related There was enterohepatic recycling of berberine,
statistically significant increase in the incidence which also had a high volume of distribution
of liver eosinophilic foci in male and female rats. (Gupta et al., 2010).
[The Working Group noted that hepatocellular In humans given berberine orally, plasma
tumours are uncommon tumours in F344/N rats concentrations of berberine were very low and
and that evidence from studies reported in the variable (Li et al., 2000a; Hua et al., 2007). For
literature showed that hepatocellular adenomas example, the plasma Cmax of berberine was only
may progress to malignant tumours in F344/N 0.4 ng/mL after a single oral dose of 400 mg of
rats.] berberine (Hua et al., 2007). This suggested that
berberine is poorly absorbed by the gastroin-
testinal wall. [The Working Group noted the
4. Mechanistic and Other discrepancy between pharmacokinetic parame-
Relevant Data ters for purified berberine compared with those
for goldenseal.]
The metabolites of berberine are mainly
4.1 Absorption, distribution, 1,3-dioxolane ring-opened, demethylated,
metabolism, and excretion demethylenated, glycosylated, and sulfonated
products. Several clinical studies have been
4.1.1 Humans performed to identify these metabolites in the
Goldenseal contains berberine and hydras- plasma and urine (Pan et al., 2002; Qiu et al.,
tine (see Section 1). Most data on the carcino- 2008).
genicity of goldenseal came from studies with In one study, five healthy male volunteers
berberine. In 11 healthy subjects treated orally (aged 2128 years) were given berberine chloride
with a single dose of goldenseal (2.7 g) containing orally (900 mg/day; 100 mg tablets) for 3 days
78 mg of hydrastine and 132 mg of berberine, (Pan et al., 2002). Urine samples were collected
both berberine and hydrastine were absorbed and the metabolites isolated and purified by
from the gastrointestinal tract, and their phase polyporous resin column chromatography. Three
I and II metabolites were rapidly detected in metabolites were recovered and identified via
the plasma and urine (Gupta et al., 2010). The electrospray ionization mass spectroscopy (ESI-
maximal plasma concentration (Cmax) of hydras- MS) and proton nuclear magnetic resonance
tine was 225 100 ng/mL, the time to Cmax (1H-NMR) spectroscopy: jatrorrhizine-3-sul-
(Tmax) was 1.5 0.3 hours, and the area under fate, demethylene berberine-2-sulfate, and
the curve (AUC) was 6.4 4.1 ngh/mLkg. thalifendine-10-sulfate. The amounts of these
The elimination half-life (t1/2) of hydrastine purified metabolites in the urine were 250, 17,
was 4.8 1.4 hours. Corresponding values for and 2 mg, respectively (Pan et al., 2002).
80
Goldenseal
In a later study with 12 healthy male volun- The metabolism of berberine in vitro and in
teers (aged 2226 years; body weight, 6080 kg), vivo has been well studied in rats (Qiu et al., 2008;
additional metabolites were discovered (Qiu Liu et al., 2009, 2010). In rat liver microsomes, or
et al., 2008). In this study, subjects were given an following intravenous administration, berberine
oral dose of berberine chloride of 300 mg, three was metabolized by several cytochromes P450
times per day for 2 days, and urine samples were (CYPs) and UDP-glucuronosyltransferases.
collected in the 72 hours after dosing. Using Oxidative demethylenation was the major meta-
nuclear magnetic resonance spectroscopy in bolic pathway and the metabolite obtained can
addition to liquid chromatographymass spec- subsequently undergo glucuronidation (Liu
trometry, the study identified the above three et al., 2009). In one study in vivo, male Wistar
conjugated metabolites, plus previously unseen rats (age, 810 weeks) were given berberine at a
conjugates: demethyleneberberine-2-O-sul- dose of 100 mg/kg bw, and urine samples were
fate, jatrorrhizine-3-O--D-glucuronide, collected for 48 hours; the five urinary metabo-
thalifendine-10-O--D-glucuronide, berberru- lites of berberine isolated and identified included
bine-9-O--D-glucuronide,jatrorrhizine-3-O-sul- berberrubine-9-O--D-glucuronide, demeth-
fate, 3-10-demethylpalmatine-10-O-sulfate, and yleneberberine-2,3-di-O--D-glucuronide,
columbamin-2-O--D-glucuronide (40 mg, demethyleneberberine-2-O-sulfate, 3,10-demeth-
6 mg, 4 mg, 6.1 mg, 2.2 mg, 1.2 mg, and 1 mg, ylpalmatine-10-O-sulfate, and thalifendine (Qiu
per 16 L of urine, respectively) (Qiu et al., 2008). et al., 2008).
In this study, both sulfates and glucuronides of Berberine was found to undergo extensive
berberine were observed. first-pass metabolism in the rat intestine. After
Using the human intestinal Caco-2 model, intragastric dosing, approximately half of the
Zhang et al. (2011) examined the intestinal absorp- intact berberine passed through the gastrointes-
tion mechanisms of berberine. They found that tinal tract and another half was biotransformed
the cellular uptake of berberine was 30.130.57 by the small intestine, resulting in an extremely
mol/g protein, at a substrate concentration of 10 low absolute oral bioavailability in rats (0.36%)
M, and that the apparent permeability (Papp) of (Liu et al., 2010).
berberine was 0.660.04 (absorptive direction) [The above findings indicate that berberine
and 15.910.61 (secretory direction) 10-6 cm/s, is poorly absorbed in both rats and humans due
with an efflux ratio of 24.28 (Zhang et al., 2011). to extensive first-pass disposition. Berberine is
extensively metabolized by intestinal and hepatic
Transport inhibition studies using cyclosporin A
phase I and II enzymes in rats and humans.]
[ciclosporin] and verapamil [both potent inhib-
itors of P-glycoprotein 1 (P-gp/multidrug resist-
ance protein 1)] indicated that berberine was a 4.1.3 Alterations in enzymes and metabolic
substrate of this transporter. capacity
Goldenseal strongly inhibits CYP2C9,
4.1.2 Experimental systems CYP3A4, CYP2D6, and CYP2C19 in vitro
The absorption of berberine in rats is very (Budzinski et al., 2000; Chatterjee & Franklin,
poor; in rats given berberine at an oral dose of 2003; Foster et al., 2003). In one experiment,
100 mg/kg bw, the plasma Cmax of berberine was Etheridge et al. (2007) examined the effects of
estimated to be only 4.0 ng/mL (Liu et al., 2009, goldenseal on the activities of human CYPs in
2010). human liver microsomes, finding that goldenseal
inhibited CYP2C8, CYP2D6, and CYP3A4 by
50% (Etheridge et al., 2007).
81
IARC MONOGRAPHS 108
In humans, goldenseal did not significantly 4.2 Genetic and related effects
affect the metabolism and pharmacokinetic of
indinavir, a protease inhibitor, which is a substrate 4.2.1 Humans
of CYP3A4, (Sandhu et al., 2003). However, gold- No data were available to the Working Group.
enseal has been shown to increase the plasma
concentration of cyclosporine [ciclosporin] in
healthy volunteers (Xin et al., 2006) and in renal
4.2.2 Experimental systems
transplant patients (Wu et al., 2005). In another See also Table4.1.
study, single time-point phenotypic metabolic
ratios of CYP probe drugs were administered to (a) Mutagenicity
12 healthy subjects (6 males and 6 females, all Goldenseal root powder (100010000g/plate)
extensive metabolizers of CYP2D6) to determine was not mutagenic in Salmonella typhimurium
whether a 28-day supplementation of golden- strains TA98 or TA100, or Escherichia coli strain
seal root extract (900 mg, three times per day, WP2 uvrA/pKM101, with or without metabolic
2.70 g/day; no standardization claim) affected activation from rat liver S9 (NTP, 2010). Berberine
the activity of CYP1A2, CYP2D6, CYP2E1, or was also not mutagenic in S. typhimurium strains
CYP3A4/5. The following probe cocktails were TA97, TA98, TA100 and TA1535, with or without
administered before and after supplementation: metabolic activation from rat or hamster liver S9.
midazolam (CYP3A4) and caffeine (CYP1A2), In three tests in TA98, berberine gave negative
followed 24 hours later by chlorzoxazone results in two tests and equivocal results in one.
(CYP2E1) and debrisoquine (CYP2D6). The
phenotypic traits of CYP3A4, CYP1A2, CYP2E1, (b) Chromosomal damage
and CYP2D6 activities were assessed using the No increase in the frequency of micronucle-
1-hydroxymidazolam/midazolam serum ratio at ated normochromatic erythrocytes or polychro-
1 hour, the paraxanthine/caffeine serum ratio at matic erythrocytes was observed in blood from
6 hours, the 6-hydroxychlorzoxazone/chlorzox- male or female mice exposed to diets containing
azone serum ratio at 2 hours, and debrisoquine goldenseal root powder at up to 50000 ppm for 3
urinary recovery ratio after an 8-hour collec- months. No increases in the frequency of micro-
tion. Phenotypic ratios, taken before and after nucleated polychromatic erythrocytes were
supplementation, show remarkable inhibition of observed in bone marrow from male B6C3F1
CYP2D6 (~40%) and CYP3A4/5 (~40%) activi- mice treated with berberine chloride at a dose
ties due to goldenseal supplementation. CYP1A2 of up to 658 mg/kg bw via three intraperitoneal
and CYP2E1 activities were virtually unaffected injections at intervals of 24 hours (NTP, 2010).
(Gurley et al., 2005). For CYP3A4/5, this clinical
observation was confirmed by a more compre- (c) Interaction with DNA
hensive study in which goldenseal caused a mean Data on the interaction of some goldenseal
increase in midazolam AUC0- and Cmax by about constituents with DNA are presented below (see
63% and 40%, respectively, while the AUC0- and Section 4.3.2).
Cmax of clarithromycin (an inhibitor of CYP3A4)
increased about 5.5-fold and 2-fold, respectively
(Gurley et al., 2008a). However, goldenseal root
extract is only a moderate inhibitor of CYP2D6
in humans (Gurley et al., 2008b).
82
Table 4.1 Genetic and related effects of goldenseal root and berberine
Test system Results Concentration or dose Reference

(LED or HID)
Without exogenous With exogenous
metabolic system metabolic
system
Goldenseal root
Salmonella typhimurium TA100, TA98 reverse 10000g/plate NTP (2010)
mutation
Escherichia coli WP2 uvrA/pKM101 reverse 10000g/plate NTP (2010)
mutation
Micronucleus formation in peripheral blood NT 50000ppm [50g/kg] in diet for NTP (2010)
erythrocytes of male and female B6C3F1 mice in 3months
vivo
Berberine
Salmonella typhimurium TA100, TA98, TA97, ()a 1000g/plate NTP (2010)
TA1535 reverse mutation
Inhibition of DNA repair activity, topoisomerase II (+)b NT 2.5M [0.9g/mL] Kobayashi et al. (1995)
inhibition, plasmid pUL402DNA
Inhibition of DNA repair activity, topoisomerase II (+)b NT 50M [18g/mL] Kim et al. (1998)
inhibition,plasmid pBS DNA
Micronucleus formation in bone-marrow NT 658mg/kg bw, i.p.3 NTP (2010)
polychromatic erythrocytes of male B6C3F1 mice
in vivo
a
One out of three tests in TA98 gave equivocal results.
b The test gave positive results with berberrubine, but weakly positive results with berberine.
+, positive; (+), weakly positive; , negative; ?, inconclusive; HID, highest ineffective dose; i.p., intraperitoneal; LED, lowest effective dose; NR, not reported; NT, not tested
83
Goldenseal
IARC MONOGRAPHS 108
4.3 Other mechanistic data relevant exhibited some sequence selectivities. Additional
to carcinogenicity studies using ESI-MS to examine noncovalent
complexes have produced new information
4.3.1 Humans (Chen et al., 2005a, b). In one study, berberine
exhibited different binding affinities to different
No data were available to the Working Group. nucleotides (Chen et al., 2005b); however,
ESI-MS and fluorescence titration experiments
4.3.2 Experimental systems with these four alkaloids indicated that sequence
Berberine alkaloids found in goldenseal root selectivity was not significant and no specific
powder inhibited topoisomerases I (Topo I) and AT- or GC-rich DNA binding preferences were
II (Topo II) (Krishnan & Bastow, 2000; Li et al., observed, in contrast to the aforementioned
2000b; Kettmann et al., 2004). The ability of reports (Mazzini et al., 2003; Chen et al., 2004).
berberine to disrupt Topo I and Topo II is related Pang et al. (2005) suggested that berberine and
to its antitumour activity (Mazzini et al., 2003). its derivatives, especially those with a primary
While the role of drugDNA interactions in the amino group, are able to bind strongly with calf-
inhibition of Topo I is unclear, the binding of thymus DNA via an intercalation mechanism.
berberine to DNA has been considered as the Substitution at the C-9 position is an important
cause for inhibition. determinant of the biological activity, as Park
It has been shown by nuclear magnetic reso- et al. (2004) demonstrated in these studies on the
nance spectroscopy that berberine binds selec- structureactivity relationships of the berberine
tively to AT-rich sequences, interacting with analogues (Park et al., 2004).
several oligonucleotides (Mazzini et al., 2003). When berberine units are bridged at the C-9
Using a two-dimensional nuclear Overhauser position with different linker lengths (termed
effect assay, several contacts were detected bridged berberine derivatives), they exhibit the
between protons of the self-complementary highest binding affinity to DNA when they form
oligomer d(AAGAATTCTT)2 and the protons a compound berberine dimer with a propyl chain
of berberine. Berberine was found on the convex (Chen et al., 2005a; Qin et al., 2006). DNA and
side of the helix groove, the minor groove of the RNA triplexes can be bound and stabilized better
nucleotide (at the level of the A4T7 and A5T6 than their respective parent duplexes when they
base pairs), presenting its positive nitrogen atom bind with the sanguinarine and berberine alka-
to the negative ionic surface of the oligonucle- loids (Das et al., 2003).
otide. The aromatic protons H-11 and H-12 are
close to the ribose of cytidine C8, while the meth- 4.4 Susceptibility
ylenedioxy and ring A group are more external
to the helix (Mazzini et al., 2003). No data were available to the Working Group.
Electrospray ionization-mass spectrometry
(ESI-MS) has been used to study noncovalent 4.5 Mechanistic considerations
complexes of protoberberine alkaloids, berberine,
palmatine, jatrorrhizine, and coptisine with Goldenseal root powder gave negative results
double-strand DNA oligonucleotides, showing in several standard bacterial assays for mutation
that berberine has the lowest binding affinity in the absence or presence of metabolic activa-
and palmatine the highest (Chen et al., 2004). tion systems. The principal alkaloid in goldenseal
These preliminary data suggested that berberine root powder, berberine, also gave negative results
84
Goldenseal
in many of these assays. Goldenseal root powder 5.2 Human carcinogenicity data
gave negative results in the test for mouse periph-
eral blood micronucleated erythrocytes (normo- No data were available to the Working Group.
chromatic and polychromatic) after 3months of
dietary exposure. Berberine also gave negative 5.3 Animal carcinogenicity data
results for induction of micronuclei in mouse
bone-marrow polychromatic erythrocytes after A well-characterized goldenseal root powder
three doses. shown to contain all the major alkaloids char-
Goldenseal produced mostly negative results acteristic of goldenseal was tested for carcino-
in assays for bacterial mutation (NTP, 2010). genicity by oral administration in one study in
However, berberine, its metabolite, berberru- mice and one study in rats.
bine, and several protoberberines have been In male mice fed a diet containing golden-
shown to inhibit the activity of topoisomerases seal root powder, there was a significant positive
(Kobayashi et al., 1995; Makhey et al., 1995; Kim trend in the incidence of hepatoblastoma and of
et al., 1998; Li et al., 2000b; Krishnan & Bastow, hepatocellular adenoma. There were no signif-
2000). Etoposide, an inhibitor of topoisomerase icant increases in the incidence of tumours in
II, was classified by the IARC Monographs as female mice.
carcinogenic to humans (Group 1) (IARC, 2012). In male and female rats fed a diet containing
goldenseal root powder, there was an increased
incidence of hepatocellular adenoma, which in
5. Summary of Data Reported F344/N rats is an uncommon tumour that is
known to progress to malignancy. In addition,
one rare hepatocellular carcinoma was observed
5.1 Exposure data
in the group of males given the highest dose.
Goldenseal (Hydrastis canadensis L.) is
an endangered plant that is widely consumed 5.4 Mechanistic and other relevant
in several countries. The root has been used
traditionally in herbal medicine. Reported uses data
include the treatment of skin disorders, digestive The major alkaloid components of golden-
disorders, anorexia, menstrual disorders, and seal, berberine and hydrastine, are absorbed
mucosal inflammations. Currently, the main from the gastrointestinal tract into the circula-
applications for this plant include the prevention and extensively metabolized in the liver after
tion and reduction of inflammation and related oral administration of goldenseal.
diseases. Goldenseal is available in the form of Goldenseal root powder gave negative results
tea, capsules containing the crude drug or the in several standard bacterial assays for mutation
extract, skin lotion, eyewash, and eardrops. In in the absence or presence of exogenous metabolic
2011, goldenseal ranked 37th among top-selling activation systems. Berberine also gave negative
dietary supplements in the USA. Other coun- results in many of these assays. Likewise, gold-
tries that reported significant sales of goldenseal enseal root powder gave negative results in the
included Canada, France, and Germany. mouse peripheral blood micronucleated eryth-
rocyte test (normochromatic and polychromatic
erythrocytes) after dietary exposure. Berberine
also gave negative results for the induction of
85
IARC MONOGRAPHS 108
micronuclei in mouse bone-marrow polychro- Borodina VM, Kirianova EE, Fedorova OV, Zelenin
matic erythrocytes. AV (1979). Cytochemical properties of interphase
chromatin condensed as a result of treatment with
Berberine and its metabolite, berberrubine, caffeine. Exp Cell Res, 122(2):3914. doi:10.1016/0014-
have been shown to inhibit DNA topoisomerases. 4827(79)90316-1 PMID:92416
Such inhibition may account for the carcino- Brown PN & Roman MC (2008). Determination of
hydrastine and berberine in goldenseal raw materials,
genicity of goldenseal in experimental animals. extracts, and dietary supplements by high-perfor-
mance liquid chromatography with UV: collaborative
study. J AOAC Int, 91(4):694701. PMID:18727526
6. Evaluation Budzinski JW, Foster BC, Vandenhoek S, Arnason JT
(2000). An in vitro evaluation of human cytochrome
P450 3A4 inhibition by selected commercial herbal
6.1 Cancer in humans extracts and tinctures. Phytomedicine, 7(4):27382.
doi:10.1016/S0944-7113(00)80044-6 PMID:10969720
CDC (2013). National Health and Nutrition Examination
There is inadequate evidence in humans for Survey. Available from: http://www.cdc.gov/nchs/
the carcinogenicity of goldenseal. nhanes.htm, accessed 10 September 2014.
Chatterjee P & Franklin MR (2003). Human cytochrome
p450 inhibition and metabolic-intermediate complex
6.2 Cancer in experimental animals formation by goldenseal extract and its methylenedioxy-
phenyl components. Drug Metab Dispos, 31(11):13917.
There is sufficient evidence in experimental doi:10.1124/dmd.31.11.1391 PMID:14570772
animals for the carcinogenicity of goldenseal Chen W-H, Chan C-L, Cai Z, Luo GA, Jiang ZH (2004).
Study on noncovalent complexes of cytotoxic proto-
root powder. berberine alkaloids with double-stranded DNA by
using electrospray ionization mass spectrometry.
Bioorg Med Chem Lett, 14(19):49559. doi:10.1016/j.
6.3 Overall evaluation bmcl.2004.07.037 PMID:15341959
Chen W-H, Pang J-Y, Qin Y, Peng Q, Cai Z, Jiang ZH
Goldenseal root powder is possibly carcino- (2005a). Synthesis of linked berberine dimers and
genic to humans (Group 2B). their remarkably enhanced DNA-binding affinities.
Bioorg Med Chem Lett, 15(10):268992. doi:10.1016/j.
bmcl.2004.10.098 PMID:15863343
Chen W-H, Qin Y, Cai Z, Chan CL, Luo GA, Jiang ZH
References (2005b). Spectrometric studies of cytotoxic protober-
berine alkaloids binding to double-stranded DNA.
Bioorg Med Chem, 13(5):185966. doi:10.1016/j.
Abourashed EA & Khan IA (2001). High-performance
bmc.2004.10.049 PMID:15698803
liquid chromatography determination of hydrastine
Chi CW, Chang YF, Chao TW, Chiang SH, Peng FK,
and berberine in dietary supplements containing gold-
Lui WY et al. (1994). Flowcytometric analysis of the
enseal. J Pharm Sci, 90(7):81722. doi:10.1002/jps.1035
effect of berberine on the expression of glucocorticoid
PMID:11458331
receptors in human hepatoma HepG2 cells. Life Sci,
AHP (2001). Golden root. In: Roy Upton, editor. American
54(26):2099107. doi:10.1016/0024-3205(94)00719-5
Herbal Pharmacopoeia and Therapeutic Compendium.
PMID:7516035
Scotts Valley (CA), USA: AHP; pp. 136.
Choudhry VP, Sabir M, Bhide VN (1972). Berberine in
American Chemical Society (2014). SciFinder Database.
giardiasis. Indian Pediatr, 9(3):1436. PMID:4555485
Available from: https://scifinder.cas.org, accessed 10
CITES (2006). Convention on International Trade in
September 2014.
Endangered Species of Wild Fauna and Flora. Fifty-
BHMA; British Herbal Medicine Association (1983)
fourth meeting of the Standing Committee, Geneva
British herbal pharmacopoeia. Bournemouth, UK:
(Switzerland). SC54 Doc. 43.5. Available from: http://
BHMA; pp 113115.
www.cites.org/eng/com/sc/54/E54-43-5.pdf, accessed
BHMA; British Herbal Medicine Association (1992).
08 December 2014.
British herbal compendium, volume I, Bournemouth,
Creasey WA (1977). Plant Alkaloids. In: Becker FF, editor.
UK: BHMA; pp 119120.
Cancer, a Comprehensive Treatise: Chemotherapy. Vol.
5. New York and London: Plenum Press; pp. 379425.
86
Goldenseal
Creasey WA (1979). Biochemical effects of berberine. in vivo. Clin Pharmacol Ther, 83(1):619. doi:10.1038/
Biochem Pharmacol, 28(7):10814. doi:10.1016/0006- sj.clpt.6100222 PMID:17495878
2952(79)90308-3 PMID:444265 Gurley BJ, Swain A, Hubbard MA, Williams DK, Barone
Croom EM Jr & Walker L (1995). Botanicals in the phar- G, Hartsfield F et al. (2008b). Clinical assessment of
macy: new life for old remedies. Drug Topics, 139:8493. CYP2D6-mediated herb-drug interactions in humans:
Das S, Kumar GS, Ray A, Maiti M (2003). Spectroscopic effects of milk thistle, black cohosh, goldenseal, kava
and thermodynamic studies on the binding of sangui- kava, St. Johns wort, and Echinacea. Mol Nutr Food
narine and berberine to triple and double helical DNA Res, 52(7):75563. doi:10.1002/mnfr.200600300
and RNA structures. J Biomol Struct Dyn, 20(5):70314. PMID:18214849
doi:10.1080/07391102.2003.10506887 PMID:12643773 Hamon NW (1990). Herbal medicine: Goldenseal. Can
Duke JA, editor (2002). Handbook of medicinal herbs. Pharm J, 123:508510.
2nd ed. Boca Raton (FL): CRC Press; pp. 340342. Hobbs C (1990). Golden seal in early American medical
Etheridge AS, Black SR, Patel PR, So J, Mathews JM (2007). botany. Pharm Hist, 32(2):7982. PMID:11622733
An in vitro evaluation of cytochrome P450 inhibi- HSDB (1997). The Hazardous Substance Data Bank.
tion and P-glycoprotein interaction with goldenseal, Berberine sulfate. Online database produced by the
Ginkgo biloba, grape seed, milk thistle, and ginseng National Library of Medicine. Last Database Update:
extracts and their constituents. Planta Med, 73(8):731 January 14, 2002.
41. doi:10.1055/s-2007-981550 PMID:17611934 Hua W, Ding L, Chen Y, Gong B, He J, Xu G (2007).
FDA (1994). Dietary Supplement Health and Education Determination of berberine in human plasma by
Act of 1994. Public Law 103-417, 103rd Congress., US liquid chromatography-electrospray ionization-mass
Food and Drug Administration. Available from http:// spectrometry. J Pharm Biomed Anal, 44(4):9317.
www.fda.gov/RegulatoryInformation/Legislation/ doi:10.1016/j.jpba.2007.03.022 PMID:17531424
FederalFoodDrugandCosmeticActFDCAct/Signif Huang JH & Johnston GA (1990). (+)-Hydrastine,
icantAmendmentstotheFDCAct/ucm148003.htm, a potent competitive antagonist at mammalian
accessed 09 December 2014 GABAA receptors. Br J Pharmacol, 99(4):72730.
Foster BC, Vandenhoek S, Hana J, Krantis A, Akhtar doi:10.1111/j.1476-5381.1990.tb12997.x PMID:2163278
MH, Bryan M et al. (2003). In vitro inhibition of Hwang BY, Roberts SK, Chadwick LR, Wu CD, Kinghorn
human cytochrome P450-mediated metabolism of AD (2003). Antimicrobial constituents from golden-
marker substrates by natural products. Phytomedicine, seal (the Rhizomes of Hydrastis canadensis) against
10(4):33442. doi:10.1078/094471103322004839 selected oral pathogens. Planta Med, 69(7):6237.
PMID:12809364 doi:10.1055/s-2003-41115 PMID:12898417
Goldenseal Buccal (2010). Natural Health Products IARC (2012). Pharmaceuticals. Volume 100 A. A review of
Ingredients Database. Health Canada. Available from: human carcinogens. IARC Monogr Eval Carcinog Risks
http://webprod.hc-sc.gc.ca/nhpid-bdipsn/monoReq. Hum, 100:Pt A: 1401. PMID:23189749
do?id=107&lang=eng, accessed 10 September 2014. IMS Health (2012) Multinational Integrated Data
Govindan M & Govindan G (2000).A convenient method Analysis (MIDAS). Plymouth Meeting, Pennsylvania:
for the determination of the quality of goldensealA IMS Health, 2012.
convenient method for the determination of the quality Junio HA, Sy-Cordero AA, Ettefagh KA, Burns JT, Micko
of goldenseal. Fitoterapia, 71(3):2325. doi:10.1016/ KT, Graf TN etal. (2011). Synergy-directed fractiona-
S0367-326X(99)00162-8 PMID:10844160 tion of botanical medicines: a case study with golden-
Gupta PK, Gurley BJ, Barone G, Hendrickson HP seal (Hydrastis canadensis). J Nat Prod, 74(7):16219.
(2010). Clinical Pharmacokinetics and Metabolism doi:10.1021/np200336g PMID:21661731
of Berberine and Hydrastine Following an Oral Dose Kettmann V, Kosflov D, Jantov S, Cernkov M,
of Goldenseal Supplement. Planta Med, 76(05):110 Drmal J (2004). In vitro cytotoxicity of berberine
doi:10.1055/s-0030-1251872 against HeLa and L1210 cancer cell lines. Pharmazie,
Gurley BJ, Gardner SF, Hubbard MA, Williams DK, 59(7):54851. PMID:15296093
Gentry WB, Khan IA et al. (2005). In vivo effects of Khin-Maung-U , Myo-Khin , Nyunt-Nyunt-Wai ,
goldenseal, kava kava, black cohosh, and valerian on Aye-Kyaw , Tin-U (1985). Clinical trial of berberine
human cytochrome P450 1A2, 2D6, 2E1, and 3A4/5 in acute watery diarrhoea. Br Med J (Clin Res Ed),
phenotypes. Clin Pharmacol Ther, 77(5):41526. 291(6509):16015. doi:10.1136/bmj.291.6509.1601
doi:10.1016/j.clpt.2005.01.009 PMID:15900287 PMID:3935203
Gurley BJ, Swain A, Hubbard MA, Hartsfield F, Thaden Kim DI, Pedersen H, Chin CK (1990). Two stage cultures for
J, Williams DK et al. (2008a). Supplementation with the production of berberine in cell suspension cultures
goldenseal (Hydrastis canadensis), but not kava kava of Thalictrum rugosum. J Biotechnol, 16(3-4):297303.
(Piper methysticum), inhibits human CYP3A activity doi:10.1016/0168-1656(90)90043-B PMID:1366939
87
IARC MONOGRAPHS 108
Kim SA, Kwon Y, Kim JH, Muller MT, Chung IK Martin EJ, Zell HC, Poon BT, editors (1978). Castor Oil to
(1998). Induction of topoisomerase II-mediated DNA Chlorosulfuric Acid. In: Kirk-Othmer Encyclopedia of
cleavage by a protoberberine alkaloid, berberrubine. Chemical Technology. 3rd ed, Vol. 5. New York (NY):
Biochemistry, 37(46):1631624. doi:10.1021/bi9810961 John Wiley and Sons; pp. 513542.
PMID:9819224 Mazzini S, Bellucci MC, Mondelli R (2003). Mode of
Kobayashi Y, Yamashita Y, Fujii N, Takaboshi K, Kawakami binding of the cytotoxic alkaloid berberine with the
T, Kawamura M etal. (1995). Inhibitors of DNA topoi- double helix oligonucleotide d(AAGAATTCTT)(2).
somerase I and II isolated from the Coptis rhizomes. Bioorg Med Chem, 11(4):50514. doi:10.1016/S0968-
Planta Med, 61(5):4148. doi:10.1055/s-2006-958127 0896(02)00466-2 PMID:12538015
PMID:7480201 McKenna DJ, Plotnikoff GA (2010). Goldenseal. In:
Koehlers Medicinal-Plants (1887). Available from: Coates PM, Betz JM, editors. Encyclopedia of Dietary
http://pharm1.pharmazie.uni-greifswald.de/allgemei/ Supplements. 2nd ed. New York, USA: Informa
koehler/koeh-eng.htm, accessed 10 September 2014. Healthcare; pp. 379390.
Krishnan P & Bastow KF (2000). The 9-position in Mike V & Dadk V (1983). Berberine derivatives as
berberine analogs is an important determinant of cationic fluorescent probes for the investigation of
DNA topoisomerase II inhibition. Anticancer Drug the energized state of mitochondria. Biochim Biophys
Des, 15(4):25564. PMID:11200501 Acta, 723(2):2319. doi:10.1016/0005-2728(83)90122-6
Kulkarni SK, Dandiya PC, Varandani NL (1972). PMID:6849903
Pharmacological investigations of berberine sulphate. Mike V & Yaguzhinskij LS (1985). Interaction of fluo-
Jpn J Pharmacol, 22(1):116. doi:10.1254/jjp.22.11 rescent berberine alkyl derivatives with respiratory
PMID:4260852 chain of rat liver mitochondria. J Bioenerg Biomembr,
Li BX, Yang BF, Hao XM, Zhou H, Sun MZ (2000a). Study 17(1):2332. doi:10.1007/BF00744986 PMID:3988724
on the pharmacokinetics of berberine in single dosage Mikkelsen SL & Ash KO (1988). Adulterants causing
and coadministration with oryzanol in rabbits and false negatives in illicit drug testing. Clin Chem,
healthy volunteers. Chin Pharm J, 35:335. 34(11):23336. PMID:3052928
Li TK, Bathory E, LaVoie EJ, Srinivasan AR, Olson WK, Mller K, Ziereis K, Gawlik I (1995). The antipsori-
Sauers RR et al. (2000b). Human topoisomerase I atic Mahonia aquifolium and its active constituents;
poisoning by protoberberines: potential roles for II. Antiproliferative activity against cell growth
both drug-DNA and drug-enzyme interactions. of human keratinocytes. Planta Med, 61(1):745.
Biochemistry, 39(24):710716. doi:10.1021/bi000171g doi:10.1055/s-2006-958005 PMID:7700998
PMID:10852708 NBJ (2012a). Market Data and Overview. NBJs Supplement
Li Y, Ren G, Wang YX, Kong WJ, Yang P, Wang YM etal. Business Report. Nutrition Business Journal. 2009
(2011). Bioactivities of berberine metabolites after trans- Penton Media.
formation through CYP450 isoenzymes. J Transl Med, NBJ (2012b). NBJs Supplement Business Report. An anal-
9(1):62 doi:10.1186/1479-5876-9-62 PMID:21569619 ysis of markets, trends, competition and strategy in the
Liu Y, Hao H, Xie H, Lv H, Liu C, Wang G (2009). Oxidative US dietary supplement industry. Penton Media, Inc.,
demethylenation and subsequent glucuronidation are Neuw York, NY, USA.
the major metabolic pathways of berberine in rats. NCCAM; National Center for Comprehensive and
J Pharm Sci, 98(11):4391401. doi:10.1002/jps.21721 Alternative Medicine (2012). Goldenseal. In: Herbs
PMID:19283771 at a glance. Available from: http://nccam.nih.gov/
Liu YT, Hao HP, Xie HG, Lai L, Wang Q, Liu CX etal. sites/nccam.nih.gov/files/Herbs_At_A_Glance_
(2010). Extensive intestinal first-pass elimination Goldenseal_071012_0.pdf?nav=gsa, accessed 10
and predominant hepatic distribution of berberine September 2014.
explain its low plasma levels in rats. Drug Metab Newall CA, Anderson LA, Phillipson JD (1996). Herbal
Dispos, 38(10):177984. doi:10.1124/dmd.110.033936 Medicines. A guide for health-care professionals. 2nd
PMID:20634337 ed, Pharmaceutical Press.
Makhey D, Gatto B, Yu C et al. (1995). Protoberberine NMCD; Natural Medicines Comprehensive Database
alkaloids and related compounds as dual inhibitors of (2013). Goldenseal. Available from http://
mammalian topoisomerase I and II. Med Chem Res, naturaldatabase.therapeuticresearch.com/nd/Search.
5:112. aspx?cs=&s=ND&pt=9&Product=Goldenseal&btnSe
Marin-Neto JA, Maciel BC, Secches AL, Gallo Jnior arch.x=15&btnSearch.y=3, accessed 09 December 2014
L (1988). Cardiovascular effects of berberine in NTP (2010). Toxicology and carcinogenesis studies of gold-
patients with severe congestive heart failure. Clin enseal root powder (Hydrastis Canadensis) in F344/N
Cardiol, 11(4):25360. doi:10.1002/clc.4960110411 rats and B6C3F1 mice (feed studies). Natl Toxicol
PMID:3365876 Program Tech Rep Ser, 562(562):1188. PMID:21372858
88
Goldenseal
ONeilMJeditor. (2013). The Merck Index. 15th ed. abuse. J Anal Toxicol, 17(1):147. doi:10.1093/jat/17.1.14
Cambridge: The Royal Society of Chemistry; pp. 2023, PMID:8429620
3012, 8823. Sinclair A & Catling PM (2000). Status of goldenseal,
Oldfield S (2005). IUCN/SSC Medicinal Plant Specialist Hydrastis canadensis (Ranunculaceae), in Canada. Can
Group PC 15 Inf. 3. Available from: http://www.cites. Field-Nat, 114:111120.
org/common/com/pc/15/X-PC15-03-Inf.pdf, accessed Sun Y, Xun K, Wang Y, Chen X (2009). A systematic
10 September 2014. review of the anticancer properties of berberine, a
Palmer EL (1975). Goldenseal, Orangeroot. Field book of natural product from Chinese herbs. Anticancer Drugs,
Natural History. New York (NY): McGraw-Hill Book 20(9):75769. doi:10.1097/CAD.0b013e328330d95b
Co.; p. 173. PMID:19704371
Pan J-F, Yu C, Zhu D-Y, Zhang H, Zeng JF, Jiang SH USP; United States Pharmacopeia (2013). 36th rev. and
etal. (2002). Identification of three sulfate-conjugated National Formulary. 31st ed. Rockville (MD): United
metabolites of berberine chloride in healthy volunteers States Pharmacopeial Convention, Inc.
urine after oral administration. Acta Pharmacol Sin, Vennerstrom JL & Klayman DL (1988). Protoberberine
23(1):7782. PMID:11860742 alkaloids as antimalarials. J Med Chem, 31(6):10847.
Pang J-Y, Qin Y, Chen W-H, Luo GA, Jiang ZH (2005). doi:10.1021/jm00401a006 PMID:3286870
Synthesis and DNA-binding affinities of monomod- Vennerstrom JL, Lovelace JK, Waits VB, Hanson WL,
ified berberines. Bioorg Med Chem, 13(20):583540. Klayman DL (1990). Berberine derivatives as antileish-
doi:10.1016/j.bmc.2005.05.048 PMID:15993616 manial drugs. Antimicrob Agents Chemother, 34(5):918
Park H-S, Kim E-H, Kang M-R etal. (2004). Spectroscopic 21. doi:10.1128/AAC.34.5.918 PMID:2360830
Studies on Interaction of Protoberberines with the Wu X, Li Q, Xin H, Yu A, Zhong M (2005). Effects of
Deoxyoligonucleotide d(GCCGTCGTTTTACA)2. berberine on the blood concentration of cyclosporin
Bull Korean Chem Soc, 25(10):155963. doi:10.5012/ A in renal transplanted recipients: clinical and phar-
bkcs.2004.25.10.1559 macokinetic study. Eur J Clin Pharmacol, 61(8):56772.
Peng WH, Hsieh MT, Wu CR (1997). Effect of long-term doi:10.1007/s00228-005-0952-3 PMID:16133554
administration of berberine on scopolamine-in- Xin HW, Wu XC, Li Q, Yu AR, Zhong MY, Liu YY
duced amnesia in rats. Jpn J Pharmacol, 74(3):2616. (2006). The effects of berberine on the pharmacoki-
doi:10.1254/jjp.74.261 PMID:9268086 netics of cyclosporin A in healthy volunteers. Methods
Pengelly A, editor (2012) Goldenseal. Hydrastis canadensis Find Exp Clin Pharmacol, 28(1):259. doi:10.1358/
L. In: Appalachian Plant Monographs. Frostburg mf.2006.28.1.962774 PMID:16541194
(MD): Tai Sophia Institute for the Appalachian Center Zeiger E and Tice R (1997). Goldenseal (Hydrastis
for Ethnobotanical Studies; pp. 141. canadensis L.) and two of its constituent alkaloids
Qin Y, Pang J-Y, Chen W-H, Cai Z, Jiang ZH (2006). Berberine [2086-83-1] and hydrastine [118-08-1].
Synthesis, DNA-binding affinities, and binding mode Integrated Laboratory Systems. Available from: http://
of berberine dimers. Bioorg Med Chem, 14(1):2532. ntp.niehs.nih.gov/ntp/htdocs/chem_background/
doi:10.1016/j.bmc.2005.07.069 PMID:16169735 exsumpdf/goldenseal_508.pdf. Accessed 08 December
Qiu F, Zhu Z, Kang N, Piao S, Qin G, Yao X (2008). Isolation 2014.
and identification of urinary metabolites of berberine Zhang X, Qiu F, Jiang J, Gao C, Tan Y (2011). Intestinal
in rats and humans. Drug Metab Dispos, 36(11):2159 absorption mechanisms of berberine, palmatine, jate-
65. doi:10.1124/dmd.108.021659 PMID:18703644 orhizine, and coptisine: involvement of P-glycoprotein.
Ridler PJ & Jennings BR (1983). Electro-optical fluores- Xenobiotica, 41(4):2906. doi:10.3109/00498254.2010.5
cence studies on the DNA binding of medically active 29180 PMID:21319959
drugs. Phys Med Biol, 28(6):62532. doi:10.1088/0031- Zieger E, Tice R (1997). Goldenseal (Hydrastis canadensis
9155/28/6/002 PMID:6308687 L.) and Two of Its Constituent Alkaloids. Integrated
Sabir M, Akhter MH, Bhide NK (1978). Further studies on Laboratory Systems. pp. 152. Available from: http://
pharmacology of berberine. Indian J Physiol Pharmacol, ntp.niehs.nih.gov/ntp/htdocs/Chem_Background/
22(1):923. PMID:680944 ExSumPdf/GoldenSeal_508.pdf, accessed 10
Sabir M & Bhide NK (1971). Study of some pharmacolog- September 2014.
ical actions of berberine. Indian J Physiol Pharmacol,
15(3):11132. PMID:4109503
Sandhu RS, Prescilla RP, Simonelli TM, Edwards DJ
(2003). Influence of goldenseal root on the pharmacoki-
netics of indinavir. J Clin Pharmacol, 43(11):12838.
doi:10.1177/0091270003258660 PMID:14551183
Schwarzhoff R & Cody JT (1993). The effects of adulter-
ating agents on FPIA analysis of urine for drugs of
89
GINKGO BILOBA
1. Exposure Data 1.1 Identification of the agent
Ginkgo biloba is one of the worlds oldest 1.1.1 Botanical data

living tree species. It has survived for more than (a) Nomenclature
200 million years and has become popular as
an ornamental tree in parks, gardens and city Botanical name: Ginkgo biloba L.
streets. Originating from China, Ginkgo biloba Family: Ginkgoaceae
is now found all over the world (Gilman & Genus: Ginkgo
Watson, 1993; ABC, 2000). Ginkgo seeds can be
Plant part: Leaf
cooked and eaten as food. They have also been
adopted in traditional Chinese medicine for Common names: Fossil tree; Kew tree;
many years. Some of the historical ethnomed- Japanese silver apricot; Maidenhair tree
ical applications of ginkgo leaf extract include From ABC (2000)
the treatment of a variety of ailments and condi-
tions, such as asthma, bronchitis, and fatigue (b) Description
(Rai et al., 1991). Nowadays, ginkgo leaf extracts Ginkgo is a perennial plant with little inva-
are promoted for the improvement of memory, sive potential, which is resistant to insects and
to treat or help prevent Alzheimer disease and disease. Gingko grows slowly up to a height of
other types of dementia, and to decrease inter- about 40 m. The ginkgo plant is deciduous, with
mittent claudication (Kleijnen & Knipschild, green leaves that turn golden in autumn. The
1992; Kanowski et al., 1996, 1997; Le Bars et al., leaves are simple, with alternate arrangement and
1997; Morgenstern & Biermann, 1997; Nicola lobed margins, fan-shaped with parallel vena-
et al., 2009; Snitz et al., 2009; Herrschaft et al., tion, and a blade length of 24 inches [510 cm].
2012; Vellas et al., 2012). Some of these extracts The female trees bear an inedible foul-smelling
are also used to treat multiple sclerosis, tinnitus, fruit containing a hard edible seed (Gilman &
sexual dysfunction, and other health conditions Watson, 1993; ABC, 2000).
(Oken et al., 1998; Peters et al., 1998; Lovera et al.,
2012; Herrschaft et al., 2012; Evans, 2013; Nicola 1.1.2 Chemical constituents and their
et al., 2009; Hilton et al., 2013). properties
The major bioactive constituents found in the
leaves of ginkgo are reported to be flavonoids and
terpene lactones, with the flavonoids present
91
IARC MONOGRAPHS 108
primarily as glycosides (Ding et al., 2006; van (16)--D-glucopyranosyloxy]-4H-chromen

Beek & Montoro, 2009). Major and minor flavo- -4-one
noids are described below. Standardized extracts Description: Yellow-brownish powder (Ding
of ginkgo leaves (CAS No. 122933-57-7; 123009- et al., 2006; van Beek & Montoro, 2009;
84-7; 401901-81-3) are frequently formulated to ONeil, 2013)
contain ~24% flavonoids and ~6% lactones (van Melting-point: 195 C (ONeil, 2013)
Beek & Montoro, 2009). Other important constit-
uents found in ginkgo include biflavonoids and Solubility: Soluble in water (ONeil, 2013)
traces of alkylphenols, such as ginkgolic acids
(Wagner & Bladt, 1996; DeFeudis, 1991; Schtz, (b) Minor flavonoids
2002; Siegers, 1999; van Beek & Montoro, 2009). (i) Quercetin
Ginkgo also contains ginkgotoxin, which has
been reported to be structurally related to Chem. Abstr. Serv. Reg. No.: 117-39-5
vitamin B6 (Leistner & Drewke, 2010; Fig.1.1). IUPAC name: 2-(3,4-dihydroxyphenyl)-3,5,7-
CAS numbers and IUPAC names of the major trihydroxy-4H-chromen-4-one
components found in ginkgo are presented below Description: Yellow crystalline substance
(Chemical Abstracts Service, 2014). (Ding et al., 2006; van Beek & Montoro, 2009;
ONeil, 2013)
(a) Major flavonoids
Melting-point: 316 C
(i) Quercetin-3--D-glucoside
Solubility: Insoluble in water
Chem. Abstr. Serv. Reg. No.: 482-35-9 Quercetin was previously evaluated by IARC
IUPAC name: 2-(3,4-dihydroxyphenyl)- (IARC, 1999) as not classifiable as to its carcino-
5,7-dihydroxy-3-[(2S,3R,4S,5S,6R)-3,4,5- genicity to humans (Group 3).
trihydroxy-6-(hydroxymethyl)oxan-2-yl]
(ii) Kaempferol
oxychromen-4-one
(ii) Quercitrin
IUPAC name: 3,5,7-trihydroxy-2-(4-hydroxy-
Chem. Abstr. Serv. Reg. No.: 522-12-3 phenyl)-4H-chromen-4-one
IUPAC name: 2-(3,4-dihydroxyphenyl)-5,7- Description: Yellow powder (Ding et al., 2006;
dihydroxy-3-[(2S,3S,4R,5R,6S)-3,4,5-trihy van Beek & Montoro, 2009; ONeil, 2013)
droxy-6-methyloxan-2-yl]oxychromen-4-one Melting-point: 276 C
Description: Yellow crystalline substance Solubility: Slowly soluble in water
(Ding et al., 2006; van Beek & Montoro, 2009;
ONeil, 2013) (iii) Isorhamnetin
Melting-point: 174 C Chem. Abstr. Serv. Reg. No.: 480-19-3
Solubility: Insoluble in cold water IUPAC name: 3,5,7-trihydroxy-2-(4-hydroxy-
3-methoxyphenyl)chromen-4-one
(iii) Rutin
Description: Yellow powder (Ding et al., 2006;
Chem. Abstr. Serv. Reg. No.: 153-18-4 van Beek & Montoro, 2009; ONeil, 2013)
IUPAC name: 2-(3,4-dihydroxyphenyl)- Melting-point: 307 C
5,7-dihydroxy-3-[-L-rhamnopyranosyl-
92
Ginkgo biloba
Fig.1.1 Structural and molecular formulae and relative molecular mass of the major constituents
found in Ginkgo biloba
Flavonoids found in the leaves of Ginkgo biloba
OH OCH 3
OH OH OH
HO O HO O HO O
OH OH OH
OH O OH O OH O
Quercet in Kaempfer ol Isor hamnet in

C 15H10O 7 C 15H10O 6 C 16H12O 7
RMM = 3 02.24 RMM = 2 86.24 RMM = 3 16.26
OH OH OH
OH OH OH
HO O HO O HO O CH3
O
OH O HO
CH3
O O O OH
O O O
OH O OH O OH O OH
OH OH OH
OH OH OH
Quercet in -3- - D -glu co si d e Quercit ri n Rutin
C 21H20O 12 C 21H20O 11 C 27H30O 16
RMM = 4 64.38 RMM = 4 48.38 RMM = 6 10.52
Terpene lactones found in the leaves of Ginkgo biloba

O O
HO O HO O
O
C(CH3) 3 C(CH3) 3
O R1 O O
OH
R3 R2 O O
H 3C O
O
O
Ginkgoli de A R1=R 2=H, R3=OH Biloba lide

Ginkgoli de B R1=R 3=OH, R 2=H
Ginkgoli de C R1=R 2=R 3=OH
Ginkgoli de A- RMM=408.40; C20H 24O 9; Ginkgoli de B- RMM=424.40; C20H 24O 10
Ginkgoli de C- RMM=440.40; C20H 24O 11; Bilobalide - RMM=326.3; C 15H 18O 8
Ginkgotoxin found in Ginkgo biloba seeds

O CH3
HO
OH
H 3C N
4-O-me thyl pyridoxi ne

C 9H13NO 3
RMM=183.20
RMM, relative molecular mass

From Ding et al. (2006) and Leistner & Drewke (2010)
93
IARC MONOGRAPHS 108
(c) Lactone components (iv) Bilobalide

(i) Ginkgolide A Chem. Abstr. Serv. Reg. No.: 33570-04-6
Chem. Abstr. Serv. Reg. No.: 15291-75-5 IUPAC name: (5aR-(3aS*,5a,8b,8aS*,9a,10a
IUPAC name: 9H-1,7a-(epoxymethano)-1H, ))-9-(1,1-dimethylethyl)-10,10a-dihydro-
6aH-c yclopenta[c]furo[2,3-b]furo[3,2:3,4]-8,9-dihydrox y-4H,5a H,9H-f uro[2,3-b]
cyclopenta[1,2-days]furan-5,9,12(4H)-trione, f u r o [ 3 , 2 : 2 , 3 ] c y c l o p e n t a [1 , 2 - c ]
3-(1,1-dimet hylet hyl)hexah ydro-4,7b- furan-2,4,7(3H,8H)-trione
dihydroxy-8-methyl-, [1R-(1,3,3aS*,4,6a
(v) Ginkgotoxin
,7a,7b,8,10a,11aS*)]-)
Description: White crystalline substance Chem. Abstr. Serv. Reg. No.: 1464-33-1
(Ding et al., 2006; van Beek & Montoro, 2009; IUPAC name: 5-Hydroxy-3-(hydroxymethyl)-
ONeil, 2013) 4-(methoxymethyl)-6-methylpyridine)
Melting-point: 280 C
(ii) Ginkgolide B 1.1.3 Technical and commercial products,

and impurities
Products containing dried ginkgo leaf, dried
IUPAC name: 9H-1,7a-(epoxymethano)- ginkgo leaf extract, and standardized dried
1H, 6 a H-c yclopent a[c]f u ro[2 , 3 -b]f u ro ginkgo leaf extract are sold worldwide as herbal
[3,2:3,4]-cyclopenta[1,2-d]furan-5,9,12(4H)- medicinal products, dietary supplements, and
trione, 3-(1,1-dimethylethyl)hexahydro- food additives. Tablets and capsules containing
4,7b,11-trihydroxy-8-methyl-, [1R-(1,3,3aS between 4060 mg of extract are sold in the USA
*,4,6a,7a,7b,8,10a,11,11aR*)]-) as dietary supplements (Hnsel, 1991; Brestel &
Description: White crystalline substance Van Dyke, 1991). In Europe, the extract has been
(Ding et al., 2006; van Beek & Montoro, 2009; sold primarily as a phytopharmaceutical in a
ONeil, 2013) variety of dosage forms such as tablets, liquids,
Melting-point: ~300 C and parenteral preparations (Hnsel, 1991;
Brestel & Van Dyke, 1991), and is now available
(iii) Ginkgolide C as a herbal medicinal product and food supple-
ment (Lachenmeier et al., 2012). Commercial
products have been shown to contain variable
IUPAC name: 9H-1,7a-(epoxymethano)- amounts of the active constituents, and some
1H, 6 a H-c yclopent a[c]f u ro[2 , 3 -b]f u ro also contain high amounts of ginkgolic acids
[3,2:3,4]-cyclopenta[1,2-d]furan-5,9,12(4H)- (McKenna et al., 2002; Consumer Council,
trione, 3-(1,1-dimethylethyl)hexahydro-2,4, 2000; Harnly et al., 2012). The flavonoids present
7b,11-tetrahydroxy-8-methyl-, [1R-(1,2,3 in ginkgo are primarily in the glycoside form
,3aS*,4,6a,7a,7b,8,10a,11,11aR*)]-) (Ding et al., 2006; van Beek & Montoro, 2009).
Description: White crystalline substance The most widely used quality assurance assays
(Ding et al., 2006; van Beek & Montoro, 2009; (Association of Analytical Communities, AOAC;
ONeil, 2013) United States Pharmacopeia, USP, etc) are based
Melting-point: ~300 C on the measurement of free flavonoids obtained
directly from the hydrolysis of the flavonoid
94
Ginkgo biloba
glycosides. Addition of cheaper plant mate- lactones includes toluene, ethyl acetate, acetone
rial containing large amounts of rutin or other and methanol (20:10:10:1.2). Acetic anhydride is
flavonoids and flavonoid glycosides to boost the used as spraying reagent (USP, 2013).
apparent content of flavonol glycosides has been The USP requires that a dry extract of Ginkgo
reported, and more recent tests call for evalua- biloba dried leaf is characterized by containing
tion of flavonoid ratios to detect such adultera- not less than 22% and not more than 27% of
tion (Harnly et al., 2012). flavonoids calculated as flavonol glycosides
Also, enzyme-assisted extraction has led to via high-performance liquid chromatography
an increase in the amount of impurities in the (HPLC). The extract should also contains not
extract (van Beek & Montoro, 2009). Commercial less than 5.4% and not more than 12% of terpene
standardized Ginkgo biloba extracts are now lactones. Ginkgo biloba leaf extract is required to
prepared through a complex series of extractions have a ratio of crude plant material to powdered
and back-extractions using different solvents. extract between 35:1 and 67:1. A mobile phase
The purpose is to purify the flavonol glycosides composed of methanol, water, and phosphoric
and to remove unwanted compounds (Harnly acid (100:100:1) is used for the content of flavonol
et al., 2012). glycosides. A gradient eluent mixture of meth-
Kressmann et al. (2002) investigated the anol and water (25:7590:10) is used for the
pharmaceutical quality and phytochemical content of terpene lactones (USP, 2013).
composition of several different brands of Ginkgo Validated HPLC methods for flavonoids and
biloba sold as dietary supplements in the USA. terpenes in ginkgo have been published (Gray
In-vitro dissolution characteristics and phyto- et al., 2007; Croom et al., 2007). Quantitative
chemical content varied between products, in determination of the major components of ginkgo
some cases dramatically (as for the ginkgolic acid have also been reported using combination
content). [This study highlighted the difficulty of methods including HPLC, gas chromatography
estimating exposure to botanical products and (GC), nuclear magnetic resonance (NMR), and
their constituent phytochemicals, which may be mass spectrometry (MS) (Ding et al., 2006; van
marketed under the same name, but have very Beek & Montoro, 2009). Methods that have been
different compositions.] recently standardized for the quantitation of the
major components of ginkgo include reversed-
phase HPLC/ESI-MS (high-performance liquid
1.2 Analysis chromatography/electrospray ionization-mass
Identification tests for Ginkgo biloba in the spectrometry), HPLC/MS-MS, GC-MS, and
1H-NMR (Ding et al., 2006; van Beek & Montoro,
USP are based on high-performance thin-layer
chromatography (HPTLC). Botanical identity 2009; Li et al., 2004).
and composition are confirmed by HPTLC, as Pharmacopeial standards are mandatory
well as macroscopic and microscopic exami- in registered drug products, but not in dietary
nation (USP, 2013). A standardized developing supplement or food products, so it is difficult to
solvent system for flavonoids includes ethyl evaluate the composition of marketed non-drug
acetate, anhydrous formic acid, glacial acetic products.
acid, and water (100:11:11:26). The spraying
reagents include 5 mg/mL of 2-aminoethyl
diphenylborinate in methanol and 50 mg/mL
of polyethylene glycol 400 in alcohol. A stand-
ardized developing solvent system for terpene
95
IARC MONOGRAPHS 108
1.3 Uses (Blumenthal et al., 1998). These doses correspond

to an estimate of 50 fresh ginkgo leaves to yield
1.3.1 Indications one standard dose of the extract. Dried extracts
Ginkgo leaves and fruit are used medici- of leaves in the form of tablets, standardized to
nally for a variety of conditions. Among the uses contain 24% flavone glycosides and 6% terpenes,
reported for ginkgo are treatment of asthma, are available commercially (Brestel & Van Dyke,
bronchitis, cardiovascular diseases, improve- 1991; McKenna et al., 2002; Hilton et al., 2013;
ment of peripheral blood flow, and reduction of Kleijnen & Knipschild, 1992).
cerebral function (Perry, 1984; Mouren et al.,
1994). Additional uses include allergies, bron- 1.3.3 Trends in use
chitis, tinnitus, dementia, and memory issues According to USA National Health and
(Morgenstern & Biermann, 1997; Wang et al., Nutrition Survey (NHANES) data, there has been
2010; Holgers et al., 1994). The World Health a steady decline in the prevalence of use in men
Organization and the German Commission E and women from 19992002 (3.9%), and 2003
have included additional uses for peripheral arte- 2006 (3.0%), to 20072010 (1.6%) (NHANES,
rial occlusive diseases (WHO, 1999). 19992010). Barnes et al. (2008) reported that in
Previous studies performed using standard- a survey of users of complementary and alter-
ized ginkgo extracts have also found therapeutic native medicine that 11.3% of supplement-users
benefits for early stages of dementia, peripheral had used ginkgo in the previous 30 days.
arterial occlusive diseases, cerebral insufficiency
due to lack of adequate blood flow, and for other
related ailments (Wesnes et al., 2000; Oken et al., 1.4 Production, sales, and
1998; Pittler & Ernst, 2000; Hopfenmller, 1994). consumption
Lately, several clinical trials to evaluate cognitive
performance for Alzheimer disease, multiple 1.4.1 Production
sclerosis, and for peripheral arterial diseases Ginkgo has been planted on a large scale
have been conducted with mixed results (Vellas in France and USA since the 1980s and planta-
et al., 2012; Schneider, 2012; Weinmann et al., tions are found in the south eastern USA with a
2010; Snitz et al., 2009; Herrschaft et al., 2012; density of 10 million ginkgo trees per 1000 acres
Nicola et al., 2009; Hilton et al., 2013). (Del Tredici, 1991, 2005). A large amount of the
ginkgo sold in the USA comes from plantations
1.3.2 Dosage in China (Schmid & Balz, 2005).
According to the Commission E Monographs,
120240 mg of standardized dry extract in liquid 1.4.2 Sales
or solid pharmaceutical form for oral intake, is Ginkgo biloba is the most frequently
given in two or three daily doses for dementia prescribed herbal medicine in Germany and one
syndromes, such as primary degenerative of the most commonly used over-the-counter
dementia, vascular dementia, and mixed forms herbal preparations in the USA (Diamond et al.,
of both. Doses of 120160 mg of native dry 2000). The use of dietary supplements in the
extract is given in two or three daily doses for USA has significantly increased in the past few
improvement of pain-free walking distance in years. According to the 2012 Nutrition Business
peripheral arterial occlusive disease, and vertigo Journal Annual report (Nutrition Business
and tinnitus of vascular and involutional origin
96
Ginkgo biloba
Journal, 2012), ginkgo was the 37th best-selling 1.6 Regulations and guidelines
herbal dietary supplement in the USA in 2011,
and the 13th best-selling (by dollar volume) According to the 1994 Dietary Supplement
herbal dietary supplement at US$ 90 million. An Health and Education Act (DSHEA) in the USA,
estimated 3 million individuals in the USA used ginkgo is considered a dietary supplement under
ginkgo in 2007, a relatively steady decline from the general umbrella of foods (FDA, 1994). In
approximately 7.7 million users, and sales of US$ the USA, dietary supplements put on the market
151 million in 2002 (Wu et al., 2011). According before 15 October 1994 do not require proof
to IMS Health MIDAS data (IMS Health, of safety; however, the labelling recommenda-
2012), total worldwide sales of Ginkgo biloba tions for dietary supplements include warnings,
products in 2012 were US$ 1.26 billion. China dosage recommendations, and substantiated
accounted for 46% of sales (US$ 578 million). structure or function claims. The product
Fifty-six individuals per 1000 reported using label must declare prominently that the claims
natural health products containing ginkgo in a have not been evaluated by the Food and Drug
Canadian survey in 2006 (Singh & Levine, 2006). Administration, and bear the statement This
Other countries with substantial sales included product is not intended to diagnose, treat, cure,
Germany (US$ 152 million), Australia (US$ 61 or prevent any disease (Croom & Walker, 1995).
million), France (US$ 53 million), Brazil (US$ 48 In Europe, ginkgo is available either as
million), Republic of Korea (US$ 40 million), and food supplement or as medicinal product. The
Viet Nam (US$ 37 million). Commission E approved different types of ginkgo
preparations for human consumption (Diamond
1.4.3 Consumption et al., 2000) and published a monograph dedi-
cated to standardized ginkgo leaf extract (Mills
Consumption of ginkgo occurs orally or topi- & Bone, 2005). In Europe, most herbal products,
cally in pharmaceutical or dietary-supplement including ginkgo, were marketed as medic-
formulations (Hnsel, 1991; Brestel & Van Dyke, inal products. This changed when Directive
1991; Blumenthal et al., 1998). Consumers may 65/65/EEC (Council of the European Economic
also be exposed through products containing Community, 1965) was implemented, with the
ginkgo, such as teas, yogurts, and cosmetics that requirement of quality, efficacy, and safety data
are sold over the internet. [The Working Group on medicinal products. Many medicinal prod-
also noted that besides the use in medicinal ucts did not satisfy those requirements and were
products and supplements, ginkgo has also been nevertheless marketed as food supplements, often
used in Europe as ingredient in foods, as in the just changing the label. This is currently the case
so-called wellness beverages.] for ginkgo, which is widely marketed as a food
Ginkgo biloba extract is one of four herbal supplement in Europe (Lachenmeier et al., 2012).
preparations refunded by health insurance in
Germany, on prescription by a medical doctor
(AMR, 2013). 2. Cancer in Humans
1.5 Occupational exposure 2.1 Background

No data were available to the Working The available epidemiological studies on
Group. Workers on ginkgo plantations and in ginkgo and cancer consisted of one rand-
ginkgo-processing plants are probably exposed. omized controlled trial (the Ginko Evaluation of
97
IARC MONOGRAPHS 108
Memory, GEM study), four nested casecontrol period, beginning 1 year after randomization,
studies from the Vitamins and Lifestyle (VITAL) were generally similar, but with a statistically
cohort, and one population based casecontrol significant increase in the risk of cancer of the
study of cancer of the ovary. breast (HR, 2.50; 95% CI, 1.036.07) and lower
risk of cancer of the bladder (HR, 0.99; 95% CI,
0.522.61). The hazard ratios from a sensitivity
2.2 Randomized controlled trial analysis excluding participants who reported
See Table2.1 cancer 5years before baseline were within 20%,
The Ginkgo Evaluation of Memory (GEM) with the exception of an increase for cancer of
study is a randomized, double-blind, placebo-con- the colorectum reaching statistical significance
trolled clinical trial on Ginkgo biloba extract (EGb (HR, 2.15; 95% CI, 1.114.15) and a attenuation
761) for the prevention of dementia (DeKosky of risk of cancer of the breast, losing statistical
et al., 2008; Biggs et al., 2010). Participants aged significance (HR, 1.55; 95% CI, 0.663.63). [The
75 years and older (n=3069) from four clinical major strengths of the study were randomiza-
centres in the USA were enrolled between 2000 tion of treatment groups and adequate follow-up
and 2002, and were randomized to either the procedures for identifying cancer cases hospital-
placebo group, or the group receiving ginkgo ized during follow-up; however, incident cancers
as two daily doses of 120 mg of ginkgo extract. not resulting in hospitalization may have been
Participants were followed until 2008, with a missed and the follow-up was short, so the statis-
median follow-up of 6.1years. [The authors did tical power for cancer was low, and long-term
not specify the period of treatment.] Invasive carcinogenic effects due to exposure could not
cancer (excluding non-melanoma cancer of the be evaluated. The study findings were difficult to
skin) was evaluated as a secondary outcome, and interpret since the sites with statistically signifi-
identified from hospital admission and discharge cant associations were not consistent in different
records. Analyses were performed with inten- analyses. Furthermore, the generalizability of the
tion to treat beginning at randomization and at findings was limited by the clinical trial design.]
1 year after randomization. Adherence to study
protocol was 64% for the placebo group and 59% 2.3 Casecontrol studies
for the group receiving gingko.
Of the 310 hospitalizations for cancer, 162 See Table2.2
occurred in the group receiving gingko, and 148 The VITAL cohort includes 77738 men and
occurred in the group receiving placebo (adjusted women, aged 5076 years, in Washington state,
hazard ratio, HR, 1.09; 95% CI, 0.871.36). For USA, who completed a postal questionnaire on
the observation period beginning at random- supplement use, diet, health history, and risk
ization, elevated hazard ratios were reported factors between 2000 and 2002 (White et al.,
for cancers of the colon and rectum (HR, 1.62; 2004). The overall response rate was about 23%.
95% CI, 0.922.87), urinary bladder (HR, 1.21; Detailed data on use of vitamins, mineral supple-
95% CI, 0.652.26) and breast (HR, 2.15; 95% ments, herbal preparations and related products
CI, 0.974.80) and a decreased hazard ratio was in the past 10 years were obtained via questions
found for cancer of the prostate (HR, 0.71; 95% regarding brand, current and past use, frequency
CI, 0.431.17). The hazard ratios for cancer of the and duration of use. Dose information was not
lung and combined leukaemia and lymphoma obtained because of lack of accurate information
were close to unity (see Table2.1 for values). In on potency. Cohort members were followed for
general, hazard ratios for the second observation 56years, with incident cancer cases or deaths
98
Table 2.1 Randomized intervention trial on exposure to Ginkgo biloba extract
Reference, Total Study design Organ site Exposure categories Exposed Hazard ratio Covariates
study subjects cases (95% CI) Comments
location and
period
Biggs et al. 3069 Randomized Any Ginkgo extract: 162 1.09 (0.871.36) Adjusted for clinical centre.
(2010), USA double-blind, Prostate 120 mg, 2 per day; 27 0.71 (0.431.17) GEM study, participants aged
20008 placebo-controlled Lung randomization 26 0.90 (0.531.52) 75 yrs, intention to treat
clinical trial observation period % compliance: placebo, 64%;
Colorectal 31 1.62 (0.922.87) ginkgo extract, 59%
Urinary bladder 22 1.21 (0.652.26) Sensitivity analysis performed
excluding participants with a
Breast 18 2.15 (0.974.80)
5-yr history of cancer before
Leukaemia and 13 1.07 (0.492.34) enrolment.
lymphoma Median follow-up, 6.1 yr
Any 1 yr post-randomization 150 1.07 (0.851.35) End-point assessment from
observation period hospital admissions and
Prostate 25 0.68 (0.411.14) discharge records
Lung 27 1.00 (0.581.70)
Colorectal 22 1.27 (0.682.40)
Urinary bladder 18 0.99 (0.521.91)
Breast 16 2.50 (1.036.07)
Leukaemia and 13 1.17 (0.522.61)
lymphoma
GEM, Ginkgo Evaluation of Memory; yr, year
99
Ginkgo biloba
100
Table 2.2 Casecontrol studies of cancer and exposure to Ginkgo biloba extract
Reference, Total No. Control source Exposure Organ site Exposure Exposed Relative risk Covariates
study location cases (hospital, assessment categories cases (95% CI) Comments
and period Total No. population)
controls
Satia Lung Nested case Mailed Lung cancer Any pill per day 80 1.04 (0.821.32) Age, sex, education, smoking
et al. (2009), cancer, 665 control study: baseline Colorectal previous 10 yrs 49 0.83 (0.591.17) Age, sex, education, physical
Washington 76460 VITAL cohort questionnaire; cancer activity, fruit and vegetable
State, USA Colorectal supplement consumption, BMI, NSAID use,
cancer, 428 use 10 yrs and sigmoidoscopy
IARC MONOGRAPHS 108
76084 before Age 5076 yr; low response

baseline rate (21.8%); cases identified by
cancer registry (SEER); follow-
up until 2006 (mean, 5 yr)
Hotaling 330 Nested case Same as Satia Urothelial 10 yr daily NR NR (no Age, sex, race, education,
et al. (2011) 76720 control study: et al. (2009) carcinoma average statistical family history of bladder
Washington VITAL cohort of the signficant cancer, smoking, and fruit and
State, USA bladder association in vegetable intake
multivariate Age 5076 yr; cases identified
analysis) by cancer registry (SEER);
follow-up, 6 yr
Brasky 880 Nested case Same as Satia Breast Former 47 1.06 (0.771.45) Adjusted for age, race,
et al. (2010), 34136 control study: et al. (2009) cancer education, BMI, height, fruit &
Washington VITAL cohort Current 56 0.85 (0.641.13) vegetable consumption, alcohol
State, USA 10 yr daily consumption, physical activity,
average reproductive history, history
of hysterectomy, hormone
low (<4 days/wk 82 0.98 (0.771.25)
therapy, family history of
or any use <3 yr)
breast cancer and benign breast
high (4 days/wk 40 0.88 (0.631.24) biopsy, mammography, use of
for 3 yr) aspirin, ibuprofen, naproxen
Trend P = 0.51 and multivitamins.
Postmenopausal women;
excluded women with history
of breast cancer or with in-situ
breast cancer; participation
rate, 23%; mean follow-up, 6 yr
Reference, Total No. Control source Exposure Organ site Exposure Exposed Relative risk Covariates
study location cases (hospital, assessment categories cases (95% CI) Comments
and period Total No. population)
controls
Brasky 1602 Cancer-registry Same as Satia Prostate Use 165 1.03 (0.871.22) Age, race, education, BMI, PSA
et al. (2011), 33637 study: VITAL et al. (2009) cancer Average, 10 yr test, history of prostate disease,
Washington cohort (invasive) Low (<4 days/wk 127 1.08 (0.901.31) family history of prostate
State, USA or any use <3 yr) cancer, multivitamin use,
memory loss, diabetes.
High (4 days/ 85 1.04 (0.821.31)
Males: age 5076 yr.; low
wk for 3 yr)
response rate (~22%); cases
Trend P=0.52 identified by cancer registry
(SEER); follow-up; 6 yr
Ye et al. (2007) 668 Population In-person Ovarian Weekly; at least 6 11 0.41 (0.200.84) Age, study centre, oral
MA, NH, USA, 721 interviews cancer months contraceptive use, parity, and
19982003 and self- (epithelial) Ginkgo alone 6 0.36 (0.140.91) Jewish ethnic background
administered (not other drugs) RR similar for current versus
dietary Tumour type no-longer-using, no association
questionnaires with exposure duration;
Mucinous 3 1.17 (0.344.05)
however, small number of
Non-mucinous 8 0.33 (0.150.74) exposed cases. Participation
rate: cases, 52%; controls, 39%
BMI, body mass index; NR, not reported; NSAID, nonsteroidal anti-inflammatory drugs; PSA, prostate-specific antigen; RR, relative risk; SEER, Surveillance, Epidemiology, and End
Results programme of the National Cancer Institute; VITAL; VITamins and Lifestyle Cohort Study; wk, week; yr, year
101
Ginkgo biloba
IARC MONOGRAPHS 108
identified via linkage system to the SEER cancer cell-culture analyses showing antiproliferative
registry, state death files, and other databases. effects of G. biloba extract in serous (non-muci-
Associations between ginkgo intake and nous), but not in mucinous ovarian cancer cells.
cancer were investigated in nested casecontrol [The consistency of findings in cell culture and
studies with the large VITAL cohort for cancers humans was a strength of this study. Limitations
of the lung and colorectum (Satia et al., 2009), of the study were low statistical power, low
prostate (Brasky et al., 2011), breast (Brasky et al., participation rates, exposure assessment only for
2010), and urothelial carcinoma of the bladder 1 year before diagnosis, and lack of information
(Hotaling et al., 2011) (see Table 2.2 for more on dose.]
information). No statistically significant increase
or decrease in the occurrence of any of the cancers
examined was observed in these studies. Hazard 3. Cancer in Experimental Animals
ratios were near unity for all associations reported,
except cancer of the colorectum and any use of 3.1 Mouse
ginko (HR, 0.83; 95% CI, 0.591.17) and cancer
of the breast and current use (HR, 0.85; 95% CI, See Table3.1
0.651.13) or high average use (HR, 0.88; 95% CI, In one study of oral administration, groups
0.631.24). [The strengths of these studies were of 50 male and 50 female B6C3F1 mice (age,
the prospective design, adequate case-ascertain- 67weeks) were given Ginkgo biloba extract at a
ment, broad age range of the study participants, dose of 0 (corn oil vehicle, 5mL/kg body weight,
and large size. The major limitations were use of bw), 200, 600, or 2000 mg/kg bw by gavage,
self-reported exposure information, which was 5 days per week, for 104 weeks. The G. biloba
not updated after baseline; short follow-up time; extract contained 31.2% flavonol, 15.4% terpene
low response rates, and inadequate evaluation of lactones (bilobalide, 6.94%; ginkgolide A, 3.74%;
exposureresponse and exposuretime relation- ginkgolide B, 1.62%; ginkgolide C, 3.06%), and
ships. These limitations would most likely bias ginkgolic acid at 10.45 ppm. Survival of males
the findings towards the null.] at 600 and 2000 mg/kg bw was significantly
Ye et al. (2007) conducted a population-based less than that of controls; survival of females
casecontrol study of 668 incident cases of at 600mg/kg bw was significantly greater than
cancer of the ovary (identified from state regis- that of controls. Mean body weights of males at
tries and tumour boards) and 721 age- and 600 and 2000mg/kg bw were less (10% or more)
residence-matched general population controls. than those of the controls after weeks85 and 77,
Approximately 53% of eligible cases and 39% of respectively; mean body weights of females at
eligible controls participated. Intake of ginkgo 2000mg/kg bw were generally less (10% or more)
and other herbal remedies, dietary information than those of the controls between weeks17 and
and other factors was assessed via in-person 69, and after week93.
interview and self-administered questionnaire. In males, the incidence of hepatocellular
Ginkgo intake at least weekly for 6 months or carcinoma, hepatocellular adenoma or carci-
more was associated with a decreased risk of noma (combined), and hepatoblastoma was
cancer of the ovary (adjusted odds ratio, OR, significantly increased in the groups receiving
0.41; 95% CI, 0.200.84). However, the reduced the lowest, intermediate, and highest dose, and
risk was restricted to women with non-mucinous had a significant positive trend. Hepatocellular
ovarian tumours (OR, 0.33; 9%% CI, 0.150.74; carcinoma and hepatoblastoma were observed
eight exposed cases). This study also included in the same animal on multiple occasions. The
102
Table 3.1 Studies of carcinogenicity with Ginkgo biloba extracts in mice
Strain (sex) Dosing regimen, Incidence of tumours Significancea Comments

Duration Animals/group at
Reference start
B6C3F1 (M,F) 0 (control), 200, 600, Hepatocellular adenoma: *P<0.001 (trend) Ginkgo biloba study material contained:
104105wk 2000mg/kg bw, 17/50*, 37/50**, 41/50**, 48/50** (F) **P0.001 31.2% flavonol, 15.4% terpene lactones (bilobalide,
NTP (2013), by gavage in corn Hepatocellular carcinoma: ***P0.05 6.94%; ginkgolide A, 3.74% ; ginkgolide B, 1.62%;
Hoenerhoff et al. oil, 5days/wk for 22/50*, 31/50***, 41/50**, 47/50** (M) ****P0.01 ginkgolide C, 3.06%), ginkgolic acid, 10.45ppm
(2013) 104105wk 9/50*, 10/50, 15/50, 44/50** (F) HPLC/UV profiles identified 37 components:
50 M and 50 F/group Hepatocellular adenoma or carcinoma quercetin, 34.08%; kaempferol, 27.7%; isorhamnetin,
(age, 67 wk) (combined): 5.43%.
39/50*, 46/50****, 46/50****, 49/50** (M) HPLC/ELS profiles identified 18 components:
20/50*, 39/50**, 41/50**, 49/50** (F) bilobalide, 17.31%; ginkgolide C, 3.25%; ginkgolide
A, 9.06%; ginkgolide B, 2.05%; quercetin, 28.74%;
Hepatoblastoma:
kaempferol, 12.58%; isorhamnetin, 2.24%.
3/50*, 28/50**, 36/50**, 38/50** (M)
Mean body weight of males at 600 and 2000mg/kg
1/50*, 1/50, 8/50***, 11/50****(F)
bw were 10% less than the vehicle-control groups
Thyroid gland follicular cell adenoma: after wk85 and 77, respectively.
0/49, 0/49, 2/50, 2/50 (4%) (M)b Mean body weight of females at 2000mg/kg bw was
10% less than the vehicle-control group between
wk17 and 69 and after wk 93
a The Poly-3 test was used for all statistical analysis in this table.
b Historical incidence for 2-year gavage studies with corn oil vehicle-control groups (meanstandard deviation): 1/349 (0.3%0.8%), range, 02%; all routes: 7/1143 (0.6%1.0%),
range, 02%.
bw, body weight; ELS, evaporative light scattering; F, female; HPLC/UV, high-performance liquid chromatography/ultraviolet; M, male; wk, week
103
Ginkgo biloba
104
Table 3.2 Studies of carcinogenicity with Ginkgo biloba extracts in rats
Strain Dosing regimen, Incidence of tumours Significancea Comments

Duration
Reference
F344/N 0 (control), 100, 300, Mononuclear cell leukaemia: *P=0.004 (trend) Ginkgo biloba study material contained:
(M,F) 1000mg/kg bw, by gavage 9/50*, 12/50, 22/50**, 21/45** **P0.01 31.2% flavonol, 15.4% terpene lactones (bilobalide, 6.94%,
104 in corn oil, 5days/wk for (M) ***P=0.04 ginkgolide A, 3.74%; ginkgolide B, 1.62%; ginkgolide C,
105wk 104105wk Thyroid follicular cell adenoma: 3.06%), ginkgolic acid. 10.45ppm
NTP 50 M and 50 F/group (age, 2/50***, 1/50, 3/50, 5/45 (M) HPLC/UV profiles identified 37 components: quercetin,
IARC MONOGRAPHS 108
(2013) 67 wk) 1/49, 0/50, 3/49, 1/49 (F)b 34.08%; kaempferol, 27.7%; isorhamnetin, 5.43%
Thyroid follicular cell HPLC/ELS profiles identified 18 components: bilobalide,
carcinoma: 17.31%; ginkgolide C, 3.25%; ginkgolide A, 9.06%;
0/49, 0/50, 1/49, 1/49 (F)c ginkgolide B, 2.05%; quercetin, 28.74%; kaempferol, 12.58%;
isorhamnetin, 2.24%
Nose respiratory epithelium
Mean body weight of males at 300mg/kg bw, 10% less than
adenoma:
the vehicle-control group after wk 93; at 1000mg/kg bw,
0/49, 0/49, 2/50, 0/46 (F)d
10% less than the vehicle-control group after wk89.
Mean body weight of females: at 300mg/kg bw: 10% less
than the vehicle-control group after wk 93; at 1000mg/kg bw,
10% less than the vehicle-control group after wk89.
Body-weight suppression at highest dose could have reduced
tumour incidences.
a The Poly-3 test was used for all statistical analyses in this table.
b Historical incidence for 2-year gavage studies with corn-oil vehicle-control groups (meanstandard deviation): 3/298 (1.0%1.1%), range 02%; all routes: 8/1186 (0.7%1.0%),
range, 02%
c Historical incidence for 2-year gavage studies with corn-oil vehicle-control groups (meanstandard deviation): 1/298 (0.3%0.8%), range 02%; all routes: 5/1186 (0.4%1.0%),
range, 04%
d Historical incidence 2-year gavage studies with corn-oil vehicle-control groups (meanstandard deviation): 0/299; all routes: 1/1196 (0.1%0.4%), range, 02%
bw, body weight; ELS, evaporative light scattering; F, female; HPLC/UV, high-performance liquid chromatography/ultraviolet; M, male; wk, week
Ginkgo biloba
incidence of thyroid follicular cell adenoma was In males, the incidence of mononuclear cell
higher in the group receiving the intermediate leukaemia was significantly higher in groups
dose (2 out of 50; 4%) and highest dose (2 out at the intermediate and highest dose, and had
of 50; 4%) groups than among the historical a significant positive trend. The incidence of
controls in the National Toxicology Program thyroid follicular cell adenoma (2 out of 50, 4%;
(NTP) Technical Report study series (historical 1 out of 50, 2%; 3 out of 50, 6%; and 5 out of 45,
incidence range, 02%; corn oil vehicle controls, 10%) had a significant positive trend. In females,
1 out of 349; all routes, 7 out of 1449). the incidences of thyroid follicular cell adenoma
In females, the incidence of hepatocellular (1 out of 49, 0 out of 50, 3 out of 49, 1 out of 49),
adenoma was significantly higher in the groups thyroid follicular cell carcinoma (0 out of 49,
receiving the lowest, intermediate, and highest 0 out of 50, 1 out of 49, 1 out of 49), and nose
dose, and had a significant positive trend. The respiratory epithelium adenoma (0 out of 49, 0
incidence of hepatocellular carcinoma was out of 49, 2 out of 50, 0 out of 46) in the dosed
significantly higher in the group receiving the groups were higher than the ranges for histor-
highest dose, and had a significant positive ical controls (all routes: 8 out of 1186 [02%],
trend. The incidence of hepatocellular adenoma 5 out of 1186 [04%] and 1 out of 1186 [02%],
or carcinoma (combined) was significantly respectively) in the NTP Technical Report study
higher in the groups receiving the lowest, inter- series (NTP, 2013). [The Working Group noted
mediate, and highest dose, and had a significant that body-weight suppression at the highest dose
positive trend. The incidence of hepatoblastoma could have reduced the tumour incidence, and
was significantly higher at the intermediate and that nose respiratory epithelium adenomas are
highest dose, and had a significant positive trend rare spontaneous tumours in F344/N rats.]
(Hoenerhoff et al., 2013; NTP, 2013).
3.2 Rat 4. Mechanistic and Other

Relevant Data
See Table3.2
In one study of oral administration, groups
of 50 male and 50 female F344/N rats (age,
4.1 Absorption, distribution,
67weeks) were given G. biloba extract at a dose metabolism, and excretion
of 0 (corn oil vehicle, 2.5mL/kg bw), 100, 300, or 4.1.1 Humans
1000mg/kg bw by gavage, 5days per week, for
104weeks for males or 105weeks for females. The Several pharmacokinetics studies on the
G. biloba extract contained 31.2% flavonol, 15.4% main active components of Ginkgo biloba in
terpene lactones (bilobalide, 6.94%; ginkgolide humans have been reported (Fourtillan et al.,
A, 3.74%; ginkgolide B, 1.62%; ginkgolide C, 1995; Wjcicki et al., 1995; Pietta et al., 1997;
3.06%), and ginkgolic acid at 10.45ppm. Survival Mauri et al., 2001; Drago et al., 2002; Wang et al.,
of males at 1000mg/kg bw was significantly less 2003).
than that of the controls. Mean body weights of Wjcicki et al. (1995) investigated the phar-
males and females at 300mg/kg bw were less (10% macokinetics of flavonoid glycosides in 18
or more) than those of the controls after week93, healthy volunteers given one of three different
and those of males and females at 1000mg/kg bw formulations of Ginkgo biloba (capsules, drops,
were less (10% or more) after week89. and tablets) by oral administration. The area-
under-the-curve values were similar for all
105
IARC MONOGRAPHS 108
formulations. Drago et al. (2002) evaluated the (Griffiths & Smith, 1972). Excretion of quercetin
pharmacokinetics of ginkgolide B, the main or its conjugates in human urine ranged from
active ingredient in G. biloba extract, in 12 0.07% to 17.4% of intake. Only quercetin glucu-
healthy volunteers given different doses (80 mg ronides, but not free quercetin, could be detected
once daily, or 40 mg twice daily, for 7days). The in the plasma (Prior, 2003).
results indicated that the maximum concentra-
tion time (Tmax) was 2.3hours for both dosages, 4.1.2 Experimental systems
and the elimination half-life (t1/2) and mean
residence time (MRT) were longer for the dose Several studies have addressed the issue of
of 40 mg given twice daily than that for a single absorption and excretion of Ginkgo biloba in
80 mg dose, although the latter had a higher rats and mice (Moreau et al., 1986; Chen et al.,
concentration peak (Cmax) (Drago et al., 2002). 2007; Ude et al., 2013). In one study in which
14C-labelled G. biloba extract (360 mg/kg bw) was
After oral administration of a tablet containing
G. biloba extract in humans, two polyphenols, administered orally to rats, the amount expired
quercetin and kaempferol were found in urine as carbon dioxide only accounted for 16% of the
mainly as glucuronides, and to a lesser extent, original dose within the first 3 hours. About
sulfates (Wang et al., 2003). 38% of the administered dose was exhaled as
As part of a clinical study to determine carbon dioxide after 72 hours; 22% was excreted
the metabolites of G. biloba after human oral in the urine, and 29% was excreted in the faeces.
consumption, an extract of G. biloba leaves was Absorption reached at least 60%. A half-life of
given to six healthy volunteers [sex, age, and 4.5 hours and peak of 1.5 hours marked the
weight not specified] as a single dose of 4.0 g pharmacokinetics of G. biloba in serum, with
per day (Pietta et al., 1997). Urine samples were the characterization of a first-order phase kinetic
collected for 2days, and blood samples were with- model. After 48 hours of gradual uptake, radi-
drawn every 30 minutes for 5hours. The samples olabel was primarily found in the plasma. The
were purified through solid-phase extraction activity in the erythrocytes was similar to that in
with C18 cartridges (SPE C18 cartridges) and the plasma. It was also present in the neuronal,
analysed by reversed-phase liquid chromatog- glandular, and ocular tissue, and it was suggested
raphydiode array detection for the presence that the upper gastrointestinal tract plays a role
of metabolites. Only urine samples contained in absorption (Moreau et al., 1986). In rats but not
detectable amounts of substituted benzoic acids, in humans, phenylacetic acid or phenylpropionic
i.e. 4-hydroxybenzoic acid conjugate, 4-hydrox- acid derivatives were found in the urine after oral
yhippuric acid, 3-methoxy-4-hydroxyhippuric administration of G. biloba leaf extract (Pietta
acid, 3,4-dihydroxybenzoic acid, 4-hydroxyben- et al., 1995). Recent studies in rats have shown
zoic acid, hippuric acid and 3-methoxy-4-hy- that significant amounts of terpene trilactones
droxybenzoic acid (vanillic acid). No metabolites (ginkgolides A and B, and bilobalide) and flavo-
were detected in the blood (Pietta et al., 1997; see noids (quercetin, kaempferol, and isorhamnetin)
Fig. 4.1). cross the bloodbrain barrier and enter the
The pharmacokinetics of quercetin and its central nervous system after intravenous and
glycosides (components of G. biloba) have been oral administration of G. biloba extract (Chen
extensively studied in humans (Prior, 2006). et al., 2007; Rangel-Ordez et al., 2010).
Quercetin glycoside was originally assumed to be
absorbed from the small intestine after cleavage
of the -glycoside linkage by colonic microflora
106
Fig.4.1 Metabolites of Ginkgo biloba in humans
O O O
HO
OH OR OH
HO HO HO
4-H ydroxybenzoic acid and its conjugate 3,4-D i hydroxybenzoic acid
O O
O
H 3C OH OH
R
HO
3-M ethoxy-4-hydroxybenzoic acid O ther benzoic acid derivatives
(vanillic acid)
H uman urine
HO
O
H
N
OH
O HO
O
4-H ydroxyhippuric acid H
H 3C N
O ral O OH
Ex tract of G inkgo biloba
administration O
O
H 3-M ethoxy-4-hydroxyhippuric acid
N
OH
H ippuric acid
H uman blood N on-detectable
Subjects were given Gingko biloba extract orally. Urine samples contain detectable amounts of metabolites of Ginkgo biloba, but no metabolites were detected in the blood
Compiled by the Working Group from data in Pietta et al. (1997)
107
Ginkgo biloba
IARC MONOGRAPHS 108
4.1.3 Herbdrug interactions 4.2 Genetic and related effects

Despite potential therapeutic effects, the 4.2.1 Humans
widespread use of G. biloba extract may also
cause herbdrug interactions, altering drug No data were available to the Working Group.
efficiency or leading to undesired toxic effects
of concurrent medications, especially for drugs 4.2.2 Experimental systems
with narrow therapeutic indices. A growing body
See Table4.2
of literature has shown that G. biloba extract and
its constituents may influence the pharmacoki- (a) Mutagenicity
netics of coadministered drugs via altering the
Ginkgo biloba extract (up to 10000 g/plate)
expression and activity of drug-metabolizing
was mutagenic in Salmonella typhimurium
enzymes and transporters. However, the results
strains TA98 and TA100 and Escherichia coli
from in-vitro studies were not always in agree-
strain WP2 uvrA pKM101 with or without meta-
ment with those from in-vivo studies; and two
bolic activation from rat liver S9 (NTP, 2013).
similar clinical studies also showed disparities.
Two components of G. biloba extract, quercetin
For instance, in-vitro studies with human micro-
and kaempferol, were also found to give posi-
somes demonstrated that G. biloba extract inhib-
tive results in these assays and in other assays
ited cytochrome P450 (CYP) CYP2C9 (Mohutsky
for mutagenicity with and without metabolic
et al., 2006; Etheridge et al., 2007), while clin-
activation (NTP, 1992, 2013). G. biloba constitu-
ical studies showed that G. biloba extract had
ents that induce mutagenicity include quercetin
no significant effect on CYP2C9 (Greenblatt
(Bjeldanes & Chang, 1977; Hardigree & Epler,
et al., 2006; Mohutsky et al., 2006; Uchida et al.,
1978; Carver et al., 1983; NTP, 1992; Zeiger et al.,
2006). Robertson et al. (2008) reported that G.
1992; Chan et al., 2007), kaempferol (Silva et al.,
biloba extract induced the activity of CYP3A4
1997), and ginkgolic acids (Westendorf & Regan,
in a clinical study, while other studies reported
2000).
that G. biloba extract did not alter CYP3A4
(Gurley et al., 2002; Zadoyan et al., 2012), or (b) Chromosomal damage
decreased CYP3A4 activity (Uchida et al., 2006).
[The discrepancy was probably due to multiple No increase in the frequency of micronucleus
factors, for example, different approaches for formation in peripheral blood erythrocytes was
assessing enzyme activity were applied, different observed in male B6C3F1 mice, but results were
formulations of herbal materials were used, some equivocal in female B6C3F1 mice exposed to G.
studies did not include a sufficient sample size to biloba extract at a dose of up to 2.0 g/kg bw per day
meet statistical requirements; and the sensitivity by gavage for 3 months (NTP, 2013). Quercetin
of detection methods was different; or the studies and kaempferol were also found to produce
were conducted in different ethnic groups.] chromosomal alteration in various types of cells
Clinical studies and case reports have iden- (NTP 1992, 2013; Gaspar et al., 1994; Caria et al.,
tified herbdrug interactions potentiated by the 1995; Silva et al., 1997).
concurrent use of G. biloba extract and prescrip- (c) DNA damage
tion drugs, many of which are substrates of CYPs
and/or transport P-glycoprotein 1 (multidrug A study has examined and compared the
resistance protein 1); these studies are reviewed genotoxicity of flavonoids in human somatic cells
by Chen et al. (2011, 2012). and germ cells. Human somatic cells (human
108
Ginkgo biloba
Table 4.2 Genetic and related effects of Ginkgo biloba extract
Test system Results Concentration or Reference

dose
Without With (LED or HID)
exogenous exogenous
metabolic metabolic
system system
Salmonella typhimurium TA100 and TA98, reverse + + 10000 g/plate NTP (2013)
mutation
Escherichia coli WP2 uvrA/pKM101 reverse mutation + + 10000 g/plate NTP (2013)
Micronucleus formation in peripheral blood (males), EQ NT Up to 2.0 g/kg bw NTP (2013)
erythrocytes of male and female B6C3F1 mice in vivo (females) per day by gavage
for 3months
+, positive; (+), weakly positive; , negative; bw, body weight; LED, lowest effective dose; NT, not tested; NR, not reported; EQ, equivocal
lymphocytes) and germ cells (human sperm exert estrogenic activity by directly binding
cells) were treated with flavonoids (including both estrogen receptors and (Oh & Chung,
quercetin, kaempferol, and rutin, which are 2004, 2006). The proliferation of MCF-7 cells
present in G. biloba extract) at a concentration in response to G. biloba extract was biphasic
of 50500 M and their DNA damage potentials depending on the concentrations of extract and
were examined using the comet assay (Anderson E2 (17-estradiol) via estrogen receptor-de-
et al., 1997). Results show that DNA damage was pendent and independent pathways. In MCF-7
observed in lymphocytes and sperm cells over a cells, G. biloba extract induced cell proliferation
similar dose range. [The Working Group noted at low concentrations of E2, which has little or
that genotoxic responses occurred in somatic no estrogenic activity, but blocked the cell prolif-
and germ cells in approximately a one-to-one eration caused by higher concentrations of E2,
ratio.] In another study Duthie et al. (1997) which shows high estrogenic activity (Oh &
showed DNA strand breaks (as measured by Chung, 2006).
comet assay) in human cell lines Caco-2 (colon), Most studies have focused on the pharmaco-
HepG2 (liver), HeLa (epithelium) and normal logical effects of G. biloba. G. biloba extract and
lymphocytes after treatment with quercetin. its constituents have been shown to be involved
in many cellular activities, such as anti-oxidant
activity, anti-platelet activating factor, anti-in-
4.3 Other mechanistic data relevant flammatory effect, inhibition of mitochondrial
to carcinogenicity dysfunction, inhibition of amyloid aggrega-
tion in neuroblastoma cells, and anti-apoptosis
4.3.1 Effects on cell physiology
activity (Smith & Luo, 2004; Chan et al., 2007;
Increased levels of thyroid stimulating Shi et al., 2010a, b).
hormone (TSH) were observed in rats after treat-
ment with G. biloba extract for 14 weeks (NTP, 4.3.2 Effects on cell function
2013). In a 3-month study, follicular cell hyper-
trophy was observed in male and female rats No data were available to the Working Group.
(NTP, 2013).
G. biloba extract and its constituents including
quercetin, kaempferol, and isorhamnetin, may
109
IARC MONOGRAPHS 108
4.4 Susceptibility the USA and other parts of the world, including
China. Major reported indications are for asthma,
No data were available to the Working Group. bronchitis, cardiovascular diseases, improvement
of peripheral blood flow, and reduction of cere-
4.5 Mechanistic considerations bral insufficiency, allergies, tinnitus, dementia,
and memory issues. The main parts of the plant
The genotoxicity of G. biloba extract could used for these applications are the leaves and
be one of the mechanisms responsible for its the seeds. Ginkgo seeds are cooked and eaten
possible carcinogenicity. In addition, quercetin as food. Various forms of processed and unpro-
and kaempferol, the two flavonoid constituents cessed ginkgo leaf are present in dietary supple-
that are present in high levels in G. biloba extract, ments, herbal medicinal products and in foods.
are mutagenic as examined in several assays in Considerable sales were reported from China,
vitro and may thus contribute to the genotoxicity the USA, Germany, Australia, France, Brazil, the
of the G. biloba extract. Republic of Korea, Viet Nam, and Canada.
Quercetin and kaempferol have also been
shown to suppress the activities of DNA topoi-
somerases (Lpez-Lzaro et al., 2010; Russo et al.,
5.2 Human carcinogenicity data
2012). Although some topoisomerase suppres- The potential carcinogenicity of G. biloba
sors have therapeutic efficacy in human cancer, extract has been evaluated in few epidemiolog-
the clinical use of topoisomerase inhibitors can ical studies: one randomized controlled trial (the
also cause formation of secondary tumours, and Ginko Evaluation of Memory study, GEM), four
increased maternal consumption of flavonoids nested casecontrol studies from the VITamins
(some are topoisomerase inhibitors) during And Lifestyle (VITAL) cohort, and one popula-
pregnancy may be associated with infant acute tion-based casecontrol study of cancer of the
leukaemia (Ross et al., 1996; Strick et al., 2000; ovary.
Mistry et al., 2005; Ezoe, 2012) by interfering The GEM study was considered to be inform-
with DNA repair processes and inducing chro- ative because of its randomized design; however,
mosomal aberrations. It is possible that an inhib- the population was limited to people aged >75
itory effect on topoisomerase is the underlying years and compliance in the placebo and ginkgo
mechanism for quercetin- or kaempferol-associ- treatment groups was only about 60%. The
ated chromosomal damage. strengths of the VITAL study were its prospec-
Genotoxicity or topoisomerase inhibition tive design, the large number of exposed cases,
may be mechanisms of G. biloba-associated and the broad age range of the study participants;
carcinogenicity. however, no information was available on use of
ginkgo after enrolment.
The VITAL and GEM studies both reported
5. Summary of Data Reported risk estimates for cancers of the breast, colorectum,
lung, prostate, and for urothelial cell carcinoma,
5.1 Exposure data and the GEM study also reported a risk estimate for
leukaemia and lymphoma combined. Increased
Ginkgo biloba, also known as the fossil tree, risk for cancers of the breast and colorectum
is the oldest living tree. Products containing was reported in the GEM random clinical trial.
ginkgo leaf extract have been widely consumed The findings for cancer of the colorectum were
in Europe for many years, and are also popular in considered to be stronger in this study because,
110
Ginkgo biloba
in a sensitivity analysis that excluded partici- 5.4 Mechanistic and other relevant
pants reporting a history of cancer 5years before data
baseline, the relative risk increased and reached
statistical significance, while the relative risk of Components of G. biloba extract are exten-
cancer of the breast was attenuated and no longer sively metabolized in rodents and humans after
statistically significant in the sensitivity analysis. oral administration.
The VITAL cohort did not corroborate the GEM G. biloba extract gave positive results in
findings for cancers of the colorectum or breast, standard bacterial assays for mutation in the
finding somewhat decreased risk for both types absence or presence of exogenous metabolic
of cancer and ginkgo intake. The differences in activation. Two components of G. biloba extract,
findings between the two studies could be due to quercetin and kaempferol, also gave posi-
differences in age or window of exposure. Risks tive results in standard tests for genotoxicity.
for other types of cancers were either null or Quercetin, kaempferol, and rutin, a third compo-
somewhat decreased in both studies. The popu- nent of G. biloba extract, produced chromosomal
lation casecontrol study found a decreased risk damage.
of non-mucinous ovarian cancer, but not muci- Quercetin and kaempferol are inhibitors of
nous ovarian cancer, but the analysis was based DNA topoisomerases.
on small numbers of exposed cases. Genotoxicity and/or topoisomerase inhibi-
tion may be mechanisms of carcinogenicity asso-
ciated with G. biloba extract.
5.3 Animal carcinogenicity data
A G. biloba extract was tested for carcino-
genicity in two studies of oral administration 6. Evaluation
in mice and rats. In male and female mice
treated by gavage, a G. biloba extract containing 6.1 Cancer in humans
31.2% flavonol, 15.4% terpene lactones (bilo-
balide, 6.94%; ginkgolide A, 3.74%; ginkgolide There is inadequate evidence in humans for
B, 1.62%; ginkgolide C, 3.06%), and ginkgolic the carcinogenicity of Ginkgo biloba extract.
acid at a concentration of 10.45 ppm, produced
a significant increase in the incidences of hepa-
tocellular adenoma or carcinoma (combined),
6.2 Cancer in experimental animals
hepatocellular carcinoma and hepatoblastoma. There is sufficient evidence in experimental
In male mice receiving the extract, the incidence animals for the carcinogenicity of Ginkgo biloba
of thyroid follicular cell adenoma exceeded the extract.
range for historical controls in the study series. In
male rats given G. biloba extract by gavage, there
was a significant positive trend in the incidence 6.3 Overall evaluation
of thyroid follicular cell adenoma and a signif- Ginkgo biloba extract is possibly carcinogenic
icant increase in the incidence of mononuclear to humans (Group 2B).
cell leukaemia. In female rats, the incidences of
thyroid follicular cell adenoma, thyroid follicular
cell carcinoma, and nasal respiratory epithe-
lium adenoma exceeded the range for historical
controls in the study series.
111
IARC MONOGRAPHS 108
References Carver JH, Carrano AV, MacGregor JT (1983). Genetic

effects of the flavonols quercetin, kaempferol, and
galangin on Chinese hamster ovary cells in vitro. Mutat
ABC; American Botanical Council (2000). Boston: Res, 113(1):4560. doi:10.1016/0165-1161(83)90240-6
Integrative Medicine Communications. American PMID:6828043
Botanical Counsel. Integrative Medicine Chan PC, Xia Q, Fu PP (2007). Ginkgo biloba leave
Communications. Copyright. extract: biological, medicinal, and toxicological effects.
AMR (2013). Guidelines of the Federal Committee of J Environ Sci Health C Environ Carcinog Ecotoxicol
Physicians and Health Insurances on the prescrip- Rev, 25(3):21144. doi:10.1080/10590500701569414
tion of medicines within the Statutory Health Care PMID:17763047
(Arzneimittel-Richtlinien/AMR). Available from: Chemical Abstracts Service (2014). Scifinder. Columbus,
http://www.g-ba.de/downloads/83-691-314/AM-RL- OH, USA. American Chemical Society. Available at:
I-OTC_2013-04-05.pdf, accessed 09 December 2014 https://scifinder.cas.org:443/, accessed 09 December
Anderson D, Basaran N, Dobrzyska MM, Basaran AA, 2014
Yu TW (1997). Modulating effects of flavonoids on food Chen WD, Liang Y, Xie L, Lu T, Liu XD, Wang GJ
mutagens in human blood and sperm samples in the (2007). Pharmacokinetics of the ginkgo B following
comet assay. Teratog Carcinog Mutagen, 17(2):4558. intravenous administration of ginkgo B emulsion in
doi:10.1002/(SICI)1520-6866(1997)17:2<45::AID- rats. Biol Pharm Bull, 30(1):15. doi:10.1248/bpb.30.1
TCM1>3.0.CO;2-E PMID:9261919 PMID:17202649
Barnes PM, Bloom B, Nahin RL (2008). Complementary Chen XW, Serag ES, Sneed KB, Liang J, Chew H, Pan SY etal.
and alternative medicine use among adults and chil- (2011). Clinical herbal interactions with conventional
dren: United States, 2007. Natl Health Stat Report, drugs: from molecules to maladies. Curr Med Chem,
12(12):123. PMID:19361005 18(31):483650. doi:10.2174/092986711797535317
Biggs ML, Sorkin BC, Nahin RL, Kuller LH, Fitzpatrick PMID:21919844
AL (2010). Ginkgo biloba and risk of cancer: secondary Chen XW, Sneed KB, Pan SY, Cao C, Kanwar JR, Chew
analysis of the Ginkgo Evaluation of Memory (GEM) H et al. (2012). Herb-drug interactions and mecha-
Study. Pharmacoepidemiol Drug Saf, 19(7):6948. nistic and clinical considerations. Curr Drug Metab,
doi:10.1002/pds.1979 PMID:20582906 13(5):64051. doi:10.2174/1389200211209050640
Bjeldanes LF & Chang GW (1977). Mutagenic activity PMID:22292789
of quercetin and related compounds. Science, Consumer Council (2000). Test casts doubt on clinical
197(4303):5778. doi:10.1126/science.327550 benefits of Ginkgo leaf products with nonstandardized
PMID:327550 extract. Hong Kong, Choice No. 289 pp. 15.
Blumenthal M, Busse WR, Goldberg A et al., editors Klein Council of the European Economic Community (1965).
K, Rister RS (trans.). (1998). The Complete German Council Directive 65/65/EEC of 26 January 1965 on
Commission E Monographs Therapeutic Guide to the approximation of provisions laid down by Law,
Herbal Medicines. Austin. Regulation or Administrative Action relating to
Brasky TM, Kristal AR, Navarro SL, Lampe JW, Peters proprietary medicinal products. Off. J. 022, 03690373.
U, Patterson RE et al. (2011). Specialty supplements Croom E, Pace R, Paletti A, Sardone N, Gray D (2007).
and prostate cancer risk in the VITamins and Lifestyle Single-laboratory validation for the determination of
(VITAL) cohort. Nutr Cancer, 63(4):57382. doi:10.108 terpene lactones in Ginkgo biloba dietary supplement
0/01635581.2011.553022 PMID:21598177 crude materials and finished products by high-per-
Brasky TM, Lampe JW, Potter JD, Patterson RE, White formance liquid chromatography with evaporative
E (2010). Specialty supplements and breast cancer light-scattering detection. J AOAC Int, 90(3):64758.
risk in the VITamins And Lifestyle (VITAL) Cohort. PMID:17580616
Cancer Epidemiol Biomarkers Prev, 19(7):1696708. Croom EM Jr & Walker L (1995). Botanicals in the phar-
doi:10.1158/1055-9965.EPI-10-0318 PMID:20615886 macy: New life for old remedies. Drug Topics, 139:8493.
Brestel EP, Van Dyke K (1991). Chapter 42 in Modern DeFeudis FV (1991). Ginkgo biloba extract (EGb 761):
Pharmacology, 3rd ed., C.R. Craig and R. E. Stitzel, eds., pharmacological activities and clinical applications.
Little, Brown, Boston, pp. 567569. Amsterdam, Elsevier, pp 1013.
Caria H, Chaveca T, Laires A, Rueff J (1995). Genotoxicity DeKosky ST, Williamson JD, Fitzpatrick AL, Kronmal
of quercetin in the micronucleus assay in mouse bone RA, Ives DG, Saxton JA et al.; Ginkgo Evaluation of
marrow erythrocytes, human lymphocytes, V79 cell Memory (GEM) Study Investigators (2008). Ginkgo
line and identification of kinetochore-containing biloba for prevention of dementia: a randomized
(CREST staining) micronuclei in human lympho- controlled trial. JAMA, 300(19):225362. doi:10.1001/
cytes. Mutat Res, 343(23):8594. doi:10.1016/0165- jama.2008.683 PMID:19017911
1218(95)90075-6 PMID:7791812
112
Ginkgo biloba
Del Tredici P (1991). Ginkgos and people A thousand Gray D, LeVanseler K, Meide P, Waysek EH (2007).
years of interaction. Arnoldia, 51:215. Evaluation of a method to determine flavonol agly-
Del Tredici P (2005). The evolution, ecology, and culti- cones in Ginkgo biloba dietary supplement crude
vation of Ginkgo biloba. In: Ginkgo biloba. Van Beek materials and finished products by high-performance
TA, Harwood Academic Publishers, p. 20. Available liquid chromatography: collaborative study. J AOAC
from: http://books.google.com/books?id=g_MPcQ1 Int, 90(1):4353. PMID:17373435
gu3MC&dq=ginkgo+plantations&lr=&source=gbs_ Greenblatt DJ, von Moltke LL, Luo Y, Perloff ES, Horan
navlinks_s, accessed 09 December 2014 KA, Bruce A et al. (2006). Ginkgo biloba does
Diamond BJ, Shiflett SC, Feiwel N, Matheis RJ, Noskin not alter clearance of flurbiprofen, a cytochrome
O, Richards JA et al. (2000). Ginkgo biloba extract: P4502C9 substrate. J Clin Pharmacol, 46(2):21421.
mechanisms and clinical indications. Arch Phys Med doi:10.1177/0091270005283465 PMID:16432273
Rehabil, 81(5):66878. PMID:10807109 Griffiths LA & Smith GE (1972). Metabolism of myricetin
Ding S, Dudley E, Plummer S, Tang J, Newton RP, Brenton and related compounds in the rat. Metabolite forma-
AG (2006). Quantitative determination of major active tion in vivo and by the intestinal microflora in vitro.
components in Ginkgo biloba dietary supplements Biochem J, 130(1):14151. PMID:4655415
by liquid chromatography/mass spectrometry. Rapid Gurley BJ, Gardner SF, Hubbard MA, Williams DK,
Commun Mass Spectrom, 20(18):275360. doi:10.1002/ Gentry WB, Cui Y et al. (2002). Cytochrome P450
rcm.2646 PMID:16921563 phenotypic ratios for predicting herb-drug interac-
Drago F, Floriddia ML, Cro M, Giuffrida S (2002). tions in humans. Clin Pharmacol Ther, 72(3):27687.
Pharmacokinetics and bioavailability of a Ginkgo doi:10.1067/mcp.2002.126913 PMID:12235448
biloba extract. J Ocul Pharmacol Ther, 18(2):197202. Hnsel R. (1991). Phytopharmaka, 2nd ed., Berlin, Springer-
doi:10.1089/108076802317373941 PMID:12002672 Verlag, pp. 5972.
Duthie SJ, Johnson W, Dobson VL (1997). The effect of Hardigree AA & Epler JL (1978). Comparative mutagen-
dietary flavonoids on DNA damage (strand breaks esis of plant flavonoids in microbial systems. Mutat
and oxidised pyrimdines) and growth in human Res, 58(23):2319. doi:10.1016/0165-1218(78)90014-9
cells. Mutat Res, 390(12):14151. doi:10.1016/S0165- PMID:370575
1218(97)00010-4 PMID:9150762 Harnly JM, Luthria D, Chen P (2012). Detection of adulter-
Etheridge AS, Black SR, Patel PR, So J, Mathews JM (2007). ated Ginkgo biloba supplements using chromatographic
An in vitro evaluation of cytochrome P450 inhibi- and spectral fingerprints. J AOAC Int, 95(6):157987.
tion and P-glycoprotein interaction with goldenseal, doi:10.5740/jaoacint.12-096 PMID:23451372
Ginkgo biloba, grape seed, milk thistle, and ginseng Health IMS (2012). Multinational Integrated Data
extracts and their constituents. Planta Med, 73(8):731 Analysis (MIDAS). Plymouth Meeting, Pennsylvania:
41. doi:10.1055/s-2007-981550 PMID:17611934 IMS Health.
Evans JR (2013). Ginkgo biloba extract for age-related Herrschaft H, Nacu A, Likhachev S, Sholomov I, Hoerr
macular degeneration. Cochrane Database Syst Rev, R, Schlaefke S (2012). Ginkgo biloba extract EGb
1:CD001775 PMID:23440785 761 in dementia with neuropsychiatric features: a
Ezoe S (2012). Secondary leukemia associated with the randomised, placebo-controlled trial to confirm the
anti-cancer agent, etoposide, a topoisomerase II inhib- efficacy and safety of a daily dose of 240 mg. J Psychiatr
itor. Int J Environ Res Public Health, 9(7):244453. Res, 46(6):71623. doi:10.1016/j.jpsychires.2012.03.003
doi:10.3390/ijerph9072444 PMID:22851953 PMID:22459264
FDA (1994). Dietary Supplement Health and Education Hilton MP, Zimmermann EF, Hunt WT (2013). Ginkgo
Act of 1994. Public Law 103417, 103rd Congress. biloba for tinnitus. Cochrane Database Syst Rev,
Available from: http://www.fda.gov/opacom/laws/ 3:CD003852 PMID:23543524
dshea.html, accessed 09 December 2009. Hoenerhoff MJ, Pandiri AR, Snyder SA, Hong HH, Ton
Fourtillan JB, Brisson AM, Girault J, Ingrand I, Decourt TV, Peddada S etal. (2013). Hepatocellular carcinomas
JP, Drieu K et al. (1995). [Pharmacokinetic proper- in B6C3F1 mice treated with Ginkgo biloba extract
ties of Bilobalide and Ginkgolides A and B in healthy for two years differ from spontaneous liver tumors in
subjects after intravenous and oral administration of cancer gene mutations and genomic pathways. Toxicol
Ginkgo biloba extract (EGb 761)] Therapie, 50(2):137 Pathol, 41(6):82641. doi:10.1177/0192623312467520
44. PMID:7631288 PMID:23262642
Gaspar J, Rodrigues A, Laires A, Silva F, Costa S, Monteiro Holgers KM, Axelsson A, Pringle I (1994). Ginkgo biloba
MJ et al. (1994). On the mechanisms of genotoxicity extract for the treatment of tinnitus. Audiology,
and metabolism of quercetin. Mutagenesis, 9(5):4459. 33(2):8592. doi:10.3109/00206099409071870
doi:10.1093/mutage/9.5.445 PMID:7837978 PMID:8179518
Gilman EF & Watson DG (1993). Ginkgo biloba Maidenhair Hopfenmller W (1994). [Evidence for a therapeutic effect
Tree. Fact Sheet,, ST-273:13. of Ginkgo biloba special extract. Meta-analysis of 11
113
IARC MONOGRAPHS 108
clinical studies in patients with cerebrovascular insuf- inhibitors of DNA topoisomerases I and II in cells. Mutat
ficiency in old age] Arzneimittelforschung, 44(9):1005 Res, 696(1):417. doi:10.1016/j.mrgentox.2009.12.010
13. PMID:7986236 PMID:20025993
Hotaling JM, Wright JL, Pocobelli G, Bhatti P, Porter MP, Lovera JF, Kim E, Heriza E, Fitzpatrick M, Hunziker J,
White E (2011). Long-term use of supplemental vita- Turner AP etal. (2012). Ginkgo biloba does not improve
mins and minerals does not reduce the risk of urothe- cognitive function in MS: a randomized placebo-con-
lial cell carcinoma of the bladder in the VITamins And trolled trial. Neurology, 79(12):127884. doi:10.1212/
Lifestyle study. J Urol, 185(4):12105. doi:10.1016/j. WNL.0b013e31826aac60 PMID:22955125
juro.2010.11.081 PMID:21334017 Mauri P, Simonetti P, Gardana C, Minoggio M, Morazzoni
IARC (1999). Some chemicals that cause tumours of the P, Bombardelli E etal. (2001). Liquid chromatography/
kidney or urinary bladder in rodents and some other atmospheric pressure chemical ionization mass spec-
substances. IARC Monogr Eval Carcinog Risks Hum, trometry of terpene lactones in plasma of volunteers
73:1674. dosed with Ginkgo biloba L. extracts. Rapid Commun
Kanowski S, Hermann WM, Stephan K etal. (1997). Proof Mass Spectrom, 15(12):92934. doi:10.1002/rcm.316
of the efficacy of the Ginkgo biloba special extract EGb PMID:11400198
761 in outpatients suffering from mild to moderate McKennaOJJonesKHughesKTylerVM2002). Botanical
dementia of the Alzheimers type or multi-infarct medicines: the desk reference for major herbal supple-
dementia. Phytomedicine, 4:313. doi:10.1016/S0944- ments. 2nd ed. New York: The Haworth Herbal Press;
7113(97)80021-9 PMID:23195239 pp. 480-8.
Kanowski S, Herrmann WM, Stephan K, Wierich MillsSBoneK2005). Ginkgo. In: MillsSBoneKeditors.
W, Hrr R (1996). Proof of efficacy of the ginkgo Safety Monographs. Part II. The essential guide to
biloba special extract EGb 761 in outpatients herbal safety. Elsevier; pp. 425-32.
suffering from mild to moderate primary degener- Mistry AR, Felix CA, Whitmarsh RJ, Mason A, Reiter
ative dementia of the Alzheimer type or multi-in- A, Cassinat B et al. (2005). DNA topoisomerase II in
farct dementia. Pharmacopsychiatry, 29(2):4756. therapy-related acute promyelocytic leukemia. N Engl
doi:10.1055/s-2007-979544 PMID:8741021 J Med, 352(15):152938. doi:10.1056/NEJMoa042715
Kleijnen J & Knipschild P (1992). Ginkgo biloba for cere- PMID:15829534
bral insufficiency. Br J Clin Pharmacol, 34(4):3528. Mohutsky MA, Anderson GD, Miller JW, Elmer GW
doi:10.1111/j.1365-2125.1992.tb05642.x PMID:1457269 (2006). Ginkgo biloba: evaluation of CYP2C9
Kressmann S, Mller WE, Blume HH (2002). drug interactions in vitro and in vivo. Am J Ther,
Pharmaceutical quality of different Ginkgo 13(1):2431. doi:10.1097/01.mjt.0000143695.68285.31
biloba brands. J Pharm Pharmacol, 54(5):6619. PMID:16428919
doi:10.1211/0022357021778970 PMID:12005361 Moreau JP, Eck CR, McCabe J, Skinner S (1986).
Lachenmeier DW, Steffen C, el-Atma O, Maixner S, [Absorption, distribution and elimination of a labelled
Lbell-Behrends S, Kohl-Himmelseher M (2012). extract of Ginkgo biloba leaves in the rat] Presse Med,
What is a food and what is a medicinal product in the 15(31):145861. PMID:2947082
European Union? Use of the benchmark dose (BMD) Morgenstern C & Biermann E (1997). Ginkgo biloba
methodology to define a threshold for pharmacolog- special extract EGb 761 in the treatment of tinnitus
ical action. Regul Toxicol Pharmacol, 64(2):28695. aurium: Results of a randomized, double-blind, place-
doi:10.1016/j.yrtph.2012.08.017 PMID:22960033 bo-controlled study. Fortschr Med, 115:711
Le Bars PL, Katz MM, Berman N, Itil TM, Freedman Mouren X, Caillard P, Schwartz F (1994). Study of
AM, Schatzberg AF (1997). A placebo-controlled, the antiischemic action of EGb 761 in the treat-
double-blind, randomized trial of an extract of ment of peripheral arterial occlusive disease by
Ginkgo biloba for dementia. North American EGb TcPo2 determination. Angiology, 45(6):4137.
Study Group. JAMA, 278(16):132732. doi:10.1001/ doi:10.1177/000331979404500601 PMID:8203766
jama.1997.03550160047037 PMID:9343463 Nicola SP, Kruidenier LM, Bendermacher BL, Prins MH,
Leistner E & Drewke C (2010). Ginkgo biloba and ginkgo- Teijink JA (2009). Ginkgo biloba for intermittent clau-
toxin. J Nat Prod, 73(1):8692. doi:10.1021/np9005019 dication. Cochrane Database Syst Rev, 2(2):CD006888
PMID:20041670 PMID:19370657
Li CY, Lin CH, Wu CC, Lee KH, Wu TS (2004). Efficient NTP (1992). Toxicology and carcinogenesis studies of
1H nuclear magnetic resonance method for improved quercetin (CAS No. 117-39-5) in F344/N rats (feed
quality control analyses of Ginkgo constituents. J studies). Natl Toxicol Program Tech Rep Ser, 409:
Agric Food Chem, 52(12):37215. doi:10.1021/jf049920h NTP (2013). Toxicology and carcinogenesis studies of
PMID:15186088 Ginkgo biloba extract (CAS No. 90045-36-6) in F344/N
Lpez-Lzaro M, Willmore E, Austin CA (2010). The rats and B6C3F1/N mice (gavage studies). Natl Toxicol
dietary flavonoids myricetin and fisetin act as dual Program Tech Rep Ser, 578(578):1183. PMID:23652021
114
Ginkgo biloba
Nutrition Business Journal (2012). Nutrition Business Med, 76(15):168390. doi:10.1055/s-0030-1249962

Journal Annual Report. Available from: http:// PMID:20486074
nutritionbusinessjournal.com/, accessed 09 December Robertson SM, Davey RT, Voell J, Formentini E, Alfaro
2014 RM, Penzak SR (2008). Effect of Ginkgo biloba
ONeil MJ (2013). An Encyclopedia of Chemicals, Drugs, extract on lopinavir, midazolam and fexofenadine
and Biologicals. 15th Ed., Merck Sharp & Dohme Corp. pharmacokinetics in healthy subjects. Curr Med Res
Whitehouse, NJ, 2013, RSC Publishing. Opin, 24(2):5919. doi:10.1185/030079908X260871
Oh SM & Chung KH (2004). Estrogenic activities of Ginkgo PMID:18205997
biloba extracts. Life Sci, 74(11):132535. doi:10.1016/j. Ross JA, Potter JD, Reaman GH, Pendergrass TW, Robison
lfs.2003.06.045 PMID:14706564 LL (1996). Maternal exposure to potential inhibitors of
Oh SM & Chung KH (2006). Antiestrogenic activities of DNA topoisomerase II and infant leukemia (United
Ginkgo biloba extracts. J Steroid Biochem Mol Biol, States): a report from the Childrens Cancer Group.
100(45):16776. doi:10.1016/j.jsbmb.2006.04.007 Cancer Causes Control, 7(6):58190. doi:10.1007/
PMID:16842996 BF00051700 PMID:8932918
Oken BS, Storzbach DM, Kaye JA (1998). The efficacy Russo P, Del Bufalo A, Cesario A (2012). Flavonoids
of Ginkgo biloba on cognitive function in Alzheimer acting on DNA topoisomerases: recent advances and
disease. Arch Neurol, 55(11):140915. doi:10.1001/ future perspectives in cancer therapy. Curr Med Chem,
archneur.55.11.1409 PMID:9823823 19(31):528793. doi:10.2174/092986712803833272
Perry LM (1984). Medicinal Plants of East and Southeast PMID:22998568
Asia: Attributed Properties and Uses, MIT Press: Satia JA, Littman A, Slatore CG, Galanko JA, White E
Cambridge, MA. (2009). Associations of herbal and specialty supple-
Peters H, Kieser M, Hlscher U (1998). Demonstration of ments with lung and colorectal cancer risk in the
the efficacy of ginkgo biloba special extract EGb 761 VITamins and Lifestyle study. Cancer Epidemiol
on intermittent claudicationa placebo-controlled, Biomarkers Prev, 18(5):141928. doi:10.1158/1055-9965.
double-blind multicenter trial. Vasa, 27(2):10610. EPI-09-0038 PMID:19423520
PMID:9612115 Schmid W & Balz JP (2005). Cultivation of Ginkgo
Pietta PG, Gardana C, Mauri PL (1997). Identification of biloba L. on three continents. [ISHS]Acta Hortic,
Gingko biloba flavonol metabolites after oral admin- 676:177180.Available fromhttp://www.actahort.org/
istration to humans. J Chromatogr B Biomed Sci Appl, books/676/676_23.htmaccessed 09 December 2014
693(1):24955. doi:10.1016/S0378-4347(96)00513-0 Schneider LS (2012). Ginkgo and AD: key negatives and
PMID:9200545 lessons from GuidAge. Lancet Neurol, 11(10):8367.
Pietta PG, Gardana C, Mauri PL, Maffei-Facino R, Carini doi:10.1016/S1474-4422(12)70212-0 PMID:22959218
M (1995). Identification of flavonoid metabolites Schtz K (2002). Detection of allergenic urushiols in
after oral administration to rats of a Ginkgo biloba Ginkgo biloba leaves. Pharmazie, 57(7):50810.
extract. J Chromatogr B Biomed Appl, 673(1):7580. PMID:12168544
doi:10.1016/0378-4347(95)00252-E PMID:8925077 Shi C, Liu J, Wu F, Yew DT (2010a). Ginkgo biloba extract
Pittler MH & Ernst E (2000). Ginkgo biloba extract for the in Alzheimers disease: from action mechanisms to
treatment of intermittent claudication: a meta-anal- medical practice. Int J Mol Sci, 11(1):10723. doi:10.3390/
ysis of randomized trials. Am J Med, 108(4):27681. ijms11010107 PMID:20162004
doi:10.1016/S0002-9343(99)00454-4 PMID:11014719 Shi C, Wu F, Yew DT, Xu J, Zhu Y (2010b). Bilobalide
Prior RL (2003). Fruits and vegetables in the prevention of prevents apoptosis through activation of the PI3K/Akt
cellular oxidative damage. Am J Clin Nutr, 78(3):Suppl: pathway in SH-SY5Y cells. Apoptosis, 15(6):71527.
570S8S. PMID:12936951 doi:10.1007/s10495-010-0492-x PMID:20333467
Prior RL (2006). Phytochemicals. In: Shils ME, Shike Siegers CP (1999). Cytotoxicity of alkylphenols from
M, Ross AC, Caballero B Cousins RJ (eds) 10th ed. Ginkgo biloba. Phytomedicine, 6(4):2813. doi:10.1016/
Modern Nutrition in Health and Disease. Philadelphia: S0944-7113(99)80021-X PMID:10589448
Lippincott, Williams and Wilkins, 582593. Silva ID, Rodrigues AS, Gaspar J, Maia R, Laires A, Rueff
Rai GS, Shovlin C, Wesnes KA (1991). A double-blind, J (1997). Involvement of rat cytochrome 1A1 in the
placebo controlled study of Ginkgo biloba extract biotransformation of kaempferol to quercetin: rele-
(tanakan) in elderly outpatients with mild to moderate vance to the genotoxicity of kaempferol. Mutagenesis,
memory impairment. Curr Med Res Opin, 12(6):3505. 12(5):38390. doi:10.1093/mutage/12.5.383
doi:10.1185/03007999109111504 PMID:2044394 PMID:9379919
Rangel-Ordez L, Nldner M, Schubert-Zsilavecz M, Singh SR & Levine MAH (2006). Natural health product
Wurglics M (2010). Plasma levels and distribution of use in Canada: analysis of the National Population
flavonoids in rat brain after single and repeated doses Health Survey. Can J Clin Pharmacol, 13(2):e24050.
of standardized Ginkgo biloba extract EGb 761. Planta PMID:16921199
115
IARC MONOGRAPHS 108
Smith JV & Luo Y (2004). Studies on molecular mech- Wesnes KA, Ward T, McGinty A, Petrini O (2000). The
anisms of Ginkgo biloba extract. Appl Microbiol memory enhancing effects of a Ginkgo biloba/Panax
Biotechnol, 64(4):46572. doi:10.1007/s00253-003- ginseng combination in healthy middle-aged volun-
1527-9 PMID:14740187 teers. Psychopharmacology (Berl), 152(4):35361.
Snitz BE, OMeara ES, Carlson MC, Arnold AM, Ives doi:10.1007/s002130000533 PMID:11140327
DG, Rapp SR et al.; Ginkgo Evaluation of Memory Westendorf J & Regan J (2000). Induction of DNA strand-
(GEM) Study Investigators (2009). Ginkgo biloba for breaks in primary rat hepatocytes by ginkgolic acids.
preventing cognitive decline in older adults: a rand- Pharmazie, 55(11):8645. PMID:11126011
omized trial. JAMA, 302(24):266370. doi:10.1001/ White E, Patterson RE, Kristal AR, Thornquist M, King
jama.2009.1913 PMID:20040554 I, Shattuck AL et al. (2004). VITamins And Lifestyle
Strick R, Strissel PL, Borgers S, Smith SL, Rowley JD (2000). cohort study: study design and characteristics of supple-
Dietary bioflavonoids induce cleavage in the MLL gene ment users. Am J Epidemiol, 159(1):8393. doi:10.1093/
and may contribute to infant leukemia. Proc Natl Acad aje/kwh010 PMID:14693663
Sci USA, 97(9):47905. doi:10.1073/pnas.070061297 WHO; World Health Organization (1999). Ginkgo folium.
PMID:10758153 WHO Monographs on Selected Medicinal Plants, Vol. 1.
Uchida S, Yamada H, Li XD, Maruyama S, Ohmori Y, Oki Geneva: World Health Organization.
T etal. (2006). Effects of Ginkgo biloba extract on phar- Wjcicki J, Gawroska-Szklarz B, Bieganowski W,
macokinetics and pharmacodynamics of tolbutamide Patalan M, Smulski HK, Samochowiec L etal. (1995).
and midazolam in healthy volunteers. J Clin Pharmacol, Comparative pharmacokinetics and bioavailability of
46(11):12908. doi:10.1177/0091270006292628 flavonoid glycosides of Ginkgo biloba after a single oral
PMID:17050793 administration of three formulations to healthy volun-
Ude C, Schubert-Zsilavecz M, Wurglics M (2013). Ginkgo teers. Mater Med Pol, 27(4):1416. PMID:9000837
biloba extracts: a review of the pharmacokinetics of the Wu C-H, Wang C-C, Kennedy J (2011). Changes in herb
active ingredients. Clin Pharmacokinet, 52(9):72749. and dietary supplement use in the U.S. adult popu-
doi:10.1007/s40262-013-0074-5 PMID:23703577 lation: a comparison of the 2002 and 2007 National
USP; United States Pharmacopeia (2013). 36th rev. and Health Interview Surveys. Clin Ther, 33(11):174958.
National Formulary, 31th ed., Rockville, MD: United doi:10.1016/j.clinthera.2011.09.024 PMID:22030445
States Pharmacopeial Convention, Inc. Ye B, Aponte M, Dai Y, Li L, Ho MC, Vitonis A et al.
van Beek TA & Montoro P (2009). Chemical analysis and (2007). Ginkgo biloba and ovarian cancer preven-
quality control of Ginkgo biloba leaves, extracts, and tion: epidemiological and biological evidence. Cancer
phytopharmaceuticals. J Chromatogr A, 1216(11):2002 Lett, 251(1):4352. doi:10.1016/j.canlet.2006.10.025
32. doi:10.1016/j.chroma.2009.01.013 PMID:19195661 PMID:17194528
Vellas B, Coley N, Ousset PJ, Berrut G, Dartigues JF, Dubois Zadoyan G, Rokitta D, Klement S, Dienel A, Hoerr R,
B etal.; GuidAge Study Group (2012). Long-term use Gramatt T etal. (2012). Effect of Ginkgo biloba special
of standardised Ginkgo biloba extract for the preven- extract EGb 761 on human cytochrome P450 activity:
tion of Alzheimers disease (GuidAge): a randomised a cocktail interaction study in healthy volunteers. Eur J
placebo-controlled trial. Lancet Neurol, 11(10):8519. Clin Pharmacol, 68(5):55360. doi:10.1007/s00228-011-
doi:10.1016/S1474-4422(12)70206-5 PMID:22959217 1174-5 PMID:22189672
Wagner H, Bladt S (1996). Plant drug analysis: a thin layer Zeiger E, Anderson B, Haworth S, Lawlor T, Mortelmans
chromatography atlas, 2nd Ed., Berlin, SpringerVerlag, K (1992). Salmonella mutagenicity tests: V. Results from
p 237. the testing of 311 chemicals. Environ Mol Mutagen,
Wang BS, Wang H, Song YY, Qi H, Rong ZX, Wang BS 19(S21):Suppl 21: 2141. doi:10.1002/em.2850190603
et al. (2010). Effectiveness of standardized ginkgo PMID:1541260
biloba extract on cognitive symptoms of dementia
with a six-month treatment: a bivariate random effect
meta-analysis. Pharmacopsychiatry, 43(3):8691.
doi:10.1055/s-0029-1242817 PMID:20104449
Wang FM, Yao TW, Zeng S (2003). Determination of
quercetin and kaempferol in human urine after orally
administrated tablet of ginkgo biloba extract by HPLC.
J Pharm Biomed Anal, 33(2):31721. doi:10.1016/S0731-
7085(03)00255-3 PMID:12972097
Weinmann S, Roll S, Schwarzbach C, Vauth C, Willich SN
(2010). Effects of Ginkgo biloba in dementia: system-
atic review and meta-analysis. BMC Geriatr, 10(1):14
doi:10.1186/1471-2318-10-14 PMID:20236541
116
KAVA
1. Exposure Data (WHO, 2004; ONeil et al., 2006; NTP, 2012;

and SciFinder, 2013)
The kava plant is indigenous to Oceania According to WHO (2007), some other
(Lebot et al., 1997; Ramzan & Tran, 2004) and species such as Piper wichmanni, Piper aduncum
has been used both ceremonially and recreation- and Piper auritum have also been marketed as
ally in certain cultures of the South Pacific for kava.
at least 1500 years. Europeans documented its Common names: Kava; Kava-kava; Ava-ava;
use when they travelled to Polynesia in the 18th Antares; Ava; Ava pepper; Ava pepper shrub;
century (WHO, 2007). The cultural history of the Ava root; Awa; Fijian kava; Gea; Gi; Grog;
use of kava has been reviewed by Singh (1992). Intoxicating long pepper; Intoxicating pepper;
In the past, traditional use of kava was Kao; Kava kava rhizome; Kava root; Kavapiper;
widespread, but, in certain cultures, custom Kavapyrones; Kavarod; Kavasporal forte; Kave-
determined who could use kava and for what kave; Kawa; Kawa kawa; Kawa pepper; Kawa
purposes. In recent years, as part of the processes Pfeffer; Kew; Long pepper; Macropiper latifo-
of modernization, major changes have occurred lium; Malohu; Maluk; Maori kava; Meruk; Milik;
with regard to who uses kava, and where and Pepe kava; Piperis methystici rhizome; Rhizoma
how it is consumed. In some places, kava is now piperis methystici; Sakaua; Sakau; Tonga; Yagona;
being consumed much like alcohol in western Yangona; Yaqona; Yongona
countries, as a beverage that is drunk socially (b) Description
(McDonald & Jowitt, 2000).
See Fig.1.1
The tropical shrub Piper methysticum is a
1.1 Identification of the agent hardy, fairly succulent, slow-growing perennial
that is widely cultivated in Oceania. The species
1.1.1 Botanical data
is sterile and reproduces asexually. Due to its
(a) Nomenclature traditional use as a ritual beverage known for
promoting relaxation and a sense of well-being,
Chem. Abstr. Serv. Reg. No.: 9000-38-8 the kava plant spread widely throughout Oceania,
Chem. Abstr. Name: Kava-kava resin (8Cl) in Polynesia, Melanesia, and the Federated States
Botanical name: Piper methysticum G. Forst of Micronesia (Norton & Ruze, 1994; NTP, 2012).
Family: Piperaceae The leaves are heart-shaped, pointed, 825 cm
in length, and smooth and green on both sides.
Genus: Piper
Kava is cultivated for its rootstock (rhizome), also
Plant part: Rhizome
117
IARC MONOGRAPHS 108
Fig.1.1 Piper methysticum G. Forst between stem joints), colour of stems, intensity of
leaf colour, and quality of the root. Different vari-
eties are classified, named, and used for different
purposes by the indigenous people (NTP, 2012).
The dried rhizome consists of irregular,
transverse and longitudinal pieces, varying
considerably in size and shape: 320 cm in length
and 15 cm in diameter. The outer surface is light
yellowish or greyish-brown, longitudinally wrin-
kled, with large, whitish, circular root scars. The
fracture is coarsely fibrous, the inner surface is
yellow-white, with thin bark, radiate xylem, and
large pith (WHO, 2004).
1.1.2 Chemical constituents and their

properties
Analysis of the composition of kava rhizome
indicates that the fresh material is on average
80% water. When dried, the rhizome consists of
approximately 43% starch, 20% fibres, 12% water,
3.2% sugars, 3.6% proteins, 3.2% minerals, and
15% kavalactones, although the kavalactone
component can vary between 3% and 20% of
the dry weight of the rhizome, depending on
the age of the plant and the cultivar. The bioac-
tive principles of kava rhizome are mostly, if not
From Spohn (2013)
Roland Spohn
entirely, contained in the lipid-soluble resin. The
compounds of greatest pharmacological interest
referred to as the stump. The stump is knotty, are the substituted -pyrones or kavapyrones,
thick, and sometimes tuberous and often contains commonly known as kavalactones. At least 15
holes or cracks created by partial destruction of lactones have been isolated from kava rhizome.
the parenchyma. A fringe of lateral roots up to 23 The following six compounds are present in
m in length extends from the pithy rhizome. The the highest concentrations and account for
roots comprise a multitude of ligneous fibres and approximately 96% of the lipid resin: kavain,
consist of >60% starch. Rhizome colour varies dihydrokavain, yangonin, desmethoxyyan-
from white to dark yellow, depending upon the gonin, methysticin, and dihydromethysticin
amount of kavalactones contained in the lemon- (see Fig.1.2). Other constituents of kava include
yellow resin. The plant is usually harvested when chalcones and other flavanones, and conjugated
it is about 22.5 m in height (Singh, 1992; Lebot diene ketones (Shulgin, 1973; Dentali, 1997;
et al., 1997; NTP, 2012). WHO, 2004; NTP, 2012).
The cultivation and selection of kava has In the past, kavain has been used to indi-
produced numerous varieties or cultivars recog- cate a racemic mixture resulting from chem-
nized by differences in the internodes (space ical synthesis, and kawain for the naturally
118
Kava
Fig.1.2 Structures of the major kavalactones occurring in kava rhizome

O O O
O
O O O O O O
O
O
K avain Y angonin M ethysticin
O O O
O
O O O O O O
D ihydrokavain D esmethoxyyangonin D ihydromethysticin

(7,8-D ihydrokavain) (5,6-D ehydrokavain)
occurring compound, which is a dextro-isomer. to the extent that they only contain 7178% of
Currently, the two terms, kavain and kawain, are the expected active constituents (Clough et al.,
frequently used interchangeably in the scientific 2000). Singh (2004a) mentioned adulteration
literature, but the term kavain has started to with sawdust, flour, or soil. Adulteration of kava
supersede kawain (Singh, 2004a). with plants resembling the genuine kava, but
The chemistry of kava and kavalactones lacking the kavalactones and the distinct kava
has been reviewed in detail by Ramzan & Tran odour has also been reported. The main false
(2004). kava species are P. auritum and P. aduncum
(Singh, 2004b).
1.1.3 Technical and commercial products Very few data exist on kava contamination
by bacteria or with mycotoxins (Teschke et al.,
Kava biomass is normally sold as the rhizome, 2011). A study on ochratoxin A contamination
with the periderm and roots removed. The peeled found concentrations of 3.0 ng/g in one sample
rhizome is also the desired material for solvent of kava root (Trucksess et al., 2006). The level of
extraction to produce kava extracts. Kava may contamination with aflatoxin B1 in four samples
also be sold as an unpeeled rhizome covered of ground kava was 0.5ng/g (Weaver & Trucksess,
with the cork or with the roots attached. Peelings 2010).
from the root and stump have also been used in The part of the plant used, processing tech-
commerce (Morgan et al., 2005). Powdered forms niques, and specifically the extraction solvent
of rhizome are available in commercial markets and the ratio between solvent/plant material in
in Fiji and have been described to be adulterated the case of kava extracts, may have considerable
119
IARC MONOGRAPHS 108
influence on the chemical composition of the end 1.3 Use

product. For example, the alkaloid pipermethys-
tine was not detectable in some commercial kava 1.3.1 Indications
extracts (Teschke et al., 2011). (a) Medicinal use
[The Working Group noted that the influences
on composition mentioned above may hinder the According to WHO (2004), the medicinal
comparison of studies, especially if the applied uses supported by clinical data are short-term
kava material was not specified exactly.] symptomatic treatment of mild states of anxiety
or insomnia due to nervousness, stress or
tension; the medicinal uses described in pharma-
1.2 Analysis copoeias and in traditional systems of medicine
The chemical analysis and quality control of are to induce relaxation, reduce weight, and treat
both kava and its extracts obtained by aqueous fungal infections. Uses described in traditional
acetone or aqueous methanol, and supercritical medicine, but not supported by experimental or
fluid extraction typically with carbon dioxide clinical data, are treatment of asthma, common
modified with methanol as solvent were cold, cystis, gonorrhoea, headaches, menstrual
reviewed by Bilia et al. (2004). Both gas chro- irregularities, urinary infections, and warts.
matography (GC) and high-performance liquid The German Commission E has approved
chromatography (HPLC) can be used for the kava for use in conditions of nervous anxiety,
analysis of kavalactones with some advantages stress, and restlessness (Anonymous, 2000).
and disadvantages for each method. Using GC (b) Traditional food and recreational use
analysis, methysticin and yangonin, which are
two of the major components, are generally not A local traditional drink, also known under
separated. In addition, the high temperature the name kava, is obtained from the root or
of the injection port causes the decomposition rhizome of the kava plant. The kava drink is
of methysticin. Concerning HPLC analyses, made from water extracts, with water-insoluble
reversed-phase separation is generally better substances made available to the drinker by
because it is highly reproducible with a very emulsification, which may be accomplished by
low detection limit for all compounds even if pounding or chewing of the rhizome (WHO,
the quantitative analysis of the kavalactones by 2007).
HPLC needs to be carried out in the absence On some islands in the South Pacific, fresh
of light to prevent the cis/trans isomerization kava root or rhizome is used to prepare the
of yangonin (Bilia et al., 2004). Besides various traditional drink, while on others it is the dried
chromatographic approaches reviewed by Bilia and ground roots or rhizomes that are used.
et al. (2004), near infrared spectroscopy and For fresh preparations, the root is chewed by
nuclear magnetic resonance spectroscopy have young women, who spit the juice into the kava
been suggested to directly determine kavalac- bowl without swallowing it themselves. The juice
tones without the need for separation (Table1.1). is then mixed with water or coconut milk and
further processed. Most people drink only the
water extracts of kava. This is obtained by adding
water to kava roots which are finely ground and
then filtered using cheese-cloth (WHO, 2007).
The kava drink has been described to have
a psychoactive activity, and potency can vary
120
Table 1.1 Selected methods of analysis of constituents of kava in various matrices
Sample matrix Analyte/purpose of analysis Sample preparation Assay Detection Reference

method limit
Serum and urine Kavain and metabolites/metabolism Glucuronidase treatment, extraction with HPLC-DAD 1 ng/mL Tarbah et al. (2003)
dichloromethane:diethylether (7:3, v:v) and LC-MS
Urine Kavalactones/metabolism Chloroform extraction GC-MS, NA Duffield et al. (1989)
HPLC
Kava root from Ochratoxin A/contamination Immunoaffinity column cleanup HPLC-FD NA Trucksess et al.
botanical supplier (2006)
Dried kava roots Kavain and other major Ethanol extraction HSCCC NA Schfer &
kavalactones/isolation Winterhalter (2005)
Commercial kava Kavalactones and a range of other Solution with DMSO-d6 NMR NA Bilia et al. (2002)
extract, and kava finely compounds/quality assessment
powdered
Kava root extract Kavalactones/structural elucidation Extraction with methylene chloride NMR NA Dharmaratne et al.
(2002)
Kava dry extracts Kavain and total kavalactones/ None NIRS NA Gaub et al. (2004)
routine quality control
Food supplements Total kavalactones/regulatory Solution in ethanol, buffer addition NMR NA Monakhova et al.
containing kava control (2013)
DMSO, dimethyl sulfoxide; GC, gas chromatography; HPLC, high-performance liquid chromatography; HPLC-DAD, high-performance liquid chromatography with diode-array
detection; HPLC-FD, high performance liquid chromatography with fluorescence detection; HSCCC, high-speed counter-current chromatography; LC, liquid chromatography; MS,
mass spectrometry; NA, not applicable; NIRS, near infrared spectroscopy; NMR, nuclear magnetic resonance spectroscopy; v:v, volume:volume
121
Kava
IARC MONOGRAPHS 108
considerably. Kava drinking initially produces 1.3.2 Dosage

a slight numbing of the tongue. Delayed effects
have been described as relief of fatigue, reduc- (a) Medicinal use
tion of anxiety, and production of a pleasant, The comminuted crude drug and extracts are
cheerful, and sociable attitude in the drinker used for oral use. Daily dosage for crude drug
(WHO, 2007). and extracts is equivalent to 60210 mg of kava-
The consumption of kava is part of everyday lactones (WHO, 2004). The recommended oral
life on islands such as Fiji, Tonga, and Vanuatu, dose for use of commercial kava extracts as an
and occurs during important events or social anxiolytic is 5070 mg of kavalactones, two to
gatherings (Singh, 1992). four times per day and, as a hypnotic, 150210 mg
It is difficult to compare the psychopharmaco- in a single oral dose before bedtime (Bilia et al.,
logical effects of kava between published studies 2002b).
as methods of preparation, means of ingestion, The pharmaceutical industry was primarily
and the potency and quantity of dosages actu- interested in the organic solvent (such as 95%
ally consumed vary considerably (Cairney et al., ethanol or acetone) extracts of kava containing
2002). the organic compounds of commercial interest.
Kava bars, at which prepared kava can be Some marketed products, referred to as
purchased to drink on the spot or to take away, synthetic, consist of a single kavalactone, l-ka-
are an increasingly common feature throughout vain (WHO, 2007).
Oceania (McDonald & Jowitt, 2000). A review of standardized kava brands in the
USA found an approximate equivalence of actual
(c) Non-traditional food use [measured] and labelled amounts of kavalactones
Non-traditional kava products are marketed in 13 products that listed amounts of constitu-
in Europe and North America typically as food ents. Kavalactones per tablet or capsule ranged
or dietary supplements in tablet form (Morris & from 50 to 110 mg. Two brands that did not label
Avorn, 2003; Teschke & Lebot, 2011). Interestingly, amounts of constituents contained 1015 mg per
these food supplements are often marketed over tablet or capsule (Ulbricht et al., 2005).
the internet (Morris & Avorn, 2003). In some Typical usage has ranged from 70 to 280 mg
countries this may be due to difficulties with of kavalactones per day as a single bedtime dose
regulatory acceptance. For example, in Europe or divided doses (60120 mg of kavalactones per
this practice is illegal, but kava products are day). Many practitioners allegedly start at a lower
nevertheless available (Monakhova et al., 2013). dose and titrate up as needed (Ulbricht et al.,
2005).
(d) Cosmetic use
(b) Traditional food and recreational use
Kava extracts from various parts of the
plant may be used as skin-conditioning agents Only rough estimations exist on the dosage
in cosmetics. However, the USA Cosmetic of traditional food and in recreational use of
Ingredient Review expert panel concluded that kava. Heavy consumers may drink the equiva-
the available data were insufficient to support the lent of at least 610 g/week of kava powder, which,
safety of kava extracts for cosmetic use (Robinson with an estimated kavalactone content of 12.5%,
et al., 2009). may equate to approximately 76 g of lactones per
week or more than 50 times the recommended
therapeutic dose (Cairney et al., 2002).
122
Kava
In Arnhem Land, Australia, weekly per capita (b) Production volume

consumption was estimated as 145 g of powder Kava was one of the most extensively used
for 198990 and 368 g of powder for 199091. herbal products in the USA in the 1990s (NTP,
When seven cups of 100 mL are consumed in 2012). According to Morris & Avorn (2003), sales
1hour, about 3.8 g of lactones may be consumed. of kava were US$ 69 million in 2000. In 2003, 62
In a detailed review of the literature on weekly retail sites were identified that sold kava over the
consumption levels and possible lactone contents, internet (Morris & Avorn, 2003).
the estimations encompassed a wide variation In Australia, trade in kava rhizome in
from 39 to 1840 g of kava powder consumed, and Arnhem Land was approximately 28 tonnes in
from 4.1 g to 188.6 g of lactones consumed per 1992, and between 27 and 36 tonnes in 1997. At
week (Clough et al., 2000). the end of 1999, by which time trade in kava was
Typical dosage of dried root or by decoction illegal, trade was estimated to be 20 tonnes, while
was reported to be 612 g per day (Morgan et al., in 2000 the trade was approximately 15 tonnes
2005). (Clough, 2003).
(c) Non-traditional food use
1.4.2 Sales
The Dietary Supplements Label Database lists
11 products that contain kava as active ingredient By the mid-1990s, North Americans,
in amounts of 601000 mg. Of the 11 products, 4 Europeans, and Australians had begun using
are listed as discontinued (NLM, 2012). kava products as an alternative medicine and
Kava food supplements, illegally sold over the herbal relaxant. Commercial kava bars promoted
internet in Germany, contained 810 mg of kava- recreational kava drinking, which can often
lactones per capsule (Monakhova et al., 2013). occur for extended periods. Drug stores and
supermarkets offered a variety of kava products
(d) Cosmetic use in pill, capsule, tea, and liquid form. In addition,
The Cosmetic, Toiletry, and Fragrance powdered kava root was available by mail order
Association (CTFA) provided a use concentra- from several internet sites. Most of this exported
tion of 0.00010.01% for leaf/root/stem extract, kava derived from Fiji and Vanuatu, and to a
and of 0.1% for root extract (Robinson et al., lesser extent, Samoa and Tonga (Lindstrom,
2009). 2004). Kava abuse has been reported, especially
in Pacific Island nations, leading to significant
health and social problems (McDonald & Jowitt,
1.4 Production, sales, and 2000; Rychetnik & Madronio, 2011).
consumption Current use in North America, Europe, and
Australia may have been influenced by regula-
1.4.1 Production tory measures (see Section 1.6) and reports of
(a) Production process adverse events in the popular press.
Kava production including cultivation, According to the 2012 Nutrition Business
diseases and pests, harvesting and processing Journal Annual Report, kava was the 36th best-
has been reviewed by Singh (2004b). selling dietary supplement in the USA in 2011.
There has been a considerable decline in kava
sales in the USA from US$ 52 million in 2000
to US$ 17 million in 2004. Sales remained at
a similar level between US$ 18 and 22 million
123
IARC MONOGRAPHS 108
Fig.1.3 Sales of kava as a dietary supplement in the USA
60
50
Kava sales (million US$)
40
30
20
10
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
Compiled by the Working Group from data from Nutrition Business Journal (2010, 2012).
during 20052011, and then increased to US$31 1.4.3 Consumption

million in 2011 (Fig. 1.3; Nutrition Business
Journal, 2010, 2012). Total global sales of kava Consumers of products specified in Section
(Piper methysticum) as an herbal supplement 1.3 are exposed to kava. No literature about the
were US$8 million in 2012, and appreciable sales degree of population-based exposure to kava was
volumes occurred in the USA (US $3 million), available to the Working Group. [The Working
Brazil (US$ 2 million), and Hungary (US$1 Group estimated that current exposure to kava
million) (IMS Health, 2012). was expected to be a fraction of what it was in the
[The Working Group suggested that prohi- previous decade due to withdrawal of marketing
bition in some countries may have resulted in authorization in many countries (see Section
increases in unrecorded sales of kava, e.g. unre- 1.6).]
corded individual imports, or illegal sales.]
124
Kava
Table 1.2 Guidelines for dried kava rhizome
Characteristic Guideline
Content Not less than 3.5% kavapyrones [kavalactones], as determined by IR spectroscopy
Identity tests Macroscopic, microscopic and microchemical examinations, and TLC for the presence of
characteristic unsaturated -pyrones known as kavapyrones [kavalactones]
Microbiological purity, heavy Limits according to WHO guidelines on quality control methods for medicinal plants
metals, radioactive residues
Foreign organic matter Not more than 2%
Total ash Not more than 8%
Acid-insoluble ash Not more than 1.5%
Water-soluble extractive Not less than 5%
Loss on drying Not more than 12%
Pesticide residues Aldrin and dieldrin not more than 0.05 mg/kg. For other pesticides, general guidelines
apply.
IR, infrared; TLC, thin-layer chromatography
From WHO (2004)
1.5 Occupational exposure et al., 2003; Teschke et al., 2003; Ulbricht et al.,
2005; NTP, 2012).
No specific studies on occupational exposure Although sales of kava were not regulated
were available to the Working Group. It can be or controlled in the USA in 2012 (NTP, 2012),
assumed that workers in kava production for the Food and Drug Administration (FDA) had
food, cosmetic, or medicinal use may be exposed. issued a public warning in 2001 that kava might
be be associated with serious liver damage,
1.6 Regulations and guidelines including hepatitis, cirrhosis, and liver failure
(FDA, 2002). The regulatory action taken by
Several cases of liver damage have been asso- various countries around the world from the
ciated with exposure to kava in Europe, and have year 2000 after concerns about hepatotoxicity is
led to withdrawal of the product license (NTP, summarized in WHO (2007). Current regula-
2012). Reviews on the cases of adverse effects tory status was summarized by Teschke & Lebot
potentially caused by exposure to kava have been (2011), and included suggested chemical stand-
compiled by Schmidt et al. (2005) (detailed anal- ards and agricultural standardizations. WHO
ysis of 83 cases), as by WHO (2007) (analysis of (2004) provided some guidelines for the quality
93 cases). Speculations about the causes of the control of kava (see Table1.2).
adverse effects included the use of less expensive
stem peelings in commercial materials instead of
the usual peeled rhizomes (Teschke et al., 2011). 2. Cancer in Humans
Sales of kava have been suspended or
withdrawn in several countries, including Steiner (2000) investigated the association
Australia, Canada, France, Germany, Spain, and between cancer incidence and consumption of
Switzerland, and due to reported association kava in an ecological study of six countries in
with hepatotoxicity in humans (Russmann et al., the South Pacific. Exposure was estimated by
2001, 2003; Campo et al., 2002; De Smet, 2002; a surrogate measure of consumption based on
Parkman, 2002; Clough et al., 2003; Humberston the number of kava plants under cultivation in
125
IARC MONOGRAPHS 108
each country. Exposure estimates and data on receiving the intermediate dose. The incidence
cancer incidence for men in the 1980s were used of eosinophilic hepatocyte foci, a preneoplastic
in the analysis on the assumption that all kava hepatocyte lesion, was significantly higher in
produced before 1990 was consumed locally, and the groups receiving the intermediate or highest
primarily by men. An inverse correlation was dose.
observed between the incidence of all cancers In females, the incidence of hepatocellular
in men and estimated exposure, but no test of carcinoma was significantly higher in the group
statistical significance or confidence intervals, receiving the lowest dose. The incidence of hepato-
was reported. [The Working Group considered cellular adenoma or carcinoma (combined) was
this study as uninformative because of its ecolog- significantly higher in the groups receiving the
ical design, the use of crude measures of expo- lowest and intermediate doses. The incidence
sure and outcome, and inadequate assessment of of hepatocellular carcinoma and hepatoblas-
the role of chance.] toma (combined) was significantly higher in the
group receiving the lowest dose. [The Working
Group noted that reduced body weight may
3. Cancer in Experimental Animals have reduced one tumour response in females at
the highest dose.] The incidence of eosinophilic
hepatocyte foci was significantly higher in the
3.1 Mouse
group receiving the highest dose. The incidence
See Table3.1 of squamous cell hyperplasia of the forestomach
In one study of oral administration, groups was significantly higher in the groups receiving
of 50 male and 50 female B6C3F1 mice (age, the lowest or intermediate doses (Behl et al., 2011;
56 weeks) were given kava extract at a dose NTP, 2012).
of 0 (corn oil vehicle, 10 mL/kg body weight,
bw), 0.25, 0.5, or 1.0g/kg bw per day by gavage, 3.2 Rat
5days per week, for 105weeks. The purity of the
kava extract was 98.04% by high-performance See Table3.1
liquid chromatography/ultraviolet (HPLC/UV) In one study of oral administration, groups
profiles. The extract contained 27% kavalactones of 49 or 50 male and 50 female F344/N rats (age,
identified as kavain, dihydrokavain, methysticin, 67weeks) were given kava extract at 0 (corn-oil
dihydromethysticin, yangonin, and desmethoxy- vehicle, 5mL/kg bw), 0.1, 0.3, or 1.0g/kg bw per
yangonin. In males, the mean body weight of the day by gavage, 5days per week, for 104 (male rats)
dosed groups was similar to that in the control or 105 (female rats) weeks. The purity of the kava
group. In females, the mean body weight of the extract was 98.04% by HPLC/UV profiles. The
group at 1.0g/kg bw was 11% less than that in the extract contained 27% kavalactones identified
control group after week21. The mean survival as kavain, dihydrokavain, methysticin, dihy-
time of male and female mice in the dosed groups dromethysticin, yangonin, and desmethoxyyan-
was similar to that of the controls. gonin. The mean body weight of the groups at
In males, the incidence of hepatoblastoma 1.0g/kg bw was 10% less than that of the control
was significantly higher in the groups receiving group after week65 in males and after week41
the intermediate and highest dose, and had in females. The mean survival time for rats in the
a significant positive trend. The incidence of dosed groups was similar to that of controls for
hepatocellular carcinoma and hepatoblastoma both sexes.
(combined) was significantly higher in the group
126
Table 3.1 Studies of carcinogenicity with kava extracts in mice and rats
Species, Dosing regimen, Incidence of tumours Significancea Comments

strain Animals/group at start
(sex)
Duration
Reference
Mouse, 0 (control), 0.25, 0.5, or 1.0g/kg bw Hepatocellular carcinoma: *P=0.007 Extract purity, 98.04% (HPLC/UV profiles),
B6C3F1 by gavage in corn oil, 5days/wk for 3/50, 13/50*, 8/50, 8/50 (F) containing 27% kavalactones
(M,F) 105wk Hepatocellular adenoma or carcinoma *P=0.015 Mean body weight of females at 1.0g/kg bw was
105wk 50 M and 50 F/group (age, 56 wk) (combined): **P=0.036 11% less than that in the vehicle-control group
NTP 38/50, 39/50, 39/50, 40/50 (M); after wk21
(2012), 10/50, 21/50*, 20/50**, 13/50 (F)
Behl et al. Hepatoblastoma: *P<0.001
(2011) 0/50*, 4/50, 9/50**, 12/50*** (M); (trend)
0/50, 0/50, 1/50, 0/50 (F) **P=0.002
***P<0.001
Hepatocellular carcinoma or *P<0.05
hepatoblastoma (combined): **P=0.007
20/50, 21/50, 30/50*, 25/50 (M);
3/50, 13/50**, 9/50, 8/50 (F)
Rat, F344 0 (control), 0.1, 0.3, or 1.0g/kg bw by Testis interstitial (Leydig) cell adenomab: *P=0.003 Extract purity, 98.04% (HPLC/UV profiles),
(M,F) gavage in corn oil, 5days/wk for 104 37/49 (76%)*, 44/50 (88%), 49/50 (98%)**, (trend) containing 27% kavalactones
104 (M) or 105 (F) wk 46/50 (92%)***(M) **P=0.002 Mean body weight of group at 1.0g/kg bw was
105wk 49 or 50 M, and 50 F (age, 67 wk) ***P<0.001 10% less than that of the vehicle-control group
NTP after wk65 (M) and wk41 (F)
(2012), No significant increase in the incidence of any
Behl et al. neoplasm in females
(2011)
a Poly-3 test
b Historical incidence in 2-year studies with administration by gavage with corn oil vehicle-control group (meanstandard deviation): 176/199 (88.4%8.6%), range 7694%; all
routes: 1053/1298 (81.1%13.4%), range 5498%
bw, body weight; F, female; HPLC/UV, high-performance liquid chromatography/ultraviolet; M, male; wk, week
127
Kava
IARC MONOGRAPHS 108
In males, the incidence of testis interstitial the main metabolic pathways (Kppel & Tenczer,
(Leydig) cell adenoma was significantly higher in 1991).
the groups at the intermediate or highest dose, Zou et al. (2005) identified a pyrone ring-
and had a significant positive trend. [The inci- opened product, 6-phenyl-3-hexen-2-one, a
dence of this tumour in controls was low (76%) proposed metabolite of kava, as its mercapturic
compared with that in historical controls (corn acid adduct, in urinary samples from two kava
oil vehicle controls: range, 7694%; all routes: drinkers. This metabolite was possibly formed
range, 5498%).] The incidence of renal pelvis from enzymatic demethylation of 7,8-dihy-
transitional cell hyperplasia was significantly dromethysticin, followed by ring opening of the
higher in the group receiving the highest dose. In -pyrone ring, and rearrangement (Zou et al.,
females, the incidence of renal pelvis transitional 2005).
cell hyperplasia was significantly higher in the 11,12-D i hyd roxy-7,8-dihydrokavain-o-
groups at the highest or intermediate dose. There quinone and 11,12-dihydroxykavain-o-quinone,
was no significant increase in the incidence of two electrophilic metabolites, were identified
any neoplasm in females (Behl et al., 2011; NTP, as glutathione conjugates when kava extract
2012). was incubated with human liver microsomes.
The glucuronic acid and sulfate conjugates of
these two urinary metabolites were detected in
4. Mechanistic and Other a human volunteer who ingested a single dose
Relevant Data of a dietary supplement containing kava extract
(about 90 mg of kavalactones) (Johnson et al.,
2003).
metabolism, and excretion 4.1.2 Experimental systems
4.1.1 Humans (a) Absorption, distribution, and excretion
The metabolism of kava and individual kava- Few studies have been published on the
lactones has been studied in humans (Duffield absorption, distribution, and excretion of the
et al., 1989; Kppel & Tenczer, 1991; Johnson constituents of kava (kavain, dihydrokavain,
et al., 2003; Zou et al., 2005). Demethylation and methysticin, dihydromethysticin, yangonin,
hydroxylation products were found in human desmethoxyyangonin, and dihydroyangonin).
urine after ingestion of kava extract (Duffield Kavain is rapidly absorbed from the gastro-
et al., 1989), or its constituent kavain. The metab- intestinal tract, distributed to tissues, and
olites were mainly excreted as conjugates (Kppel eliminated.
& Tenczer, 1991). In male F344 rats given kavain at a single
Ten urinary metabolites were identified when oral dose of 100 mg/kg bw, the maximum blood
kavain was given as a therapeutic oral dose of concentration of kavain was measured at 0.88
200 mg to five healthy volunteers. The struc- hours, after which plasma concentrations declined
tures of kavain and its metabolites are shown in with a mean terminal half-life of 1.3hours. The
Fig.4.1. The major metabolite was a hydroxydi- mean oral bioavailability of kavain in F344 rats
hydrokavain. Hydroxylation of the phenyl ring, was about 50% (Mathews et al., 2005).
reduction of the 7,8 double bond, hydroxylation In male F344 rats given kavain orally for
of the lactone ring with subsequent dehydration, 7 days, kavain was primarily excreted in the
and opening of the lactone ring appeared to be urine, with about 77% recovered during the
128
Kava
Fig.4.1 Structures of kavain and its eight identified metabolites
OCH 3
COOH OH O
O O CH 3
OH
Kavain m-Hydroxybenzoic acid 4-Hydroxy-6-phenyl-5-hexen-2-one
OCH 3
O OH O
HO
N COOH CH 3 O O
H
HO
Hippuric acid 4-Hydroxy-6-hydroxyphenyl-5-hexen-2-one 12-Hydroxydihydrokavain
OCH 3 OCH 3 OCH 3
HO
O O O O O O
HO HO
x-Hydroxykavain 12-Hydroxykavain 12-Hydroxy-5,6-dehydrokavain
72 hours after the last dose. Faecal excretion methysticin and dihydromethysticin (3045
accounted for about 14% of the administered minutes). Yangonin and desmethoxyyangonin
dose. Only 0.4% of the kavain was retained in the were poorly absorbed, and rapid elimination
tissues, and kavain did not accumulate preferen- occurred (Meyer, 1967; Robinson et al., 2009).
tially in any particular tissue. In addition, there In male F344 rats given an intravenous injec-
were no differences in the pharmacokinetics of tion of kavain at a dose of 7 mg/kg bw, kavain was
kavain when administered as a single dose or as rapidly eliminated from the systemic circulation,
repeated doses (Mathews et al., 2005). with a terminal half-life of 0.63 hours. Systemic
Oral absorption of kavain, dihydrokavain, clearance and volume of distribution were 89 mL/
methysticin, dihydromethysticin, yangonin, minutes per kg and 2.70 L/kg, respectively, indi-
and desmethoxyyangonin was investigated in cating that a significant amount of kavain was
mice (Meyer, 1967). Kavain and dihydrokavain rapidly distributed out of the plasma into tissues
were rapidly absorbed from the gastrointestinal and quickly cleared from the body (Mathews
tract (with a peak at 10 minutes), followed by et al., 2005).
129
IARC MONOGRAPHS 108
Keledjian et al. (1988) observed a peak With kavain, a total of 10 metabolites were
concentration at 5 minutes in brain for kavain formed in very small amounts. Eight were deter-
and 7,8-dihydrokavain; the compounds were mined structurally and two remained uniden-
rapidly eliminated after intraperitoneal admin- tified. Both hydroxylated and ring-opened
istration (100 mg/kg bw) of individual kava products were formed (Fig.4.1).
constituents in male Balb/c mice. The maximum With methysticin, only small amounts of
concentrations of kavain and 7,8-dihydroka- two metabolites (11,12-dihydroxykavain and
vain were 64.7 and 29.3 ng/mg wet brain tissue, 11,12-dihydroxydihydrokavain) were formed by
respectively. The maximum concentrations of demethylenation of the methylenedioxyphenyl
desmethoxyyangonin and yangonin were 10.4 moiety (Fig.4.3).
and 1.2 ng/mg wet brain tissue, lower than Metabolites of yangonin and dihydroyan-
those of kavain or 7,8-dihydrokavain. When gonin were formed via O-demethylation. No
kava extract was given intraperitoneally to male ring-opened products were detected (Fig.4.4 and
Balb/c mice, the maximum concentrations of Fig.4.5).
kavain and yangonin increased in the brain,
while the concentrations of 7,8-dihydrokavain 4.1.3 Effects on drug-metabolizing enzymes
and desmethoxyyangonin were similar to those
measured after they were injected separately Studies in vivo and in vitro have shown
(Keledjian et al., 1988). that kava extract and its constituents altered
drug-metabolizing enzymes. Table 4.1 lists the
(b) Metabolism major enzymes affected by kava.
Rasmussen et al. (1979) investigated the
metabolism of five kavalactones (kavain, 4.2 Genetic and related effects
dihydrokavain, methysticin, yangonin, and
dihydroyangonin) in male albino rats. The indi- 4.2.1 Humans
vidual kavalactones were administered orally No data were available to the Working Group.
(400 mg/kg bw) or intraperitoneally (100 mg/kg
bw), the metabolites and the recovered parent 4.2.2 Experimental systems
substrate in the urine were then identified.
Kavalactones were metabolized to several prod- See Table4.2
ucts via demethylation, mono- and dihydroxyl-
ation, and reduction and pyrone ring-opening (a) Mutagenicity
(Rasmussen et al., 1979; NTP, 2012). Kava extract (up to 10000g/plate) was not
With 7,8-dihydrokavain, large amounts of the mutagenic in Salmonella typhimurium strains
parental compound were found in the urine. Nine TA97, TA98, TA1535 or TA100, or Escherichia
metabolites were identified, with 12-hydroxydi- coli strain WP2 uvrA pKM101 with or without
hydrokavain being the most abundant. About metabolic activation (rat liver S9) (NTP, 2012). In
two thirds of the metabolites were hydroxylated one trial out of three of which two of the results
forms, and one third of the metabolites were were negative, kava extract tested equivocal in
formed by scission of the 5,6-dihydro--pyrone TA97 with metabolic activation (NTP, 2012).
ring (Fig.4.2). The proposed metabolic pathways Kava extracts were not mutagenic in an assay in
for 7,8-dihydrokavain are depicted in Fig. 4.2 L5178Y mouse lymphoma cells (Whittaker et al.,
(adapted from NTP (2012). 2008).
130
Kava
Fig.4.2 The proposed metabolic pathways for 7,8-dihydrokavain

OCH 3 OCH 3
OH OH
HO
O O O O
8-Hydroxydihydrokavain 8,x-Dihydroxydihydrokavain
OCH 3 OCH 3 OCH 3
HO HO
O O O O O O
HO
7,8-Dihydrokavain x-Hydroxydihydrokavain x,y-Dihydroxydihydrokavain
OCH 3 OCH 3
O O O O
HO HO
OH
12-Hydroxydihydrokavain 11,12-Dihydroxydihydrokavain
OH OCH 3 OH OCH 3
COOH COOH
HO
OH O OH O
CH 3 CH 3
HO
4-Hydroxy-6-phenylhexan-2-one 4-Hydroxy-6-(p-hydroxyphenyl)hexan-2-one
N COOH
H
Hippuric acid
Compiled by the Working Group using data from NTP (2012).
131
IARC MONOGRAPHS 108
Fig.4.3 Structures of methysticin and its two metabolites
OCH 3 OCH 3 OCH 3
O O O O O O
O HO HO
O OH OH
Methysticin 11,12-Dihydroxykavain 11,12-Dihydroxydihydrokavain
Compiled by the Working Group using data from Fu et al. (2008)
Fig.4.4 Structures of 7,8-dihydroyangonin and its three metabolites

OCH 3 OCH 3 OCH 3 OCH 3
OH
HO
O O O O O O O O
CH3O HO HO HO
7,8-Dihydroyangonin 12-Hydroxy-5,6-dehydro- Dihydroxy-5,6-dehydro-7,8- Dihydroxy-5,6-dehydro-7,8-

7,8-dihydrokavain dihydrokavain I dihydrokavain II
I II III
For the metabolites II and III the positioning of the second hydroxyl group (m, o or at C8) are uncertain.
Fig.4.5 Structures of yangonin and its three metabolites
OCH 3 OCH 3 OCH 3 OCH 3
HO
O O O O O O O O
CH 3 O HO HO HO
Yangonin 12-Hydroxy-5,6-dehydrokavain x,12-Dihydroxy-7,8- 12-Hydroxy-kavain

dihydro-5,6-dehydrokavain
132
Table 4.1 Effects of kava extract and kavalactones on metabolizing enzymes
Reference Species Kava preparation Dose Duration Detection method Result

or cell type of
treatment
In vivo
Russmann Human Aqueous extract 727g of kavalactones 6yr Substrate turnover CYP1A2 (inhibition)
et al. (2005) per wk, oral
Guo et al. Mouse Methanolic and aqueous extracts 0.1252.0g/kg bw per 98days Gene expression CYP4A10 (inhibition);
(2010) day, 5days/wk, gavage CYP2A5, CYP2B20,
CYP2C55, GSTA1, GSTA2
(induction)
Guo et al. Rat Methanolic and aqueous extracts 0.1252.0g/kg bw per 98days Gene expression CYP3A13, CYP17A1,
(2009) day, 5days/wk, gavage ABCB9 (inhibition);
CYP1A1, CYP1A2,
CYP3A1, CYP3A3,
ABCC3, NQO1, UGT1A6
(induction)
Yamazaki Rat Kava extracts (not specified) Kava extract 8days Rat liver microsomes/ CYP1A1, CYP1A2
et al. (2008) (kavalactones, enzyme assay, gene (induction)
380mg/kg bw per expression, protein
day), gavage expression
Lim et al. Rat Acetone kava leave extract 100mg/kg/ bw per 14days Protein expression CYP1A2, CYP2E1
(2007) day, gavage (induction)
Clayton Rat Methanolic and aqueous extracts 0.1252.0g/kg bw per 90days Protein expression CYP2D1 (inhibition);
et al. (2007) day, 5days/wk, gavage CYP1A2, CYP2B1,
CYP3A1 (induction)
Mathews Rat Methanol or acetone kava extract 256mg/kg bw, 1g/kg 7days Rat liver microsomes/ CYP2D1, 2C11 (inhibition);
et al. (2005) bw, gavage enzyme assay CYP1A2, 2B1, 2C6, 2D1,
3A1/2 (induction)
In vitro
Li et al. Hepa1c1c7 Methanolic and aqueous Various 24h Cell-based enzymatic CYP1A1 (induction)
(2011) extracts, methysticin, concentrations assay/enzyme assay,
7,8-dihydromethysticin to 25M of kava gene expression, protein
constituents or expression
6.25g/mL of kava
extract
Mathews Human liver Methanol or acetone kava extract, 1, 10, 100M 10min Recombinant protein/ CYP2C9, 2C19, 2D6, 3A4
et al. (2005) microsomes yangonin, dihydrokavain, enzyme assay (inhibition); P-glycoprotein
methysticin, dihydromethysticin, ATPase (inhibition)
composite kavalactones
133
Kava
134
Reference Species Kava preparation Dose Duration Detection method Result

or cell type of
treatment
Weiss et al. P388 and Methanol aqueous Various NR Calcein uptake ABCB1 (inhibition)
(2005) P388/dx cell kava extracts, kavain, concentrations (to
lines dihydrokavain, methysticin, the highest soluble
dihydromethysticin, yangonin, concentration)
and desmethoxyyangonin
IARC MONOGRAPHS 108
Zou et al. Baculovirus/ Ethanol extract, methysticin, Various 15 Recombinant enzyme/ CYP1A2, 2C9, 2C19, 2E1,
(2004) insect cell desmethoxyyangonin, yangonin concentrations up to 45min enzyme assay 3A4 (inhibition)
system and 100M
cryopreserved
human
hepatocytes
Ct et al. Human liver Methanol, acetone, ethanol or Various 5min Enzyme assay CYP3A4, CYP1A2,
(2004) microsomes/ aqueous kava extracts concentrations up to CYP2C9, CYP2C19
200g/mL (inhibition)
Raucy (2003) Primary Kava extract (not specified) 100g/mL 48h Gene expression, report CYP3A4 (induction)
human gene assay
hepatocytes
and HepG2
Mathews Human liver Methanol or acetone extract, Kava extract 15min Enzyme assay CYP1A2, CYP2C9,
et al. (2002) microsomes/ desmethoxyyangonin, normalized to 100M CYP2C19, CYP2D6,
methysticin, dihydromethysticin kavalactones CYP3A4, CYP4A9/11
(induction)
Unger et al. Baculovirus/ Methanol, acetone and ethyl 1100mg/mL 30min Recombinant enzyme/ CYP3A4 (inhibition)
(2002) insect cell acetate extracts enzyme assay
system
Zou et al. cDNA human Methysticin, Various 30 Recombinant enzyme/ CYP1A2, CYP2C9,
(2002) CYP isoforms desmethoxyyangonin, concentrations up to 45min enzyme assay CYP2C19, CYP2D6,
dihydromethysticin, kavain, ~200 M CYP3A4 (inhibition)
dihydrokavain
ABC, ATP-binding cassette; CYP, cytochrome; GST, glutathione-S-transferase; min, minute; NR, not reported; NQO, NAD(P)H quinone oxidoreductase; UGT, UDP
glycosyltransferase; wk, week; yr, year
Kava
Table 4.2 Genetic and related effects of kava extract
Test system Resultsa Concentration or Reference

dose
Without exogenous With exogenous (LED or HID)
metabolic system metabolic system
Salmonella typhimurium TA100, TA98, 10000g/plate NTP (2012)
TA1535, reverse mutation
Salmonella typhimurium TA97, reverse Equivocal in 1 out 10000g/plate NTP (2012)
mutation of 3 tests
L5178Y mouse lymphoma mutation NR 300g/mL Whittaker et
assay al. (2008)
umu point mutation assay +a +a 2330g/mL extract Jhoo et al.
300M pure (2007)
kavalactones
Escherichia coli WP2 uvrA/pKM101, 10000g/plate NTP (2012)
reverse mutation
Micronucleus induction in peripheral NT Up to 2.0g/kg bw NTP (2012)
blood erythrocytes of male and female per day by gavage for
B6C3F1 mice in vivo 3months
+, positive; (+), weakly positive; , negative; HID, highest ineffective dose; LED, lowest effective dose; NT, not tested; NR, not reported
a These data were not analysed statistically.
The only report of positive mutagenic activity (Russmann et al., 2001; Bujanda et al., 2002;
with kava extracts (two positive results, but six Campo et al., 2002; Brauer et al., 2003; Gow et al.,
negative results) involved the umu point muta- 2003; Humberston et al., 2003; Stickel et al., 2003;
tion assay (Jhoo et al., 2007). [The Working Teschke et al., 2003, 2008; Thomsen et al., 2004).
Group noted that these data were not analysed
statistically.]
4.4 Susceptibility
(b) Chromosomal damage
Genetic polymorphisms
In male or female mice given kava extract at
a dose of up to 2.0 g/kg bw per day by gavage for Deficiency in CYP2D6, the major kavalac-
3months, there was no increase in the frequency tone-metabolizing enzyme, was detected in two
of micronucleated normochromatic or polychro- patients with liver failure (Russmann et al., 2001).
matic erythrocytes in blood (NTP, 2012). Genetic polymorphism of CYP2D6 has a prev-
alence of 79% in Caucasian populations, but
<1% in Polynesian populations (Wanwimolruk
4.3 Other mechanistic data relevant et al., 1998; Ingelman-Sundberg, 2005). Severe
to carcinogenesis liver failure has not been observed in people
using kava in the traditional way in islands in
Effects on hepatic cell physiology the South Pacific (Moulds & Malani, 2003; Anke
Case reports of liver injury associated with & Ramzan, 2004).
kava intake have been described. The types
of liver damage reported include fulminant
hepatitis, necrosis, cirrhosis, and liver failure
requiring liver transplantation or causing death
135
IARC MONOGRAPHS 108
4.5 Mechanistic considerations significance were reported. The Working Group

regarded the study as uninformative because the
Kava extract is not mutagenic based on the ecological design provided only weak support for
results of numerous studies of genotoxicity, causal inference at the individual level, the meas-
including tests for mutagenicity in bacteria, ures of exposure and outcome were crude, and
induction of micronuclei in vivo (NTP, 2012), the role of chance was not evaluated.
and the mouse lymphoma assay (Whittaker et al.,
2008). The reported carcinogenicity in mice is
most probably mediated through nongenotoxic 5.3 Animal carcinogenicity data
mechanisms. A kava extract was tested for carcinogenicity
in one study in mice and one study in rats treated
by gavage. In mice, the extract caused a signifi-
5. Summary of Data Reported cant increase in the incidence of hepatoblastoma
in males, and of hepatocellular adenoma or
5.1 Exposure data carcinoma (combined), and hepatocellular carci-
noma, in females. In male rats, the same extract
The kava (or kava kava) plant Piper caused a significant increase in the incidence of
methysticum is a perennial tropical shrub that testis interstitial (Leydig) cell adenoma; however,
is widely cultivated in Oceania. The rhizome the incidence in controls was low compared with
of the plant was originally used as an ingre- that in historical controls. There was no signifi-
dient in local traditional drinks with psychop- cant increase in the incidence of any neoplasm
harmacological properties, and as traditional in female rats.
folk medicine. More recently, rhizome extracts
have been used in medicinal products, food or
dietary supplements, and cosmetics. Important 5.4 Mechanistic and other relevant
chemical constituents of the resin contained in data
the kava rhizome are kavalactones, kavain being
the major compound. The medicinal uses of kava The major components of kava extract, kava-
supported by clinical data are short-term symp- lactones, are extensively metabolized in humans
tomatic treatment of mild states of anxiety or and experimental animals. Among the numerous
insomnia due to nervousness, stress, or tension. metabolites are products from demethylation,
Use of kava was popular worldwide, but several hydroxylation, and ring-opening.
case reports of liver damage associated with Kava extract gave negative results in several
exposure to kava reduced sales, and caused kava standard bacterial assays for mutation in the
to be banned in several countries. absence or presence of exogenous metabolic
activation. Kavalactones gave negative results in
most of these assays.
5.2 Human carcinogenicity data The reported carcinogenicity of kava in mice
is most likely to be mediated through a nongen-
The Working Group was able to identify
otoxic mechanism.
only one epidemiological study of cancer and
kava consumption. This ecological study found
an inverse correlation between all cancers in
men and a proxy measure of kava consumption,
but no confidence intervals or test of statistical
136
Kava
6. Evaluation Brauer RB, Stangl M, Stewart JR, Pfab R, Becker K (2003).

Acute liver failure after administration of herbal
tranquilizer kava-kava (Piper methysticum). J Clin
6.1 Cancer in humans Psychiatry, 64(2):2168. doi:10.4088/JCP.v64n0215c
PMID:12633134
There is inadequate evidence in humans for Bujanda L, Palacios A, Silvario R, Snchez A, Muoz
C (2002). [Kava-induced acute icteric hepatitis]
the carcinogenicity of kava extract. Gastroenterol Hepatol, 25(6):4345. doi:10.1016/S0210-
5705(02)70281-1 PMID:12069710
Cairney S, Maruff P, Clough AR (2002). The neurobehav-
6.2 Cancer in experimental animals ioural effects of kava. Aust N Z J Psychiatry, 36(5):65762.
doi:10.1046/j.1440-1614.2002.01027.x PMID:12225450
There is sufficient evidence in experimental Campo JV, McNabb J, Perel JM, Mazariegos GV,
animals for the carcinogenicity of kava extract. Hasegawa SL, Reyes J (2002). Kava-induced fulminant
hepatic failure. J Am Acad Child Adolesc Psychiatry,
41(6):6312. doi:10.1097/00004583-200206000-00001
6.3 Overall evaluation PMID:12049436
Clayton NP, Yoshizawa K, Kissling GE, Burka LT, Chan
PC, Nyska A (2007). Immunohistochemical analysis of
Kava extract is possibly carcinogenic to expressions of hepatic cytochrome P450 in F344 rats
humans (Group 2B). following oral treatment with kava extract. Exp Toxicol
Pathol, 58(4):22336. doi:10.1016/j.etp.2006.08.002
PMID:17059882
Clough A (2003). Enough! or too much. What is excessive
References kava use in Arnhem Land? Drug Alcohol Rev, 22(1):43
51. doi:10.1080/0959523021000059820 PMID:12745358
Anke J & Ramzan I (2004). Kava hepatotoxicity: Are Clough AR, Bailie RS, Currie B (2003). Liver function test
we any closer to the truth? Planta Med, 70(3):1936. abnormalities in users of aqueous kava extracts. J Toxicol
doi:10.1055/s-2004-815533 PMID:15114493 Clin Toxicol, 41(6):8219. doi:10.1081/CLT-120025347
Anonymous (2000). Kava kava rhizome (root). In: PMID:14677792
Blumenthal M, Goldberg A, Brinckmann J, editors. Clough AR, Burns CB, Mununggurr N (2000). Kava
Expanded Commission E Monographs. Herb in Arnhem Land: a review of consumption and its
Monographs, based on those created by a special social correlates. Drug Alcohol Rev, 19(3):31928.
expert committee of the German Federal Institute for doi:10.1080/713659370
Drugs and Medicinal Devices. Newton (MA), USA: Ct CS, Kor C, Cohen J, Auclair K (2004). Composition and
Integrative Medicine Communications biological activity of traditional and commercial kava
Behl M, Nyska A, Chhabra RS, Travlos GS, Fomby LM, extracts. Biochem Biophys Res Commun, 322(1):14752.
Sparrow BR et al. (2011). Liver toxicity and carcino- doi:10.1016/j.bbrc.2004.07.093 PMID:15313185
genicity in F344/N rats and B6C3F1 mice exposed De Smet PA (2002). Safety concerns about kava not
to Kava Kava. Food Chem Toxicol, 49(11):28209. unique. Lancet, 360(9342):1336 doi:10.1016/S0140-
doi:10.1016/j.fct.2011.07.067 PMID:21871523 6736(02)11347-X PMID:12414243
Bilia AR, Bergonzi MC, Lazari D, Vincieri FF (2002). Dentali SJ (1997). Herb safety review: Kava: Piper
Characterization of commercial kava-kava herbal methysticum Forster F. (Piperaceae). Boulder (CO),
drug and herbal drug preparations by means of nuclear USA: Herb Research Foundation.
magnetic resonance spectroscopy. J Agric Food Chem, Dharmaratne HR, Nanayakkara NP, Khan IA (2002).
50(18):501625. doi:10.1021/jf020049j PMID:12188601 Kavalactones from Piper methysticum, and their
Bilia AR, Gallon S, Vincieri FF (2002b). Kava-kava and 13C NMR spectroscopic analyses. Phytochemistry,
anxiety: growing knowledge about the efficacy and 59(4):42933. doi:10.1016/S0031-9422(01)00443-5
safety. Life Sci, 70(22):258197. doi:10.1016/S0024- PMID:11830162
3205(02)01555-2 PMID:12269386 Duffield AM, Jamieson DD, Lidgard RO, Duffield PH,
Bilia AR, Scalise L, Bergonzi MC, Vincieri FF (2004). Bourne DJ (1989). Identification of some human
Analysis of kavalactones from Piper methysticum urinary metabolites of the intoxicating beverage kava.
(kava-kava). J Chromatogr B Analyt Technol Biomed Life J Chromatogr A, 475(2):27381. doi:10.1016/S0021-
Sci, 812(1-2):20314. doi:10.1016/j.jchromb.2004.07.038 9673(01)89682-5 PMID:2777959
PMID:15556499 FDA (2002). Kava-containing dietary supplements may be
associated with severe liver injury. US Food and Drug
137
IARC MONOGRAPHS 108
Administration. Available at: http://www.fda.gov/ A, 562(1-2):20711. doi:10.1016/0378-4347(91)80578-Z

food/resourcesforyou/consumers/ucm085482.htm, PMID:2026693
accessed 18/04/2012. Lebot V, Merlin M, Lindstrom L (1997). Kava-the Pacific
Fu PP, Xia Q, Guo L, Yu H, Chan PC (2008). Toxicity of kava elixir: the definitive guide to its ethnobotany, history,
kava. J Environ Sci Health C Environ Carcinog Ecotoxicol and chemistry. Rochester (VT), USA: Healing Arts
Rev, 26(1):89112. doi:10.1080/10590500801907407 Press.
PMID:18322868 Li Y, Mei H, Wu Q, Zhang S, Fang JL, Shi L etal. (2011).
Gaub M, Roeseler Ch, Roos G, Kovar KA (2004). Analysis Methysticin and 7,8-dihydromethysticin are two
of plant extracts by NIRS: simultaneous determina- major kavalactones in kava extract to induce CYP1A1.
tion of kavapyrones and water in dry extracts of Piper Toxicol Sci, 124(2):38899. doi:10.1093/toxsci/kfr235
methysticum Forst. J Pharm Biomed Anal, 36(4):859 PMID:21908763
64. doi:10.1016/j.jpba.2004.06.030 PMID:15533680 Lim ST, Dragull K, Tang CS, Bittenbender HC, Efird JT,
Gow PJ, Connelly NJ, Hill RL, Crowley P, Angus PW Nerurkar PV (2007). Effects of kava alkaloid, piper-
(2003). Fatal fulminant hepatic failure induced methystine, and kavalactones on oxidative stress and
by a natural therapy containing kava. Med J Aust, cytochrome P450 in F-344 rats. Toxicol Sci, 97(1):214
178(9):4423. PMID:12720510 21. doi:10.1093/toxsci/kfm035 PMID:17329236
Guo L, Li Q, Xia Q, Dial S, Chan PC, Fu P (2009). Analysis Lindstrom L (2004). History, folklore, traditional and
of gene expression changes of drug metabolizing current uses of kava. In: Singh YN, editor. Kava: from
enzymes in the livers of F344 rats following oral treat- ethnology to pharmacology. Boca Raton (FL), USA:
ment with kava extract. Food Chem Toxicol, 47(2):433 CRC Press; pp. 1028.
42. doi:10.1016/j.fct.2008.11.037 PMID:19100306 Mathews JM, Etheridge AS, Black SR (2002). Inhibition
Guo L, Shi Q, Dial S, Xia Q, Mei N, Li QZ etal. (2010). of human cytochrome P450 activities by kava extract
Gene expression profiling in male B6C3F1 mouse and kavalactones. Drug Metab Dispos, 30(11):11537.
livers exposed to kava identifieschanges in drug doi:10.1124/dmd.30.11.1153 PMID:12386118
metabolizing genes and potential mechanisms linked Mathews JM, Etheridge AS, Valentine JL, Black SR,
to kava toxicity. Food Chem Toxicol, 48(2):68696. Coleman DP, Patel P et al. (2005). Pharmacokinetics
doi:10.1016/j.fct.2009.11.050 PMID:19948201 and disposition of the kavalactone kawain: interac-
IMS Health (2012) Multinational Integrated Data Analysis tion with kava extract and kavalactones in vivo and in
(MIDAS). Plymouth Meeting, 2012, Pennsylvania: IMS vitro. Drug Metab Dispos, 33(10):155563. doi:10.1124/
Health. dmd.105.004317 PMID:16033948
Humberston CL, Akhtar J, Krenzelok EP (2003). McDonald D & Jowitt A (2000). Kava in the Pacific
Acute hepatitis induced by kava kava. J Toxicol Clin Islands: a contemporary drug of abuse? Drug Alcohol
Toxicol, 41(2):10913. doi:10.1081/CLT-120019123 Rev, 19(2):21727. doi:10.1080/713659319
PMID:12733846 Meyer HJ (1967). Pharmacology of kava. 1. Psychopharmacol
Ingelman-Sundberg M (2005). Genetic polymorphisms Bull, 4(3):101. PMID:5616309
of cytochrome P450 2D6 (CYP2D6): clinical conse- Monakhova YB, Kuballa T, Lbell-Behrends S, Maixner
quences, evolutionary aspects and functional diver- S, Kohl-Himmelseher M, Ruge W et al. (2013).
sity. Pharmacogenomics J, 5(1):613. doi:10.1038/ Standardless 1H NMR determination of pharmaco-
sj.tpj.6500285 PMID:15492763 logically active substances in dietary supplements
Jhoo JW, Ang CY, Heinze TM, Deck J, Schnackenberg LK, and medicines that have been illegally traded over
Beger RD et al. (2007). Identification of C-glycoside the internet. Drug Test Anal, 5(6):40011. doi:10.1002/
flavonoids as potential mutagenic compounds in dta.1367 PMID:22550015
kava. J Food Sci, 72(2):C1205. doi:10.1111/j.1750- Morgan M, Bone K, Mills S et al. (2005). Kava. Safety
3841.2007.00278.x PMID:17995826 monograph. In: Mills S, Bone K, editors. The essential
Johnson BM, Qiu SX, Zhang S, Zhang F, Burdette JE, guide to herbal safety. St. Louis (MO), USA: Elsevier
Yu L etal. (2003). Identification of novel electrophilic Churchill Livingstone; pp. 484492.
metabolites of Piper methysticum Forst (kava). Chem Morris CA & Avorn J (2003). Internet marketing of
Res Toxicol, 16(6):73340. doi:10.1021/tx020113r herbal products. JAMA, 290(11):15059. doi:10.1001/
PMID:12807356 jama.290.11.1505 PMID:13129992
Keledjian J, Duffield PH, Jamieson DD, Lidgard RO, Moulds RF & Malani J (2003). Kava: herbal panacea or
Duffield AM (1988). Uptake into mouse brain of four liver poison? Med J Aust, 178(9):4513. PMID:12720513
compounds present in the psychoactive beverage kava. NTP (2012). Toxicology and carcinogenesis studies of
J Pharm Sci, 77(12):10036. doi:10.1002/jps.2600771203 kava kava extract (CAS No. 9000388) in F344/N
PMID:3244102 rats and B6C3F1 mice (gavage studies). Natl Toxicol
Kppel C & Tenczer J (1991). Mass spectral characteriza- Program Tech Rep Ser, 571(571):1186. PMID:22441424
tion of urinary metabolites of D,L-kawain. J Chromatogr
138
Kava
NLM (2012). Products that contain active ingredient Clin Pharmacol Ther, 77(5):4534. doi:10.1016/j.
- Kava Kava. Dietary supplements labels database. clpt.2005.01.021 PMID:15900292
United States National Library of Medicine. Available Russmann S, Lauterburg BH, Helbling A (2001).
from: http://www.dsld.nlm.nih.gov/dsld/rptQSearch. Kava hepatotoxicity. Ann Intern Med, 135(1):689.
jsp?item=Kava+Kava&db=adsld, accessed 7 July 2014. doi:10.7326/0 0 03-4819-135-1-20 0107030- 0 0 036
Norton SA & Ruze P (1994). Kava dermopathy. J Am PMID:11434754
Acad Dermatol, 31(1):8997. doi:10.1016/S0190- Rychetnik L & Madronio CM (2011). The health and social
9622(94)70142-3 PMID:8021378 effects of drinking water-based infusions of kava: a
Nutrition Business Journal (2010). NBJs Supplement review of the evidence. Drug Alcohol Rev, 30(1):7483.
Business Report. An analysis of markets, trends, doi:10.1111/j.1465-3362.2010.00184.x PMID:21219501
competition and strategy in the U.S. dietary supple- Schfer K & Winterhalter P (2005). Application of high
ment industry. New York (NY), USA: Penton Media, speed countercurrent chromatography (HSCCC) to the
Inc. isolation of kavalactones. J Liquid Chromatogr Relat
Nutrition Business Journal (2012). NBJs Supplement Technol, 28(11):170316. doi:10.1081/JLC-200060451
Business Report. An analysis of markets, trends, Schmidt M, Morgan M, Bone K et al. (2005). Kava: a
competition and strategy in the U.S. dietary supple- risk-benefit assessment. In: Mills S, Bone K, editors.
ment industry. . New York (NY), USA: Penton Media, The essential guide to herbal safety. St. Louis (MO),
Inc. Available from: http://newhope360.com/2012- USA: Elsevier Churchill Livingstone; pp. 155221.
supplement-business-report; accessed 4 September SciFinder (2013). CAS Registry Number 9000-38-8
2014. (accessed: 15/01/2013). Columbus, Ohio, USA: Chemical
ONeil MJ, Heckelman PE, Koch CB et al. (2006). The Abstracts Service, American Chemical Society.
Merck Index - An Encyclopedia of Chemicals, Drugs, Shulgin AT (1973). The narcotic pepper - the chemistry
and Biologicals. 14th ed. Version 14.6. Whitehouse and pharmacology of Piper methysticum and related
Station (NJ), USA: Merck & Co., Inc. species. Bull Narc, 25:5974.
Parkman CA (2002). Another FDA warning: Kava Singh YN (1992). Kava: an overview. J Ethnopharmacol,
supplements. Case Manager, 13(4):268. doi:10.1067/ 37(1):1345. doi:10.1016/0378-8741(92)90003-A
mcm.2002.126437 PMID:12131903 PMID:1453702
Ramzan I, Tran VH (2004). Chemistry of kava and kava- Singh YN (2004a). An introduction to Kava Piper
lactones. In: Singh YN, editor. Kava: from ethnology methysticum. In: Singh YN, editor. Kava: from
to pharmacology. Boca Raton (FL), USA: CRC Press; ethnology to pharmacology. Boca Raton (FL), USA:
pp. 76103. CRC Press; pp. 19.
Rasmussen AK, Scheline RR, Solheim E, Hnsel R Singh YN (2004b). Kava: production, marketing and
(1979). Metabolism of some kava pyrones in the rat. quality assurance. In: Singh YN, editor. Kava: from
Xenobiotica, 9(1):116. doi:10.3109/00498257909034699 ethnology to pharmacology. Boca Raton (FL), USA:
PMID:760318 CRC Press; pp. 2949.
Raucy JL (2003). Regulation of CYP3A4 expression in Spohn R (2013). Water colour of Piper methysticum.
human hepatocytes by pharmaceuticals and natural Available from: http://www.spohns.de/
products. Drug Metab Dispos, 31(5):5339. doi:10.1124/ heilpflanzenillus/pipermethysticum.html, accessed 13
dmd.31.5.533 PMID:12695340 March 2013.
Robinson V, Bergfeld WF, Belsito DV, Klaassen CD, Steiner GG (2000). The correlation between cancer
Marks JG Jr, Shank RC et al.; Cosmetic Ingredient incidence and kava consumption. Hawaii Med J,
Review Expert Panel (2009). Final report on the safety 59(11):4202. PMID:11149250
assessment of Piper methysticum leaf/root/stem extract Stickel F, Baumller HM, Seitz K, Vasilakis D, Seitz G,
and Piper methysticum root extract. Int J Toxicol, Seitz HK et al. (2003). Hepatitis induced by Kava
28(6):Suppl: 175S88S. doi:10.1177/1091581809350934 (Piper methysticum rhizoma). J Hepatol, 39(1):627.
PMID:19966149 doi:10.1016/S0168-8278(03)00175-2 PMID:12821045
Russmann S, Barguil Y, Cabalion P, Kritsanida M, Tarbah F, Mahler H, Kardel B, Weinmann W, Hafner D,
Duhet D, Lauterburg BH (2003). Hepatic injury Daldrup T (2003). Kinetics of kavain and its metab-
due to traditional aqueous extracts of kava root olites after oral application. J Chromatogr B Analyt
in New Caledonia. Eur J Gastroenterol Hepatol, Technol Biomed Life Sci, 789(1):11530. doi:10.1016/
15(9):10336. doi:10.1097/00042737-200309000-00015 S1570-0232(03)00046-1 PMID:12726850
PMID:12923378 Teschke R, Gaus W, Loew D (2003). Kava extracts:
Russmann S, Lauterburg BH, Barguil Y, Choblet E, safety and risks including rare hepatotoxicity.
Cabalion P, Rentsch K etal. (2005). Traditional aqueous Phytomedicine, 10(5):4406. doi:10.1078/0944-7113-
kava extracts inhibit cytochrome P450 1A2 in humans: 00314 PMID:12834011
Protective effect against environmental carcinogens?
139
IARC MONOGRAPHS 108
Teschke R & Lebot V (2011). Proposal for a kava quality Geneva, Switzerland: World Health Organization.
standardization code. Food Chem Toxicol, 49(10):2503 Available at http://apps.who.int/medicinedocs/en/d/
16. doi:10.1016/j.fct.2011.06.075 PMID:21756963 Js4927e/23.html.
Teschke R, Qiu SX, Lebot V (2011). Herbal hepatotoxicity WHO (2007). Assessment of the risk of hepatotoxicity
by kava: update on pipermethystine, flavokavain B, and with kava products. Geneva, Switzerland: World
mould hepatotoxins as primarily assumed culprits. Dig Health Organization.
Liver Dis, 43(9):67681. doi:10.1016/j.dld.2011.01.018 Yamazaki Y, Hashida H, Arita A, Hamaguchi K, Shimura
PMID:21377431 F (2008). High dose of commercial products of kava
Teschke R, Schwarzenboeck A, Hennermann KH (2008). (Piper methysticum) markedly enhanced hepatic
Kava hepatotoxicity: a clinical survey and critical anal- cytochrome P450 1A1 mRNA expression with liver
ysis of 26 suspected cases. Eur J Gastroenterol Hepatol, enlargement in rats. Food Chem Toxicol, 46(12):37328.
20(12):118293. doi:10.1097/MEG.0b013e3283036768 doi:10.1016/j.fct.2008.09.052 PMID:18930106
PMID:18989142 Zou L, Harkey MR, Henderson GL (2002). Effects of
Thomsen M, Vitetta L, Schmidt M, Sali A (2004). Fatal herbal components on cDNA-expressed cytochrome
fulminant hepatic failure induced by a natural therapy P450 enzyme catalytic activity. Life Sci, 71(13):157989.
containing kava. Med J Aust, 180(4):1989, author reply doi:10.1016/S0024-3205(02)01913-6 PMID:12127912
199. PMID:14960147 Zou L, Harkey MR, Henderson GL (2005). Synthesis,
Trucksess M, Weaver C, Oles C, DOvidio K, Rader J in vitro, reactivity, and identification of 6-phenyl-3-
(2006). Determination of aflatoxins and ochratoxin hexen-2-one in human urine after kava-kava (Piper
A in ginseng and other botanical roots by immuno- methysticum) ingestion. Planta Med, 71(2):1426.
affinity column cleanup and liquid chromatography doi:10.1055/s-2005-837781 PMID:15729622
with fluorescence detection. J AOAC Int, 89(3):62430. Zou L, Henderson GL, Harkey MR, Sakai Y, Li A
PMID:16792061 (2004). Effects of kava (Kava-kava, Awa, Yaqona,
Ulbricht C, Basch E, Boon H, Ernst E, Hammerness P, Piper methysticum) on c-DNA-expressed
Sollars D et al. (2005). Safety review of kava (Piper cytochrome P450 enzymes and human cryopre-
methysticum) by the Natural Standard Research served hepatocytes. Phytomedicine, 11(4):28594.
Collaboration. Expert Opin Drug Saf, 4(4):77994. doi:10.1078/0944711041495263 PMID:15185840
doi:10.1517/14740338.4.4.779 PMID:16011454
Unger M, Holzgrabe U, Jacobsen W, Cummins C, Benet LZ
(2002). Inhibition of cytochrome P450 3A4 by extracts
and kavalactones of Piper methysticum (Kava-Kava).
Planta Med, 68(12):10558. doi:10.1055/s-2002-36360
PMID:12494328
Wanwimolruk S, Bhawan S, Coville PF, Chalcroft SC (1998).
Genetic polymorphism of debrisoquine (CYP2D6)
and proguanil (CYP2C19) in South Pacific Polynesian
populations. Eur J Clin Pharmacol, 54(5):4315.
doi:10.1007/s002280050488 PMID:9754989
Weaver CM & Trucksess MW (2010). Determination of
aflatoxins in botanical roots by a modification of AOAC
Official Method 991.31: single-laboratory validation. J
AOAC Int, 93(1):1849. PMID:20334179
Weiss J, Sauer A, Frank A, Unger M (2005). Extracts and
kavalactones of Piper methysticum G. Forst (kava-
kava) inhibit P-glycoprotein in vitro. Drug Metab
Dispos, 33(11):15803. doi:10.1124/dmd.105.005892
PMID:16051732
Whittaker P, Clarke JJ, San RH, Betz JM, Seifried HE, de
Jager LS etal. (2008). Evaluation of commercial kava
extracts and kavalactone standards for mutagenicity
and toxicity using the mammalian cell gene mutation
assay in L5178Y mouse lymphoma cells. Food Chem
Toxicol, 46(1):16874. doi:10.1016/j.fct.2007.07.013
PMID:17822821
WHO (2004). Rhizoma Piperis Methystici. In: WHO
monographs on selected medicinal plants. Vol. 2,
140
PULEGONE
1. Exposure Data Croteau, 2004). It is found in different varieties

of M. piperita oils at a range of 0.5% to 4.6%, and
Pulegone is a monoterpene ketone present in in M. arvensis oils at a range of 0.2% to 4.9%;
the leaves and flowering tops of several members oils in natural form contain lower concentrations
of the mint family Lamiaceae. Two enantiomeric of pulegone than those that have been partially
forms occur in nature, the R-(+)-enantiomer dementholized (Smith & Levi, 1961). Pulegone is
being the most abundant in the essential oils also found in various concentrations in Buchu
(Hayes et al., 2007; Barceloux, 2008). leaf oils (Barosma betulina and B. crenulata with
3% and 50%, respectively) (Kaiser et al., 1975).
1.1 Identification of the agent 1.1.2 Structural and molecular formulae and
1.1.1 Classification relative molecular mass
(a) Nomenclature H
Chem. Abstr. Serv. Reg. No: 89-82-7

IUPAC systematic name: (R)-5-Methyl-2-(1-
methylethylidine) cyclohexanone O
From Scifinder (2014).
(b) Description of origin C10H16O

Pulegone is a major constituent of the volatile Relative molecular mass: 152.23
oils of European pennyroyal (Mentha pulegium From Farley & Howland (1980), Thomassen
L.) and American pennyroyal (Hedeoma pulegi- et al. (1988), and Da Rocha et al. (2012).
oides L.), where it comprises 8597% (w/v) and
about 30% (w/v) of the respective oil (Guenther, 1.1.3 Chemical and physical properties of the
1949; Smith & Levi, 1961; Smith et al., 1963; Von
pure substance
Hefendehl & Ziegler, 1975; Farley & Howland,
1980). The compound is also a minor component Description: Colourless oil with a strong
of several other edible mint (Mentha) species pungent aromatic mint smell.
and their derived volatile oils, including pepper-
mint (Mentha piperita) and spearmint (Mentha Boiling point: 224C
spicata) (Virmani & Datta, 1968; Turner & Density: 0.9346 g/mL at 25C
141
IARC MONOGRAPHS 108
Optical activity ([]20D): +22 (a) Medicinal indications

Solubility: Insoluble in water (Farley & The aerial parts of both American and
Howland, 1980; ONeil, 2013; Sigma Aldrich, European pennyroyal plants have tradition-
2013). ally been used internally as a tea for non-ulcer
dyspepsia, primary dysmenorrhoea, secondary
1.1.4 Analysis amenorrhoea and oligomenorrhoea, as abortifa-
cient, and as a diaphoretic. Pennyroyal oils have
Physical properties such as density and been used for these same medicinal indications
optical rotation are used to characterize essential (Hoppe, 1975; List & Hrhammer, 1976; Foster
oils. Gas chromatography with flame-ionization & Duke, 1990; Skenderi, 2003). Today, recorded
detection has been the standard method of anal- uses for Mentha piperita and Mentha pulegium L.
ysis for essential oil composition. Petrakis et al. are for common cold, headache, and as diuretic,
(2009) have developed a direct and rapid method spasmolytic, anti-convulsive, anti-emetic, heart
to quantify pulegone using Fourier transform stimulant, and sedative (Karousou et al., 2007).
mid-infrared spectroscopy, which showed equiv- Peppermint oil is used for the treatment of the
alent results to those obtained when using gas symptoms of inflammatory bowel syndrome
chromatography. (Cappello et al., 2007).
(b) Dosage
1.2 Production and use
No typical dose was found in the literature
1.2.1 Production for pulegone. According to literature reports on
levels of pulegone present in M. piperita oils, two
According to the United Nations Commodity
capsules of 225 mg of oil taken twice per day
Trade Statistics Database (United Nations
could contain an amount of pulegone of between
Comtrade, 2013), 3.43.7 tonnes of essential oils
4.5 and 41.4 mg (0.54.6%) (Smith & Levi, 1961).
of mint other than peppermint were imported by
According to the European Union, the highest
the China, Germany, Japan, Singapore, and USA
recommended daily dose is 1.2 mL of pepper-
and in recent years. No separate data were avail-
mint oil; 1080 mg of peppermint oil contains a
able for spearmint, peppermint, or pennyroyal
maximum of 140 mg of pulegone, a daily intake
oil from this source.
of 2.3 mg/kg body weight (bw) for a person of 60
kg (EMEA, 2005).
1.2.2 Use
Aerial parts [leaves and flowering tops] of 1.3 Occurrence and exposure
plant species containing pulegone have been
used as a traditional remedy, as flavouring, as 1.3.1 Occurrence
spice, and for brewing teas. Pennyroyal oil has
Pulegone is naturally found in plants of the
been used as a traditional medicine. It is also
Lamiaceae (or Labiatae) family. The amount of
used to flavour alcoholic beverages, baked goods,
pulegone in the various oils varies depending
candies, ice creams, as a fragrance component of
on several factors such as origin of the plant,
detergents, cosmetics and oral hygiene products,
yearly weather conditions, harvest date, plant
and as an insect repellent (Karousou et al., 2007;
age, fertilization, location and planting time
Da Rocha et al., 2012).
(Farley & Howland, 1980; Weglarz & Zalecki,
1985; Murray et al., 1988; Voirin et al., 1990;
142
Pulegone
Marotti et al., 1994; Kumar et al., 2000; Misra & 2012). According to the Cosmetic Ingredient
Srivastava, 2000; Dolzhenko et al., 2010). Review Expert Panel, the concentration of pule-
gone in cosmetic formulations should not exceed
1.3.2 Consumer exposure 1% (Nair, 2001).
In addition to the use in medication, humans
are exposed to pulegone as a constituent of the 2. Cancer in Humans
essential oil in flavourings, confectionery, and
cosmetics (Karousou et al., 2007; Barceloux,
2008). According to the Joint FAO/WHO Expert
Committee on Food Additives (JECFA), the esti-
mated intake of pulegone per capita is 2 g per
day and 0.04 g/kg bw per day for Europe, and 3. Cancer in Experimental Animals
12 g per day and 0.03 g/kg bw per day for the
USA (IPCS, 2001). 3.1 Mouse
See Table3.1
1.3.3 Occupational exposure
In one study of oral administration, groups
No data were available to the Working Group. of 50 male and 50 female B6C3F1 mice (age,
Workers from the flavouring, confectionary, and 67 weeks) were given pulegone at a dose of
cosmetic industries are probably exposed to 0 (corn oil only, 10 mL/kg bw), 37.5, 75, or
pulegone. 150 mg/kg body weight (bw) by gavage, 5 days
per week for 105weeks. The purity of pulegone
was approximately 96%. Survival in all dosed
1.4 Regulations and guidelines groups was similar to that in the vehicle-control
Limits in the use of pulegone in food prod- group. Mean body weights of males and females
ucts have been issued for different applications. at 150 mg/kg bw were lower than those in the
According to regulation (EC) 1334/2008, the use vehicle-control group after weeks 25 and 33,
of pulegone in food and beverages has limits respectively.
of: 100 mg/kg for mint/peppermint containing In males, the incidence of hepatoblastoma
alcoholic beverages; 20 mg/kg for mint/ was significantly higher in the group at the
peppermint containing non-alcoholic bever- intermediate dose. The incidences of hepato-
ages; 2000 mg/kg for micro breath freshening cellular adenoma, and hepatocellular adenoma
confectionery; 350 mg/kg for chewing gum; or carcinoma (combined) were also signifi-
and 250 mg/kg for mint/peppermint containing cantly higher in the group at the intermediate
confectionery, except the micro breath. As a dose. The incidences of hepatocellular adenoma,
pure ingredient, pulegone may not be added hepatocellular carcinoma, or hepatoblastoma
to foodstuffs. According to the Committee of (combined) showed a significant positive trend.
Experts on Flavoring Substances (CEFS), provi- [The Working Group noted that the lower
sional consumption limits were established for average body weight in the group at the highest
pulegone at 20 mg/kg in food and beverages dose may have reduced the incidences of liver
(European Commission, 2002, 2008). tumours (Haseman et al., 1997).] The incidence of
In the USA, pulegone is not authorized as liver clear cell foci was significantly higher in all
a synthetic flavouring substance (DHHS-FDA, dosed groups; the incidence of eosinophilic liver
143
144
Table 3.1 Studies of carcinogenicity with pulegone in mice and rats
Species, Dosing regimen Incidence of tumours Significancea Comments

Duration
Reference
Mouse, 0 (control), 37.5, 75, 150mg/kg bw, by Hepatoblastoma: *P=0.040 Purity, 96% (approximate)
B6C3F1 gavage in corn oil, 5days per wk, for 1/50, 3/50, 7/50*, 2/50 (M); Body weight of group at 150mg/kg bw
(M,F) 105wk 0/50, 1/50, 2/50, 2/50 (F) was 10% less than that of the vehicle-
105wk 50 M and 50 F/group (age, 56 wk) Hepatocellular adenoma: *P=0.008 control group after week 25 for males
NTP (2011) 22/50, 31/50, 35/50*, 28/50 (M); **P<0.001 (trend) and after week33 for females
IARC MONOGRAPHS 108
13/49**, 15/50, 13/50, 27/50*** (F) ***P<0.002

13/50, 11/50, 18/50, 15/50 (M);
5/50, 1/50, 4/50, 8/50 (F)
Hepatocellular adenoma or carcinoma *P=0.004
(combined): **P=0.001 (trend)
29/50, 36/50, 42/50*, 35/50 (M); ***P=0.002
17/49**, 15/50, 15/50, 32/50*** (F)
Hepatocellular adenoma or carcinoma *P=0.038 (trend)
or hepatoblastoma (combined): **P=0.004
29/50*, 37/50, 42/50**, 36/50 (M); ***P<0.001
17/49***, 15/50, 15/50, 33/50**** (F) (trend)
****P<0.001
Osteoma or osteosarcoma (combinedb): # Greater
0/49, 0/50, 3/50#, 1/50 (F) thanhistorical

controls
Rat, F344 0 (control), 18.75 (M only), 37.5, 75 or Urinary bladder papilloma: *P=0.044 Purity, 96% (approximate)
(M,F) 150(F only) mg/kg bw, by gavage in corn 0/50, 0/49, 1/50 (2%), 3/47* (F) Mean body weight of males and females
104wk oil, 5days per wk, for 104wk Urinary bladder carcinoma: at 75mg/kg bw was 10% less than that of
NTP (2011) Dosing of males at the highest dose 0/50, 0/49, 0/50, 2/47 (F) the vehicle-control group after week13
(75mg/kg bw) and females at the highest Urinary bladder papilloma or **P=0.005 and 21 respectively; and mean body
dose (150mg/kg bw) was stopped after carcinoma (combined)c: weight in females at 150mg/kg bw was
wk60 because of high morbidity and 0/50, 0/49, 1/50, 5/47**(F) 12% less than that in the vehicle-control
mortality. Surviving animals at these group after week9.
doses were treated with corn oil only from No treatment-related increases in
wk60 to end of study. tumour incidences were found in males.
50 M and 50 F/group (age, 67wk)
a Poly-3 test was used for all the statistical analyses in this table
b Historical incidence for 2-year gavage studies with corn oil vehicle-control groups: 1/248 (0.4%0.9%); range 02%; for all routes: 8/1498 (0.5%1.0%), range 04%
c Historical incidence for 2-year gavage studies with corn oil vehicle-control groups: 0/200; for all routes, 0/1347

Pulegone
cell foci was significantly higher in the groups were given the corn-oil vehicle until the end of
receiving the intermediate and highest doses; the study. Survival of males at 37.5 mg/kg bw
and the incidence of mixed liver cell foci was was significantly lower than that of the vehicle
significantly higher in the group at the highest controls; only two males in the group receiving
dose. The incidence of forestomach squamous 75mg/kg bw and stop-exposure survived to the
cell hyperplasia was significantly higher in the end of the study, and no females in the group
group at the highest dose. receiving 150 mg/kg bw and stop-exposure.
In females, the incidence of hepatocellular Compared with those of the rats in the vehi-
adenoma was significantly higher in the group at cle-control group, mean body weights of males
the highest dose. The incidence of hepatocellular in the group receiving 75mg/kg bw and stop-ex-
adenoma or carcinoma (combined) was signifi- posure, and of females in the group receiving
cantly higher in the group at the highest dose and 75mg/kg bw, and of females receiving 150mg/kg
had a significant positive trend. The incidence of bw and stop-exposure were lower after weeks13,
hepatocellular adenoma, hepatocellular carci- 21 and 9, respectively.
noma, or hepatoblastoma (combined) was signifi- In females, the incidence of urinary bladder
cantly higher in the group at the highest dose and papilloma was significantly higher in the group
had a significant positive trend. The incidence of at the highest dose. The incidence of urinary
liver clear cell foci was significantly higher in all bladder papilloma or carcinoma (combined) was
dosed groups; the incidence of eosinophilic liver significantly higher in the group at the highest
cell foci was significantly higher in the high-dose dose (150 mg/kg bw with stop-exposure). [The
group; and the incidence of mixed liver cell foci Working Group noted that no urinary bladder
was significantly higher in the groups receiving papillomas or carcinoma were observed in 1347
the intermediate and highest doses. The incidence control animals from previous studies (all routes
of osteoma or osteosarcoma (combined) was of exposure) by the National Toxicology Program
increased in the group at the intermediate dose (NTP).] In males, no treatment-related increases
(3 out of 50) compared with historical controls in tumour incidences were found (NTP, 2011).
(historical incidence for gavage studies was 1 out
of 248 (0.4%0.9%); range, 02%). The incidence
of forestomach squamous cell hyperplasia was 4. Mechanistic and Other
significantly higher in the groups receiving the Relevant Data
intermediate and highest doses (NTP, 2011).
3.2 Rat metabolism, and excretion
See Table3.1 Fig.4.1 describes the metabolism of pulegone
In one study of oral administration, groups in humans and rodents.
of 50 male and 50 female F344/N rats (age,
67 weeks) were given pulegone at 0 (corn oil
4.1.1 Humans
only, 5 mL/kg bw), 18.75 (males only), 37.5, 75,
or 150(females only) mg/kg bw by gavage, 5days A haematological post-mortem examination
per week for up to 104weeks. Due to excessive of a woman who ingested pennyroyal extract
morbidity and mortality, males at 75mg/kg bw used as an abortifacient, indicated that both
and females at 150mg/kg bw were not given pule- serum concentrations of pulegone at 18 ng/mL
gone after week60 (stop-exposure); these groups and menthofuran at 1 ng/mL possibly resulted
145
Fig.4.1 Metabolism of pulegone in humans and rodents
146
Pulegol
H H OH
or
2,8-Dihydroxymenthone
OH OH O
HO
H
OH
H Glucuronide conjugates and/or glutathione conjugates

CYP Pulegone
O
IARC MONOGRAPHS 108
O H
CYP CYP
H O H 2C H
10-Hydroxypulegone OH H H H O
H
CYP CYP
O
H O O O O O (+)-Isomintlactone
CH 2 O H +
OH
O 5-Hydroxypulegone 9-Hydroxypulegone Menthofuran
O H
CYP Epoxidementhofuran 2-Hydroxymenthofuran
H
CH 2 O H H
9-Hydroxy-p-menthan-3-one O
Pr otein binding
O O O O
O
C H 2O H CH O ( )-Mintlactone
Piperitenone 9-Hydroxypiperitenone H -Ketoenal
(8-pulegone aldehyde)
4-Methyl-2-cyclohexenone
4-,5-,7-, and 10-Hydroxypiperitenone OH O
O OH O
p-Cresol
Most of the metabolites of pulegone are derived from menthofuran and piperitenone. A -ketoenal is generated as a major electrophilic metabolite from both pulegone and menthofuran.
CYP, cytochrome P450
Adapted from Chen et al. (2011) with permission from Elsevier
Pulegone
in fatal poisoning. Serum samples were analysed or rearranged under conditions found in the
for both metabolites at 26 hours post mortem, human body, forming ,,4-trimethyl-1-cy-
72hours after ingestion (Anderson et al., 1996). clohexene-1-methanol (3-p-menthen-8-ol) as
In another case, serum was found to contain a minor metabolite (Engel, 2003). It was also
menthofuran at 40 ng/mL, with no detectable deduced that the reduction of the carbonyl group
pulegone, 10 hours after ingestion (Anderson in menthone leads to the formation of menthol,
et al., 1996). Menthofuran is considered the major while oxidation at C-5 also yields the menthone
proximate toxic metabolite; however, pulegone metabolite. The formation of ,,4-trime-
oxidation produces other metabolites that may thyl-1-cyclohexene-1-methanol from pulegol
also be toxic (see Fig.4.1; Anderson et al., 1996). occurs at a body pH of6.5. However, the formation
Pulegone is metabolized by multiple human of the same metabolite from pulegol as an artefact
liver CYPs to menthofuran, a toxic metabolite during enzymatic hydrolysis with sulfatase and
of pulegone (Khojasteh-Bakht et al., 1999). In a glucuronidase cannot be totally excluded (Engel,
study by Khojasteh-Bakht et al. (1999), pulegone 2003). The major metabolite, 9-hydroxy-p-men-
(200 M) was separately incubated with indi- than-3-one, is formed through the oxidation of
vidual human CYPs, namely CYP1A2, CYP2A6, 10-hydroxypulegone via the reduction of the
CYP2C9, CYP2C19, CYP2D6, CYP2E1, or exocyclic double bond (Engel, 2003). Although
CYP3A4. It was found that CYP2E1, CYP1A2, the results of this study suggested that mentho-
and CYP2C19 metabolized the oxidation of furan is not a relevant metabolite in humans,
pulegone to menthofuran with the following Km menthofuran was found to be present in the
and Vmax: CYP2E: Km, 29 M; Vmax, 8.4 nmol/ serum of two individuals, hours after ingestion
min per nmol protein; CYP1A2: Km, 94 mM; of a large amount of pennyroyal oil in a study
Vmax, 2.4 nmol/min per nmol protein; and for by Anderson et al. (1996). Additionally, it should
CYP2C19: Km, 31 M; and Vmax, 1.5 nmol/min be noted that toxicity seems to be stereospecific,
per nmol protein. in that for (R)-(+)-pulegone both reduction steps
Menthofuran was metabolized by the same (to menthone and from 10-hydroxypulegone
human liver CYPs involved in the metabolism to 9-hydroxy-p-menthan-3-one) are much less
of pulegone except for the addition of CYP2A6 favoured by the conformation of the terpene,
(Khojasteh-Bakht et al., 1999). as opposed to the (S)-()-pulegone enantiomer
After oral administration of (R)-(+)-pulegone (Engel, 2003). Subsequently, 10-hydroxypule-
at 0.5mg/kg bw or (S)-()-pulegone at 1mg/kg gone is accumulated and further oxidation at the
bw to six volunteers, six metabolites were identi- hydroxyl group leads to pulegon-10 aldehyde, a
fied in the urine. The major metabolite of (R)-(+)- toxic metabolite.
pulegone was 10-hydroxypulegone (Engel, 2003). In summary, from a general perspective,
Although 10-hydroxypulegone was shown to minimal data existed on the excretion of pule-
convert to menthofuran in vitro, menthofuran gone in humans.
and its metabolites were found in relative small
amounts in the urine. Alternatively, pulegone 4.1.2 Experimental systems
may be reduced to menthone, a product that is
detected in small amounts in the urine. It might The majority of experimental studies used
be possible that pulegone is also reduced at the toxic stereoisomer (R)-(+)-pulegone, which is
carbonyl group first; however, no trace of pulegol the natural component of pennyroyal oil, but the
was found in the urine. Consequently, it was also (S)-()-isomer is also metabolized in the same
proposed that pulegol can be reduced to menthol manner (Speijers, 2001). A total of approximately
147
IARC MONOGRAPHS 108
14 phase I metabolites exist in rats in vivo, with terms of ring- and side-chain oxidation to obtain
approximately 10 identified phase II metabolites numerous hydroxylated by-products (Speijers,
(Thomassen et al., 1991; Chen et al., 2001; Zhou 2001). For the predominant pathway, the isopro-
et al., 2005). Observed metabolites account for pylidene substituent of pulegone is subjected to
only 3% of total radiolabel typically excreted in regiospecific allylic oxidation to yield 9-hydroxy-
bile, with glucuronide conjugates and minimal pulegone, which forms menthofuran cyclically
glutathione conjugates being found in highest (Gordon et al., 1987; Madyastha & Raj, 1993).
quantities (Speijers, 2001). The most common As a minor pathway, it is presumed that the
biliary metabolites observed were the glucuro- exocyclic alkene of pulegone is oxidized (with
nide conjugates of hydroxylated pulegone and the assumption of an epoxide intermediate) to
pulegol (Speijers, 2001). yield 2,8-dihydroxymenthone, (Speijers, 2001).
The metabolism of pulegone involves three Additionally, pulegone is reduced to pulegol
major metabolic pathways: (i) hydroxylation which is then rearranged to isopulegone with
to give monohydroxylated pulegones at C-5 or the aid of a supposed free-radical intermediate
methyl (9- or 10-), followed by conjugation with (Gordon et al., 1987; Speijers, 2001).
glucuronic acid or with glutathione; the conju- When pulegone is converted to menthofuran,
gates being further metabolized; (ii) reduction it undergoes a reaction where an oxycarbonium
of the carboncarbon double bond that leads ion is created via CYP-mediated oxidation
to the formation of menthofuran; and (iii) the of menthofuran, generating an intermediate,
formation of piperitenone after 5-hydroxyl- -ketoenal (one of the primary reactive metab-
ation, followed by dehydration (see Fig. 4.1) olites), but also another intermediate epoxy-
(Thomassen et al., 1990; Speijers, 2001; Chen furan (Chen et al., 2011). Additionally, p-cresol
et al., 2011). Most of the metabolites of pulegone is also generated via pulegone metabolism and
are derived from menthofuran and piperitenone, also depletes glutathione with minor hepatotoxic
and 4-methyl-2-cyclohexenone is one of these effects (Chen et al., 2011).
metabolites. A -ketoenal is generated as a major In addition to cyclizing to obtain mentho-
electrophilic metabolite from both pulegone and furan, 9-hydroxypulegone can also be oxidized
menthofuran (Thomassen et al., 1992; Speijers, in a secondary detoxication pathway to
2001). This reactive enonal may be derived directly 9-carboxy-pulegone, also called 5-methyl-2-(1-
from incipient oxycarbonium ions formed in methyl-1-carboxyethylidene) cyclohexanone
the oxidation of menthofuran by cytochrome (Speijers, 2001). This product is then partially
P450 (CYP), or from an epoxyfuran interme- cyclized to hydroxylactone or is assumed to be
diate (Thomassen et al., 1992). Mintlactones are oxidized and hydrated to hydroxyacids that are
formed as stable products of the -ketoenal, but eliminated through urine (Speijers, 2001). Studies
also may be formed by direct proton loss from an of oral administration in rats have shown piperi-
incipient oxycarbonium ion (Chen et al., 2011). tenone is hydroxylated: metabolites found in the
As shown experimentally, pulegone is specif- urine were isolated and found to be hydroxylated
ically metabolized to menthofuran and para- at the 4, 5, 7, 9, and 10 positions (Speijers, 2001).
mentha-1,4(8)-dien-3-one, commonly known 9-Piperitenone can be further converted to an
as piperitenone (Speijers, 2001). In subsequent analogous furan metabolite and to the -ketoenal
reactions, the tertiary ring carbon (C-5) is (Chen et al., 2011).
hydroxylated to obtain 5-hydroxypulegone According to studies in experimental animals
(Speijers, 2001). This product is then dehydrated treated orally, gastrointestinally absorbed pule-
to piperitenone, which is further metabolized in gone is excreted in the urine within 24 hours.
148
Pulegone
Additional studies have demonstrated that pule- (ii) Drosophila melanogaster

gone is excreted and eliminated at 624% in Pulegone was reported to be weakly muta-
faecal matter, and a small amount in expired air genic in the Drosophila melanogaster somatic
(NTP, 2011). mutation and recombination test. However, a
sample of pennyroyal oil that was reported to
4.2 Genetic and related effects contain pulegone at 75.7% was not mutagenic in
this assay (Franzios et al., 1997).
4.2.1 Humans
(b) Cytogenetic effects
In B6C3F1 mice given pulegone at doses up
to 150mg/kg bw per day by gavage for 3months,
4.2.2 Experimental systems there was no increase in the frequency of micro-
A limited number of studies of genotoxicity nucleus formation in peripheral blood erythro-
have been conducted with pulegone (Table4.1). cytes (NTP, 2011).
(a) Mutation
4.3 Other mechanistic data
(i) Bacteria
Pulegone was not mutagenic in Salmonella 4.3.1 Humans
typhimurium strains TA97, TA98, TA100, In a review of published reports, exces-
TA1535, or TA1537, with or without metabolic sive consumption of pennyroyal oil has been
activation (Andersen & Jensen, 1984). shown to induce moderate to severe toxicity
Three additional assays for gene mutation in (Anderson et al., 1996; Speijers, 2001; NTP, 2011).
bacteria were conducted and results were mixed Consumption of amounts greater than 15mL or
(NTP, 2011). In the first two studies, pulegone approximately 250mg/kg bw may result in death
was not mutagenic with or without metabolic (Anderson et al., 1996; Speijers, 2001; NTP, 2011).
activation. Bacterial strains tested in the first Adverse physiological reactions following exces-
study included S. typhimurium TA97, TA98, sive consumption lead to massive centrilobular
TA100, and TA1535, with and without metabolic hepatic necrosis, pulmonary oedema, internal
activation (10% or 30% S9 from Syrian hamster bleeding, and body weight loss (Anderson et al.,
or Sprague-Dawley rat liver). Strains tested in the 1996; Speijers, 2001).
second study included S. typhimurium strains Eighteen cases of hepatotoxicity were
TA98 and TA100, and Escherichia coli strain described in people who had ingested 10mL or
WP2uvrA/pKM101, with and without metabolic more of pennyroyal oil (Anderson et al., 1996;
activation (10% S9 from rat liver S9). The third NTP, 2011). Symptoms of toxicity included
study also tested pulegone in S. typhimurium coma, seizures, liver, and kidney effects, while
and E. coli; results were positive in S. typhimu- consumption of less than 10 mL of pennyroyal
rium strain TA98 and E. coli strain WP2uvrA/ oil resulted in gastritis and mild toxicity of the
pKM101 in the presence of metabolic activation. central nervous system (Anderson et al., 1996;
[There was no explanation for the discrepancy NTP, 2011). There was no clear correlation
between the multiple assays for mutagenicity.] between dose and toxicological effect (NTP,
2011).
Functionally, reactive metabolites of pule-
gone and menthofuran bind to cellular proteins.
149
150
Table 4.1 Genetic and related effects of pulegone
Test system Resultsa Dose or concentration Reference

(LED or HID)
Salmonella typhimurium TA100, TA1535, TA1537, TA98, 800g/plate Andersen &
TA97, reverse mutation Jensen (1984)
Salmonella typhimurium TA100, TA1535, TA98, TA97, b 2167g/plate NTP (2011)
reverse mutation
Salmonella typhimurium TA100, TA98, Escherichia coli b 3500 g/plate NTP (2011)
IARC MONOGRAPHS 108
WP2, uvrA pKM101

Salmonella typhimurium TA98, Escherichia coli WP2, uvrA +c 500g/plate NTP (2011)
pKM101
Salmonella typhimurium TA100 c 1500g/plate NTP (2011)
Drosophila melanogaster, somatic mutation (and NT 2.1L, applied to filter paperd Franzios et al.
recombination) (1997)
Drosophila melanogaster, somatic mutation (and + NT 0.2L, applied to filter paperd Franzios et al.
recombination) (1997)
Micronucleus formation in peripheral blood lymphocytes, NT 150 mg/kg bw per day by NTP (2011)
B6C3F1 male and female mice in vivo gavage, for 3months
a +, positive; , negative
b 10% or 30% S9 from Syrian hamster or Sprague-Dawley rat liver
c 10% S9 from Sprague-Dawley rat liver
d Larval feeding for 18 hours
HID, highest ineffective dose; LED, lowest effective dose; NT, not tested
Pulegone
Reactive metabolites of pulegone deplete hepatic Urinary bladders from treated rats showed super-
glutathione concentrations, whereas mentho- ficial cell layer necrosis and exfoliation in both
furan, only slightly diminishes these concen- treated groups, and a significant increase in the
trations. On a molecular level, diminished incidence of cellular proliferation in the group
concentrations of pulegone-induced glutathione at the highest dose (150 mg/kg bw) (Da Rocha
result from the generation of electrophili- et al., 2012). Examination of urine collected
cally metabolites that form covalent adducts during week 6 of treatment revealed the pres-
with glutathione (Anderson et al., 1996). Thus, ence of pulegone, piperitone, piperitenone, and
N-acetylcysteine serves as a protectant against menthofuran. Piperitenone was concentrated in
pennyroyal oil poisoning within the first few the urine at cytotoxic levels in rats treated with
hours after ingestion and may also protect pulegone at the highest dose.
cells from further damage in late-stage injuries
(Anderson et al., 1996).
Bakerink et al. (1996) have reported two
4.4 Susceptibility
cases of infant death following the ingestion of No relevant data were available to the
mint tea containing pulegone. The male patient Working Group.
(aged 6months) was given mint tea rich in pule-
gone along with two crushed aspirin tablets.
Presentation of this case indicated hepatic fulmi- 4.5 Mechanistic considerations
nation with cerebral oedema and necrosis, and Studies in humans and rodents indicated
the patient died with a serum concentration of that the metabolism of pulegone to menthofuran
menthofuran of 10 ng/mL. Most importantly, generates electrophilic species that can bind to
characteristic hepatotoxicity findings included proteins. This may result in chronic regenera-
hepatomegaly, poor perfusion, and dark blood tive cell proliferation that may be related to the
from the nasogastric tube and rectum. The carcinogenicity in the liver and urinary bladder
second case presented with hepatic dysfunction that is observed in experimental animals (see
and severe epileptic encephalopathy and had Section 3).
serum concentrations of pulegone at 25ng/mL,
and menthofuran at 41 ng/mL. Overall, find-
ings were similar for both cases, suggesting that 5. Summary of Data Reported
pulegone intake is associated with marked hepa-
totoxicity (Bakerink et al., 1996).
5.1 Exposure data
4.3.2 Experimental systems Pulegone is present as a major constituent in
Many studies in experimental animal have pennyroyal oils, and to a minor extent in the oil
observed that p-cresol, a pulegone metabolite, of several other species of mint. Pulegone is also
induces diminished hepatic function, increased found in Buchu leaf oils. Pennyroyal oil has been
liver and kidney weight, gastrointestinal and used as flavouring in confectionery, as a spice,
nasal epithelial irritation, and atrophy of female and in brewing teas. Pennyroyal leaves and oil
reproductive organs (Chen et al., 2011). are used in traditional medicine applications,
Female Fischer rats (age, 6 weeks) were given for the treatment of dyspepsia and menstrual
pulegone orally at a dose of 0, 75, or 150 mg/kg disorders, and as a diaphoretic. Pennyroyal oil
bw per day, 5 days per week, for 4 or 6 weeks. has also been used as a fragrance in foods and
151
IARC MONOGRAPHS 108
in cosmetics, and as a flea repellent. Given the which is then hydroxylated at the 9-position and
wide range of uses of mint, there is a possibility further converted to an analogous furan metab-
of exposure to pulegone on a daily basis. Limits olite and to the -ketoenal. Further metabolism
to the use of pulegone have been issued for foods of the -ketoenal produces 4-methyl-2-cyclohex-
and beverages. Synthetic pulegone is not author- enone and p-cresol.
ized as a flavouring substance in the USA or Pulegone was not mutagenic in standard
Europe. bacterial assays, either with or without exoge-
nous metabolic activation.
Studies in humans and rodents indicated that
5.2 Human carcinogenicity data some of the pulegone metabolites deplete hepatic
No data were available to the Working Group. levels of glutathione and can bind to cellular
proteins. This may result in chronic regenerative
cell proliferation, which may be related to the
5.3 Animal carcinogenicity data carcinogenicity observed in the liver and urinary
Pulegone was tested for carcinogenicity after bladder in experimental animals.
oral administration in one study in mice and one
study in rats. 6. Evaluation
In male and female mice given pulegone by
gavage, there was a significant increase in the inci-
dences of hepatocellular adenoma, and hepato- 6.1 Cancer in humans
cellular adenoma and carcinoma (combined) in There is inadequate evidence in humans for
males and females, and of hepatoblastoma in the carcinogenicity of pulegone.
males. In female mice, the incidence of osteoma
or osteosarcoma (combined) was higher than
that in historical controls. 6.2 Cancer in experimental animals
In female rats given pulegone by gavage,
There is sufficient evidence in experimental
there was an increase in the incidence of urinary
animals for the carcinogenicity of pulegone.
bladder papilloma and of urinary bladder papil-
loma or carcinoma (combined). In males, there
were no treatment-related increases in tumour 6.3 Overall evaluation
incidences.
Pulegone is possibly carcinogenic to humans
(Group 2B).
5.4 Mechanistic and other relevant
data
References
Pulegone is readily absorbed in humans. It is
metabolized in humans and rodents to isomers Andersen PH & Jensen NJ (1984). Mutagenic investigation
of hydroxypulegone, predominantly by hepatic of peppermint oil in the Salmonella/mammalian-mi-
oxidation at the 5-, 9-, and 10-positions. In crosome test. Mutat Res, 138:1720. doi:10.1016/0165-
1218(84)90080-6 PMID:6387476
rodents, 9-hydroxypulegone is further oxidized Anderson IB, Mullen WH, Meeker JE et al. (1996).
to menthofuran, which is converted to a reactive Pennyroyal toxicity: measurement of toxic metabolite
epoxide and a reactive aldehyde (-ketoenal). levels in two cases and review of the literature. Ann
5-Hydroxypulegone is converted to piperitenone, Intern Med, 124:726734. doi:10.7326/0003-4819-124-
8-199604150-00004 PMID:8633832
152
Pulegone
Bakerink JA, Gospe SM Jr, Dimand RJ, Eldridge MW Available from: http://ec.europa.eu/food/fs/sc/scf/
(1996). Multiple organ failure after ingestion of penny- out133_en.pdf. Accessed 18 June 2014.
royal oil from herbal tea in two infants. Pediatrics, European Commission (2008). Regulation (EC) No
98:944947. PMID:8909490 1334/2008 of the European Parliament and of the
Barceloux DG (2008). Pennyroyal and Pulegone (Mentha Council of 16 December 2008 on flavourings and
pulegium L.). In: Medical Toxicology of Natural certain food ingredients with flavouring properties for
Substances: Foods, Fungi, Medicinal Herbs, Plants and use in and on foods and amending Council Regulation
Venomous Animals. Hoboken, NJ: John Wiley & Sons, (EEC) No 1601/91, Regulations (EC) No 2232/96 and
Inc.; pp. 563567. (EC) No 110/2008 and Directive 2000/13/EC. Off J Eur
Cappello G, Spezzaferro M, Grossi L, Manzoli L, Marzio Union L, 354:34
L (2007). Peppermint oil (Mintoil) in the treatment Farley DR & Howland V (1980). The natural variation of
of irritable bowel syndrome: a prospective double the pulegone content in various oils of peppermint. J Sci
blind placebo-controlled randomized trial. Dig Food Agric, 31(11):114351. doi:10.1002/jsfa.2740311104
Liver Dis, 39(6):5306. doi:10.1016/j.dld.2007.02.006 Foster S, Duke J (1990). A field guide to medicinal plants.
PMID:17420159 Boston (MA): Houghton Mifflin Co.; p. 190.
Chen LJ, Lebetkin EH, Burka LT (2001). Metabolism of Franzios G, Mirotsou M, Hatziapostolou E etal. (1997).
(R)-(+)-pulegone in F344 rats. Drug Metab Dispos, Insecticidal and Genotoxic activities of mint essential
29:15671577. PMID:11717176 oils. J Agric Food Chem, 45:26902694. doi:10.1021/
Chen XW, Serag ES, Sneed KB, Zhou SF (2011). Herbal jf960685f
bioactivation, molecular targets and the toxicity rele- Gordon WP, Huitric AC, Seth CL etal. (1987). The metab-
vance. Chem Biol Interact, 192:161176. doi:10.1016/j. olism of the abortifacient terpene, (R)-(+)-pulegone, to
cbi.2011.03.016 PMID:21459083 a proximate toxin, menthofuran. Drug Metab Dispos,
Da Rocha MS, Dodmane PR, Arnold LL, Pennington 15:589594. PMID:2891472
KL, Anwar MM, Adams BR et al. (2012). Mode of Guenther E (1949). The Essential Oils. New York (NY): D
action of pulegone on the urinary bladder of F344 Van Nostrand Company Inc.
rats. Toxicol Sci, 128(1):18. doi:10.1093/toxsci/kfs135 Haseman JK, Young E, Eustis SL, Hailey JR (1997). Body
PMID:22499580 weight-tumor incidence correlations in long-term
DHHS-FDA (2012). Synthetic flavouring substances and rodent carcinogenicity studies. Toxicol Pathol, 25:256
adjuvants. 21CFR182.60. Department of Health and 263. doi:10.1177/019262339702500302 PMID:9210256
Human Services, Food and Drug Administration. Hayes JR, Stavanja MS, Lawrence BM (2007). Biological
Available from: http://www.gpo.gov/fdsys/ and toxicological properties of mint oils and their
granule/CFR-2011-title21-vol3/CFR-2011-title21- major isolates: Safety assessment. In: Lawrence BM,
vol3-sec182-60/content-detail.html. Accessed 18 June editor. Mint: The Genus Mentha. CRC Press, Taylor &
2014. Francis Group; pp. 462477.
Dolzhenko Y, Bertea CM, Occhipinti A, Bossi S, Maffei Hoppe HA (1975). Drogenkunde, 8th Ed. Berlin: Walter de
ME (2010). UV-B modulates the interplay between Gruyter; Vol 1, p. 764.
terpenoids and flavonoids in peppermint (Mentha x IPCS (2001).WHO food additives series 46: Pulegone
piperita L.). J Photochem Photobiol B, 100(2):6775. and related substances. Available from: http://www.
doi:10.1016/j.jphotobiol.2010.05.003 PMID:20627615 inchem.org /documents/jecfa/jecmono/v46je10.
EMEA; European Medicines Agency (2005). Public htm#_46101200. Accessed 19 June 2014.
Statement on the use of herbal medicinal products Kaiser R, Lamparsky D, Schudel P (1975). Analysis of
containing pulegone and menthofuran. EMEA/ buchu leaf oil. J Agric Food Chem, 23(5):94350.
HMPC/138386/2005. London, United Kingdom: doi:10.1021/jf60201a021
Committee on Herbal Medicinal products (HMPC). Karousou R, Balta M, Hanlidou E, Kokkini S (2007).
Available from: http://www.ema.europa.eu/docs/en_ Mints, smells and traditional uses in Thessaloniki
GB/document_library/Scientific_guideline/2010/04/ (Greece) and other Mediterranean countries. J
WC500089958.pdf. Accessed 18 June 2014. Ethnopharmacol, 109(2):24857. doi:10.1016/j.
Engel W (2003). In vivo studies on the metabolism of the jep.2006.07.022 PMID:16962274
monoterpene pulegone in humans using the metabo- Khojasteh-Bakht SC, Chen W, Koenigs LL et al.
lism of ingestion-correlated amounts (MICA) approach: (1999). Metabolism of (R)-(+)-pulegone and (R)-(+)-
explanation for the toxicity differences between (S)-(-)- menthofuran by human liver cytochrome P-450s:
and (R)-(+)-pulegone. J Agric Food Chem, 51:65896597. evidence for formation of a furan epoxide. Drug Metab
doi:10.1021/jf034702u PMID:14558782 Dispos, 27:574580. PMID:10220485
European Commission (2002). Opinion of the Scientific Kumar S, Bahl JR, Bansal RP, Kukreja AK, Garg SN, Naqvi
Committee on Food on pulegone and menthofuran. AA (2000). Profits of Indian menthol mint Mentha
Brussels, Belgium: Scientific Committee on Food.
153
IARC MONOGRAPHS 108
arvensis cultivars at different stages of crop growth in peppermint oils by gas chromatographic analysis. J
northern plains J Med Arom Plant Sci, 22:774786. Agric Food Chem, 9(3):23044. doi:10.1021/jf60115a017
List PH, Hrhammer L, editors (1976). Hagers handbook Smith DM, Skakum W, Levi L (1963). Determination of
of pharmaceutical practice. 4th Ed. Berlin: Springer- botanical and geographical origin of spearmint oils by
Verlag; Vol 5, p. 20. gas chromatographic and ultraviolet analysis. J Agric
Madyastha KM & Raj CP (1993). Studies on the metab- Food Chem, 11(3):26876. doi:10.1021/jf60127a010
olism of a monoterpene ketone, R-(+)-pulegonea Speijers GJA (2001). WHO Food Additives Series 46:
hepatotoxin in rat: isolation and characterization Pulegone and related substances. Bilthoven, the
of new metabolites. Xenobiotica, 23(5):50918. Netherlands: National Institute of Public Health and
doi:10.3109/00498259309059391 PMID:8342298 the Environment, Section on Public Health, Centre
Marotti M, Piccaglia R, Giovanelli E, Deans SG, Eaglesham for Substances and Risk Assessment. Available from:
E (1994). Effects of planting time and mineral fertili- http://www.inchem.org/documents/jecfa/jecmono/
zation on peppermint (Mentha piperita L.) essential v46je10.htm. Accessed 22 July 2014.
oil composition and its biological activity. Flavour Thomassen D, Knebel N, Slattery JT etal. (1992). Reactive
Fragrance J, 9(3):1259. doi:10.1002/ffj.2730090307 intermediates in the oxidation of menthofuran by
Misra A & Srivastava NK (2000). Influence of water stress cytochromes P-450. Chem Res Toxicol, 5:123130.
on Japanese mint. J Herbs Spices Med Plants, 7(1):518. doi:10.1021/tx00025a021 PMID:1581528
doi:10.1300/J044v07n01_07 Thomassen D, Pearson PG, Slattery JT, Nelson SD (1991).
Murray MJ, Marble P, Lincoln D, Hefendehl FW (1988). Partial characterization of biliary metabolites of pule-
Peppermint oil quality differences and the reason for gone by tandem mass spectrometry. Detection of
them. In: Lawrence BM, Mookherjee BD, Willis BJ, glucuronide, glutathione, and glutathionyl glucuro-
editors. Flavor and Fragrances: A World Perspective. nide conjugates. Drug Metab Dispos, 19(5):9971003.
Amsterdam, the Netherlands: Elsevier Press; pp. PMID:1686249
189210. Thomassen D, Slattery JT, Nelson SD (1988). Contribution
Nair B (2001). Final report on the safety assessment of menthofuran to the hepatotoxicity of pulegone:
of Mentha Piperita (Peppermint) Oil, Mentha assessment based on matched area under the curve
Piperita (Peppermint) Leaf Extract, Mentha Piperita and on matched time course. J Pharmacol Exp Ther,
(Peppermint) Leaf, and Mentha Piperita (Peppermint) 244(3):8259. PMID:3252034
Leaf Water. Int J Toxicol, 20(4):Suppl 3: 6173. Thomassen D, Slattery JT, Nelson SD (1990). Menthofuran-
doi:10.1080/10915810152630747 PMID:11766133 dependent and independent aspects of pulegone hepa-
NTP; National Toxicology Program (2011). Toxicology totoxicity: roles of glutathione. J Pharmacol Exp Ther,
and carcinogenesis studies of pulegone (CAS No. 253:567572. PMID:2338648
89827) in F344/N rats and B6C3F1 mice (gavage Turner GW & Croteau R (2004). Organization of mono-
studies). Natl Toxicol Program Tech Rep Ser, 563:1201. terpene biosynthesis in Mentha. Immunocytochemical
PMID:21921962 localizations of geranyl diphosphate synthase,
ONeil MJ (2013). An Encyclopedia of Chemicals, Drugs, limonene-6-hydroxylase, isopiperitenol dehydroge-
and Biologicals. 15th Ed. Whitehouse (NJ): Merck nase, and pulegone reductase. Plant Physiol, 136(4):4215
Sharp & Dohme Corp., RSC Publishing. 27. doi:10.1104/pp.104.050229 PMID:15542490
Petrakis EA, Kimbaris AC, Pappas CS, Tarantilis PA, United Nations Comtrade (2013). United Nations
Polissiou MG (2009). Quantitative determination of Comtrade Database. Available from: http://comtrade.
pulegone in pennyroyal oil by FT-IR spectroscopy. un.org/, accessed 17 December 2014.
J Agric Food Chem, 57(21):100448. doi:10.1021/ Virmani OP & Datta SC (1968). Oil of spearmint Perfume
jf9026052 PMID:19817373 Essential Oil Rec, 59:351362.
Scifinder (2014). CAS Database: Pulegone. American Voirin B, Brun N, Bayet C (1990). Effects of day length
Chemical Society. Available from: https://scifinder.cas. on the monoterpene. composition of leaves of
org Mentha piperita. Phytochemistry, 29(3):74955.
Sigma Aldrich (2013). Pulegone. MSDS Analytical Standard. doi:10.1016/0031-9422(90)80012-6
Available from: http://www.sigmaaldrich.com/catalog/ Von Hefendehl FW & Ziegler E (1975). Analysis of pepper-
product/a ldrich/w296309?lang=en&region=US. mint oil. Deut Lebens-Ritnds, 8:287290.
Accessed 19 June 2014. Weglarz Z & Zalecki R (1985). Investigations of depend-
Skenderi G (2003). Herbal Vade Mecum. Rutherford (NJ): ence of the crop season of peppermint (Mentha piperita
Herbacy Press; p. 287. L.) herb upon the crop itself and the quality of the raw
Smith DM & Levi L (1961). Essential oils, treatment of material. Herba Pol, 31:175180.
compositional data for the characterization of essen- Zhou S, Chan E, Duan W etal. (2005). Drug bioactiva-
tial oils. Determination of geographical origins of tion, covalent binding to target proteins and toxicity
relevance. Drug Metab Rev, 37:41213. PMID:15747500
154
METHYLENE BLUE
1. Exposure Data methylene blue independent of hydra-

tion state. Due to its hygroscopic nature,
Methylene blue was originally synthesized commercial methylene blue is typically sold
in 1876 as an aniline-based dye for the textile as the hydrate, but is sometimes incorrectly
industry (Berneth, 2008), but scientists such as presented as the trihydrate.]
Robert Koch and Paul Ehrlich were quick to Chem. Abstr. Serv. Name: Phenothiazin-
realize its potential for use in microscopy stains 5-ium, 3,7-bis(dimethylamino)-, chloride
(Ehrlich, 1881; Oz et al., 2011). The observation (ONeil et al., 2006)
of selective staining and inactivation of micro- IUPAC Systematic Name: [7-(Dimethyl-
bial species led to the testing of aniline-based amino)phenothiazin-3-ylidene]-dimethyla-
dyes against tropical diseases (Oz et al., 2011). zanium chloride (PubChem, 2013)
Methylene blue was the first such compound to Synonyms: Aizen methylene blue; Basic blue
be administered to humans, and was shown to be 9 (8CI); Calcozine blue ZF; Chromosmon;
effective in the treatment of malaria (Guttmann C.I. 52 015; Methylthionine chloride;
& Ehrlich, 1891; Oz et al., 2011). Methylene blue Methylthioninium chloride; Phenothiazine-
was also the first synthetic compound ever used 5-ium,3,7-bis, (dimethylamino)-, chloride;
as an antiseptic in clinical therapy, and the first Swiss blue; Tetramethylene blue; Tetramethyl
antiseptic dye to be used therapeutically. In fact, thionine chloride (NTP, 2008; PubChem,
the use of methylene blue and its derivatives was 2013).
widespread before the advent of sulfonamides
and penicillin (Oz et al., 2011).
1.1.2 Structural and molecular formulae and
relative molecular mass
1.1 Chemical and physical data
N
1.1.1 Nomenclature
Chem. Abstr. Serv. Reg. No.: 61-73-4 (anhy- H3C CH 3

N S+ N
drous); 7220-79-3 (methylene blue trihydrate)
CH 3 Cl- CH 3
According to recent research, methylene
blue occurs in the form of several different
hydrates, but not as trihydrate (Rager et al., C16H18ClN3S
2012). [The Working Group noted that most Relative molecular mass (anhydrous form):
of the scientific literature refers only to 319.85 (PubChem, 2013)
155
IARC MONOGRAPHS 108
1.1.3 Chemical and physical properties of the 3-(Amino)-7-(methylamino)phenothiazin-

pure substance 5-ium chloride (azure C) (PubChem, 2013)
N
Description: Dark green crystals or crystal-
line powder with bronze lustre, odourless, CH 3
stable in air, deep blue solution in water or S +
N
alcohol, forms double salts (PubChem, 2013) H2 N
Cl-
Melting point: 100110 C (decomposition) H
(PubChem, 2013)
Density: 1.0 g/mL at 20 C (ChemNet, 2013) 1.2 Analysis
Solubility: 43.6 g/L in water at 25 C; also
soluble in ethanol (PubChem, 2013) There are several compendial and non-com-
Vapour pressure: 1.30107 mm Hg at 25 C pendial methods for the analysis of methylene
(estimated) (PubChem, 2013). blue (Table 1.1). To quantify methylene blue in
formulations, ultraviolet-visible spectroscopy
can be conducted. For the quantification of
1.1.4 Technical products and impurities methylene blue in biological specimens, liquid
chromatography coupled with different detec-
(a) Trade names
tors seems to be the method of choice.
Desmoid piller; desmoidpillen; panatone;
urolene blue; vitableu (NTP, 2008)
(b) Impurities
1.3.1 Production
3-Amino-7-(dimethylamino)phenothiazin- Methylene blue is synthesized commercially
5-ium chloride (azure A) (PubChem, 2013) by oxidation of N,N-dimethyl-phenylenediamine
N with sodium dichromate (Na2Cr2O7) in the
presence of sodium thiosulfate (Na2S2O3),
H3C
followed by further oxidation in the presence of
N S+ NH 2 N,N-dimethylaniline (NTP, 2008). Methylene
Cl- blue hydrochloride is isolated by addition of 30%
CH 3
hydrochloric acid and of a saturated common
salt solution to the dye solution; after filtration,
3 -(Di met hyla m i no)-7-(met hyla m i no) the product is washed with a 2% common salt
phenothiazin-5-ium chloride or N,N,N'- solution. Instead of sodium dichromate, manga-
trimethylthionin (azure B) (PubChem, 2013) nese dioxide, and catalytic amounts of copper
sulfate can be used for the oxidation (Berneth,
N 2008).
Methylene blue of high purity can be obtained
H3C CH 3 by chloroform extraction of impurities from
N S+ N
solutions of raw dye in borate buffer at pH 9.510,
H Cl- followed by acidification of the aqueous solution
CH 3
and isolation of the dye (Berneth, 2008).
156
Table 1.1 Some compendial and non-compendial methods for the analysis of methylene blue
Matrix Sample preparation Assay method Detection limit Reference

Compendial methods
Assay UV-visible spectroscopy US Pharmacopeial
Wavelength: 663 nm Convention (2013)
Assay Iodimetric titration British
Titration with sodium thiosulfate using starch Pharmacopoeia
solution as indicator Commission (2005)
Related LC-UV British
substance test Column: C18 Pharmacopoeia
Mobile phase: acetonitrile and phosphoric acid Commission (2005)
(3.4mL in 1000mL of water) (27:73, v/v)
Flow rate: 1mL/min
Wavelength: 246 nm
Non-compendial methods
Human Addition of NaCl and dichloroethane, UV-visible spectroscopy 0.02 g/mL (LOD) DiSanto & Wagner
blood centrifugation, analysis of dichloroethane Wavelength: 660 nm (1972)
layer
Human urine Addition of NaCl and dichloroethane, UV-visible spectroscopy 0.02 g/mL (LOD) DiSanto & Wagner
centrifugation, analysis of dichloroethane Wavelength: 660 nm (1972)
layer
Rat tissue Blotting on filter paper, addition of 0.1N UV-visible spectroscopy 0.02 g/mL (LOD) DiSanto & Wagner
hydrochloric acid, homogenization, Wavelength: 660 nm (1972)
addition of NaCl and dichloroethane,
centrifugation, analysis of dichloroethane
layer
Human Haemolysis, addition of sodium UV-visible spectroscopy 0.1 g/mL (LOQ) Belaz-David et al.
blood hexanesulfonate, extraction Wavelength: 657 nm (1997)
(dichloroethane), centrifugation, analysis of
organic layer
Human Addition of sodium hexanesulfonate, UV-visible spectroscopy 0.1 g/mL (LOQ) Belaz-David et al.
plasma extraction (dichloroethane), centrifugation, Wavelength: 657 nm (1997)
analysis of organic layer
Human urine Reduction of leucomethylene blue UV-visible spectroscopy 3 g/mL (LOQ) Belaz-David et al.
into methylene blue, addition of Wavelength: 657 nm (1997)
sodium hexanesulfonate, extraction
(dichloroethane), centrifugation, analysis of
organic layer
157
Methylene blue
158

Human Mixing with sodium hexanesulfonate, LC-UV 9 nmol/L (LOQ) Peter et al. (2000)
blood extraction (dichloroethane), centrifugation,
Column: cyano
evaporation Mobile phase: ammonium dihydrogen phosphate,
acetonitrile and methanol
pH 2.75
Flow rate: 0.7 mL/min
Wavelength: 660 nm
IARC MONOGRAPHS 108
Human urine Reduction of leucomethylene blue LC-UV 9 nmol/L (LOQ) Peter et al. (2000)
into methylene blue, mixing with Column: cyano
sodium hexanesulfonate, extraction Mobile phase: ammonium dihydrogen phosphate,
(dichloroethane), centrifugation, acetonitrile and methanol
evaporation pH 2.75
Wavelength: 660 nm
Human blood Precipitation with acetonitrile, LC-ESI-MS 0.5 ng/mL (LOQ) Rengelshausen et al.
and plasma centrifugation, and analysis of clear Column: C18 (2004)
supernatant Mobile phase: 0.1% acetic acid in 5 mM acetate buffer
and acetonitrile
Human blood Acidic protein precipitation, centrifugation, IEX-MS 75 ng/mL (LOQ) Burhenne et al.
and plasma analysis of clear supernatant Column: uptisphere mixed mode (2008)
Mobile phase: 0.1% acetic acid including 100mM
ammonium acetate (solvent A) and 2.5% formic acid/
acetonitrile (1 : 1, v/v) including 500 mM ammonium
acetate (solvent B)
Dried blood Cutting of paper sheet, soaking in IEX-MS 75 ng/mL (LOQ) Burhenne et al.
demineralized water, ultrasonication, Column: uptisphere mixed mode (2008)
protein precipitation, and analysis of clear Mobile phase: 0.1% acetic acid including 100mM
supernatant ammonium acetate (solvent A) and 2.5% formic acid/
acetonitrile (1 : 1, v/v) including 500 mM ammonium
acetate (solvent B)
Human urine Dilution of urine FIA-PIF 16 ng/mL (LOD) Laassis et al. (1994)
Wavelength: ex at 345 nm and em at 485 nm 0.06 g/mL (LOQ)
pH 13
Flow rate: 2 mL/min

Human urine Addition of sodium hexanesulfonate, CE-UV 1 g/mL (LOQ) Borwitzky et al.
extraction (dichloromethane), evaporation, Extended light path(bubble) capillary (2005)
reconstitution in water Mobile phase: 100 mM phosphate buffer with 25%
acetonitrile
pH 2.5
Wavelength: 292 and 592 nm
Rat urine and Addition of 1 M sodium chloride solution, LC-UV 3.9 ng/mL (LOD) Gaudette & Lodge
mouse urine mixing, addition of dichloroethane, Column: C18 13 ng/mL (LOQ) (2005)
centrifugation, collection of dichloroethane Mobile phase: acetonitrile and 0.1% trifluoroacetic
layer, evaporation, reconstitution in 0.1% acid in water
trifluoroacetic acid and acetonitrile pH adjusted to ~2.74 with triethylamine
Flow rate: 1 mL/min
Wavelength: 660 nm
Rat blood Addition of p-toluene sulfonic acid, CE-ESI-MS 0.22 g/mL (LOD) Yang et al. (2011)
buffering at pH 3 with ammonium Fused silica capillary 0.5 g/mL (LOQ)
acetate buffer, addition of acetonitrile Electrolyte: 2 mol/L acetic acid
and ultrasonic extraction, defatting Sheath liquid: methanol : water (80:20, v/v)
of liquid phase with hexane, addition
of dichloromethane, centrifugation,
evaporation, reconstitution in water
Cows milk Addition of acetonitrile, centrifugation, LC-UV 2.5 ppb [ng/mL] Munns et al. (1992)
transferring of liquid into separating Column: cyano (LOD)
funnel, addition of NaCl, extraction with Mobile phase: acetonitrile and acetate buffer 5 ppb [ng/mL]
chloroform twice, collection of lower pH 4.5 (LOQ)
layer, evaporation, dissolve in acetonitrile, Flow rate: 1 mL/min
column clean-up with CBA column, Wavelength: 627 nm
evaporation of eluent, reconstitution in
methanol
Muscle of Addition of McIlvaine buffer (pH 3.0), LC-UV 3 g/kg (LOD) Kasuga et al. (1991)
fish (rainbow homogenization, addition of acetonitrile, Column: C18
trout) centrifugation, washing of supernatant Mobile phase: 0.1 M citrate buffer, acetonitrile
with n-hexane twice, addition of 10% NaCl pH 3.0
solution and dichloromethane, addition of Flow rate: 0.8 mL/min
sodium sulfate to dichloromethane layer, Wavelength: 636 nm
filtration, evaporation, reconstitution with
methanol
159
Methylene blue
160

Fish tissue Homogenization with ammonium acetate LC-ESI-MS 23.8 g/kg Tarbin et al. (2008)
(pH 4.5) and acetonitrile, addition of Column: C18 (LOD)
basic aluminium oxide, centrifugation, Mobile phase: ammonium acetate and acetonitrile
transferring of supernatant into separating pH 4.5
funnel, re-extraction of solid residue Flow rate: 0.3 mL/min
in the same manner, further extraction
(dichloromethane), addition of DDQ and
formic acid to dichloromethane layer,
IARC MONOGRAPHS 108
clean-up with isolute strong cation-

exchange cartridge
Edible Addition of p-toluene sulfonic acid, LC-ESI-MS 0.1 g/kg (LOD) Xu et al. (2009)
aquatic buffering at pH 4.5 with sodium Column: C18 0.5 g/kg (LOQ)
products acetate buffer, extraction (acetonitrile, Mobile phase: methanol, 0.1% formic acid
(eel, shrimp) dichloromethane and diglycol), pH 4.5
centrifugation, evaporation, reconstitution Flow rate: 250 L/mL
in acetonitrile, clean-up with neutral
alumina and weak cation-exchange
cartridges, evaporation, reconstitution in
3:7 (v/v) methanol : water solution
Formulation LC-ED 3 pmol (LOD) Roybal et al. (1989)
Column: cyano
Mobile phase: methanol, 0.1 M sodium acetate
pH 4.5
Formulation First derivative UV spectroscopy 6 g/mL (LOQ) Onur & Acar (1992)
Wavelength: 273 nm
Formulation HPLC-PO-CL 120 fmol (LOD) Kimoto et al. (1996)
Colum: C18
Mobile phase: acetonitrile and 25 mM imidazole
buffer containing 10 mM sodium 1-propanesulfonate
pH 6.5
CL reaction solution: 0.25 mM TDPO and 25 mM
H2O2 in acetonitrile
Flow rate: for eluent, 1 mL/min; and for CL solution,
1.3mL/min
ex, excitation; em, emission; CBA, carboxylic acid; CE-ESI-MS, capillary electrophoresis/electrospray ionization mass spectrometry; CE-UV, capillary electrophoresis ultraviolet
spectroscopy; CL, chemiluminescence; DDQ, 2,3-dichloro-5,6-dicyano-1,4-benzoquinone; FIA-PIF, flow injection analysis photochemically induced fluorescence; HPLC-PO-CL, high-
performance liquid chromatography peroxyoxalate chemiluminescence; IEX-MS, ion exchange chromatography mass spectrometry; LC-ED, liquid chromatography electrochemical
detection; LC-ESI-MS, liquid chromatography electrospray ionization mass spectrometry; LC-UV, liquid chromatography ultraviolet spectroscopy; LOD, limit of detection; LOQ, limit
of quantitation; ppb, parts per billion; NaCl, sodium chloride; TDPO, bis(4-nitro-2-(3,6,9-trioxadecyloxycarbonyl)phenyloxalate
Methylene blue
1.3.2 Medical use malaria in children in Africa. In malaria combi-

nation therapy, methylene blue is also advanta-
(a) Indications geous because the blue colour of the urine can be
Methylene blue is used in human and veter- used as an indicator that the drug combination
inary medicine for several therapeutic and containing methylene blue has not been counter-
diagnostic procedures, including as a stain in feited, which is a serious problem in developing
bacteriology, as a redox colorimetric agent, as a countries (Schirmer et al., 2011). Some phase II
targeting agent for melanoma, as an antihaemo- trials have shown promising results, especially
globinaemic, as an antidote, and as an antiseptic when methylene blue is combined with a more
and disinfectant (ONeil et al., 2006; NTP, 2008). rapidly acting partner drug (Zoungrana et al.,
Methylene blue is used clinically in a wide 2008; Coulibaly et al., 2009; Bountogo et al.,
range of indications, including the emergency 2010).
treatment of methaemoglobinemia, ifosfa-
mid-induced encephalopathy or poisoning by (b) Dosage
cyanide, nitrate or carbon monoxide, and for In clinical use, methylene blue is either
intraoperative tissue staining (Oz et al., 2011; dissolved in sterile water to a concentration of
Schirmer et al., 2011). 10 mg/mL (1%) injectable solution or adminis-
One of the most common clinical applica- tered orally in gelatin capsules to avoid staining
tions of methylene blue is for the treatment of of the oral mucous membranes and to ensure
methaemoglobinaemia induced by overexposure complete gastrointestinal delivery (Oz et al., 2011).
to drugs, to industrial chemicals such as nitro- The dosage depends on the therapeutic indication
phenols (ATSDR, 1992), or to environmental (Schirmer et al., 2011). For inherited methaemo
poisons such as excessive nitrate in well-water, globinaemia, the suggested oral dosage was
or cyanide compounds (Sills & Zinkham, 1994; 150250mg/day (for a lifetime), while for acute
Christensen et al., 1996). methaemoglobinaemia the suggested dosage
Methylene blue is used in the treatment of was 12 1.3 mg/kg body weight (bw), given
some psychiatric disorders because of the anxio- intravenously over 20 minutes. In ifosfamid-
lytic and antidepressant properties attributed to induced neurotoxicity, oral or intravenous doses
its ability to block activation of guanyl cyclase of 4 50 mg/day were used. For prevention of
by nitric oxide (Naylor et al., 1986; Erolu & urinary-tract infections in elderly patients, a dose
Calayan, 1997). In 2011, however, the Food and of 365mg/day was given orally. In Alzheimer
Drug Administration of the United States issued disease, the dosage was 360mg/day, and for
a safety warning concerning the risk of serotonin paediatric malaria it was 212mg/kg bw orally
syndrome when methylene blue is given concur- for 3days (Schirmer et al., 2011). In a controlled
rently with serotonergic psychiatric medications trial in semi-immune adults with uncomplicated
(FDA, 2011). falciparum malaria, the oral dosage was 390mg
Recent studies suggested that methylene blue twice per day (Bountogo et al., 2010). According
may have beneficial effects in the treatment of to Medscape (2013), a solution (10mg/mL) may
Alzheimer disease and memory improvement be injected at the following intravenous dosages:
(Oz et al., 2011). 12mg/kg bw over 510minutes for methaemo-
The use of methylene blue as a candidate globinaemia, and 50mg every 6 to 8hours until
antimalarial drug was revived in 1995, with the symptoms resolve for prevention of ifosfamid-
major goal to develop an affordable, available, and induced encephalopathy.
accessible therapy for uncomplicated falciparum
161
IARC MONOGRAPHS 108
(c) Sales volume in samples of dried whole blood on paper spots

Worldwide sales of methylene blue totalled ranged between 531 and 2645 ng/mL within
US$44million in 2012, with 59% occurring in the 1 hour after administration (Burhenne et al.,
USA. The only other nation to report substantial 2008). In a phase 1 study of malaria treatment,
sales volumes was Brazil (US$11million) (IMS mean plasma concentrations after a single
Health, 2012). dose of methylene blue in healthy adults were
748ng/mL (50mg, intravenous injection; n = 16)
1.3.3 Other uses and 3905 ng/mL (500 mg, oral administration;
n = 16) (Walter-Sack et al., 2009).
Methylene blue is used as a disinfectant and No systematic data on other exposures, e.g.
biological stain (NTP, 2008; Oz et al., 2011). As a
environmental contamination, were available to
disinfectant, methylene blue is sold to end-con-
sumers as an aquarium fungicide (Schirmer et al., the Working Group. While methylene blue may
2011). Most recently, methylene blue has been hypothetically enter the food chain after appli-
used as an optical probe in biophysical systems, cation in veterinary medicine (which would be
as an intercalator in nanoporous materials, as illegal in most jurisdictions), or as a contaminant
a redox mediator, and in photoelectrochromic in drinking-water, no systematic data on residue
imaging (NTP, 2008). levels in food or water were available. In the few
Methylene blue is used to dye paper and office available studies, it was found that metabolites
supplies, but also to tone up silk colours (Berneth, rather than methylene blue itself were detect-
2008). In analytical chemistry, methylene blue is able, e.g. in milk from dairy cattle treated with
applied to determine anionic surfactants, which methylene blue (Roybal et al., 1996).
are termed methylene blue active substances
(Kosswig, 2000). Methylene blue is also used in
pH and redox indicator reagents (Sabnis et al., 1.5 Regulations and guidelines
2009).
No permissible exposure limits for methylene
blue have been established in the USA by the
1.4 Occurrence and exposure Occupational Safety and Health Administration,
1.4.1 Natural occurrence the National Institute for Occupational Safety
and Health, or the American Conference of
Methylene blue is a synthetic substance and Governmental Industrial Hygienists (NTP, 2008).
does not occur naturally. In the European Union, the use of methylene
blue in food-producing animals is not allowed.
1.4.2 Occupational exposure According to Xu et al. (2009), Japan has estab-
A National Occupational Exposure Survey lished a maximum residue limit of 10g/kg for
in the USA indicated that an estimated 69 563 methylene blue in aquatic products, because it is
workers were potentially exposed to methylene used as a replacement for other antifungal dyes
blue in the workplace between 1981 and 1983 in aquaculture.
(NTP, 2008). Specifications for methylene blue are
published in several official pharmacopoeias
1.4.3 General population and consumers (Table1.2).
In 20 paediatric patients in Burkina Faso who
were treated for malaria with methylene blue at
an oral dose of 20mg/kg bw, the concentrations
162
Table 1.2 Specifications for methylene blue
Parameter WHO International Pharmacopoeia, United States European Pharmacopoeia 7.0

4th edition Pharmacopoeia 36
Content C16H18ClN3S (dried 97.0101.0% 98.0103.0% 95.0101.0%
substance)
Identity tests A. IR IR A. UV/VIS
B. Colour reaction with hydrochloric B. TLC
acid and zinc powder C. Colour reaction with glacial acetic acid and zinc
C. General identification test as powder
characteristic of chlorides D. Reaction of chlorides
Copper or zinc Absence of zinc; copper, max. Absence of zinc; copper Zinc, max. 100 ppm; copper, max. 300 ppm
0.20 mg/g max. 0.02%
Metals besides copper and zinc Iron, max. 0.10 mg/g Arsenic, max. 8 ppm Max. contents: aluminium, 300 ppm; cadmium,
1 ppm; chromium, 100 ppm; tin, 10 ppm; iron, 200 ppm;
manganese, 10 ppm; mercury, 1 ppm; molybdenum, 10
ppm; nickel, 10 ppm; lead, 10 ppm
Sulfated ash Max. 10 mg/g Max. 0.25%
Loss on drying 80220 mg/g 8.018.0% 8.022.0%
Foreign substances/chromatographic TLC: no spots besides the TLC: max. four spots HPLC: detailed specification of max. peak areas of
purity/related substances characteristic spots impurities
Residue on ignition Max. 1.2%
Organic volatile impurities Meets the requirements
Bacterial endotoxins Max. 2.5 IU of endotoxin per mg
Methanol-insoluble substances Max. 10.0 mg (1.0%)
HPLC, high-performance liquid chromatography; IR, infrared; IU, international unit; max., maximum; TLC, thin-layer chromatography; UV/VIS, ultraviolet and visible absorption
spectrophotometry
From EDQM (2008), WHO (2011), US Pharmacopeial Convention (2013)
163
Methylene blue
IARC MONOGRAPHS 108
2. Cancer in Humans for all routes of administration (151/1507; range,

424%) and the incidence in controls in the
No data were available to the Working Group. current study was below the range for historical
controls. [The Working Group considered that
the significantly increased incidence and signifi-
3. Cancer in Experimental Animals cant positive trend in the incidence of bronchiolo-
alveolar carcinoma was therefore not related to
treatment with methylene blue.] In females, the
3.1 Mouse incidences of bronchiolo-alveolar carcinoma
In a study of oral administration, groups of 50 were decreased in all groups of treated mice
male and female B6C3F1 (age, 6 weeks) received (5/50, 0/50, 0/50, 1/50), and the decreases were
methylene blue (in a 0.5% aqueous methylcellu- significant (P 0.05, poly-3 test) in the groups
lose solution) at a dose of 0 (control), 2.5, 12.5, receiving the lowest and intermediate dose.
or 25mg/kg bw per day by gavage on 5days per The incidence of malignant lymphoma in
week for up to 106weeks. There was an increase females occurred with a significant positive
in mean body weight in females at the interme- trend: 6/50 (12%), 4/50 (8%), 9/50 (18%), 12/50
diate and highest doses compared with controls. (24%); P=0.025, poly-3 trend test. However, the
Survival of treated groups was similar to that of incidence in females at the highest dose (24%)
controls. was well within the range for historical controls
In males, there was a significant positive (308/1508; range, 658%) for this neoplasm with
increase in the trend in the incidence of carci- a highly variable incidence. In males, the inci-
noma (P = 0.027, poly-3 trend test) and of adenoma dences were 2/50 (4%), 2/50 (4%), 2/50 (4%), 5/50
or carcinoma (combined) of the small intestine (10%). While the incidence in the group at the
(P = 0.029, poly-3 trend test). The incidences highest dose was higher than in controls, it was
of carcinoma were: 0/50 (0%), 1/50 (2%), 2/50 not significantly increased, and barely exceeded
(4%), 4/50 (8%); and the incidences of adenoma the range for historical controls (70/1508; range,
or carcinoma (combined) were: 1/50 (2%), 2/50 08%) (NTP, 2008; Auerbach et al., 2010).
(4%), 4/50 (8%), 6/50 (12%). The incidences in
the dosed groups were not significant by pair- 3.2 Rat
wise comparison. The incidence of adenoma or
carcinoma (combined) in the group receiving the In a study of oral administration, groups of
highest dose (12%) exceeded the range for histor- 50 male and 50 female F344/N rats (age, 6 weeks)
ical controls (39/1508; range, 010%); while the received methylene blue in a 0.5% aqueous
incidence in controls (2%) was consistent with methylcellulose solution at a dose of 0 (control),
the range for historical controls. 5, 25, or 50mg/kg bw, by gavage once per day on
In males, the incidence of bronchiolo-alveolar 5days per week for up to 106weeks. The mean
carcinoma of the lung occurred with a signifi- body weights of males and females in groups at
cant positive trend: 1/50 (2%), 4/50 (8%), 5/50 the intermediate and highest dose were decreased
(10%), 7/50 (14%); P = 0.043, poly-3 trend test); compared with controls at the end of the study.
and the incidence was significantly increased in There was no effect on body weight in groups at
the group at the highest dose (P=0.039; poly-3 the lowest dose. Survival of treated groups was
test). The incidence in males receiving methylene similar to that of the controls.
blue were within the range for historical controls In males, the trend in the incidence of
pancreatic islet cell adenoma and of adenoma
164
Methylene blue
or carcinoma (combined) were non-significantly Methylene blue is well absorbed, reduced,

increased. The incidences of adenoma were: 4/50 and excreted largely in the urine as the reduced
(8%), 9/50 (18%), 12/50 (24%), and 8/50 (16%); leucomethylene blue (colourless) form (DiSanto
and the incidences of adenoma or carcinoma & Wagner, 1972a; Fig.4.1). The N-demethylated
(combined) were: 4/50 (8%), 9/50 (18%), 14/50 metabolites azure A (minor), azure B, and azure
(28%), and 8/50 (16%). The incidences were signif- C (minor), which have the potential to undergo
icantly increased only in the group receiving the deprotonation to a neutral quinone imine, have
intermediate dose (adenoma, P=0.037; adenoma been reported (Munns et al., 1992; Schirmer et al.,
or carcinoma (combined), P=0.013; poly3-test), 2011; Fig.4.2), but their pharmacokinetic charac-
and the incidence of islet cell carcinoma of teristics do not appear to have been investigated.
the pancreas (2/50; 4%) in the group receiving One study mentioned the presence of azure B in
the intermediate dose was within the range autopsied peripheral organs from a patient who
for historical controls (26/1448; range, 08%). had received 200 mg of methylene blue intra-
[Although the incidence of pancreatic islet cell venously, at levels (4752943 ng/g) higher than
hyperplasia was significantly increased in the those (74208 ng/g) of methylene blue in the
group at the highest dose versus controls (26/50 same tissues (Warth et al., 2009). [The Working
versus 13/50; P0.01), and in view of the fact that Group noted that the metabolites of methylene
islet cell hyperplasia, adenoma, and carcinoma blue are anticipated to have greater lipophilicity
are thought to constitute a morphological and than the parent compound and may accumulate
biological continuum in the progression of islet in tissues.]
cell proliferation, the Working Group consid- When administered orally to seven healthy
ered that the positive trend in the incidence of human subjects at a dose of 10 mg in capsule
adenoma or carcinoma (combined) was mainly form, the total urinary recovery ranged from
the result of the increased trend in the incidence 53% to 97% of the administered dose, with an
of adenoma]. average of 74%. Of the material recovered, an
There was no increase in the incidence of average of 78% was excreted as leucomethylene
any neoplasm in exposed females (NTP, 2008; blue and the remainder as methylene blue.
Auerbach et al., 2010). Excretion ratetime plots for methylene blue
and leucomethylene blue suggested a circadian
rhythm (DiSanto & Wagner, 1972a).
4. Mechanistic and Other In another study, the concentration of
Relevant Data methylene blue in whole blood was measured
in healthy individuals, before and after oxida-
tion, following intravenous (n = 7) or oral (n =
4.1 Absorption, distribution, 7) administration of 100 mg of methylene blue.
metabolism, and excretion The concentration of methylene blue in whole
blood after intravenous administration showed
4.1.1 Humans a multiphasic time course with an estimated
After an intravenous bolus injection of 100 mg, terminal half-life of 5.25 hours. The area under
the mean plasma concentration of methylene the curve (AUC) was 0.1340.025 mol/mL.min
blue was reported to be 5 M in healthy volun- and the systemic clearance was 3.00.7 L/min.
teers [number not specified] (Aeschlimann et al., After oral administration (in capsule form),
1996). maximum concentrations were reached within
12hours; the AUC (0.010.004mol/mL.min)
165
IARC MONOGRAPHS 108
Fig.4.1 Structures of methylene blue and leucomethylene blue
H3C CH 3 Methylene blue

N S+ N
CH 3 CH 3
+H+ -H +
+2e- -2e -
H
N
H3C CH 3 Leucomethylene blue

N S N
CH 3 CH 3
was one order of magnitude lower than upon (time to peak, 23minutes) and an average peak
intravenous administration. The urinary excre- serum concentration of 71.3 ng/mL. The elim-
tion of total methylene blue (methylene blue and ination was slow (t1/2=11.1hours), and 32% of
leucomethylene blue), between 4 and 14 hours, the initial dose was recovered within 48hours.
was significantly (P < 0.01) higher after intra- The highest serum concentration was 280ng/mL
venous administration than after oral admin- (Pruthi et al., 2011). Of note, methylene blue
istration (28.6 3.0% and 18.4 2.4% of the concentrations have been found to be four- to
administered dose, respectively). In this study, fivefold higher in whole blood than in plasma
approximately one third of the methylene blue (Peter et al., 2000; Rengelshausen et al., 2004).
excreted in the urine was in the leuco form (Peter [The Working Group noted that leuco-
et al., 2000). methylene blue is readily oxidized in air and
Another study compared the administra- forms stable complexes in the urine, but not
tion of single doses of methylene blue: 50 mg blood (DiSanto & Wagner, 1972b, c). It is not
intravenously (n = 16) versus 500 mg orally clear whether or not discrepancies in the relative
(n = 16). The mean plasma AUCs were estimated proportions of methylene blue and the leuco form
to be 7.63.4g/mL.h and 51.217.1g/mL.h between studies may be due to different aeration
after intravenous and oral administration, conditions during sample processing.]
respectively. The absolute bioavailability was
72.323.9% (Walter-Sack et al., 2009). 4.1.2 Experimental animals
The pharmacokinetics of methylene blue were
investigated in the setting of lymphatic mapping In one male and one female dog given
of cancer of the breast. A subareolar injection of methylene blue orally at a dose of 15mg/kg bw,
4mL of a methylene blue solution at 1.25mg/mL methylene blue was not detectable in the blood. The
(total dose, 5 mg) resulted in rapid absorption female was catheterized and urine was collected
166
Methylene blue
Fig.4.2 Structures of the methylene blue metabolites azure B, azure A, and azure C
N
Azure B: R = C H 3
R CH 3 Azure A: R = H
N S+ N
H CH 3
- H+ + H+
R CH 3 Quinone imine
N S N
CH 3
CH 3 Azure C
H2N S+ N
H
- H+ + H+
N
Quinone imine
CH 3
H2N S N
for 10hours after dosing; the recovery was 2.4% extensive uptake of methylene blue by tissues; the
of the administered dose. When the female was uptake was best described by a nonlinear model
given methylene blue orally at a dose of 10mg/kg (DiSanto & Wagner, 1972c).
bw, 3.8% of the administered dose was recovered The distribution of total methylene blue in
in the urine within 14hours (DiSanto & Wagner, different tissues of male Wistar rats was meas-
1972a). In comparison with the data obtained ured after intravenous or intraduodenal admin-
for humans in the same study (see Section 4.1.1), istration of a single dose at 10 mg/kg bw. The rats
this low recovery indicated that methylene blue were killed after 1hour and samples from several
is well absorbed in humans and poorly absorbed different tissues were collected. The concentra-
in dogs after oral administration. tions of the drug in the blood and brain were
In another study, male Sprague-Dawley rats significantly higher (P<0.05) after intravenous
were treated intravenously with methylene blue than after intraduodenal administration. In
at a dose of 225 mg/kg bw and killed 3minutes contrast, the concentrations in the intestinal
after dosing; lungs, liver, kidneys, and heart wall and in the liver were significantly (P<0.05)
were removed and assayed for methylene blue. higher after intraduodenal administration, while
An average of 29.8% of the administered dose concentrations in bile, and biliary excretion, were
(range, 25.235.8%) was recovered in the four not affected by the route of administration. Less
tissues, which is consistent with very rapid and than 3% of the administered dose was found in
167
IARC MONOGRAPHS 108
the intestinal lumen 1hour after intraduodenal oesophageal mucosal sites that were treated with
administration (Peter et al., 2000). methylene blue, and at adjacent sites not treated
When a 10% solution of methylene blue with methylene blue. Comet assays revealed that
was administered by intramammary infusion elevated levels of DNA damage were observed in
to lactating goats, the drug passed quickly into oesophageal mucosal cells exposed to methylene
systemic circulation, peaked at 3hours, and was blue in all 15 patients, while samples adjacent
still detectable in the blood 12hours after infu- to the methylene blue-exposed sites had signif-
sion (Ziv & Heavner, 1984). icantly lower levels of DNA damage, despite
Azure B, together with methylene blue and photosensitization with white light from the
leucomethylene blue, was reported to be present endoscope (Olliver et al., 2003). Exposure in
in the urine of male and female Fischer 344 rats vitro of normal oesophageal tissue, obtained by
(n = 5) given methylene blue as a single intrave- biopsy, to methylene blue (0.5% for 1 minute) in
nous dose of 2.5mg/kg bw, or a single oral dose the absence of light did not result in an increase
of either 2.5 or 50mg/kg bw. The methylene blue in DNA damage (Olliver et al., 2003), confirming
used in the experiment was contaminated with the role of white light-activated methylene blue
azure B at approximately 15%; metabolism of in the induction of DNA damage. Similarly, an
methylene blue through N-demethylation was increase in DNA damage (alkali-labile sites and
inferred from a time-dependent increase in the FPG-sensitive sites) was seen in biopsied colonic
amount of azure B present in the urine, but quan- epithelium sprayed with methylene blue dye
tification of azure B was not provided (Gaudette (0.1%) during colonoscopy (which used illumi-
& Lodge, 2005). nation with white light) compared with colonic
Methylene blue was reported to bind epithelial cells sampled in the same region before
strongly to rabbit plasma (7177% of bound spraying with methylene blue (Davies et al.,
drug). Extensive tissue and protein binding was 2007).
proposed to account for the high apparent volume
of distribution (21 L/kg) in rabbits (Kozaki & 4.2.2 Experimental systems
Watanabe, 1981).
(a) Mutation
(i) Assays in bacteria or yeast
4.2 Genetic and related effects
Methylene blue was shown to be muta-
See Table4.1 genic without photoactivation in a variety of
Salmonella typhimurium tester strains, inducing
4.2.1 Humans both base-substitution and frameshift muta-
tions, with and without metabolic activation
In mucosal cells from Barrett oeosophagus in (Chung et al., 1981; Yamaguchi, 1981; Lunn &
humans undergoing endoscopy, methylene blue Sansone, 1991; NTP, 2008); mutagenic activity
dye (0.5% solution) (which was used to identify or induction of DNA damage was also reported
specific areas of interest for biopsy) induced DNA in several strains of Escherichia coli (McCarroll
damage as detected by the alkaline comet assay et al., 1981; Mohn et al., 1984; Webb & Hass,
and the modified comet assay using the enzyme 1984; NTP, 2008). In contrast, photoactivated
formamide pyrimidine-DNA glycosylase (664 nm) methylene blue did not induce gene
(FPG) to detect damage associated with reac- conversion in the yeast Saccharomyces cerevisiae
tive oxygen species (Olliver et al., 2003). Fifteen (Ito & Kobayashi, 1977), and no induction of
patients undergoing endoscopy were biopsied at gene mutation was seen in S. cerevisiae treated
168
Table 4.1 Genetic and related effects of methylene blue and its metabolites
Test system Resultsa Dose Reference

(LED or HID)
metabolic system metabolic systemb
Methylene blue
Bacteriophage PM2 cell-free, DNA damage, in the presence of white- + NT 10g/mL Epe et al. (1988)
light activation
Bacteriophage pAQ1 in Salmonella typhimurium TA1535 and TA1978, +c NT 10M Epe et al. (1989)
DNA damage, in the presence of white-light activation
Bacteriophage PM2 cell-free, DNA damage, in the presence of white- +c NT 27 M Epe et al. (1993)
light activation
Bacteriophage pAQ1 in Salmonella typhimurium TA1978, DNA +c NT 27M Epe et al. (1993)
damage in PM2 with white-light activation
Single-stranded M13mp2 bacteriophage, DNA damage with + NT 2.5M McBride et al. (1992)
photoactivationd
Calf thymus DNA, intercalation, with photoactivation + NT 1.83M Lee et al. (1973)
Calf thymus DNA, intercalation, with photoactivation + NT NRe Nordn & Tjerneld
(1982)
DNAprotein crosslinks, calf thymus DNA, calf thymus histone type + NT 5M Villanueva et al. (1993)
II, with photoactivation
Salmonella typhimurium TA100, TA1535, TA1537, TA1538, TA98, + (TA98) + (TA98) 5g/plate Chung et al. (1981)
reverse mutation
Salmonella typhimurium TA100, reverse mutation + + 20g/plate Yamaguchi (1981)
Salmonella typhimurium TA100, TA1530, TA1535, TA98, reverse + (TA1530, TA98) + (TA98) 1000g/plate Lunn & Sansone (1991)
mutation
Salmonella typhimurium TA100, reverse mutation () +f 33g/plate NTP (2008)
Salmonella typhimurium TA98, reverse mutation + +f 33g/plate, S9 NTP (2008)
3.3g/plate, +S9
Salmonella typhimurium TA100, reverse mutation + + 0.25g/plate, S9 NTP (2008)
10g/plate, +S9
Salmonella typhimurium TA98, reverse mutation + + 1g/plate, S9 NTP (2008)
10 g/plate, +S9
Salmonella typhimurium TA1535, TA1538, reverse mutation, with and + (TA1535)g NT 20 g/plate Gutter et al. (1977)
without photoactivation
Salmonella typhimurium TA1535, TA2638, TA100, TA104, reverse + NT 10 g/mL Epe et al. (1989)
mutation, with photoactivation
169
Methylene blue
170

(LED or HID)
Escherichia coli WP2, WP2 uvrA, WP67 uvrA polA, CM611 uvrA + (CM611, WP100, NT 160g/well McCarroll et al. (1981)
lexA, WP100 uvrA recA, W3110 polA+, p3478 polA, DNA damage p3478) (p3478 polA)
Escherichia coli AB1157, B/r, WP2, WP2s,WP10, WP6 (polA1), + (AB1157, WP2s, NT 2M Webb & Hass (1984)
resistance to bacteriophage T5 WP10)
IARC MONOGRAPHS 108
Escherichia coli K-12/343/113, reverse mutation to Arg+, with white- + NT 1040M Mohn et al. (1984)
light activation (LED, NR)
Escherichia coli WP2 uvrA pKM101, reverse mutation + + 0.5 g/plate, S9 NTP (2008)
25g/plate, +S9
Saccharomyces cerevisiae, gene conversion, with white light NT 0.95 (ODmax)h Ito & Kobayashi (1977)
photoactivation (max 664 nm)
Saccharomyces cerevisiae 507.4/2b, MT182/8d, CM106/5a, gene NT 20g/mL Tuite et al. (1981)
mutations, no photoactivation
Bacteriophage Serratia phage kappa, mutagenicity, with + NT NR Brendel (1973)
photoactivation
DNA damage (alkali-labile sites) (comet assay), male Sprague-Dawley + NT 0.31M2min Lbaj et al. (2007)
rat, primary hepatocytes, with visible light activation in vitro
DNA damage (FPG-sensitive sites) (comet assay), male Sprague- + NT 0.31M2min Lbaj et al. (2007)
Dawley rat, primary hepatocytes, with visible light activation in vitro
DNA damage (alkali-labile sites; FPG-sensitive sites) (comet assay), NT 0.31M3min Lbaj et al. (2007)
male Sprague Dawley rat, primary hepatocytes, in vitro
DNA damage (alkali-labile sites; FPG-sensitive sites) (comet assay), + NT 0.31M3min Horvthov et al.
male Sprague-Dawley rat, primary hepatocytes, in vitro (2012)
DNA damage (alkali-labile sites) (comet assay), male Sprague Dawley + NT 0.31M3min Horvthov et al.
rat, primary hepatocytes, with visible light activation in vitro (2012)
DNA damage (FPG-sensitive sites) (comet assay), male Sprague- + NT 0.31M3min Horvthov et al.
Dawley rat, primary hepatocytes, with visible light activation in vitro (2012)
DNA damage (alkali-labile sites) (comet assay), MCF-7 cells, with + NT 0.1%5min Masannat et al. (2009)
visible light activation in vitro
DNA damage (FPG-sensitive sites) (comet assay), MCF-7 cells, with NT 1.0%5min Masannat et al. (2009)
DNA damage (alkali-labile sites) (comet assay), HB-2 cells, with visible + NT 1.0%5min Masannat et al. (2009)
light activation in vitro
DNA damage (FPG-sensitive sites) (comet assay), HB-2 cells, with NT 1.0%5min Masannat et al. (2009)
DNA damage (comet assay), CaCo-2 cells, in vitro NT 0.1%2min Davies et al. (2007)

(LED or HID)
DNA damage (alkali-labile sites) (comet assay), CaCo-2 cells, with + NT 0.1%2min Davies et al. (2007)
DNA damage (FPG-sensitive sites) (comet assay), CaCo-2 cells, with + NT 0.1%2min Davies et al. (2007)
DNA damage (alkali-labile sites) (comet assay), human colonic mucosa + 0.1% Davies et al. (2007)
cells, with visible light activation during colonoscopy in vivo
DNA damage (FPG-sensitive sites) (comet assay), human colonic + 0.1% Davies et al. (2007)
mucosa cells, with visible light activation during colonoscopy in vivo
DNA damage (comet assay), human Barrett oesophagus cells (biopsy), NT 0.5%1min Olliver et al. (2003)
in vitro
DNA damage (alkali-labile sites) (comet assay), human Barrett + 0.5% Olliver et al. (2003)
oesophagus cells, with visible light activation during endoscopy in
vivo
DNA damage (FPG-sensitive sites) (comet assay), human Barrett + 0.5% Olliver et al. (2003)
oesophagus cells, with visible light activation during endoscopy in
vivo
DNA damage (alkali-labile sites) (comet assay), human OE33 cells, + NT 15mM Sturmey et al. (2009)
with white-light activation in vitro (0.5%)5min
DNA damage (alkali-labile sites) (comet assay), human OE33 cells, + NT 15mM Sturmey et al. (2009)
with red light activation in vitro (0.5%)5min
DNA damage (FPG-sensitive sites) (comet assay), human OE33 cells, + NT 1.5mM5min Sturmey et al. (2009)
with red light activation in vitro
DNA damage (alkali-labile sites) (comet assay), human OE33 cells, NT 15mM Sturmey et al. (2009)
with green light activation in vitro (0.5%)3min
with blue light activation in vitro (0.5%)3min
with filtered white light (to remove 580800 nm red spectrum) (0.5%)3min
activation in vitro
Drosophila melanogaster, sex-linked recessive lethal mutation, in germ 0.1% in feed Clark (1953)
cells, larval feeding
Drosophila melanogaster, somatic mutation and recombination test + 0.01mM in feed Smijs et al. (2004)
(SMART), with photoactivation
Sister-chromatid exchange, Chinese hamster V79 cells, in vitro NT 1.0g/mL Popescu et al. (1977)
171
Methylene blue
172

(LED or HID)
Sister-chromatid exchange, Chinese hamster V79 cells, in vitro, no + NT 0.1g/mL Speit & Vogel (1979)
photoactivation
Sister-chromatid exchange, Chinese hamster V79 cells, in vitro, with NT 1.0g/mL Speit & Vogel (1979)
photoactivation
IARC MONOGRAPHS 108
Sister-chromatid exchange, Syrian hamster BHK-1 cells, with/without NT 27g/mL MacRae et al. (1980)
photoactivation in vitro
Sister-chromatid exchange, Chinese hamster ovary cells, in vitro + + 0.63g/mL (S9) NTP (2008)
4.7g/mL (+S9)
Chromosomal aberrations, Chinese hamster ovary cells, in vitro NT 20Mi Au & Hsu (1979)
Chromosomal aberrations, Chinese hamster V79 cells, in vitro 1.0g/mL Popescu et al. (1977)
Chromosomal aberrations, Chinese hamster ovary cells, in vitro + + 7.5g/mL (S9) NTP (2008)
4.7g/mL (+S9)
Sister chromatid exchanges, Chinese hamster bone-marrow cells, in 12mg/kg bw, Speit (1982)
vivo ip1
Micronucleus formation, male B6C3F1 mice, bone-marrow cells or 150mg/kg bw, NTP (2008)
peripheral blood erythrocytes, in vivo ip1
Micronucleus formation, male and female B6C3F1 mice, peripheral 200mg/kg bw per NTP (2008)
blood erythrocytes, in vivo day,
gavage14 wk
Azure A
50g/plate, +S9
100g/plate, +S9
Escherichia coli WP2 uvrA pKM101, reverse mutation + + 50g/plate, S9 NTP (2008)
250g/plate, +S9
Chromosomal aberrations, Chinese hamster ovary cells, in vitro + NT 10Mj Au & Hsu (1979)
Azure B
Salmonella typhimurium TA100, TA98, reverse mutation + + 10g/plate NTP (2008)
100g/plate, +S9

(LED or HID)
Azure C
100g/plate, +S9
250g/plate, +S9
100g/plate, +S9
a +, positive; , negative; (), equivocal
b S9 from Aroclor 1254-treated Sprague-Dawley rats, unless otherwise noted
c DNA damage was in the form of base modifications consistent with singlet oxygen generation
d 8-hydroxydeoxyguanosine and SOS-induced mutations implicating generation of lesions (ionic) other than 8-hydroxydeoxyguanosine in methylene blue plus white light oxidative
DNA damage
e Intercalation orientation is changed by ionic strength; at low ionic strength, methylene blue is oriented co-planar with the DNA bases and at higher ionic strength, orientation changes
f S9 from Aroclor 1254-treated Sprague-Dawley rats or Syrian hamsters
g Photoactivation required; no increase in mutations in the absence of photoactivation with white light. Doseresponse observed in the presence of white light (2-hour exposure) over a
range of 10100g/plate
h Concentrated stock solution was diluted with 0.067 M phosphate buffer to give a final concentration of OD 1 at its absorption peak
I Not possible to accurately interpret the data; duration of exposure was only 5hours, only 50 cells were evaluated for aberrations per concentration tested, gaps were included in the
overall assessment of chromosomal damage, and data were presented as total aberrations rather than percentage of aberrant cells
j Not possible to accurately interpret the data; high levels of cytotoxicity were noted at 10M for azure A. For azure B and C, only the cytotoxic concentration (20M) was tested
bw, body weight; HID, highest ineffective dose; ip, intraperitoneal; LED, lowest effective dose; min, minute; NR, not reported; NT, not tested; po, oral; wk, week
173
Methylene blue
IARC MONOGRAPHS 108
with methylene blue at a single concentration HB-2 normal human breast cells (Masannat
of 20 g/mL in the absence of photoactivation et al., 2009), cultured colonic adenocarcinoma
(Tuite et al., 1981). It was suggested that the nega- CaCo-2 cells (Davies et al., 2007), and Barrett-
tive results in the yeast assays resulted from the associated adenocarcinoma OE33 cells (Sturmey
inability of methylene blue to penetrate the yeast et al., 2009). Masannat et al. (2009) reported no
cell wall (Ito & Kobayashi, 1977). increase in the number of FPG-sensitive sites in
MCF-7 cells treated with 1% methylene blue for
(ii) Drosophila melanogaster
5minutes in the presence of white light, but alka-
No increase in the frequency of sex-linked li-labile sites were significantly increased by this
recessive lethal mutation was detected in germ treatment, as was total DNA damage. Similar
cells of male Drosophila melanogaster given results were reported by Sturmey et al. (2009)
methylene blue via a larval feeding regimen with OE33 cells treated with methylene blue
(Clark, 1953). However, when photoactivated and white light (significant increase in alkali-
with white light, methylene blue induced high labile sites, but not FPG-sensitive sites). In all
levels of homologous mitotic recombination other cell lines, DNA damage in the form of
in a somatic mutation and recombination test both alkali-labile sites and FPG-sensitive sites)
(SMART) in D. melanogaster (Smijs et al., 2004). was observed after treatment with methylene
blue in the presence of white light. To determine
(b) DNA damage
if one particular portion of the spectrum was
Positive results were reported in several involved in the photoactivation of methylene
in-vitro tests for mutagenicity or DNA damage blue, Sturmey et al. (2009) conducted a series
induction with photoactivated methylene blue, of experiments using white light and filtered
presumably the result of singlet oxygen produc- light to activate methylene blue and assess DNA
tion (Brendel, 1973; Gutter et al., 1977; Epe et al., damage levels in OE33 cells. The concentrations
1988, 1989, 1993; McBride et al., 1992). of methylene blue ranged from 0.015 to 15mM
Methylene blue was shown to intercalate into (0.00050.5%), with the highest concentration
calf thymus DNA (Lee et al., 1973), and to bind equal to the clinically relevant concentration
to calf thymus DNA in an orientation perpen- used in colonoscopies to visualize suspicious
dicular to the helix axis, coplanar with the bases, areas for biopsy. Only the highest concentration
at low methylene blue:DNA binding ratios and of methylene blue induced significant increases
low ionic strengths (Nordn & Tjerneld, 1982). in DNA damage in OE33 cells with white-light
Villanueva et al. (1993) reported that methylene activation. However, red light (580700 nm)
blue induced light-dose-dependent increases in induced DNA damage at a lower concentra-
DNAprotein crosslinks (calf thymus DNA, calf tion of methylene blue (1.5 mM or 0.05%) and
thymus histone type II), which was attributed to increased the frequency of both alkali-labile sites
the production of singlet oxygen. and FPG-sensitive sites; no increases in DNA
Several studies of DNA damage using the damage were seen when light was filtered to allow
comet assay have been conducted with the only the blue or the green portions of the spec-
majority demonstrating a requirement for trum to interact with methylene blue. Lowering
methylene blue activation by visible (white) light the concentration of methylene blue used in the
to induce both alkali-labile and FPG-sensitive clinic, and/or eliminating the red portion of the
(oxidized guanine) sites. Studies were conducted white-light spectrum used to illuminate colonic
in male Sprague-Dawley rat primary hepato- epithelium during colonoscopy might thus result
cytes (Lbaj et al., 2007; Horvthov et al., 2012), in reduction of DNA damage in sensitive tissues
MCF-7 breast cancer cells (Masannat et al., 2009), during these medical procedures.
174
Methylene blue
(c) Chromosomal damage Similarly, no increases in the frequency of micro-

(i) In vitro nucleated erythrocytes were observed in bone-
marrow cells or peripheral blood erythrocytes of
The results of tests measuring induction of male B6C3F1 mice given a single intraperitoneal
sister-chromatid exchange in cultured Chinese dose of methylene blue, or in peripheral blood
hamster lung V79 cells (Popescu et al., 1977), and erythrocytes of male B6C3F1 mice treated by
Syrian hamster fibroblast (baby hamster kidney) gavage with methylene blue for 5days per week
BHK-1 cells (MacRae et al., 1980) treated with for 3months (NTP, 2008).
methylene blue in the absence of photoactivation
were generally negative. One exception was
reported, where Chinese hamster V79 cells
4.2.3 Metabolites of methylene blue
showed significant increases in the frequency of (a) Azure A
sister-chromatid exchange in the absence, but not Azure A was mutagenic in Salmonella typhi-
in the presence, of photoactivation (Speit & Vogel, murium strains TA98 and TA100, and Escherichia
1979). No induction of chromosomal aberration coli strain WP2 uvrA pKM101, with and without
was seen in Chinese hamster V79 cells treated exogenous metabolic activation (NTP, 2008).
with methylene blue in the absence of photo Azure A also induced chromosomal damage
activation (Popescu et al., 1977). Negative results in cultured Chinese hamster ovary cells in the
were also reported in another test for chromo- absence of exogenous metabolic activation at
somal aberration in Chinese hamster ovary cells doses (10 and 20 M) that produced marked
(Au & Hsu, 1979). [The Working Group noted cytotoxicity (Au & Hsu, 1979).
that caution should be used in interpreting the
results of Au & Hsu (1979) due to the inadequate (b) Azure B
description of the protocol and other deficien-
Azure B was mutagenic in Salmonella typhi-
cies, including the brief exposure time and the
murium strains TA98 and TA100, and Escherichia
small number of cells scored.] In a study by
coli strain WP2 uvrA pKM101, with and without
the National Toxicology Program (NTP, 2008),
exogenous metabolic activation (NTP, 2008).
induction of sister-chromatid exchange and
Azure B also induced chromosomal damage
of chromosomal aberration with and without
in cultured Chinese hamster ovary cells in the
metabolic activation was observed in Chinese
absence of exogenous metabolic activation at a
hamster ovary cells treated with methylene blue.
dose (20M) that produced marked cytotoxicity
(ii) In vivo (Au & Hsu, 1979).
Despite extensive evidence for mutagenicity
(c) Azure C
and induction of DNA damage by methylene
blue in vitro, particularly with white-light acti- Azure C was mutagenic in Salmonella typhi-
vation, no evidence for genotoxicity has been murium strains TA98 and TA100, and Escherichia
observed in a limited number of standard tests coli strain WP2 uvrA pKM101, with and without
in vivo, all of which investigated some aspect of exogenous metabolic activation (NTP, 2008).
chromosomal damage. No significant increase Azure C also induced chromosomal damage
in the frequency of sister-chromatid exchange in cultured Chinese hamster ovary cells in the
was seen in bone-marrow cells of adult Chinese absence of exogenous metabolic activation at a
hamsters given a single intraperitoneal injection dose (20M) that produced marked cytotoxicity
of methylene blue at 12mg/kg bw (Speit, 1982). (Au & Hsu, 1979).
175
IARC MONOGRAPHS 108
4.3 Other relevant mechanisms Given that glucose-6-phosphate dehy-

drogenase is required for the enzymatic pentose
4.3.1 General adverse effects phosphate pathway that produces NADPH,
In humans, large intravenous doses of patients with glucose-6-phosphate dehydroge-
methylene blue (~500 mg) have been reported nase deficiency have depleted NADPH levels. In
to cause nausea, abdominal and chest pain, these patients, methylene blue may exacerbate
cyanosis, methaemoglobinaemia, sweating, haemolytic anaemia, and haemolysis favours the
dizziness, headache, and confusion (Clifton & formation of methylene blue-induced methaemo
Leikin, 2003; Oz et al., 2011). Toxicity in infants globin (Smith & Thron, 1972; Bilgin et al., 1998).
exposed to methylene blue during prenatal or A study compared the responses of several
perinatal diagnostic or therapeutic procedures species to a single intraperitoneal injection of
is well documented: hyperbilirubinaemia, methylene blue (20100mg/kg bw in cats, dogs,
haemolytic anaemia, formation of Heinz bodies, and guinea-pigs; 20200 mg/kg bw in mice,
erythrocytic blister cells, skin discoloration, rabbits, and rats). Although the tolerance for
and photosensitization are the most commonly methylene blue varied considerably, most species
reported adverse effects (Sills & Zinkham, 1994; had a decrease in erythrocytes and haemoglobin,
Porat et al., 1996; Cragan, 1999). and an increase in reticulocytes within a few days
A series of acute toxic effects have been after treatment. Cats and dogs were the most
described in animals exposed to methylene blue, sensitive species, with Heinz bodies detected 4
including haemoconcentration, hypothermia, and 6hours, respectively, after administration of
acidosis, hypercapnia, hypoxia, increases in blood methylene blue. Heinz bodies were also detected
pressure, changes in respiratory frequency and in mice (100% incidence, at 200mg/kg bw after
amplitude, corneal injury, conjunctival damage, 24 hours), rats (12% incidence, at 200 mg/kg
and formation of Heinz bodies (Auerbach et al., bw after 96 hours), rabbits (70% incidence, at
2010). 200mg/kg bw after 96hours), and guinea-pigs
(incidence was 4%, at 100 mg/kg bw, after
72hours) (Rentsch & Wittekind, 1967).
4.3.2 Haematological toxicity In a 90-day study of toxicity by the NTP,
Severe toxic methaemoglobinaemia can methylene blue was administered at doses of
be treated by intravenous administration of 0, 25, 50, 100, and 200 mg/kg bw by gavage to
methylene blue (12mg/kg bw). In the presence F344/N rats and B6C3F1 mice. The treatment
of reduced nicotinamide adenine dinucleotide resulted in methaemoglobin formation, oxida-
phosphate (NADPH), the dye is converted by tive damage to erythrocytes, and dose-related
methaemoglobin reductases in erythrocytes regenerative Heinz-body anaemia in rats and
to leucomethylene blue, which then reduces mice. Splenomegaly and an increase in splenic
methaemoglobin nonenzymatically, restoring haematopoiesis occurred in treated rats and
functional haemoglobin and methylene blue. mice. Splenic congestion and bone-marrow
This redox cycle is sustained by regeneration of hyperplasia were also observed in treated rats.
NADPH via the hexose monophosphate shunt Mice showed increased liver haematopoiesis
(pentose phosphate pathway). However, at higher (100 mg/kg bw and above) and an accumulation
concentrations, methylene blue oxidizes ferrous of haemosiderin in Kupffer cells (50 mg/kg bw
iron in haemoglobin to the ferric state, producing and above). These observations suggested the
methaemoglobin (Bradberry et al., 2001). development of haemolytic anaemia. There was
also a dose-related increase in the reticulocyte
176
Methylene blue
count in treated rats and mice, suggesting a activity of cytochrome oxidase in the brain
compensatory response to anaemia (Hejtmancik (reviewed in Oz et al., 2009).
et al., 2002; NTP, 2008). Methylene blue and its metabolite, azure B,
The haematological toxicity documented in are reversible inhibitors of monoamine oxidase.
the 90-day study by the NTP (see above) served as This inhibition may underlie adverse effects,
the basis for selecting the doses of methylene blue but also psycho- and neuromodulatory actions
for a long-term bioassay (0, 5, 25, and 50mg/kg associated with methylene blue taken as a drug
bw per day for rats; 0, 2.5, 12.5, and 25 mg/kg (Ramsay et al., 2007; Petzer et al., 2012).
bw per day for mice; 5days per week for 2 years).
Similarly to the 90-day study, development
of methaemoglobinemia, formation of Heinz
4.4 Susceptibility
bodies, and macrocytic responsive anaemia No data were available to the Working Group.
were observed in treated rats, while methaemo-
globinaemia and formation of Heinz bodies also
occurred in treated mice (NTP, 2008; Auerbach 4.5 Mechanistic considerations
et al., 2010). Methylene blue absorbs energy directly from
a light source and then transfers this energy to
4.3.3 Additional mechanisms molecular oxygen, generating singlet oxygen
Amino acids can undergo photo-oxidation by (1O2). Singlet oxygen is electrophilic and can
methylene blue and methylene blue derivatives oxidize electron-rich double bonds in bio(macro)
(Knowles & Gurnani, 1972); multiple studies molecules (Tardivo et al., 2005).
have been conducted on the photoinactivation of Two mechanisms of action, involving photo-
a variety of enzymes by methylene blue (reviewed activation, can also be envisaged. Excitation of
in Moura & Cordeiro, 2003). methylene blue can produce both a singlet and
a triplet species; the excess triplet energy can be
In pharmacological studies, methylene blue
transferred through electrons (type I mechanism)
(110 M) is used routinely to inhibit soluble
or energy (type II mechanism) (Tardivo et al.,
guanylate cyclase for the analysis of cyclic
2005). Both mechanisms can damage bio(macro)
guanosine monophosphate (cGMP)-mediated
molecules. Energy transfer can cause strand
processes. Methylene blue also inhibits constitu-
breaks in nucleic acids, thereby leading to DNA
tive and inducible forms of nitric oxide synthase
damage. Electron transfer can produce reactive
by oxidation of ferrous iron bound to the enzyme,
oxygen species, including hydroxyl radicals and
and inactivates nitric oxide by generation of
hydroperoxides, which can be detrimental to the
superoxide anions (reviewed in Oz et al., 2011).
integrity of nucleic acids, proteins, and lipids.
Methylene blue penetrates cellular and
Although the carcinogenicity of methylene
mitochondrial membranes, accumulates within
blue may partly arise via photoactivation, the
mitochondria, and improves mitochondrial
rodent biossays were conducted without light
respiration at low concentrations (0.52 M)
activation. Therefore other mechanisms are
by shuttling electrons to oxygen in the electron
likely to operate. It is currently unclear whether
transport chain. When acting as an alternative
the effects of methylene blue upon enzyme-medi-
electron acceptor in mitochondria, methylene
ated processes, such as inhibition of nitric oxide
blue also inhibits the production of superoxide
synthase, with possible generation of superoxide
by competing with molecular oxygen. Methylene
anions, are a factor in the process.
blue has been described to increase the enzymatic
177
IARC MONOGRAPHS 108
5. Summary of Data Reported the incidence of pancreatic islet cell adenoma in

males at the intermediate dose. The incidence
of pancreatic islet cell adenoma or carcinoma
5.1 Exposure data (combined) in males at the intermediate dose was
Methylene blue is a synthetic chemical dye. significantly increased only as the result of the
Methylene blue has a variety of medical uses, increased incidence of adenoma; the incidence
including use as an antidote to methaemoglobin of carcinoma was within the range for historical
aemia induced by environmental poisons such controls. No significant increase in the incidence
as excessive nitrate in well-water or cyanide of any neoplasm was observed in females.
compounds. Other indications include treatment
of psychiatric disorders. Recent studies have 5.4 Mechanistic and other relevant
investigated its use in Alzheimer disease and
therapy for malaria. Other uses include staining
data
in bacteriology, and uses as a redox colorimetric Methylene blue is well absorbed, reduced,
agent, as a contrast agent in medical procedures, and is excreted largely in the urine as the reduced
as a dye, or as a disinfectant. Occupational expo- form, leucomethylene blue.
sure has been documented. Overall, data on Methylene blue and its N-demethylated
exposure are limited, but substantial sales have metabolites, azure A, azure B, and azure C, have
been reported in the USA and Brazil. given positive results in an extensive series of
standard in-vitro assays for genotoxicity, both in
5.2 Human carcinogenicity data the absence and presence of exogenous metabolic
activation.
No data were available to the Working Group. At high doses, methylene blue oxidizes
ferrous iron in haemoglobin to the ferric state,
producing methaemoglobin. Exposure to
5.3 Animal carcinogenicity data methylene blue results in haematological toxicity,
Methylene blue was tested for carcinogenicity including formation of Heinz bodies and haemo-
in one study in mice treated by gavage for 2 years, lytic anaemia, in several species.
and one study in rats treated by gavage for 2 years. Photoactivation of methylene blue produces
In the study in mice, methylene blue caused a high-energy species that have the potential
significant positive trend in the incidence of carci- to damage DNA, proteins, and lipids, either
noma, and of adenoma or carcinoma (combined), directly or through the production of reactive
of the small intestine in males. In males, a signif- oxygen species. In the absence of light activation,
icant positive trend and a significant increase in the carcinogenicity of methylene blue is likely to
the incidence of bronchiolo-alveolar carcinoma arise from other mechanisms. A potential mech-
of the lung at the highest dose were considered anism is the inhibition of nitric oxide synthase,
not to be related to treatment. Treatment with with possible generation of superoxide anions.
methylene blue caused the incidence of malig-
nant lymphoma in females to increase with a
significant positive trend, but all incidences were
well within the range for historical controls.
In the study in rats treated by gavage,
methylene blue caused a significant increase in
178
Methylene blue
6. Evaluation Borwitzky H, Haefeli WE, Burhenne J (2005). Analysis of

methylene blue in human urine by capillary electro-
phoresis. J Chromatogr B Analyt Technol Biomed Life
6.1 Cancer in humans Sci, 826:244251. doi:10.1016/j.jchromb.2005.09.013
PMID:16182616
No data were available to the Working Group. Bountogo M, Zoungrana A, Coulibaly B et al. (2010).
Efficacy of methylene blue monotherapy in semi-im-
mune adults with uncomplicated falciparum malaria:
a controlled trial in Burkina Faso. Trop Med Int Health,
6.2 Cancer in experimental animals 15:713717. doi:10.1111/j.1365-3156.2010.02526.x
PMID:20374561
There is limited evidence for the carcino- Bradberry SM, Aw T-C, Williams NR, Vale JA (2001).
genicity of methylene blue in experimental Occupational methaemoglobinaemia. Occup Environ
animals. Med, 58:611615, quiz 616. doi:10.1136/oem.58.9.611
PMID:11511749
Brendel M (1973). Different photodynamic action of
6.3 Overall evaluation proflavine and methylene blue on bacteriophage. II.
Mutation induction in extracellularly treated Serratia
phage kappa. Mol Gen Genet, 120:171180. doi:10.1007/
Methylene blue is not classifiable as to its BF00267245 PMID:4568530
carcinogenicity in humans (Group 3). British Pharmacopoeia Commission (2005). British
Pharmacopoeia 2005. London: Medicines and
Healthcare products Regulatory Agency.
Burhenne J, Riedel KD, Rengelshausen J et al. (2008).
References Quantification of cationic anti-malaria agent meth-
ylene blue in different human biological matrices
Aeschlimann C, Cerny T, Kpfer A (1996). Inhibition of using cation exchange chromatography coupled to
(mono)amine oxidase activity and prevention of ifosfa- tandem mass spectrometry. J Chromatogr B Analyt
mide encephalopathy by methylene blue. Drug Metab Technol Biomed Life Sci, 863:273282. doi:10.1016/j.
Dispos, 24:13361339. PMID:8971139 jchromb.2008.01.028 PMID:18258499
ATSDR (1992). Toxicological profile for nitrophenols. ChemNet (2013). Methylene blue. Royal Society of
Atlanta (GA): Agency for Toxic Substances and Disease Chemistry. Available from: http://chemnet.rsc.org.
Registry, United States Public Health Service. Christensen CM, Farrar HC, Kearns GL (1996).
Au W, Hsu TC (1979). Studies on the clastogenic effects Protracted methemoglobinemia after phenazopyridine
of biologic stains and dyes. Environ Mutagen, 1:2735. overdose in an infant. J Clin Pharmacol, 36:112116.
doi:10.1002/em.2860010109 PMID:95447 doi:10.1002/j.1552-4604.1996.tb04175.x PMID:8852386
Auerbach SS, Bristol DW, Peckham JC et al. (2010). Chung KT, Fulk GE, Andrews AW (1981). Mutagenicity
Toxicity and carcinogenicity studies of methylene testing of some commonly used dyes. Appl Environ
blue trihydrate in F344N rats and B6C3F1 mice. Food Microbiol, 42:641648. PMID:7039509
Chem Toxicol, 48:169177. doi:10.1016/j.fct.2009.09.034 Clark AM (1953). Mutagenic activity of dyes in Drosophila
PMID:19804809 melanogaster. Am Nat, 87:295305. doi:10.1086/281787
Belaz-David N, Decosterd LA, Appenzeller M etal. (1997). Clifton J 2nd, Leikin JB (2003). Methylene blue. Am J Ther,
Spectrophotometric determination of methylene blue in 10:289291. doi:10.1097/00045391-200307000-00009
biological fluids after ion-pair extraction and evidence PMID:12845393
of its adsorption on plastic polymers. Eur J Pharm Sci, Coulibaly B, Zoungrana A, Mockenhaupt FP etal. (2009).
5:335345. doi:10.1016/S0928-0987(97)00061-4 Strong gametocytocidal effect of methylene blue-based
Berneth H (2008). Azine dyes. In: Ullmanns Encyclopedia combination therapy against falciparum malaria:
of Industrial Chemistry. Weinheim, Germany: a randomised controlled trial. PLoS ONE, 4:e5318
Wiley-VCH Verlag GmbH & Co. KGaA, pp. 475514. doi:10.1371/journal.pone.0005318 PMID:19415120
doi:10.1002/14356007.a03_213.pub3. Cragan JD (1999). Teratogen update: methylene blue.
Bilgin H, zcan B, Bilgin T (1998). Methemoglobinemia Teratology, 60:4248. doi:10.1002/(SICI)1096-
induced by methylene blue pertubation during 9926(199907)60:1<42::AID-TER A12>3.0.CO;2-Z
laparoscopy. Acta Anaesthesiol Scand, 42:594595. PMID:10413340
doi:10.1111/j.1399-6576.1998.tb05173.x PMID:9605379 Davies J, Burke D, Olliver JR et al. (2007). Methylene
blue but not indigo carmine causes DNA damage
to colonocytes in vitro and in vivo at concentrations
179
IARC MONOGRAPHS 108
used in clinical chromoendoscopy. Gut, 56:155156. Gutter B, Speck WT, Rosenkranz HS (1977). A study of the
doi:10.1136/gut.2006.107300 PMID:17172595 photoinduced mutagenicity of methylene blue. Mutat
DiSanto AR, Wagner JG (1972). Pharmacokinetics of Res, 44:177181. doi:10.1016/0027-5107(77)90075-6
highly ionized drugs. I. Methylene bluewhole blood, PMID:331101
urine, and tissue assays. J Pharm Sci, 61(4):598602. Guttmann P, Ehrlich P (1891). Ueber die Wirkung des
doi:10.1002/jps.2600610422 PMID:5014319 Methylenblau bei Malaria. Berl Klin Wochenschr,
DiSanto AR, Wagner JG (1972a). Pharmacokinetics of 28:953956. [German]
highly ionized drugs. II. Methylene blueabsorption, Hejtmancik MR, Ryan MJ, Toft JD et al. (2002).
metabolism, and excretion in man and dog after oral Hematological effects in F344 rats and B6C3F1 mice
administration. J Pharm Sci, 61:10861090. doi:10.1002/ during the 13-week gavage toxicity study of methylene
jps.2600610710 PMID:5044807 blue trihydrate. Toxicol Sci, 65:126134. doi:10.1093/
DiSanto AR, Wagner JG (1972b). Pharmacokinetics of toxsci/65.1.126 PMID:11752692
highly ionized drugs. I. Methylene bluewhole blood, Horvthov E, Kozics K, Srankov A et al. (2012).
urine, and tissue assays. J Pharm Sci, 61:598602. Borneol administration protects primary rat hepat-
doi:10.1002/jps.2600610422 PMID:5014319 ocytes against exogenous oxidative DNA damage.
DiSanto AR, Wagner JG (1972c). Pharmacokinetics of Mutagenesis, 27:581588. doi:10.1093/mutage/ges023
highly ionized drugs. III. Methylene blueblood levels PMID:22544524
in the dog and tissue levels in the rat following intra- IMS Health (2012). Multinational Integrated Data
venous administration. J Pharm Sci, 61:10901094. Analysis (MIDAS). IMS Health, Plymouth Meeting,
doi:10.1002/jps.2600610711 PMID:5044808 2012, Pennsylvania, USA.
EDQM (2008). Methylthionimium chloride. In: European Ito T, Kobayashi K (1977). A survey of in vivo photo-
Pharmacopoeia. Strasbourg, France: European dynamic activity of xanthenes, thiazines, and acri-
Directorate for the Quality of Medicines & HealthCare. dines in yeast cells. Photochem Photobiol, 26:581587.
Ehrlich P (1881). Ueber das Methylenblau und seine klin- doi:10.1111/j.1751-1097.1977.tb07536.x
isch-bakterioskopische Verwerthung. Z Klin Med, Kasuga Y, Hishida M, Tanahashi N (1991). Simultaneous
2:710713. [German] determination of malachite green and methylene blue
Epe B, Hegler J, Wild D (1989). Singlet oxygen as an in cultured fishes by high performance liquid chro-
ultimately reactive species in Salmonella typhimu- matography. Shokuhin Eiseigaku Zasshi, 32:137141.
rium DNA damage induced by methylene blue/visible doi:10.3358/shokueishi.32.137
light. Carcinogenesis, 10:20192024. doi:10.1093/ Kimoto K, Gohda R, Murayama K etal. (1996). Sensitive
carcin/10.11.2019 PMID:2680144 detection of near-infrared fluorescent dyes using
Epe B, Mtzel P, Adam W (1988). DNA damage by oxygen high-performance liquid chromatography with perox-
radicals and excited state species: a comparative study yoxalate chemiluminescence detection system. Biomed
using enzymatic probes in vitro. Chem Biol Interact, Chromatogr, 10:189190. doi:10.1002/(SICI)1099-
67:149165. doi:10.1016/0009-2797(88)90094-4 0801(199607)10:4<189::AID-BMC585>3.0.CO;2-P
PMID:2844422 PMID:8831965
Epe B, Pflaum M, Boiteux S (1993). DNA damage Knowles A, Gurnani S (1972). A study of the methylene
induced by photosensitizers in cellular and cell-free blue-sensitized oxidation of amino acids. Photochem
systems. Mutat Res, 299:135145. doi:10.1016/0165- Photobiol, 16:95108. doi:10.1111/j.1751-1097.1972.
1218(93)90091-Q PMID:7683082 tb07341.x PMID:5052681
Erolu L, Calayan B (1997). Anxiolytic and antidepres- Kosswig K (2000). Surfactants. In: Ullmanns Encyclopedia
sant properties of methylene blue in animal models. of Industrial Chemistry. Weinheim, Germany:
Pharmacol Res, 36:381385. doi:10.1006/phrs.1997.0245 Wiley-VCH Verlag GmbH & Co. KGaA, pp. 487505.
PMID:9441729 doi:10.1002/14356007.a25_747
FDA (2011). Drug Safety Communication: Serious CNS Kozaki A, Watanabe J (1981). Dose dependency of apparent
reactions possible when methylene blue is given to volumes of distribution for methylene blue in rabbits.
patients taking certain psychiatric medications. Safety J Pharmacobiodyn, 4:4957. doi:10.1248/bpb1978.4.49
announcement dated 26 July 2011. Silver Spring (MD): PMID:7277192
United States Food and Drug Administration. Available Laassis B, Aaron J-J, Mahedero MC (1994). Fluorimetric
from: http://www.fda.gov/Drugs/DrugSafety/ determination of phenothiazine derivatives by
ucm263190.htm, accessed 1 October 2014. photooxidation in a flow-injection system. Talanta,
Gaudette NF, Lodge JW (2005). Determination of 41:19851989. doi:10.1016/0039-9140(94)00162-6
methylene blue and leucomethylene blue in male PMID:18966160
and female Fischer 344 rat urine and B6C3F1 mouse Lbaj J, Slamenov D, Lazarov M, Koskov B (2007).
urine. J Anal Toxicol, 29:2833. doi:10.1093/jat/29.1.28 Induction of DNA-lesions in freshly isolated rat
PMID:15808010 hepatocytes by different genotoxins and their
180
Methylene blue
reduction by lignin given either as a dietary compo- dichroism spectroscopy. Biopolymers, 21:17131734.
nent or in in vitro conditions. Nutr Cancer, 57:209215. doi:10.1002/bip.360210904 PMID:7126754
doi:10.1080/01635580701277643 PMID:17571955 NTP (2008). Toxicology and carcinogenesis studies
Lee CH, Chang CT, Wetmur JG (1973). Induced circular of methylene blue trihydrate (Cas No. 7220793)
dichroism of DNA-dye complexes. Biopolymers, in F344/N rats and B6C3F1 mice (gavage studies).
12:10991122. doi:10.1002/bip.1973.360120514 Natl Toxicol Program Tech Rep Ser, 540:1224.
PMID:4710250 PMID:18685714
Lunn G, Sansone EB (1991). Decontamination of aqueous ONeil MJ, Heckelman PE, Koch CB et al. (2006). The
solutions of biological stains. Biotech Histochem, 66:307 Merck Index: an encyclopedia of chemicals, drugs, and
315. doi:10.3109/10520299109109992 PMID:1725856 biologicals, 14th Edition (Version 14.6). Whitehouse
MacRae WD, Chan GF, Wat CK etal. (1980). Examination Station (NJ): Merck & Co., Inc.
of naturally occurring polyacetylenes and alpha-ter- Olliver JR, Wild CP, Sahay P etal. (2003). Chromoendoscopy
thienyl for their ability to induce cytogenetic damage. with methylene blue and associated DNA damage in
Experientia, 36:10961097. doi:10.1007/BF01965990 Barretts oesophagus. Lancet, 362:373374. doi:10.1016/
PMID:7418849 S0140-6736(03)14026-3 PMID:12907012
Masannat YA, Hanby A, Horgan K, Hardie LJ (2009). Onur F, Acar N (1992). Simultaneous determination of
DNA damaging effects of the dyes used in sentinel methylene blue, hexamethylene tetramine and resor-
node biopsy: possible implications for clinical practice. cinol in pharmaceutical formulations by first-deriv-
J Surg Res, 154:234238. doi:10.1016/j.jss.2008.07.039 ative UV spectrophotometry. Int J Pharm, 78:8991.
PMID:19181339 doi:10.1016/0378-5173(92)90359-A
McBride TJ, Schneider JE, Floyd RA, Loeb LA (1992). Oz M, Lorke DE, Hasan M, Petroianu GA (2011). Cellular
Mutations induced by methylene blue plus light and molecular actions of Methylene Blue in the nervous
in single-stranded M13mp2. Proc Natl Acad Sci system. Med Res Rev, 31:93117. doi:10.1002/med.20177
USA, 89:68666870. doi:10.1073/pnas.89.15.6866 PMID:19760660
PMID:1495976 Oz M, Lorke DE, Petroianu GA (2009). Methylene blue
McCarroll NE, Piper CE, Keech BH (1981). An E. coli and Alzheimers disease. Biochem Pharmacol, 78:927
microsuspension assay for the detection of DNA 932. doi:10.1016/j.bcp.2009.04.034 PMID:19433072
damage induced by direct-acting agents and promu- Peter C, Hongwan D, Kpfer A, Lauterburg BH
tagens. Environ Mutagen, 3:429444. doi:10.1002/ (2000). Pharmacokinetics and organ distribution
em.2860030404 PMID:7021147 of intravenous and oral methylene blue. Eur J Clin
Medscape (2013). Methylene blue (Rx). Dosing and uses. Pharmacol, 56:247250. doi:10.1007/s002280000124
Available from: http://reference.medscape.com/drug/ PMID:10952480
methylene-blue-343739, accessed 5 June 2013. Petzer A, Harvey BH, Wegener G, Petzer JP (2012).
Mohn GR, Kerklaan PR, van Zeeland AA et al. (1984). Azure B, a metabolite of methylene blue, is a high-po-
Methodologies for the determination of various genetic tency, reversible inhibitor of monoamine oxidase.
effects in permeable strains of E. coli K-12 differing in Toxicol Appl Pharmacol, 258:403409. doi:10.1016/j.
DNA repair capacity. Quantification of DNA adduct taap.2011.12.005 PMID:22197611
formation, experiments with organ homogenates Popescu NC, Turnbull D, DiPaolo JA (1977). Sister chro-
and hepatocytes, and animal-mediated assays. Mutat matid exchange and chromosome aberration analysis
Res, 125:153184. doi:10.1016/0027-5107(84)90067-8 with the use of several carcinogens and noncarcino-
PMID:6230533 gens. J Natl Cancer Inst, 59:289293. PMID:406414
Moura JC, Cordeiro N (2003). 3,7-Bis(dialkylamino) Porat R, Gilbert S, Magilner D (1996). Methylene blue-in-
phenothiazin-5-ium derivatives: biomedical applica- duced phototoxicity: an unrecognized complication.
tions and biological activity. Curr Drug Targets, 4:133 Pediatrics, 97:717721. PMID:8628613
141. doi:10.2174/1389450033346902 PMID:12558066 Pruthi S, Haakenson C, Brost BC et al. (2011).
Munns RK, Holland DC, Roybal JE etal. (1992). Liquid Pharmacokinetics of methylene blue dye for lymphatic
chromatographic determination of methylene blue and mapping in breast cancer-implications for use in
its metabolites in milk. J AOAC Int, 75:796800. pregnancy. Am J Surg, 201:7075. doi:10.1016/j.
Naylor GJ, Martin B, Hopwood SE, Watson Y (1986). amjsurg.2009.03.013 PMID:21167367
A two-year double-blind crossover trial of the PubChem (2013). Methylene blue. Pubchem database.
prophylactic effect of methylene blue in manic-de- National Center for Biotechnology Information.
pressive psychosis. Biol Psychiatry, 21:915920. Available from: https://pubchem.ncbi.nlm.nih.gov/
doi:10.1016/0006-3223(86)90265-9 PMID:3091097 [online database].
Nordn B, Tjerneld F (1982). Structure of methylene Rager T, Geoffroy A, Hilfiker R, Storey JMD (2012). The
blue-DNA complexes studied by linear and circular crystalline state of methylene blue: a zoo of hydrates.
181
IARC MONOGRAPHS 108
Phys Chem Chem Phys, 14:80748082. doi:10.1039/ damage in oesophageal cells: implications for chro-
c2cp40128b PMID:22481217 moendoscopy. Mutagenesis, 24:253258. doi:10.1093/
Ramsay RR, Dunford C, Gillman PK (2007). Methylene mutage/gep004 PMID:19218330
blue and serotonin toxicity: inhibition of monoamine Tarbin JA, Chan D, Stubbings G, Sharman M (2008).
oxidase A (MAO A) confirms a theoretical prediction. Multiresidue determination of triarylmethane and
Br J Pharmacol, 152:946951. doi:10.1038/sj.bjp.0707430 phenothiazine dyes in fish tissues by LC-MS/MS. Anal
PMID:17721552 Chim Acta, 625:188194. doi:10.1016/j.aca.2008.07.018
Rengelshausen J, Burhenne J, Frhlich M et al. (2004). PMID:18724993
Pharmacokinetic interaction of chloroquine and Tardivo JP, Del Giglio A, de Oliveira CS, Gabrielli DS,
methylene blue combination against malaria. Eur J Junqueira HC, Tada DB etal. (2005). Methylene blue
Clin Pharmacol, 60:709715. doi:10.1007/s00228-004- in photodynamic therapy: From basic mechanisms
0818-0 PMID:15619134 to clinical applications. Photodiagn Photodyn Ther,
Rentsch G, Wittekind D (1967). Methylene blue and 2(3):17591. doi:10.1016/S1572-1000(05)00097-9
erythrocytes in the living animal. Contribution to PMID:25048768
the toxicology of methylene blue and formation of Tuite MF, Mundy CR, Cox BS (1981). Agents that cause a
Heinz bodies. Toxicol Appl Pharmacol, 11:8187. high frequency of genetic change from [psi+] to [psi]
doi:10.1016/0041-008X(67)90029-4 PMID:6056158 in Saccharomyces cerevisiae. Genetics, 98:691711.
Roybal JE, Munns RK, Hurlbut JA, Shimoda W (1989). PMID:7037537
High-performance liquid chromatography of gentian US Pharmacopeial Convention (2013). Methylene blue.
violet, its demethylated metabolites, leucogentian United States Pharmacopeia,USP36. Rockville (MD):
violet and methylene blue with electrochemical detec- The United States Pharmacopeial Convention.
tion. J Chromatogr, 467:259266. doi:10.1016/S0021- Villanueva A, Caete M, Trigueros C et al. (1993).
9673(01)93970-6 PMID:2753937 Photodynamic induction of DNA-protein cross-linking
Roybal JE, Pfenning AP, Turnipseed SB et al. (1996). Dye in solution by several sensitizers and visible light.
residues in foods of animal origin. ACS Symposium Biopolymers, 33:239244. doi:10.1002/bip.360330206
Series, 636: 169184. doi:10.1021/bk-1996-0636-ch018 PMID:8485298
Sabnis RW, Ross E, Kthe J et al. (2009). Indicator reagents. Walter-Sack I, Rengelshausen J, Oberwittler H, Burhenne
In: Ullmanns Encyclopedia of Industrial Chemistry. J, Mueller O, Meissner P et al. (2009). High absolute
Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. bioavailability of methylene blue given as an aqueous
KGaA, pp. 953. doi:10.1002/14356007.a14_127.pub2 oral formulation. Eur J Clin Pharmacol, 65(2):17989.
Schirmer RH, Adler H, Pickhardt M, Mandelkow doi:10.1007/s00228-008-0563-x PMID:18810398
E (2011). Lest we forget you - methylene blue... Warth A, Goeppert B, Bopp C etal. (2009). Turquoise to
Neurobiol Aging, 32:2325.e72325.e16. doi:10.1016/j. dark green organs at autopsy. Virchows Arch, 454:341
neurobiolaging.2010.12.012 PMID:21316815 344. doi:10.1007/s00428-009-0734-x PMID:19189125
Sills MR, Zinkham WH (1994). Methylene blue-in- Webb RB, Hass BS (1984). Biological effects of dyes on
duced Heinz body hemolytic anemia Arch bacteria. VI. Mutation induction by acridine orange
Pediatr Adolesc Med, 148:306310. doi:10.1001/ and methylene blue in the dark with special reference
archpedi.1994.02170030076017 PMID:8130867 to Escherichia coli WP6 (polA1). Mutat Res, 137:16.
Smijs TG, Nivard MJ, Schuitmaker HJ (2004). doi:10.1016/0165-1218(84)90105-8 PMID:6379434
Development of a test system for mutagenicity of photo- WHO (2011). The International Pharmacopoeia,
sensitizers using Drosophila melanogaster. Photochem Fourth Edition. Geneva, Switzerland: World Health
Photobiol, 79:332338. doi:10.1562/2003-12-03-RA.1 Organization. Available from: http://apps.who.int/
PMID:15137509 phint/en/p/about/.
Smith RP, Thron CD (1972). Hemoglobin, methylene Xu JZ, Dai L, Wu B et al. (2009). Determination of
blue and oxygen interactions in human red cells. J methylene blue residues in aquatic products by
Pharmacol Exp Ther, 183:549558. PMID:4636392 liquid chromatography-tandem mass spectrometry.
Speit G (1982). Intercalating substances do not induce J Sep Sci, 32::41934199. doi:10.1002/jssc.200900364
sister-chromatid exchanges (SCEs) in vivo. Mutat PMID:20066681
Res, 104:261266. doi:10.1016/0165-7992(82)90154-3 Yamaguchi T (1981). Mutagenicity of low Molecular
PMID:7110164 substances in various superoxide generating
Speit G, Vogel W (1979). The effect on sister-chromatid systems. Agric Biol Chem, 45:327330. doi:10.1271/
exchanges of drugs and dyes by intercalation and photo- bbb1961.45.327
activation. Mutat Res, 59:223229. doi:10.1016/0027- Yang F, Xia S, Liu Z et al. (2011). Analysis of meth-
5107(79)90161-1 PMID:35743 ylene blue and its metabolites in blood by capil-
Sturmey RG, Wild CP, Hardie LJ (2009). Removal of lary electrophoresis/electrospray ionization mass
red light minimizes methylene blue-stimulated DNA
182
Methylene blue
spectrometry. Electrophoresis, 32:659664. doi:10.1002/

elps.201000514 PMID:21328395
Ziv G, Heavner JE (1984). Permeability of the blood-milk
barrier to methylene blue in cows and goats. J Vet
Pharmacol Ther, 7:5559. doi:10.1111/j.1365-2885.1984.
tb00879.x PMID:6708167
Zoungrana A, Coulibaly B, Si A etal. (2008). Safety and
efficacy of methylene blue combined with artesunate or
amodiaquine for uncomplicated falciparum malaria:
a randomized controlled trial from Burkina Faso.
PLoS ONE, 3:e1630 doi:10.1371/journal.pone.0001630
PMID:18286187
183
PRIMIDONE
1. Exposure Data 1.1.3 Chemical and physical properties of the

pure substance
1.1 Identification of the agent Description: White or almost white, crys-
1.1.1 Nomenclature talline powder (European Pharmacopoeia,
2008); white crystalline powder, odourless,
Chem. Abstr. Serv. Reg. No.: 125-33-7 (ONeil, very slight bitter taste, with no acidic prop-
2006) erties (Japanese Pharmacopoeia, 2007; US
Chem. Abstr. Serv. Name: 4,6(1H,5H)- Pharmacopeia, 2009)
Pyrimidinedione, 5-ethyldihydro-5-phenyl- Density: 1.138 0.06 g/cm3 (predicted)
(ONeil, 2006; US Pharmacopeia, 2007) (SciFinder, 2013)
IUPAC Systematic Name: 5-Ethyl-5-phenyl- Melting-point: 279284 C (Japanese
1,3-diazinane-4,6-dione (DrugBank, 2013) Pharmacopoeia, 2007; US Pharmacopeia,
Synonym: 2-Deoxyphenobarbital 2007); 281282 C (ONeil, 2006)
Spectroscopy Data: Data from infrared spec-
See WHO (2007) for names in other languages.
troscopy have been reported (Daley, 1973)
Solubility: Very slightly soluble in water,
slightly soluble in ethanol (96%). It dissolves
in alkaline solution (ONeil, 2006; European
H Pharmacopoeia, 2008; US Pharmacopeia,
O N
2009); soluble to 100 mM in dimethyl
sulfoxide (Tocris, 2013); soluble in dimethyl-
H 3C formamide, sparingly soluble in pyridine,
NH
and practically insoluble in diethyl ether
(Japanese Pharmacopoeia, 2007)
O
Stability data: Stable; finished product has
shelf-life of 5years (US Pharmacopeia, 2009;
MHRA, 2013)
C12H14N2O2 Octanol/water partition coefficient (log P):
Relative molecular mass: 218.25 0.91 (US Pharmacopeia, 2009)
185
IARC MONOGRAPHS 108
1.1.4 Technical products and impurities electrospray ionization mass spectrometry, with
a limit of detection of < 0.05 mg/mL (Kuhn &
(a) Trade names Knabbe, 2013).
Mysoline; Cyral; Liskantin; Majsolin; The physical properties of the substance
Midone; Mylepsinum; Mysedon; Primoline; (spectroscopy, melting point) are used for the
Primron; Prysoline; Resimatil; Sertan (NTP, identification of the substance.
2000; ONeil, 2006)
(b) Specified impurities and enantiomer 1.3 Production and use

Several impurities have been detected in the 1.3.1 Production and consumption volume
technical product (European Pharmacopoeia,
2008), including: The synthetic dug primidone is not used
frequently, with around 250000 uses in the USA
R2 per year in 20052012 mentioned by office-based
H 3C physicians in visits with patients. Based on the
R1
same source, approximately 80 000 patients in
the USA were exposed to primidone in 2012
O (IMS Health, 2012a). According to the National
Prescription Audit Plus (IMS Health, 2012b),
there were a total of 1.5 million prescriptions for
R1=NH2, R2=CO-NH2: 2-ethyl-2-phenyl- primidone dispensed in the USA in 2012, similar
propanediamide (ethylphenylmalonamide) to the 1.4 million prescriptions dispensed in
R1=NH2, R2=H: (2RS)-2-phenylbutanamide 2008. [The Working Group recognized that these
R1 = NH2, R2 = CN: (2RS)-2-cyano-2- prescription figures were larger than expected
phenylbutanamide based on the drug uses reported by office-based
physicians.]
R1 = OH, R2 = H: (2RS)-2-phenylbutanoic
Total worldwide sales of primidone in 2012
acid
were US$41 million (IMS Health, 2012c), with
Phenobarbital 60% occurring in the USA. The only other
5-Ethyl-5-phenyl-2-[(1RS)-1-phenylpropyl] country with appreciable use was Germany, with
dihydropyrimidine-4,6(1H,5H)-dione sales of US$3 million.
1.3.2 Use
1.2 Analysis
(a) Indications
Selected compendial and noncompendial
Primidone is an anticonvulsant that metab-
methods are presented in Table 1.1. Primidone
olizes to phenobarbital and phenylethylmalon-
can be quantitatively determined using ultra
amide. All three compounds are thought to be
violet spectroscopy, liquid chromatography
biologically active. Primidone is used in the treat-
using ultraviolet detection, and gas chromatog-
ment of a range of conditions, including seizure
raphy using flame ionization detection.
disorders, tremor, neuropathic pain, trigeminal
Primidone can be analysed in human plasma
neuralgia, tinnitus, and migraine headache. Its
by extraction followed by protein precipita-
use for seizure disorders has declined substan-
tion, centrifugation and finally subjecting to
tially with a shift to newer medications with
ultra-performance liquid chromatography with
186
Table 1.1 Analytical methods for primidone
Sample Sample preparation Assay method Detection limit Reference

matrix
Compendial methods
Assay UV-visible European
Wavelength: 257nm Pharmacopoeia (2008)
Assay UV-visible US Pharmacopeia
Wavelength: minima 254nm, 261nm, and maxima 257nm (2013)
Assay GC British Pharmacopoeia
(tablet and Column: glass column packed with acid-washed, silanized (2013)
suspension) diatomaceous support coated with phenyl methyl silicone fluid
Assay GC US Pharmacopeia
(tablet and Column: 10% liquid phase G3 on support S1AB (2013)
suspension) Flow rate: 40mL/min
Human Protein precipitation, vortex- UPLC-ESI-MS-MS <0.05 mg/L Kuhn & Knabbe (2013)
plasma or mixing, centrifugation, analysis Column: C18
serum of clear organic supernatant Mobile phase: solvent A, 0.1% formic acid in water containing
2mmol/L ammonium acetate; and solvent B, 0.1% formic acid
in methanol containing 2mmol/L ammonium acetate
Flow rate: 0.5mL/min
MRM: 219.0m/z reducing to 162.0m/z
Human Centrifugation, supernatant MIP-ESI-MS 0.0051g/mL Rezaei et al. (2009)
serum injected onto the anti-primidone Needle voltage: 11.40kV
column, washing with methanol Target electrode voltage: 9.00kV
and water, elution with methanol Liquid flow rate: 6L/min
and acetic acid, evaporation, Drift field: 600V/cm
sonication Desolvation field: 600V/cm
Drift gas flow (N2): 500mL/min
Desolvation gas flow (N2): 900mL/min
Drift tube length: 11cm
Shutter grid pulse: 0.3ms
Human Extraction by a liquid-liquid MECC 0.7g/mL Lanas et al. (2003)
plasma extraction system, vortex mixing Capillary: fused-silica
and centrifugation, organic layer Wavelength: 210 and 285nm
evaporated, reconstituted with Buffer: 10mM monobasic sodium phosphate, with 6mM
methanol in water for injection tetraborate, and 75mM SDS
on to the MECC system pH 9.0
187
Primidone
188
Sample Sample preparation Assay method Detection limit Reference

matrix
Tablet Nitration of primidone with Polarography Observed half-wave Daley (1973)
sulfuric-nitric acid mixture to Electrode: dropping mercury electrode potential was 0.17
form 3-nitrophenyl derivative V vs the saturated
calomel electrode
Human Acidified and extracted with LC-UV 501000ng/mL Sato et al. (1986)
serum CHCl3 and isopropanol (70:30) Column: C18 (LOQ)
IARC MONOGRAPHS 108
Mobile phase: acetonitrile and water (12:88)

Tablet Dissolved in DMSO-d6 using Proton NMR Chemical shift zden et al. (1989)
maleic acid as internal standard value: primidone,
7.53ppm; and maleic
acid, 6.50ppm
Serum or Serum or plasma + Immunoassay 0.5g/mL Thermoscientific
plasma anticoagulants, centrifugation The enzyme activity is determined spectrophotometrically at (2004)
340nm
Rat urine Extraction by LRC column LC-UV 0.5mg/mL Ferranti et al. (1998)
Column: C18
Mobile phase: 0.01M
potassium phosphate buffer, methanol and acetonitrile
(270:30:30)
pH: 4.0
Wavelength: 227nm
Rat plasma Solid phase extraction Bond-Elut LC-UV 0.1g/mL Moriyama et al. (1994)
C-18 cartridge column Column: C18
Mobile phase: acetonitrile and 0.01M KH2PO4 (25:75)
Wavelength: 210nm
DMSO, dimethylsulfoxide; GC, gas chromatography; LC, liquid chromatography; LOQ, limit of quantitation; LRC, large reserve capacity; MECC, micellar electrokinetic capillary
chromatography; MIP-ESI-IMS, molecular imprinted polymer electrospray ionization ion mobility spectrometry; MRM, multiple reaction monitoring; NMR, nuclear magnetic
resonance; SDS, sodium dodecyl sulfate; UPLC-ESI-MS-MS, ultra-performance liquid chromatography with electrospray ionization and tandem mass spectrometry; UV, ultraviolet
spectroscopy; vs, versus
Primidone
Table 1.2 Most commonly reported clinical indications for primidone in the USA, 20112012
Diagnosis ICD-9 codea Drug uses (in thousands) Percentage of total

Benign essential tremor 333.101 263 52.5
Tremor, NOS 781.005 139 27.8
Convulsion, NOS 780.301 39 7.8
Familial tremor 333.102 14 2.8
Parkinson disease 332.005 9 1.7
Migraine, NOS 346.903 7 1.5
All other diagnoses 28 5.8
Total with reported diagnoses 502 100.0
The ICD-9 code listed is a more detailed, proprietary version developed by IMS Health
a
NOS, not otherwise specified

From IMS Health (2012a)
fewer adverse effects, fewer drug interactions, [Given its use for chronic conditions, primi-
and less potential for addiction and abuse. Once done therapy would be expected to be long-term
a key medication in the management of seizure in the absence of short- or long-term adverse
disorders, primidone is now considered at best a effects.]
third-line medication for partial and tonic-clonic
seizures (MicroMedex, 2013). (b) Dosage
In the USA, primidone is currently labelled Primidone is available in tablets of 50 mg and
for use as a treatment for epilepsy in children 250 mg, with a tablet of 125 mg and an oral suspen-
and adults, either alone or as an adjunct to other sion formulation being available in some coun-
anticonvulsants (FDA, 2013; MicroMedex, 2013). tries (MicroMedex, 2013; eMC, 2013). Therapy
Most prescriptions in the USA are for off-label is initiated at lower doses and then increased,
indications (Table1.2). although lower doses may be taken when prim-
Primidone is used relatively infrequently as idone is employed as an adjunct (MicroMedex,
anticonvulsant, accounting for 0.4% of all medi- 2013). There is a wide range of dosing regimens,
cations reported as therapies for seizure disor- varying from 50 mg once per day to 500 mg twice
ders (whether alone or in combination with other per day; 50 mg once or twice daily are the most
agents) (IMS Health, 2012a). There are numerous common regimens, each representing 21% of all
other anticonvulsants with overlapping clinical uses. The mean daily dosage for primidone is
indications that have largely replaced primidone, 183 mg per day (IMS Health, 2012a).
even in cases of non-responsiveness to multiple
medications. In contrast, there are few compar-
atively effective treatments for essential tremor 1.4 Occurrence and exposure
(Zesiewicz et al., 2011). As a result, primidone Primidone has been reported in groundwater,
comprises the largest fraction (35%) of all medi- spring water and well-water (Morasch, 2013).
cations reported as therapies for essential tremor Primidone, and its metabolite phenobarbital,
(IMS Health, 2012a). were detected in groundwater within the catch-
In the European Union, primidone is indi- ment area of a drinking-water treatment plant
cated for essential tremor, and in the manage- located downstream of a former sewage farm
ment of grand mal and psychomotor (temporal in Berlin, Germany. The age of shallow ground-
lobe) epilepsy (eMC, 2013). water samples ranged from years to a decade,
189
IARC MONOGRAPHS 108
whereas the age of groundwater was up to four the observed associations between anti-epileptic
decades. Concentrations of the compounds in drugs and cancer may be attributable to detec-
groundwater increased with age. This indicated tion bias (Adelw et al., 2006).
a strong persistence of these compounds in the Few studies have conducted analysis specific
environment under anoxic aquifer conditions for individual anti-epileptic drugs such as prim-
(Hass et al., 2012). idone. The epidemiological studies available for
Human exposure is largely limited to use as evaluating exposure to primidone were limited
a medication. Workers in pharmaceutical manu- to two casecontrol studies nested in a cohort
facturing plants may be exposed, but no specific of epileptic patients conducted by Olsen and
data were available to the Working Group. colleagues in Denmark (Olsen et al., 1993, 1995).
The cohort study (Olsen et al., 1989) provided
information on the source population for the
1.5 Regulations and guidelines casecontrol studies; several anti-epileptic
Primidone has been widely approved by drug drugs were used in this cohort. A cohort study
regulatory agencies. Primidone was approved by of offspring of mothers from the Danish cohort,
the United States Food and Drug Administration which provided limited information on exposure
in 1954 (FDA, 2013). to primidone, is also briefly discussed (Olsen
There were no extraordinary regulatory et al., 1990).
restrictions on use. Primidone was listed in 1999
as a chemical known to the State to cause cancer 2.1 Cohort studies
by the Office of Environmental Health Hazard
Assessment of the State of California, requiring A cohort study of patients at the Filadelfia
public notice of potential environmental expo- epilepsy treatment community, in Dianalund,
sures (OEHHA, 2013). The basis of this listing Denmark, was the only cohort study to report
was an evaluation by the United States National on incidence of cancer after treatment with
Toxicology Program (NTP, 2000). primidone (Olsen et al., 1989). The cohort
consisted of 8004 patients admitted between
1933 and 1962, who had not died before 1943,
2. Cancer in Humans and who had hospital stays of 4 weeks or greater
and traceable records. Patients were treated
Primidone has been used to treat grand primarily with phenobarbital, phenytoin, and
seizures in epilepsy patients. Elevated risks of primidone (5001500 mg per day starting in
several types of cancers, mainly tumours of the mid-1950). Newer drugs became more common
brain and central nervous system, lymphoma, in the 1960s. The cohort was followed for cancer
myeloma, and cancers of the lung, liver, pancreas, incidence until 1984, with cases identified by
and gastrointestinal tract have been seen in some linkage to the Danish cancer registry. In the
but not all studies of epilepsy patients, suggesting analysis, hospitalization was used as a proxy
that epilepsy and long-term use of anti-epileptic for drug use, and analyses were not conducted
drugs may be risk factors for cancer (Lamminp for anti-epileptic drugs, either specifically
et al., 2002; Olsen et al., 1989). The evaluation or as a class. Standardized incidence rates
of causality was complicated because epileptic were adjusted for age, sex, and calendar year.
seizures can be early symptoms of tumours of the Among patients who were not known to have
brain, or can prompt clinical examinations, thus received Thorotrast (a radioactive compound
used as a contrast medium for radiology),
190
Primidone
statistically significant excesses were observed 2.2 Nested casecontrol studies

in the incidence of all malignant neoplasms,
and cancers of the brain and central nervous See Table2.1
system, lung, and secondary and unspecified The nested casecontrol studies on four types
sites (combined). Non-statistically significant of cancer were reported in two publications:
elevations ( 20%) were found for non-Hodgkin cancer of the lung and urinary bladder were
lymphoma, and cancers of the buccal cavity reported by Olsen et al. (1993), and malignant
and pharynx, oesophagus, larynx, liver, biliary lymphoma and cancer of the liver and biliary
tract, thyroid, testes, and unspecified sites. tract were reported by Olsen et al. (1995). The
The risk of cancer of the liver or biliary tract studies had similar methodologies and designs.
increased with increasing time since first Cancer cases identified in follow-up until 1984
admission, while no clear pattern was observed were matched with two controls each from the
for cancer of the lung. Findings for malignant cohort by sex, birth year, and survival time.
lymphoma, and cancers of the liver and biliary Detailed drug information was extracted from
tract, urinary bladder, and lung were explored medical records: between 23% and 27% of
in subsequent nested casecontrol studies. A recorded prescriptions were for primidone, but
statistically significant decrease in incidence 25% of patients had no records of prescriptions
was observed for cancer of the urinary bladder. for any anticonvulsive drugs. Smoking informa-
[The strengths of this study were adequate tion was surveyed among living controls, but not
follow-up and case ascertainment. The study among cases.
population consisted mainly of severe cases of Among patients who had ever used prim-
epilepsy and thus it was not known whether idone, non-statistically elevated relative risks
severity of disease modified the risk of cancer. were observed for malignant lymphoma (odds
The major limitation was the lack of informa- ratio, OR, 1.3; Olsen et al., 1995) and cancers of
tion on exposure to specific drugs and poten- the lung (OR, 1.3; 95% CI, 0.72.3) and urinary
tial confounders at an individual level.] bladder (OR, 1.6; 95% CI, 0.46.3) (Olsen et al.,
Olsen et al. (1990) also conducted a record- 1993). The relative risk was close to unity for
linkage study among 3727 offspring of women use of primadone and cancer of the liver and
from the Filadelfia cohort who were alive as of biliary tract (Olsen et al., 1995). Patients exposed
1968. No increased risk of any malignant cancer to Thorotrast were excluded from the reported
was found among 2579 children born after the analyses of lymphoma and cancer of the liver
mothers first hospital admission and presumably and biliary tract, while analyses of cancers of the
exposed to anti-epileptic drugs in utero (relative lung and bladder reportedly gave similar results
risk, RR, 1.0; 95% CI, 0.61.7). Mothers of 2 of when repeated excluding Thorotrast-exposed
the 14 children with cancer had taken primidone patients. [The strengths of these studies were the
and phenytoin during pregnancy. [Although the same as those of the cohort studies. Limitations
size of the cohort was relatively large and case included incomplete information on exposure to
ascertainment and follow-up were adequate, primidone (with respect to duration of use; drug
this study was not considered to be inform- exposure information was collected only during
ative because the findings were not reported the patients stay in hospital) and on potential
specifically for primidone, and few cancers were confounders, and small numbers of exposed
observed in the cohort.] cases, especially for cancers of the urinary
bladder, lymphoma, and liver and biliary tract.]
191
192
Table 2.1 Nested casecontrol studies of cancer and exposure to primidone
Reference Total No. Control source Exposure Organ site Exposure Exposed Relative Covariates
Study cases (hospital, assessment (ICD code) categories cases risk (95% Comments
location, Total No. population) CI)
period controls
Olsen et al. 104 cases Nested case Medical records Lung Ever- 29 1.3 (0.72.3) Adjusted for other anticonvulsant
(1993) 200 controls control; cohort from epilepsy exposed treatments
Denmark, 18 cases of 8004 patients centre; smoking Urinary bladder Ever- 5 1.6 (0.46.3) Controls matched to cases on
193284 33 controls with epilepsy information (living exposed sex, yr of birth and survival time;
controls only) analyses excluding patients given
IARC MONOGRAPHS 108
collected via mail Thorotrast were also conducted;

survey cohort smoked more than the
general population
Olsen et al. 39 cases Nested case Medical records Liver and biliary Ever- NR 0.9 (0.42.3) Adjusted for other anticonvulsant
(1995) 73 controls control; cohort from epilepsy tract exposed treatments
Denmark, 21 cases of 8004 patients centre Malignant (>10 g, NR 1.3 (0.35.0) Controls matched to cases on sex,
193284 98 controls with epilepsy lymphoma 40tablets) year of birth and survival time;
[non-Hodgkin analyses excluding patients given
lymphoma Thorotrast were also conducted;
and Hodgkin cohort smoked more than the
lymphoma] general population
NR, not reported; yr, year

Primidone
Table 3.1 Studies of carcinogenicity in mice and rats given diets containing primidone
Species, Dosing regimen Incidence of tumours Significance Comments

Duration
Reference
Mouse, Dietary concentrations of Hepatocellular adenoma: *P0.05 Purity, >99%
B6C3F1 0, 300, 600, or 1300ppm, 22/50*, 41/50**, 39/50**, 32/50*** (M) (trend)
(M,F) equivalent to daily 15/50****, 42/50**, 45/49**, 47/50** (F) **P0.001
104105wk doses of 0, 30, 65, or Hepatocellular carcinoma: ***P0.05
NTP (2000) 150mg/kg bw (M), or 0, 25, 12/50****, 31/50**, 35/50**, 38/50** (M) ****P0.001
50, or 100mg/kg bw (F) 3/50****, 11/50***, 19/49**, 38/50** (F) (trend)
50 M and 50 F/group (age, Hepatoblastoma:
56wk) 0/50, 17/50**, 26/50**. 7/50***(M)
1/50, 4/50, 4/49, 4/50 (F)
Hepatocellular adenoma, hepatocellular
carcinoma or hepatoblastoma
(combined):
31/50****, 49/50**, 49/50**, 46/50** (M)
16/50****, 42/50**, 46/49**, 50/50** (F)
Thyroid follicular cell adenoma:
0/49*, 3/48, 3/50, 6/50*** (M)
Rat, F344/N Dietary concentrations of Thyroid follicular cell adenoma: *P=0.047 Purity, >99%
(M,F) 0, 600, 1300, or 2500ppm, 1/50, 1/50, 6/49*, 3/49 (M) **P=0.025 No significant increase
104wk equivalent to daily doses of Renal tubule adenoma or carcinoma (trend) in the incidence of any
NTP (2000) 0, 25, 50, or 100mg/kg bw (standard and extended evaluations ***P=0.050 neoplasm in females
50 M and 50 F/group (age, combined):
6 wk) 4/50**, 2/50, 4/50, 7/50*** (M)
3. Cancer in Experimental Animals of males at the highest dose, in which survival

was less than that of controls. Primidone caused
See Table3.1 significant increases in the incidence of hepato-
Primidone was tested for carcinogenicity by cellular adenoma, of hepatocellular carcinoma,
oral administration (feed) in one study in mice and of hepatocellular adenoma, hepatocellular
and one study in rats. carcinoma, and hepatoblastoma (combined) in
all dosed groups of males and females compared
with controls. Primidone caused significant
3.1 Mouse increases in the incidence of hepatoblastoma in
all dosed groups of males. In males, there was
In one study of carcinogenicity, groups of 50
also a significant positive trend in the incidence
male and 50 female B6C3F1 mice (age, 56weeks)
of follicular cell adenoma of the thyroid in mice
were given diets containing primidone (purity,
receiving pyrimidone, with a significant increase
>99%) at a concentration of 0 (control), 300, 600,
in incidence in the group receiving the highest
or 1300ppm for 104105weeks. Primidone intake
dose. There was an increased incidence in folli-
was equivalent to average daily doses of approx-
cular cell hyperplasia of the thyroid in males and
imately 0, 30, 65, or 150mg/kg body weight (bw)
females receiving pyrimidone.
in males, and 0, 25, 50 or 100mg/kg bw in females
(NTP, 2000). Survival in exposed groups was
similar to that of controls, except for the group
193
IARC MONOGRAPHS 108
3.2 Rat Fig.4.1 Metabolism of primidone

H
In one study of carcinogenicity, groups of 50 O N
male and 50 female F344/N rats (age, 6 weeks) H 3C
were given diet containing primidone (purity, NH
> 99%) at a concentration of 0 (control), 600, O

1300, or 2500 ppm for 104 weeks. Primidone
intake was equivalent to average daily doses Primidone
of approximately 0, 25, 50, or 100 mg/kg bw
in males and females (NTP, 2000). Survival in O
H
exposed groups was similar to that in controls, H 3C C NH2

O N O
except for males at the intermediate and highest C H 3C

NH
doses, for which survival was less than that for C NH2
controls. Primidone caused a significant increase O

O
in the incidence of follicular cell adenoma of the Phenylethylmalonamide Phenobarbital

thyroid in males receiving the intermediate dose.
In the extended evaluations involving additional Compiled by the Working Group
step sections of the kidney in males, there was a
small but significant increase in the incidence of 4.1.1 Humans
renal tubule adenoma or carcinoma (single and In humans, primidone is partly eliminated
extended evaluations combined) at the highest unchanged via urinary excretion, and partly
dose; these tumours also occurred with a small metabolized by hepatic cytochrome P450 (CYP)
but significant positive trend. [The Working isozymes, principally to phenylethylmalonamide
Group noted the unusually high incidence of (PEMA) by cleavage of the pyrimidine ring, and
these uncommon tumours in the controls.] The to phenobarbital by oxidation of the methylene
incidence of renal tubule hyperplasia was also group (Baumel et al., 1972; Martines et al.,
increased in all groups of males receiving prim- 1990; Sato et al., 1992; Anderson, 1998; Tanaka,
idone. There was no significant increase in the 1999). The CYP isoenzymes responsible for
incidence of any neoplasm in females. metabolizing primidone are presently uncertain
(Anderson, 1998; Tanaka, 1999).
4. Mechanistic and Other (a) Pharmacokinetics of single doses

Relevant Data Baumel et al. (1972) reported the pharmaco
kinetics of primidone in two subjects given a
4.1 Absorption, distribution, single oral dose of 500 mg of primidone. Peak
plasma concentration of primidone, measured
metabolism, and excretion by gas-liquid chromatography, was reached at
The metabolism of primidone is shown in 0.5 hours in one subject, and 4.8 hours in the
Fig. 4.1. other. Estimated half-lives for primidone were
5.8 and 3.3hours, respectively. Two hours after
dosing, the metabolite PEMA was detected
in the plasma of both subjects, reaching peak
concentrations at 7 and 8 hours before gradu-
ally declining. Estimated half-lives were 29 and
194
Primidone
36 hours, respectively. The second metabolite, study indicated that PEMA is readily absorbed
phenobarbital, was not detectable in this study. from the gastrointestinal tract and eliminated
In a study of seven volunteers given a single predominantly unchanged in the urine (Cottrell
oral dose of 500 mg of primidone , the mean peak et al., 1982).
plasma concentration ( standard deviation)
of primidone was 41.4 5.2 mol/L, reached (b) Pharmacokinetics of repeated doses
in approximately 2 1 hours. The elimination Although phenobarbital was not detected
half-life was 172.4hours. PEMA was detected after administration of single doses of primi-
in the serum, reaching peak concentrations done, long-term administration of primidone (at
(4.10.7 mol/L) at 224 hours in these subjects. various doses) in 46 epilepsy patients showed
Phenobarbital was below the level of detection serum accumulation of phenobarbital, and PEMA
(< 2 mol/L) of gas-liquid chromatography (Baumel et al., 1972). Although there was signif-
(Pisani et al., 1984). icant inter-individual variability, concentrations
Subsequent studies using high-performance of the two metabolites showed correlation with
liquid chromatography of samples from three those of the parent drug, and concentrations
healthy volunteers given a single oral dose of of phenobarbital were consistently higher than
600 mg, showed initial slow absorption of primi- those of PEMA. Two of the subjects had been
done; peak concentration of unchanged primidone on a daily dose of primidone (750mg in divided
(mean Cmax, 41.25.4mol/mL) was achieved at doses) for more than 3years. After a single dose
12hours in each subject. Mean elimination half- of 750mg in this study, peak serum concentra-
life was 19.42.2hours. The metabolite PEMA tions of primidone were achieved rapidly (by
was detectable at 1.3 0.3 hours, and reached 0.5 hour), and declined slowly (half-lives, 5.3
peak concentration (1.70.3mol/L) at 36hours. and 7.0hours). In both subjects, peak concentra-
Elimination half-life was 26.5 1.0 hours. The tions of metabolites, PEMA (12 and 10 g/mL)
metabolite phenobarbital was detectable at and phenobarbital (33 and 11g/mL), remained
5.3 1.3 hours, and reached maximal concen- relatively constant. In the cerebrospinal fluid,
tration (1.30.2mol/L) at 5211hours, with binding to protein by PEMA and by primidone
a long (12520hours) elimination half-life (Sato was negligible, and approximately 60% by pheno-
et al., 1992). barbital (Baumel et al., 1972).
In a study of the pharmacokinetics of PEMA In a subsequent study in eight epileptic
given as a single oral dose of 400 mg to two patients (aged 1826years) receiving long-term
groups of subjects (six patients aged 1043years treatment with primidone (mean daily dose,
receiving long-term treatment with various 422115mg per day), the half-life for primidone
anti-epileptic drugs and six drug-free subjects was 14.73.5hours (Martines et al., 1990).
aged 2242years), showed no statistically signif-
icant differences between the two groups; peak (c) Absorption, distribution, and excretion
serum concentrations were normally reached under certain conditions
within 24 hours after dosing in both groups. (i) Age-dependent effects
In the drug-free subjects, recovery of unchanged The pharmacokinetics and metabolism of
PEMA in the urine gave an estimated oral primidone at steady-state were studied in 18
bioavailability of at least 80%. The elimination epileptic patients who had been receiving a
half-life ranged from 17 to 25hours in drug-free constant dose of primidone for at least 2months.
subjects, and from 10 to 23hours in patients. There Data were compared in two groups: 10 elderly
was no evidence for a glucuronide conjugate. The
195
IARC MONOGRAPHS 108
patients (age, 7081 years), and 8 young (range, 72123%) as primidone and metabolites.
patients (age, 1826 years) (Martines et al., Of the total daily dose administered, 42.3% was
1990). The mean daily doses were moderately, recovered as unchanged drug, 45.2% as PEMA,
but not significantly, higher in the elderly and 4.9% as phenobarbital. The rate of metab-
group, than the young (575 206 mg/day olism to phenobarbital showed wide variation
and 422 115 mg/day, respectively). In the (25-fold) among children, which, although not
elderly and young, respectively, the mean half- influencing the overall elimination rate constant
life of primidone was 12.1 4.6 hours and for primidine, is an important determinant of
14.73.5hours, and the mean total clearance the individual patients steady-state concen-
of primidone was 34.89.0mL/hour per kg and tration of phenobarbital. Concomitant use of
33.27.2mL/hour per kg. Differences between phenytoin had no detectable effect on half-life
the two groups were not statistically significant, or serum concentrations of phenobarbital. Of
indicating that half-life and total clearance of the total primidone daily dose, approximately
primidone were unaltered in elderly patients. equal amounts of parent drug (~40%) and PEMA
However, some differences between the two (~45%) were excreted, with phenobarbital as
groups were highlighted; serum concentrations approximately 5% (Kauffman et al., 1977).
of the metabolites PEMA and phenobarbital (ii) Pregnancy
(relative to those of parent drug) were higher in
the elderly than the young, significantly so in The placental transfer of primidone and
the case of PEMA (P<0.01). Renal clearances metabolites was investigated in 14 women
of primidone, phenobarbital, and PEMA were treated for epilepsy with primidone (and addi-
moderately decreased (again, significantly for tionally phenytoin, ethosuximide or valproate
PEMA, P<0.05) in the elderly (Martines et al., in 5 women) throughout pregnancy. Primidone,
1990). PEMA, phenobarbital, and polar metabolites
The results of this study supported previous (p-hydroxyphenobarbital and p-hydroxyphe-
suggestions that PEMA (unlike primidone and nobarbital glucuronide) were found in similar
phenobarbital) is eliminated only by renal excre- concentrations in maternal and cord blood at
tion (Cottrell et al., 1982), and so its serum accu- birth (Nau et al., 1980).
mulation in the elderly probably results from In the same study, the pharmacokinetics of
moderately reduced renal elimination accompa- primidone were studied in seven of the newborns
nied by an increase in the fraction of primidone during the first weeks of life (Nau et al., 1980).
metabolized (Cottrell et al., 1982; Pisani et al., Mean elimination half-lives were longer than
1984; Martines et al., 1990). those found in children by Kauffman et al. (1977):
The metabolism and excretion of orally 2310hours for primidine; 11340hours for
administered primidone was studied in 12 chil- phenobarbital; and 356hours for PEMA. The
dren (age, 714 years) undergoing long-term shortest half-lives for primidone (811 hours)
(> 3 months) treatment for epilepsy, and were were detected in two neonates whose mothers
assumed to be in steady state. Four children had been treated with phenytoin in addition to
were taking primidone only, and eight were primidone. Serum concentrations and elimina-
also taking phenytoin. Plasma concentrations tion rates varied among neonates, and during the
peaked at 46hours and declined exponentially period of study. For example, serum concentra-
over 624 hours, with half-lives ranging from tions of phenobarbital and PEMA increased in
4.5 to 11hours. Mean recovery of the adminis- some neonates during the first few days, due to
tered dose in the urine within 24hours was 92% neonatal metabolism of primidine, and rate of
196
Primidone
elimination increased after a few days in some increase or decrease in pharmacologically active
babies (Nau et al., 1980). species. Primidone is frequently used in combi-
Analyses of maternal milk of four of the nation with such substrates (e.g. anticonvulsants
mothers detected primidone and PEMA at such as carbamazepine, ethosuximide, valproic
approximately 75%, phenobarbital at approxi- acid, and phenytoin). A study by Sato et al. (1992)
mately 50%, and total p-hydroxyphenobarbital showed that, in patients taking both primidone
(conjugated and non-conjugated) at approx- and phenytoin, metabolites of primidone in
imately equal to concentrations measured in serum were detected earlier, elimination was
serum. Because of breastfeeding, all compounds faster, and total body clearance was increased,
were also detected in neonatal blood (Nau et al., when compared with patients taking primi-
1980). done only. In a study of seven neonates, whose
mothers were treated for epilepsy throughout
(iii) Liver disease
pregnancy, the shortest half-lives for primidone
The disposition of a single oral dose of 500 mg were reported in two neonates whose mothers
of primidone was studied in seven patients with had been treated with both phenytoin and prim-
acute viral hepatitis and in seven healthy subjects idone (Nau et al., 1980). Conversely, Kauffman
(controls). The elimination half-life and the et al. (1977) reported that there were no effects on
apparent clearance of unchanged primidone half-life or serum concentrations of phenobar-
in the patients did not differ significantly from bital in children being treated for epilepsy with
that in the controls (mean elimination half-life, phenytoin in addition to primidone.
18.03.1hours in patients, and 17.02.4hours in
controls; mean apparent clearance of unchanged
primidone, 4214mL/hour per kg in patients,
and 35 8 mL/hour per kg in controls). The PEMA and phenobarbital have been iden-
metabolite PEMA was detectable in serum of all tified as the major metabolites of primidone in
healthy subjects within 224hours, but undetect- mice (McElhatton et al., 1977), rats (Baumel et al.,
able (<2mol/L) in sera of all except one patient. 1973; Moriyama et al., 1994), rabbits (Fujimoto
In all subjects, serum concentrations of pheno- et al., 1968; Hunt & Miller, 1978), and dogs (Frey
barbital remained below the limit of detection of & Lscher, 1985).
gas-liquid chromatography. These findings indi- In a study by the NTP (2000), groups of
cated that accumulation of primidone is unlikely B6C3F1 mice were given a single dose of primi-
to occur in epilepsy patients who develop acute done (at 30, 80, or 200mg/kg bw) by gavage and
viral hepatitis (Pisani et al., 1984). blood samples were collected at various time-
points (ranging from 0.25 hour to 48 hours)
(d) Pharmacokinetic and drug interactions after administration. Plasma concentrations of
The CYP isozymes 1A2, 2C9, 2C19, and primidone in mice were dependent on dose and
3A4, and UDP-glucoronosyl transferase and time. Absorption was rapid, and for all dose
epoxide hydrolases are induced by primidone groups, plasma concentrations were detectable
and its metabolite phenobarbital [phenobar- within 15 minutes after dosing, and remained
bital also induces CYP2A6] (Riva et al., 1996; above the limit of detection for at least 30hours
Anderson, 1998; Patsalos & Perucca, 2003). (after a dose of 30 or 80 mg/kg bw) and for at
Thus pharmacokinetic interactions are likely to least 48 hours (after a dose of 200 mg/kg bw).
occur between primidone and other substrates Slightly higher plasma concentrations of prim-
for these enzymes, ultimately causing either an idone were detected in males than females.
197
IARC MONOGRAPHS 108
Plasma concentrations of phenobarbital were the plasma concentration of primidone rapidly

dose-, time- and sex-dependent; phenobarbital increased achieving maximal levels by 1 hour,
was detected within 15 minutes after dosing. but by 12hours had decreased to very low levels,
Earlier and slightly higher peak concentrations and at 24 hours was undetectable. In contrast,
were observed in males than in females, indi- concentrations of PEMA and phenobarbital
cating that, in mice, primidone is more rapidly gradually increased, reaching maximum levels
metabolized to phenobarbital in males than in after 48hours, and these metabolites were still
females (NTP, 2000). detected after 24 hours. Tmax values for primi-
Studies in pregnant mice given repeated done, PEMA, and phenobarbital were 1.36, 5.70,
intragastric doses of primidone at 100 mg/kg and 6.55 hours, respectively, and Cmax values
bw [a known teratogenic dose] over several days were 18.15g/mL, 8.11g/mL, and 9.64g/mL,
demonstrated no accumulation of the parent respectively. Thus concentrations of PEMA and
compound, or of the metabolites PEMA or phenobarbital were approximately 50% that
phenobarbital, and all were cleared rapidly from of primidone. Half-lives were reported as 1.64,
the plasma within 24hours. The relatively long 4.29, and 4.96hours for primidone, PEMA and
period of dosing with primidone resulted in phenobarbital, respectively.
its more rapid rate of metabolism, resulting in In the study by Nagaki et al. (1999), the
higher concentrations of metabolites, than after concentrations of primidone, PEMA, and pheno-
a single dose (McElhatton et al., 1977). barbital rose in a linear and dose-dependent
Studies of single doses of primidone (given by manner in serum and cerebrospinal fluid (mean
gavage) in the mouse, showed a dissimilar trend free fraction in serum [free non-protein-bound/
in results. The plasma half-life of phenobarbital total concentration ratio], 0.86, 0.97, and 0.88,
was reported to be twice that of primidone and respectively). The respective mean values for the
PEMA, and plasma:brain ratios indicated poor cerebrospinal fluid:serum ratio were 0.73, 1.06,
penetration of primidone into the brain (Leal and 0.65, suggesting rapid equilibration between
et al., 1979). blood and cerebrospinal-fluid compartments.
In contrast, in rats given primidone by Mean half-life values for primidone, PEMA and
gavage, concentrations of the parent drug peaked phenobarbital in the cerebrospinal fluid were
in the plasma after 1hour, and in the brain after similar to those reported in serum (Nagaki et al.,
2 hours (Baumel et al., 1973). This result was 1999).
supported by a subsequent study in rats given In a study by the NTP (2000), groups of
primidone by intraperitoneal injection (at a dose male and female F344/N rats were given a single
of 50, 100 or 200 mg/kg bw), which suggested dose of primidone (30, 80, or 130mg/kg bw) by
that primidone and metabolites were able to gavage, and blood samples were collected from
penetrate the bloodbrain barrier. Primidone all dose groups at various times (from 15minutes
was first detected in the serum (mean Tmax range, to 30 hours) after administration (NTP, 2000).
1.52.5 hours) and in the cerebrospinal fluid Plasma concentrations of primidone were
(mean Tmax range, 2.03.5hours), followed by its dependent on dose and time; absorption was
metabolites, PEMA and phenobarbital (Nagaki rapid at all doses, and primidone was detectable
et al., 1999). in the plasma within 15 minutes after dosing.
Moriyama et al. (1994) reported the phar- Although the time-course and doseresponse
macokinetic parameters of primidone and its profiles were similar in male and female rats,
major metabolites in the rat. After oral adminis- plasma concentrations of primidone (at most
tration of primidone (at a dose of 50 mg/kg bw), doses and time points) were consistently higher
198
Primidone
(approximately double) and half-lives greater (b) Chromosomal damage

(two- to fivefold) in females than in males. No increases in sister-chromatid exchange
Plasma concentrations of the metabolite pheno- or chromosomal aberration were noted in
barbital were also dependent on dose, time, and cultured Chinese hamster ovary cells treated
sex; although phenobarbital was detectable in with primidone at concentrations ranging from
the plasma of male rats within 15minutes after 125 to 1250 g/mL, with or without metabolic
dosing, phenobarbital was undetectable in the activation (NTP, 2000). Additional in-vitro
plasma of female rats at 15 and 30minutes, and studies showing no induction of sister-chro-
plasma concentrations of phenobarbital, for a matid exchange in Chinese hamster ovary cells,
given dose, were consistently higher in males or chromosomal aberration in human lympho-
than in females. [Thus, the metabolism of prim- cytes or Chinese hamster ovary cells, have been
idone in rats appeared to be dependent on sex, reported (Stenchever & Allen, 1973; Bishun et al.,
with males metabolizing primidone more rapidly 1975; Riedel & Obe, 1984).
than females.] Phenobarbital was still detectable In vivo, no induction of dominant lethal
in the plasma of male and female rats at 30 hours mutation was observed in germ cells of male
after dosing (NTP, 2000). mice treated with a single intraperitoneal injec-
tion of primidone at doses of up to 90mg/kg bw
4.2 Genetic and related effects (Epstein et al., 1972) or 400mg/kg bw (Zolotareva
et al., 1979). No induction of chromosomal aber-
4.2.1 Humans rations was reported in bone-marrow cells of
No data were available to the Working Group. male mice treated with primidone at doses of
up to 400 mg/kg bw by a single intraperitoneal
injection (Zolotareva et al., 1979). There was
one report of an increased frequency of micro-
See Table4.1 nucleated polychromatic erythrocytes in the
bone marrow of mice given 13.11mg of primi-
(a) Mutagenicity done [dose, approximately 500mg/kg bw] twice
Primidone (concentration range, with an interval of 24 hours (Rao et al., 1986).
3310000g/plate) was mutagenic at concentra- [The Working Group noted that the mice were
tions of 3333g/plate and higher in Salmonella sampled 6 hours after the second dose, which
typhimurium strain TA1535 in the absence of was too brief an interval to measure the effects
metabolic activation; no mutagenic activity was of the second treatment, and possibly too long
detected in TA1535 in the presence of meta- to evaluate accurately the induction of micro-
bolic activation, or in strains TA100, TA1537, nuclei after the initial treatment. These protocol
or TA98, with or without metabolic activation deficiencies hindered the interpretation of the
(Mortelmans et al., 1986). data.] Contrasting results were seen in B6C3F1
No increases in the frequencies of sex-linked male mice, in which no significant increases in
recessive lethal mutations were detected in germ the frequency of micronucleated polychromatic
cells of male Drosophila melanogaster treated erythrocytes were detected in bone marrow
as larvae by feeding on primidone solutions of after administration of primidone (dose range,
612mM (Zolotareva et al., 1979). 87.5350mg/kg bw) by intraperitoneal injection,
three times at 24-hour intervals, in each of two
replicate trials (NTP, 2000).
199
200
Table 4.1 Genetic and related effects of primidone
Test system Resultsa Concentration/dose Reference

(LED or HID)
Without With
exogenous exogenous
metabolic metabolic
system system
In vitro
Salmonella typhimurium TA100, TA1537, TA98, reverse mutation b 10000 g/plate Mortelmans et al. (1986)
Salmonella typhimurium TA1535, reverse mutation + b 3333 g/plate Mortelmans et al. (1986)
IARC MONOGRAPHS 108
Drosophila melanogaster, sex-linked recessive lethal mutation in germ cells NT 12 mM in food Zolotareva et al. (1979)
Sister-chromatid exchange, Chinese hamster ovary cells 1250 g/mL NTP (2000)
Sister-chromatid exchange, Chinese hamster ovary cells 100 g/mL Riedel & Obe (1984)
Chromosomal aberration, Chinese hamster ovary cells 1250 g/mL NTP (2000)
Chromosomal aberration, Chinese hamster ovary cells 100 g/mL Riedel & Obe (1984)
Chromosomal aberration, human lymphocytes NT 100 g/mL Stenchever & Allen (1973)
Chromosomal aberration, human lymphocytes NT 70 g/mL Bishun et al. (1975)
In vivo
Dominant lethal mutation, male ICR/Ha Swiss mouse, germ cells 90 mg/kg bw, ip 1 Epstein et al. (1972)
Dominant lethal mutation, male mouse, germ cells 400 mg/kg bw, ip1 Zolotareva et al. (1979)
Chromosomal aberration, male mouse, bone-marrow cells 400 mg/kg bw, ip1 Zolotareva et al. (1979)
Micronucleus formation, Swiss mouse, bone-marrow cells + 13.11 mg, po 2c Rao et al. (1986)
Micronucleus formation, male B6C3F1 mouse, bone-marrow cells 350 mg/kg bw, ip 3 NTP (2000)
b S9 (9000 g supernatant) from Sprague-Dawley rats and Syrian hamsters treated with Aroclor 1254
c Dose was approximately 500 mg/kg bw; four mice per treatment group. Mice were killed 6hours after the second treatment; 3000 polychromatic erythrocytes were scored per mouse
bw, body weight; LED, lowest effective dose; HID, highest ineffective dose; ip, intraperitoneal; NR, not reported; NT, not tested; po, oral
Primidone
4.2.3 Genetic and related effects of the although they did not present a consistent pattern
metabolite phenobarbital of genotoxicity. The inconsistency of the results,
the absence of any direct evidence for an interac-
In contrast to the limited information on tion with DNA, and the generally negative data
primidone, there was a significant body of litera- in vivo led to the conclusion that phenobarbital
ture describing the results of tests for genotoxicity is not genotoxic (IARC, 2001).
with phenobarbital, a major metabolite of primi- Phenobarbital transformed hamster embryo
done. The extensive literature on the genetic and cells. It inhibited gap-junctional intercellular
related effects of phenobarbital was reviewed by communication in hepatocytes of rats treated
a previous Working Group (IARC, 2001), and is in vivo, and in primary cultures of hepatocytes
summarized briefly below. from rats and mice, but not (in a single study) in
Phenobarbital did not induce sister-chromatid primary cultures of hepatocytes from humans or
exchange in patients with epilepsy receiving only rhesus monkey (IARC, 2001).
this drug (IARC, 2001).
In studies in which rodents were exposed to
phenobarbital in vivo, no covalent binding to 4.3 Other mechanistic data relevant
mouse liver DNA was observed, but the frequency to carcinogenesis
of alkali-labile damage in mouse liver cells was
increased. Gene mutation was not induced in 4.3.1 Humans
a transgenic mouse strain, and sister-chro- Toxicity associated with primidone in
matid exchange, micronucleus formation, and humans has been documented with reference to
chromosomal aberrations were not induced in side-effects after use of primidone as a drug. The
mouse bone-marrow cells. Phenobarbital did side-effects included nausea, vomiting, dizzi-
not increase the frequency of sperm-head abnor- ness, ataxia, and somnolence, and caused early
malities in mice, but spermatogonial germ-cell discontinuation of treatment. Smith et al. (1987)
chromosomal aberrations were reported in male reported that both carbamazepine and pheny-
mice in one laboratory. Further increases in the toin were associated with statistically signifi-
frequency of chromosomal aberration were found cantly lower incidences of intolerable side-effects
in liver foci cells of mice treated with phenobar- than were primidone or phenobarbital. Patients
bital after prior treatment with a genotoxic agent receiving primidone experienced the highest
(IARC, 2001). incidence of toxicity.
Chromosomal aberrations, but not gene Administration of anti-epileptic drugs,
mutations, were induced in cultured human such as primidone, and also carbamazepine,
lymphocytes treated with phenobarbital (IARC, gabapentin, oxcarbazepine, and phenytoin
2001). affected serum concentrations of folate, homo-
The numerous types of test for the genetic cysteine, and vitamin B12. In a study involving
effects of phenobarbital in vitro included assays 2730 patients treated with various anti-epileptic
for DNA damage, DNA repair induction, gene drugs, 170 untreated patients, and 200 healthy
mutation, and chromosomal aberration in controls, Linnebank et al. (2011) reported that
mammalian cells, tests for gene mutation and primidone monotherapy (10 patients) was associ-
mitotic recombination in insects and fungi, and ated with a higher frequency of folate concentra-
tests for gene mutation in bacteria. Although tions that were below the reference range when
the majority of the test results were negative, the compared with untreated patients and controls.
numerous positive results could not be ignored, This association was dose-dependent. Primidone
201
IARC MONOGRAPHS 108
monotherapy was also associated with plasma thus more sensitive to effects induced by primi-
concentrations of homocysteine that were above done metabolites.
the reference range when compared with controls
(Linnebank et al., 2011).
A review by Benedetti et al. (2005) of several
4.5 Mechanistic considerations
studies in humans suggested that therapeutic In humans, and in mice and rats, primi-
levels of primidone or phenobarbital are not done is extensively, but not totally, metabolized
associated with an increase in thyroid-stimu- to phenobarbital. Given the evidence for the
lating hormone levels. carcinogenicity of primidone (see Section 3) and
phenobarbital (IARC, 2001), the carcinogenic
4.3.2 Experimental systems activity attributable to primidone in mice can
be reasonably hypothesized to be the result of
Carl et al. (1987a, b) studied the effects of
the metabolism of primidone to phenobarbital
treatment with primidone on one-carbon metab-
considering that both cause malignant hepato-
olism by measuring levels of methylene-tetra
cellular tumours in this species.
hydrofolate reductase and related parameters
The carcinogenicity of phenobarbital was
in the brain and liver of rats given primidone
evaluated by the Working Group in 2000 (IARC,
(100mg/kg bw every 12hours) by gastric gavage
2001). Epidemiological data primarily comprised
for up to 8weeks. Primidone caused a decrease
three large cohort studies of patients with
of pteroylpentaglutamates in the liver to less than
epilepsy. On the basis of these and all other avail-
half the control value within 1week. Overall, the
able studies, the Working Group concluded that
data suggested that primidone affects concen-
there was inadequate evidence in humans for the
trations of folate in the tissue and plasma by
carcinogenicity of phenobarbital (IARC, 2001).
interfering with folate-dependent metabolic
Singh et al. (2005) reviewed studies involving risk
processes, possibly through the interaction of
of cancer in people with epilepsy, with specific
primidone with the synthesis of folylpolygluta-
reference to the role of anti-epilepsy drugs,
mates (Carl et al., 1987a).
noting studies concerning cancer of the liver,
lung, and brain. Despite considerable long-term
4.4 Susceptibility pharmaco-epidemiological data being available
for phenobarbital, evidence for carcinogenicity
No studies primarily addressing the suscep- in humans was not consistent and phenobarbital
tibility of humans to carcinogenesis induced was considered to be possibly carcinogenic to
by primidone were available to the Working humans by the authors.
Group. In a review, Singh et al. (2005) specu- The Working Group in 2000 concluded
lated that there might be a partly biological basis that phenobarbital was possibly carcinogenic to
(e.g. genetic predisposition) for the association humans (Group 2B) based solely on sufficient
between epilepsy and cancer, possibly involving evidence in experimental animals (IARC, 2001).
the tumour suppressor gene leucine-rich glioma Studies aiming to elucidate mechanisms of
inactivated 1 (LGI1). El-Masri & Portier (1998) carcinogenesis attributable to phenobarbital in
have suggested that there is wide inter-individual mice have been reported. Typically, these investi-
variation in the metabolic profile of primidone, gations exploited comparison between strains of
which may indicate the presence of people who mice that were variously sensitive and resistant
produce greater amounts of primidone metabo- to phenobarbital-induced hepatocarcinogen-
lites than the general population, and who are esis. Thus Watson & Goodman (2002) reported
202
Primidone
that there was a clear indication of more exten- respect to duration and post-discharge drug use)
sive changes in methylation in GC-rich regions and on potential confounders. The available
of DNA, primarily hypermethylation, in the studies were not informative on whether expo-
tumour-sensitive mice in response to treatment sure to primidone is a cancer hazard.
with phenobarbital.
Phillips et al. (2009) reported the effects of
treatment with phenobarbital on DNA methyl-
ation and gene expression that occurred only Primidone was tested for carcinogenicity in
in liver tumour-prone B6C3F1 mice but not one oral administration study in mice, and one
in tumour-resistant C57BL/6 mice, after 2 or oral administration study in rats. In male and
4 weeks of treatment. Differences in epigenetic female mice, feed containing primidone caused
control (e.g. DNA methylation) between species significant increases in the incidences of hepato-
could, in part, underlie the enhanced propensity cellular adenoma, of hepatocellular carcinoma,
of rodents, as compared with humans, to develop and of hepatocellular adenoma, hepatocellular
cancer. carcinoma and hepatoblastoma (combined).
Primidone also caused a significant increase in
the incidence of hepatoblastoma and of thyroid
5. Summary of Data Reported follicular cell adenoma in males. In male rats,
feed containing primidone caused a significant
5.1 Exposure data increase in the incidence of thyroid follicular
cell adenoma. Primidone also caused a small
Primidone is a synthetic drug that was used but significant increase in the incidence of renal
commonly as an oral anticonvulsant, begin- tubule adenoma or carcinoma (combined) in
ning in the 1950s. It is now only in modest use, males. There was no significant increase in the
predominantly for the treatment of essential incidence of any neoplasm in female rats.
tremor, with stable use over the past decade.
Exposure is likely to be predominantly through
use as a medication. Environmental contamina- 5.4 Mechanistic and other relevant
tion in groundwater has been reported. data
In humans, primidone is partly eliminated
5.2 Human carcinogenicity data unchanged via urinary excretion, or metabo-
lized, by hepatic cytochrome P450 isozymes
The available epidemiological studies eval- principally to phenylethylmalonamide and
uating exposure specifically to primidone were to phenobarbital, a non-genotoxic agent. The
limited to two casecontrol studies reporting data on genetic toxicity for primidone in tradi-
on several types of cancer nested in a cohort of tional assays are limited in scope and amount,
epileptic patients in Denmark. Small excesses of but suggest that any mutagenic action of the
malignant lymphoma and cancers of the lung and chemical is highly specific: clear demonstra-
urinary bladder were observed among patients tion of the mutagenic activity of primidone was
who were ever treated with primidone; however, limited to a single report of mutation induction
the findings were based on small numbers of in Salmonella typhimurium strain TA1535 in the
exposed cases. Other limitations included incom- absence of metabolic activation only, and at high
plete information on exposure to primidone (with concentrations. The majority of well-conducted
203
IARC MONOGRAPHS 108
studies of chromosomal damage available for Benedetti MS, Whomsley R, Baltes E, Tonner F (2005).
review suggested that primidone does not induce Alteration of thyroid hormone homeostasis by antie-
pileptic drugs in humans: involvement of glucuron-
chromosomal changes in vitro or in vivo. osyltransferase induction. Eur J Clin Pharmacol,
The reported carcinogenicity of primidone in 61(12):86372. doi:10.1007/s00228-005-0056-0
mice is likely to be mediated through a non-gen- PMID:16307266
Bishun NP, Smith NS, Williams DC (1975). Chromosomes
otoxic mechanism resulting from the metabo- and anticonvulsant drugs. Mutat Res, 28(1):1413.
lism of primidone to phenobarbital. doi:10.1016/0027-5107(75)90327-9 PMID:1143294
British Pharmacopoeia (2013). Primidone. London,
United Kingdom: Medicines and Healthcare products
Regulatory Agency (MHRA).
6. Evaluation Carl GF, Eto I, Krumdieck CL (1987a). Chronic treatment
of rats with primidone causes depletion of pteroylpenta-
glutamates in liver. J Nutr, 117(5):9705. PMID:3585552
6.1 Cancer in humans Carl GF, Gill MW, Schatz RA (1987b). Effect of chronic
primidone treatment on folate-dependent one-carbon
There is inadequate evidence in humans for metabolism in the rat. Biochem Pharmacol, 36(13):2139
the carcinogenicity of primidone. 44. doi:10.1016/0006-2952(87)90142-0 PMID:3606631
Cottrell PR, Streete JM, Berry DJ, Schfer H, Pisani F,
Perucca E etal. (1982). Pharmacokinetics of phenyle-
6.2 Cancer in experimental animals thylmalonamide (PEMA) in normal subjects and in
patients treated with antiepileptic drugs. Epilepsia,
There is sufficient evidence in experimental 23(3):30713. doi:10.1111/j.1528-1157.1982.tb06196.x
PMID:7084140
animals for carcinogenicity of primidone. Daley RD (1973). Primidone. In: Florey K, editor.
Analytical Profiles of Drug Substances. Academic
Press; pp. 409437.
6.3 Overall evaluation DrugBank (2013). Primidone. DrugBank: Open Data
Drug & Drug Target Database. Version 3.0. Available
Primidone is possibly carcinogenic to humans from: http://www.drugbank.ca/
(Group 2B). El-Masri HA, Portier CJ (1998). Physiologically based
pharmacokinetics model of primidone and its metab-
olites phenobarbital and phenylethylmalonamide in
humans, rats, and mice. Drug Metab Dispos, 26(6):585
References 94. PMID:9616196
eMC (2013). Primidone. Datapharm Communications
Ltd., electronic Medicines Compendium (eMC).
Adelw C, Ahlbom A, Feychting M, Johnsson F,
Available from: http://www.medicines.org.uk/emc,
Schwartzbaum J, Tomson T (2006). Epilepsy as a
accessed 4 September 2014.
risk factor for cancer. J Neurol Neurosurg Psychiatry,
Epstein SS, Arnold E, Andrea J, Bass W, Bishop Y (1972).
77(6):7846. doi:10.1136/jnnp.2005.083931
Detection of chemical mutagens by the dominant lethal
PMID:16705202
assay in the mouse. Toxicol Appl Pharmacol, 23(2):288
Anderson GD (1998). A mechanistic approach to antiepi-
325. doi:10.1016/0041-008X(72)90192-5 PMID:5074577
leptic drug interactions. Ann Pharmacother, 32(5):554
European Pharmacopoeia (2008). Primidone. European
63. doi:10.1345/aph.17332 PMID:9606477
Pharmacopoeia Work Programme; pp. 275253.
Baumel IP, Gallagher BB, DiMicco J, Goico H (1973).
Available from: http://www.edqm.eu/en/european-
Metabolism and anticonvulsant properties of primi-
pharmacopoeia-publications-1401.html, accessed 4
done in the rat. J Pharmacol Exp Ther, 186(2):30514.
September 2014.
PMID:4719783
FDA (2013). Primidone. Silver Spring (MD): United States
Baumel IP, Gallagher BB, Mattson RH (1972).
Food and Drug Administration. Available from: http://
Phenylethylmalonamide (PEMA). An important
www.accessdata.fda.gov/scripts/cder/drugsatfda/
metabolite of primidone. Arch Neurol, 27(1):34
index.cfm?fuseaction=Search.DrugDetails, accessed 4
41. doi:10.1001/archneur.1972.00490130036005
September 2014.
PMID:4626105
Ferranti V, Chabenat C, Mnager S, Lafont O (1998).
Simultaneous determination of primidone and its three
204
Primidone
major metabolites in rat urine by high-performance Leal KW, Rapport RL, Wilensky AJ, Friel PN (1979).
liquid chromatography using solid-phase extraction. Single-dose pharmacokinetics and anticonvulsant
J Chromatogr B Biomed Sci Appl, 718(1):199204. efficacy of primidone in mice. Ann Neurol, 5(5):4704.
doi:10.1016/S0378-4347(98)00356-9 PMID:9832377 doi:10.1002/ana.410050512 PMID:464548
Frey HH, Lscher W (1985). Pharmacokinetics of anti-ep- Linnebank M, Moskau S, Semmler A, Widman G, Stoffel-
ileptic drugs in the dog: a review. J Vet Pharmacol Ther, Wagner B, Weller M etal. (2011). Antiepileptic drugs
8(3):21933. doi:10.1111/j.1365-2885.1985.tb00951.x interact with folate and vitamin B12 serum levels.
PMID:3932673 Ann Neurol, 69(2):3529. doi:10.1002/ana.22229
Fujimoto JM, Mason WH, Murphy M (1968). Urinary PMID:21246600
excretion of primidone and its metabolites in rabbits. Martines C, Gatti G, Sasso E, Calzetti S, Perucca E (1990).
J Pharmacol Exp Ther, 159(2):37988. PMID:5638658 The disposition of primidone in elderly patients. Br J Clin
Hass U, Dnnbier U, Massmann G (2012). Occurrence Pharmacol, 30(4):60711. doi:10.1111/j.1365-2125.1990.
of psychoactive compounds and their metabolites tb03820.x PMID:2291873
in groundwater downgradient of a decommissioned McElhatton PR, Sullivan FM, Toseland PA (1977). Plasma
sewage farm in Berlin (Germany). Environ Sci Pollut level studies of primidone and its metabolites in the
Res Int, 19(6):2096106. doi:10.1007/s11356-011-0707-x mouse at various stages of pregnancy. Xenobiotica,
PMID:22227832 7(10):61722. doi:10.3109/00498257709038683
Hunt RJ, Miller KW (1978). Disposition of primidone, PMID:910462
phenylethylmalonamide, and phenobarbital in the MHRA (2013). UK Public Assessment Report. Mysoline
rabbit. Drug Metab Dispos, 6(1):7581. PMID:23277 50mg tablets (primidone) PL 20132/0006. London,
IARC (2001). Some thyrotropic agents. IARC Monogr United Kingdom: Medicines and Heathcare Products
Eval Carcinog Risks Hum, 79:iiv, 1725.http:// Regulatory Agency. Available from: www.mhra.gov.
monographs.iarc.fr/ENG/Monographs/vol79/index. uk/home/groups/par/documents/websiteresources/
php PMID:11766267 con088170.pdf, accessed 24 April 2015
IMS Health (2012a). National Disease and Therapeutic MicroMedex (2013). Primidone. MicroMedex 2.0. Ann
Index (NDTI). Plymouth Meeting, Pennsylvania: IMS Arbor (MI): Truven Health Analytics Inc.
Health. Morasch B (2013). Occurrence and dynamics of micro-
IMS Health (2012b). National Prescription Audit Plus pollutants in a karst aquifer. Environ Pollut, 173:1337.
(NPA). Plymouth Meeting, Pennsylvania: IMS Health. doi:10.1016/j.envpol.2012.10.014 PMID:23202643
IMS Health (2012c). Multinational Integrated Data Moriyama M, Furuno K, Oishi R, Gomita Y (1994).
Analysis (MIDAS). Plymouth Meeting, Pennsylvania: Simultaneous determination of primidone and its
IMS Health active metabolites in rat plasma by high-performance
Japanese Pharmacopoeia (2007). Primidone. In: Hosokawa liquid chromatography using a solid-phase extraction
R editor. The Japanese Pharmacopoeia. 15th ed. Tokyo, technique. J Pharm Sci, 83(12):17513. doi:10.1002/
Japan: Ministry of Health, Labour and Welfare; pp. jps.2600831220 PMID:7891306
10267. Mortelmans K, Haworth S, Lawlor T, Speck W, Tainer
Kauffman RE, Habersang R, Lansky L (1977). Kinetics of B, Zeiger E (1986). Salmonella mutagenicity tests:
primidone metabolism and excretion in children. Clin II. Results from the testing of 270 chemicals.
Pharmacol Ther, 22(2):2005. PMID:884921 Environ Mutagen, 8(S7):Suppl 7: 1119. doi:10.1002/
Kuhn J, Knabbe C (2013). Fully validated method for rapid em.2860080802 PMID:3516675
and simultaneous measurement of six antiepileptic Nagaki S, Ratnaraj N, Patsalos PN (1999). Blood and cere-
drugs in serum and plasma using ultra-performance brospinal fluid pharmacokinetics of primidone and its
liquid chromatography-electrospray ionization tandem primary pharmacologically active metabolites, pheno-
mass spectrometry. Talanta, 110:7180. doi:10.1016/j. barbital and phenylethylmalonamide in the rat. Eur J
talanta.2013.02.010 PMID:23618178 Drug Metab Pharmacokinet, 24(3):25564. doi:10.1007/
Lamminp A, Pukkala E, Teppo L, Neuvonen PJ (2002). BF03190029 PMID:10716065
Cancer incidence among patients using antiepileptic Nau H, Rating D, Huser I, Jger E, Koch S, Helge H
drugs: a long-term follow-up of 28,000 patients. Eur J (1980). Placental transfer and pharmacokinetics of
Clin Pharmacol, 58(2):13741. doi:10.1007/s00228-002- primidone and its metabolites phenobarbital, PEMA
0429-6 PMID:12012147 and hydroxyphenobarbital in neonates and infants of
Lanas FM, Sozza MA, Queiroz ME (2003). Simultaneous epileptic mothers. Eur J Clin Pharmacol, 18(1):3142.
plasma lamotrigine analysis with carbamazepine, doi:10.1007/BF00561476 PMID:7398746
carbamazepine 10,11 epoxide, primidone, phenytoin, NTP (2000). NTP Toxicology and Carcinogenesis Studies
phenobarbital, and PEMA by micellar electrokinetic of Primidone (CAS No. 125337) in F344/N Rats and
capillary chromatography (MECC). J Anal Toxicol, B6C3F1 Mice (Feed Studies). Natl Toxicol Program
27(5):3048. doi:10.1093/jat/27.5.304 PMID:12908944 Tech Rep Ser, 476:1290. PMID:12571687
205
IARC MONOGRAPHS 108
ONeil MJ (2006). The Merck Index. 14th Ed. Whitehouse Res, 138(1):714. doi:10.1016/0165-1218(84)90087-9
Station (NJ): Merck & Co., Inc. PMID:6541755
OEHHA (2013). Chemicals known to the state to Riva R, Albani F, Contin M, Baruzzi A (1996).
cause cancer or reproductive toxicity. CA: Office of Pharmacokinetic interactions between antiepileptic
Environmental Health Hazard Assessment, State of drugs. Clinical considerations. Clin Pharmacokinet,
California Environmental Protection Agency. Available 31(6):47093. doi:10.2165/00003088-199631060-00005
from: http://oehha.ca.gov/prop65/prop65_list/files/ PMID:8968658
P65single052413.pdf, accessed 30 June 2014. Sato J, Sekizawa Y, Owada E, Ito K, Sakuta N, Yoshihara
Olsen JH, Boice JD Jr, Fraumeni JF Jr (1990). Cancer in M etal. (1986). Sensitive analytical method for serum
children of epileptic mothers and the possible rela- primidone and its active metabolites for single-dose
tion to maternal anticonvulsant therapy. Br J Cancer, pharmacokinetic analysis in human subjects. Chem
62(6):9969. doi:10.1038/bjc.1990.424 PMID:2257233 Pharm Bull (Tokyo), 34(7):304952. doi:10.1248/
Olsen JH, Boice JD Jr, Jensen JP, Fraumeni JF Jr (1989). cpb.34.3049 PMID:3769107
Cancer among epileptic patients exposed to anti- Sato J, Sekizawa Y, Yoshida A, Owada E, Sakuta N,
convulsant drugs. J Natl Cancer Inst, 81(10):8038. Yoshihara M et al. (1992). Single-dose kinetics of
doi:10.1093/jnci/81.10.803 PMID:2716074 primidone in human subjects: effect of phenytoin on
Olsen JH, Schulgen G, Boice JD Jr, Whysner J, Travis LB, formation and elimination of active metabolites of
Williams GM etal. (1995). Antiepileptic treatment and primidone, phenobarbital and phenylethylmalona-
risk for hepatobiliary cancer and malignant lymphoma. mide. J Pharmacobiodyn, 15(9):46772. doi:10.1248/
Cancer Res, 55(2):2947. PMID:7812960 bpb1978.15.467 PMID:1287181
Olsen JH, Wallin H, Boice JD Jr, Rask K, Schulgen G, SciFinder (2013). SciFinder databases: a division of the
Fraumeni JF Jr (1993). Phenobarbital, drug metabo- American Chemical Society. Available from: https://
lism, and human cancer. Cancer Epidemiol Biomarkers www.cas.org/products/scifinder, accessed 30 June 2014.
Prev, 2(5):44952. PMID:8220089 Singh G, Driever PH, Sander JW (2005). Cancer risk in
zden S, zden T, Gm F, Tosun A (1989). Quantitative people with epilepsy: the role of antiepileptic drugs.
proton magnetic resonance analysis of primidone Brain, 128(Pt 1):717. doi:10.1093/brain/awh363
in solid dosage forms. Spectrosc Lett, 22(4):4716. PMID:15574465
doi:10.1080/00387018908053896 Smith DB, Mattson RH, Cramer JA, Collins JF, Novelly
Patsalos PN, Perucca E (2003). Clinically important drug RA, Craft B (1987). Results of a nationwide Veterans
interactions in epilepsy: general features and inter- Administration Cooperative Study comparing the
actions between antiepileptic drugs. Lancet Neurol, efficacy and toxicity of carbamazepine, phenobarbital,
2(6):34756. doi:10.1016/S1474-4422(03)00409-5 phenytoin, and primidone. Epilepsia, 28(s3):Suppl
PMID:12849151 3: S508. doi:10.1111/j.1528-1157.1987.tb05778.x
Phillips JM, Burgoon LD, Goodman JI (2009). The consti- PMID:3319543
tutive active/androstane receptor facilitates unique Stenchever MA, Allen M (1973). The effect of selected
phenobarbital-induced expression changes of genes antiepileptic drugs on the chromosomes of human
involved in key pathways in precancerous liver and lymphocytes in vitro. Am J Obstet Gynecol, 116(6):867
liver tumors. Toxicol Sci, 110(2):31933. doi:10.1093/ 70. PMID:4736739
toxsci/kfp108 PMID:19482888 Tanaka E (1999). Clinically significant pharmacoki-
Pisani F, Perucca E, Primerano G, DAgostino AA, Petrelli netic drug interactions between antiepileptic drugs.
RM, Fazio A etal. (1984). Single-dose kinetics of prim- J Clin Pharm Ther, 24(2):8792. doi:10.1046/j.1365-
idone in acute viral hepatitis. Eur J Clin Pharmacol, 2710.1999.00201.x PMID:10380060
27(4):4659. doi:10.1007/BF00549596 PMID:6519155 Thermoscientific (2004). Immonoassay products. Thermo
Rao KP, Shaheen S, Usha Rani MV, Rao MS (1986). Effect Scientific. Available from: http://www.thermoscientific.
of primidone in somatic and germ cells of mice. Toxicol com/en/search-results.html?keyword=primidone&ma
Lett, 34(2-3):14952. doi:10.1016/0378-4274(86)90204-3 tchDim=Y, accessed 30 June 2014.
PMID:3798475 Tocris (2013). Primidone safety data sheet. Tocris
Rezaei B, Jafari MT, Khademi R (2009). Selective separa- bioscience. Available from: http://www.tocris.com/
tion and determination of primidone in pharmaceutical literature/0830_sds.pdf?1432306358, accessed 4
and human serum samples using molecular imprinted September 2014.
polymer-electrospray ionization ion mobility spec- US Pharmacopeia (2007). Primidone. Report No.
trometry (MIP-ESI-IMS). Talanta, 79(3):66975. USP30-NF25. Rockville (MD): The United States
doi:10.1016/j.talanta.2009.04.046 PMID:19576428 Pharmacopeial Convention; pp. 3016.
Riedel L, Obe G (1984). Mutagenicity of antiepileptic drugs. US Pharmacopeia (2009). Material Safety Data Sheet
II. Phenytoin, primidone and phenobarbital. Mutat - Primidone. Rockville (MD): The United States
Pharmacopeial Convention.
206
Primidone
US Pharmacopeia (2013). Official Monographs

Primidone. Rockville (MD): The United States
Pharmacopeial Convention.
Watson RE, Goodman JI (2002). Effects of phenobarbital
on DNA methylation in GC-rich regions of hepatic
DNA from mice that exhibit different levels of suscep-
tibility to liver tumorigenesis. Toxicol Sci, 68(1):518.
doi:10.1093/toxsci/68.1.51 PMID:12075110
WHO (2007). Cumulative List No. 14 of INN in Latin,
English, French, Spanish, Arabic, Chinese and Russian
(in alphabetical order of the Latin name), with supple-
mentary information. Geneva, Switzerland: World
Health Organization. Available from: http://www.vpb.
gov.lt/files/539.pdf, accessed 30 June 2014.
Zesiewicz TA, Elble RJ, Louis ED, Gronseth GS, Ondo
WG, Dewey RB Jr etal. (2011). Evidence-based guide-
line update: treatment of essential tremor: report of
the Quality Standards subcommittee of the American
Academy of Neurology. Neurology, 77(19):17525.
doi:10.1212/WNL.0b013e318236f0fd PMID:22013182
Zolotareva GN, Akaeva EA, Goncharova RI (1979).
Antimutagenic activity of the antispasmodic repara-
tion hexamidine. Effect of hexamidine on level of spon-
taneous mutation in a number of subjects Dokl. Acad.
Nauk SSSR, 246:469471.[Russian]
207
SULFASALAZINE
1. Exposure Data 1.1.2 Structural and molecular formulae and

1.1 Identification of the agent OH
OH
1.1.1 Nomenclature
O
Chem. Abstr. Serv. Reg. No.: 599-79-1 (ONeil,
N N
2001; SciFinder, 2013)
Chem. Abstr. Serv. Name: Benzoic acid, 2-hy-
droxy-5-[2-[4-[(2-pyridinylamino)sulfonyl]
phenyl]diazenyl] (ONeil, 2001; SciFinder, N O S O
2013) NH
IUPAC systematic name: 2-Hydroxy-5-
[2-[4-[py ridine-2-ylsu lfa moyl)phenyl]
diazenyl]benzoic acid (Lide, 2005; European C18 H14 N4 O5 S
Pharmacopoeia, 2008) Relative molecular mass: 398.39
From ONeil (2001).
United States nonproprietary name (USAN):
Sulfasalazine
1.1.3 Chemical and physical properties of the
Synonyms: Benzoic acid, 2-hydroxy-5-
[[4-[(2-pyridinylamino)sulfonyl]phenyl]
pure substance
azo]-; 5-[[p-(2-Pyridylsulamoyl)phenyl]azo]
Description: Brownish-yellow, odourless crys-
salicylic acid (US Pharmacopeia, 2007, 2013)
tals (ONeil, 2001; European Pharmacopoeia,
See WHO (2007) for names in other languages. 2008).
Melting-point: Decomposes at 240245 C
(ONeil, 2001)
Density: 1.480.1g/cm3 at 20C (SciFinder,
2013)
Spectroscopy data: Ultraviolet, mass, and
nuclear magnetic resonance spectra have
been reported (McDonnell & Klaus, 1976;
European Pharmacopoeia, 2008)
209
IARC MONOGRAPHS 108
Solubility: Practically insoluble in water, 2-hydroxy-3-[2-[4-(pyridin-2-ylsulfamoyl)

ether, benzene, chloroform; very slightly phenyl]diazenyl]benzoic acid
soluble in ethanol; soluble in alkali hydrox- 5-[2-[4,5-bis(pyridin-2-ylsulfamoyl)biphe-
ides (McDonnell & Klaus, 1976; ONeil, 2001) nyl-2-yl]diazenyl]-2-hydroxy benzoic acid
Octanol/water partition coefficient: Log salicylic acid
P=3.88 (Rosenbaum, 2011) 2-hydroxy-5-[2-(4-sulfophenyl)diazenyl]
Stability data: The compound did not degrade benzoic acid
when dissolved in dimethylformamide and 4-amino-N-(pyridin-2-yl)benzenesulfona-
was subjected to thermal stress at 80C for mide (sulfapyridine).
196hours (McDonnell & Klaus, 1976; Jacoby,
2000)
1.1.5 Analysis
1.1.4 Technical products and impurities Selected compendial and non-compendial
methods of analysis are presented in Table 1.1.
(a) Trade names Sulfasalazine in human plasma can be deter-
Azulfidine EN-tabs; Azulfidine; Azaline; mined by high-performance liquid chromatog-
Sulfazine; Sulfazine EC; Apo-Sulfasalazine; raphy using ultraviolet detection (Fukino et al.,
PMS-Sulfasalazine; Salazopyrin En-Tabs; 2007). It can also be analysed through liquid
Sala
zo
pyrin; Azulfidina; Azulfin; Bomecon; chromatography-tandem mass spectrometry
Colo-Pleon; Disalazin; Falazine; Gastropyrin; in human plasma using electron spray ioniza-
Lazafin; Pyralin EN; Rosulfant; Salazine; tion techniques in multiple reaction monitoring
Salazodin; Salazopirina; Salazopyrin Entabs; mode, with a limit of quantification of 10ng/mL
Salazopyrin-EN; Salazopyrina; Salazopyrine; (Gu et al., 2011).
Salivon; Salopyr; Salopyrine; Saridine-E; In urban water, sulfasalazine can be quanti-
Sulcolon; Sulfasalazin; Sulfitis; Ulcol; Zopyrin fied by liquid chromatography-mass spectrom-
(Porter & Kaplan, 2013). etry using electron spray ionization. The limit of
quantification is 65ng/L (Tuckwell et al., 2011).
(b) Impurities
Impurities as given in European
Pharmacopoeia (2008):
1.2 Production
4,4-[(4-hydroxy-1,3-phenylene)bis(diaze 1.2.1 Production process
ne d i y l)]bi s[N - (p y r id i n-2 -y l)b e n z e ne Sulfasalazine does not occur in nature.
sulfonamide Sulfasalazine is produced by reacting sulfanila-
2-hydroxy-3,5-bis[2-[4-(pyridin-2-ylsulfa- mide with salicyclic acid through a series of steps,
moyl)phenyl]diazenyl]benzoic acid with water as a solvent (Novacek et al., 1991).
2-hydroxy-5-[2-[4-(2-iminopyridin-1(2H)-
yl)phenyl]diazenyl]benzoic acid 1.2.2 Use
4-[2-(2-hydroxyphenyl)diazenyl]-N-(pyri- (a) Indications
din-2-yl)benzenesulfonamide
Sulfasalazine is an aminosalicylate whose
2-hydroxy-4-(pyridin-2-ylsulfamoyl)-5-[2-
chief bioactive metabolite is 5-aminosalicyclic
[4-(pyridin-2-ylsulfamoyl) phenyl]diazenyl]
acid (5-ASA). Sulfasalazine, mesalazine (5-ASA),
biphenyl-3-carboxylic acid
210
Table 1.1 Analytical methods for sulfasalazine
Sample matrix Sample preparation Assay method Detection limit Reference

Compendial methods
Assay UV-visible spectroscopy European
Wavelength: 359 nm Pharmacopoeia (2008),
US Pharmacopeia (2013)
Related substances LC-UV European
Column: C18 Pharmacopoeia (2008)
Mobile phase A: sodium dihydrogen phosphate and
sodium acetate in water
Mobile phase B: mobile phase A:methanol (10:40)
Flow rate: 1 mL/min
Wavelength: 320 nm
Biological samples:
Human serum, breast LC 2.5 or 1.3mol/L Esbjrner et al. (1987)
milk (LOD)
Human plasma Centrifugation LC-UV Fukino et al. (2007)
Column: C18
Mobile phase: 35% acetonitrile in 25mM phosphate
buffer
pH 3.0
Wavelength: 365 nm
Human plasma Protein precipitation using ELISA 0.51ng/mL (sensitivity) Pastor-Navarro et al.
methanol treatment, solid Buffer solutions: phosphate buffer, carbonate/ 0.02 ng/mL (LOD) (2007)
phase extraction (1 mL of bicarbonate buffer, acetate buffer
methanol containing 5%
ammonia)
Human plasma Proteins precipitation LCESI-MS/MS 10 ng/mL (LOQ) Gu et al. (2011)
followed by centrifugation; Column: phenyl
supernatant was mixed Mobile phases: 0.2% formic acid, 2mM ammonium
with 100 L of water in acetate in water, 2mM ammonium acetate in
polypropylene tubes and methanol
transferred to the auto- MRM mode
sampler 399 m/z, 381 m/z
211
Sulfasalazine
212

Mouse plasma LC-UV 0.32 nmol/mL (LOD) Zheng et al. (1993)
Column: C18
Mobile phase: methanol and 25 mM phosphate
buffer (64:36)
pH 2.5
Flow rate: 1 mL/min
Rat plasma LC-UV 40 ng/mL (LOQ) Lee et al. (2012)
IARC MONOGRAPHS 108
Column: C18
Mobile phase: 25 mM phosphate buffer:methanol
(40:60)
pH 3.0
Wavelength: 360 nm
Food samples:
Pork meat Pressurized liquid extraction CE-ESI-MS2 6.25 g/kg (LOD) Font et al. (2007)
with hot water, clean-up Sheath liquid: methanol, waterand formic acid 21.3 g/kg (LOQ)
using oasis HLB cartridge (49.5:49.5:1)
(poly(divinylbenzene-co-N- Electrolyte: 50 mM ammonium acetate
pyrrolidone) pH 4.16
MRM mode
[M+H] + 398
317 m/z, 156 m/z, 108 m/z
Honey Added 10% trichloroacetic LC-APPI-MS/MS 0.44.5g/kg (LOD) Mohamed et al. (2007)
acid, heated at 65C, Column: C18 1.215.0g/kg (LOQ)
followed by liquidliquid Mobile phase: 0.5% formic acid (v/v) and 1 mM
extraction (acetonitrile, nonylfluoropentanoic acid (solvent A) and a mixture
dichloromethane), organic of methanol/acetonitrile (50/50, v/v), containing
phase was evaporated, 0.5% formic acid (solvent B)
reconstituted using Flow rate: 300 L/min
methanol:water (20:80) (SRM) positive ionization
Venalink blister packs Dissolve drug in LC-UV 0.1ng/mL (LOD) Elmasry et al. (2011)
(monitored dosage dimethylformamide, Column: C18 1ng/mL (LOQ)
system) dilute with methanol. Mobile phase: methanol, water, and acetic acid (70:
Internal standard was 1 29:1)
mL of 0.1% (w/v) 4-N,N- Flow rate: 1.5 mL/min
dimethylaminobenzaldehyde Wavelength: 365 nm

Environmental samples:
Water Aqueous sample was filtered, LC-ESI-MS2 955.3ng/mL (LOD) Fatta et al. (2007)
followed by SPE, and Column: C18
derivatized using acetonitrile Mobile phase A: 20 mM aqueous ammonium
and methanol acetate, 0.1% formic acid
Mobile phase B: 20 mM ammonium acetate in
acetonitrile :methanol (2:1)
Water Vacuum extraction, LC-ESI-MSn Effluent water, 150ng/L Tuckwell et al. (2011)
then evaporation under Mobile phase: water and acetonitrile with 0.1% River water, 65ng/L
gentle nitrogen stream. formic acid (LOQ)
Reconstitution with Flow rate: 0.2 mL/min
methanol Single parent ion (positive mode)
[M+H]+ 399
APPI, atmospheric pressure photospray ionization; CE-ESI-MS2, capillary electrophoresis-electrospray ionization-quadrupole ion trap-tandem mass spectrometry; ELISA, enzyme-
linked immunosorbent assay; ESI, electrospray ionization; LC, liquid chromatography; LOD, limit of detection; LOQ, limit of quantification; MRM, multiple reaction monitoring; MS,
mass spectrometry; MSn, multistage mass spectrometry; SPE, solid-phase extraction; SPME, solid-phase microextraction; SRM, selected reaction monitoring; UV, ultraviolet
213
Sulfasalazine
IARC MONOGRAPHS 108
Table 1.2 Most commonly reported clinical indications for sulfasalazine in the USA, 20112012
Diagnosis ICD code Drug uses Percentage of total

(in thousands)
Rheumatoid arthritis 714.001 367 46.2
Psoriatic arthropathy 696.001 90 11.3
Colitis, ulcerative, NOS 556.004 83 10.5
Colitis, presumed infectious 009.101 69 8.7
Spondylitis, ankylopietic 720.001 28 3.5
Crohn disease 555.901 27 3.4
Surgery after intestinal disease 67.050 25 3.1
Arthritis, spine, NOS 721.901 14 1.7
Rheumatoid arthritis plus inflammatory polyarthritis, unspecified 714.901 11 1.4
Total with reported diagnoses 795 100
but also olsalazine and balsalazide are all 5-ASA (c) Trends in use
drugs. As an anti-inflammatory and immuno- Total worldwide sales of sulfasalazine were
modulatory agent, sulfasalazine is used in the US$222million in 2012 (IMS Health, 2012b). The
treatment of autoimmune and inflammatory largest sales occurred in Japan (US$63million),
conditions, namely inflammatory bowel disease followed by the USA (US$ 32 million), the
(IBD), most prominently ulcerative colitis and United Kingdom (US$ 16 million), Germany
Crohn disease, as well as psoriatic and rheu- (US$12million), Australia (US$9million), and
matoid arthritis, including juvenile rheuma- Canada (US$8million). In the United Kingdom,
toid arthritis (IMS Health, 2012a; eMC, 2013; 49 tonnes of sulfasalazine were prescribed in
Table1.2). The use of sulfasalazine for the treat- 2007 (Tuckwell et al., 2011).
ment of urticaria has also been reported (McGirt Use of sulfasalazine has been relatively stable
et al., 2006). Sulfasalazine is recommended as a in the USA since 2005, at about 360 000 drug
third-line medication in all the above conditions uses per year. In the USA, approximately 100000
only when first- or second-line therapies have patients received sulfasalazine in 2012 (IMS
been ineffective. Health, 2012a) and 1.1million prescriptions for
(b) Dosage sulfasalazine were dispensed each year between
2008 and 2012 (IMS Health, 2012c).
Sulfasalazine is available as an oral dose at
250 or 500 mg, and as an oral suspension of
5mL. There is a wide range of dosing regimens, 1.3 Occurrence and exposure
varying from 500mg once per day to 1000mg
Sulfasalazine has been found as a persistent
four times per day, with 1000mg twice per day
residue in influent and effluent of a sewage treat-
being most common (35% of uses). The mean
ment plant, at concentrations of 0.1 to 0.4g/L
daily dosage for patients taking sulfasalazine is
(Tuckwell et al., 2011).
2150mg per day (IMS Health, 2012a; eMC, 2013).
Human exposure is largely limited to use as
a medication. Workers in plants manufacturing
sulfasalazine may be exposed.
214
Sulfasalazine
1.4 Regulations and guidelines the colorectum and exposure to dihydrofolate

reductase inhibitors (sulfasalazine, triamterene,
Sulfasalazine has been widely approved by and methotrexate) was identified (Coogan &
drug regulatory agencies around the world. In Rosenberg, 2007), but was not considered to be
the USA, it was approved by the Food and Drug informative because it did not provide a risk
Administration in 1950 (FDA, 2013). estimate specifically for sulfasalazine; for further
Sulfasalazine is listed as known to cause information, see the Monograph on triamterene
cancer by the State of Californias Office of in the present volume).
Environmental Health Hazard Assessment,
requiring public notice of potential environ-
mental exposure (OEHHA, 2013).
2.2 Longitudinal, cohort, and nested
casecontrol studies
See Table2.1 and Table2.2
2. Cancer in Humans Moody et al. (1996) evaluated long-term
treatment with sulfasalazine and risk of cancer
2.1 Background of the colorectum in a retrospective cohort of 175
patients with ulcerative colitis diagnosed between
Sulfasalazine, a member of the family of 5- 1972 and 1989 in Leicestershire, England.
ASA drugs (see Section 1.2.2), has been used since Clinical information, including compliance with
the 1950s to treat IBD (primarily ulcerative colitis) sulfasalazine treatment and history of cancer,
and, to a lesser extent, Crohn disease (Hanauer, was obtained from case records. A patient was
2004). IBD is associated with an increased risk considered to be non-compliant if there was
of dysplasia and cancer of the colorectum. clear evidence that the patient had ceased taking
Risk factors for IBD-associated cancer of the the medication or was instructed by the physi-
colorectum include duration, severity and extent cian to stop the medication without replacement
of colitis, the presence of coexistent primary by another 5-ASA drug. The crude proportion of
sclerosing cholangitis, and a family history of cases of cancer of the colorectum in the sulfasala-
cancer of the colorectum (Dyson & Rutter, 2012). zine non-compliant group (31%) was significantly
Chronic inflammation has been proposed as a higher than in the compliant group (3%), and a
mechanism for colorectal cancer associated with significant effect of compliance was observed in
IBD (or ulcerative colitis), and thus it has been survival analyses using log-ranked and Wilcoxon
suggested that 5-ASA drugs are chemopreventive methods. [This study was limited by lack of a
agents, because of their anti-inflammatory, anti- true unexposed group (the use of sulfasalazine
oxidant, and pro-apoptotic properties (Rubin in the non-compliant group was not known),
et al., 2006; Lakatos & Lakatos, 2008). small numbers, and limited information on use
The available epidemiological studies of sulfasalazine (e.g. time period, dose, dura-
included a surveillance study, two cohort studies, tion), or other medications, and information on
three nested casecontrol studies and three risk factors for cancer of the colorectum. It was
casecontrol studies of cancer of the colorectum not clear whether other 5-ASA drugs were used
among patients with IBD or ulcerative colitis. as a replacement for sulfasalazine in members
Some studies on IBD included patients with of the compliant group who developed cancer,
Crohn disease in addition to patients with ulcer- and whether the physician recommendations
ative colitis, but none of the studies stratified by for stopping treatment with sulfasalazine would
IBD subtype. A casecontrol study of cancer of affect cancer outcome.]
215
216
Table 2.1 Surveillance and cohort studies of cancer and sulfasalazine
Reference Total No. of Exposure Organ site Exposure categories Exposed Relative risk Covariates
Location, subjects assessment (ICD code) cases (95% CI) Comments
period
Moody 175 Case records Colorectum Crude proportion of cases P<0.001 Retrospective; cohort
et al. (1996) use and (compliant vs non-compliant; 2) comprised patients with
Leicester, compliance [OR for compliance] [0.07 (0.020.30)] total ulcerative colitis or
England Sulfasalazine non-compliant group 5 5/16 (31%) with limited colitis who
198192 were deceased; identified
Sulfasalazine compliant group 5 5/152 (3%)
via colitis database (case
IARC MONOGRAPHS 108
Survival analysis (cancer-free) P<0.001 ascertainment, 98%). Cancer

for compliance (log ranked and cases or dysplasia identified
Wilcoxon methods) via case records or registry;
biopsy diagnosis confirmed
on 10% sample. Survival
analysis adjusted for age and
sex
Lindberg 143 Hospital Colorectum Sulfasalazine 42 42/124 (34%) Surveillance study; patients
et al. (2001) records or (cancer or No sulfasalazine 8 8/18 (44%) with ulcerative colitis
Sweden, questionnaires dysplasia) Sulfasalazine vs no sulfasalazine P=0.38 participating in 20-yr
Follow-up, (t-test) colonoscopic surveillance
20 yr programme; 124 patients in
[OR for sulfasalazine] [0.6 (0.21.7)]
the treatment group
Cumulative risk of CRC/dysplasia P=0.40 Cumulative risk analysis
(sulfasalazine vs no treatment) adjusted for primary
Colorectum Sulfasalazine vs no sulfasalazine NR sclerosing cholangitis, sex,
(cancer (t-test) duration of ulcerative colitis,
only) Sulfasalazine 5 5/124 (4%) colectomy
Power, < 50%
No sulfasalazine 2 2/18 (11%)
[OR for sulfasalazine ] [0.3 (0.061.87)]
Reference Total No. of Exposure Organ site Exposure categories Exposed Relative risk Covariates
Location, subjects assessment (ICD code) cases (95% CI) Comments
period
van Staa 33905 with GPRD Colorectum Reference cohort (number not 116 1 Patients with 5-ASA drug
et al. (2005) 5-ASA drug (153, 154, reported); no history of IBD or prescriptions and/or IBD
United prescriptions 159) prescription for 5-ASA drug identified from GPRD;
Kingdom, 18969 5-ASA drug/IBD cohort (5-ASA 124 1.99 (1.542.56) patients with history of CRC
19872001 drug [excluding sulfasalazine] or excluded. Analysis adjusted
sulfasalazine and IBD); n=18969 for age and sex
Sulfasalazine rheumatoid arthritis 69 1.26 (0.941.70)
cohort (remaining sulfasalazine
without IBD); number, NR
Nested Sulfasalazine use 12 months before Cases selected from 5-ASA
casecontrol the index date: drug/IBD cohort, controls
analysis (100 Regular use 22 0.67 (0.361.25) randomly selected and
cases and 600 612 prescriptions 3 0.95 (0.224.11) matched for age, sex, and
controls) calendar year of index case.
1330 prescriptions 5 0.41 (0.141.20)
Analysis adjusted for BMI,
>30 prescriptions 14 0.77 (0.371.60) IBD duration, history of
Daily dose, <2 g 6 0.84 (0.292.42) colorectal polyps, NSAID,
Daily dose, 2 g 15 0.69 (0.351.37) paracetamol, aspirin,
immunosuppressive,
glucocorticoids, prior
hospitalization for
gastrointestinal condition,
physician visits, colonoscopy
5-ASA, 5-aminosalicylic acid; BMI, body mass index; CRC, colorectal cancer; GPRD, General Practice Research Database; IBD, inflammatory bowel disease; NR, not reported; NSAID,
nonsteroidal anti-inflammatory drugs; OR, odds ratio; vs, versus
217
Sulfasalazine
218
Table 2.2 Casecontrol studies of cancer of the colorectum and sulfasalazine
Reference Total Control Exposure Organ site Exposure Exposed Relative risk Covariates
Location, cases source assessment (ICD code) categories cases (95% CI) Comments
period Total (hospital,
controls population)
Pinczowski 102 Nested Medical Colorectum Sulfasalazine 48 0.38 (0.200.69) Age, number of
et al. (1994) 196 case-control; records use, one or more exacerbations/yr
Sweden, ulcerative treatment courses Controls matched by sex,
196583 colitis cohort (>3months) extent of ulcerative colitis
at diagnosis, and yr of
IARC MONOGRAPHS 108
diagnosis
Jess et al. 43 Nested Medical Colorectum Sulfasalazine NR 1.1 (1.01.3) Age and calendar yr of
(2007) 102 casecontrol: records (adenocarcinoma, use, cumulative diagnosis
Copenhagen IBD cohort; adenoma, dose: 2.9/1000g Controls from the same
Denmark, ulcerative or dysplasia (median) for cases regional cohort matched
196297; colitis or [combined]) vs 2.2 (median) for on sex, IBD (subtype,
Minnesota, Crohn disease controls duration, calendar yr, and
USA, Sulfasalazine, 36 1.1 (0.33.7) age of diagnosis). USA cohort
19402004 regular use (>2g/ followed until 2004, and
day) Danish followed until 1997.
Of the 43 cases, 23 were
CRC, 13 adenoma, and 7
dysplasia
Eaden et al. 102 IBD patients Medical Colorectum Sulfasalazine use: Possible selection bias,
(2000) 102 records <2 g/day 7 0.93 (0.223.91) source of population patients
England and 2 g/day 32 0.85 (0.322.26) from physician interested
Wales, [date, in study, controls from IBD
NR] Leicestershire database.
Controls matched for sex,
age (within 10 yr), extent and
duration of disease, but not
hospital or yr of diagnosis.
Adjusted for most influential
variables, other 5-ASA
drugs, contact with hospital
doctor, colonoscopies
diagnosis, relative with CRC
Reference Total Control Exposure Organ site Exposure Exposed Relative risk Covariates
Location, cases source assessment (ICD code) categories cases (95% CI) Comments
period Total (hospital,
controls population)
Rutter et al. 68 Hospital, Medical Sulfasalazine use: Controls matched for sex,
(2004) 136 colonoscopy records, Colorectum 10 yr 17 0.97 (0.412.26) extent and duration of
England, surveillance interviews, (cancer, adenoma, >10 yr 37 1.58 (0.713.51) ulcerative colitis, age of onset
19882002 and postal and dysplasia) of ulcerative colitis, yr of
questionnaires index colonoscopy; also had
Colorectum 10 yr 5 4.89 (0.4751.00)
to have intact colon and on
(cancer) >10 yr 8 6.59 (0.6467.92) surveillance within 5 yr of
case diagnosis
Terdiman et 18440 Population, Administrative Colorectum (ICD- Sulfasalazine use 1 64 2.33 (1.803.01) Controls free of cancer and
al. (2007) 368800 health-care claims in 9-CM) yr before diagnosis bowel surgery, matched
USA, 20013 database database of to cases by age, sex, and
364 IBD patients large health- Sulfasalazine use 1 44 1.19 (0.831.72) calendar year (20:1); CRC
1172 care insurance yr before diagnosis, but not IBD diagnosis
companies IBD patients internally validated. No
No. of adjustment in analyses
prescriptions: of any use 1yr before
diagnosis. Doseresponse
0 320 1.0 (ref.)
analysis adjusted for age, sex,
12 12 1.65 (0.803.39) colonoscopy, physician visits,
34 11 1.01 (0.462.21) ulcerative colitis or Crohn
5 21 1.10 (0.631.92) disease, hospitalization,
P for trend 0.27 NSAID, glucocorticosteroids,
and immunomodulators
5-ASA, 5-aminosalicylic acid; CRC, colorectal cancer; IBD, inflammatory bowel disease; ICD-9CM, International Classification of Diseases Ninth Revision, Clinical Modification; NR,
not reported; NSAID, nonsteroidal anti-inflammatory drugs; ref., reference; vs, versus; yr, year
219
Sulfasalazine
IARC MONOGRAPHS 108
The association between sulfasalazine intake calendar year) who had had prescriptions in the
and colorectal cancer or dysplasia was evalu- 6 months preceding the case index date. The
ated in a study of 143 patients with ulcerative type of 5-ASA drug was classified according to
colitis who underwent regular colonoscopies the last prescription issued before the index date.
and multiple biopsies in a 20-year surveillance Adjusted odds ratios (ORs) were <1 for regular
programme in Sweden (Lindberg et al., 2001). Of use or 612 prescriptions, 0.41 (0.141.20) for
the 143 patients, 124 were in the group that had 1330 prescriptions, and 0.77 (0.371.60) for
received treatment with sulfasalazine (treated > 30 prescriptions in the previous 12 months,
for at least 6months between onset of ulcerative and for daily doses of <2 g and 2 g; however,
colitis and start of surveillance by colonoscopy). they were not statistically significant and no clear
Dysplasia or cancer of the colon developed in 51 exposureresponse patterns were observed (see
patients. No statistically significant differences in Table 2.1). Duration of IBD was a strong risk
the adjusted cumulative risk analysis for devel- factor for cancer of the colorectum in the study
oping cancer or dysplasia of the colorectum, and was controlled for in the analyses. [This
or the percentage of cancers in the two treat- study had several advantages, including the
ment groups (44% in the non-treatment group prospective design, population-base selection of
compared with 34% in the treatment group) subjects, analyses by different exposure catego-
were reported. [The study was limited by small ries for specific 5-ASA drugs, and good clinical
numbers and limited exposure information.] information on each patient. However, informa-
A cohort of users of 5-ASA drugs was identi- tion on drug use was limited to the last prescrip-
fied from the General Practice Research Database tion, information on lifetime drug use was not
in the United Kingdom (van Staa et al., 2005). available, and there was limited information on
The cohort was divided into three subcohorts: follow-up procedures (e.g. no information was
(i) the 5-ASA drug/IBD cohort included 18969 provided on tracking individuals who moved out
patients who either had a prescription for an of the region of the United Kingdom database).
5-ASA drug (not including sulfasalazine) or who There was also a potential for misclassification
had taken sulfasalazine and had a diagnosis of of disease; classification appeared to be based
IBD; (ii) the remaining patients who were taking on the General Practice Research Database with
sulfasalazine but did not have IBD; and (iii) a diagnosis confirmed via physician question-
reference cohort consisting of patients without naire for a small subset of patients. The study
IBD or a prescription for a 5-ASA drug, matched had limited statistical power. Moreover, it was
by calendar year to participants receiving a unclear whether all the variables in the statistical
5-ASA drug. [The rationale for this approach models were potential confounders, which may
was that sulfasalazine is used to treat IBD and have further reduced the statistical power.]
other diseases in the United Kingdom, while the Two casecontrol studies were nested in
other 5-ASA drugs are only used to treat IBD.] population-based cohorts of patients with IBD
Relative risks (RRs) for incidence of cancer of (see Table 2.2). A Swedish study identified 102
the colorectum were 1.99 (95% CI, 1.542.56) for cases of cancer of the colorectum via linkage to the
the 5-ASA drug/IBD cohort, and 1.26 (95% CI, national Swedish cancer registry, among a cohort
0.941.70) for the sulfasalazine/non-IBD cohort. of patients with ulcerative colitis (Pinczowski
A nested casecontrol analysis was conducted et al., 1994). Living controls (n=196), with intact
among the cohort of patients receiving 5-ASA or partially intact colon, were matched to cases
drugs, which included 100 cases and 600 on sex, extent of disease, and time of diagnosis of
controls (matched to cases on age, sex, and disease. Information on pharmacological therapy
220
Sulfasalazine
(including sulfasalazine), clinical features of were identified from records of consenting

disease, smoking, and family history of IBD, gastroenterologists throughout England and
cancer of the colorectum, and other diseases, was Wales, and 102 controls matched by sex, age
collected from the patients medical records. A (categories of 10 years), extent and duration of
decreased risk of cancer of the colorectum was IBD were identified from the Leicestershire data-
found among individuals who had followed one base of IBD patients. Data were extracted from
or more treatment course of sulfasalazine (at least medical records. Sulfasalazine therapy at both
3months) (adjusted RR, 0.38; 95% CI, 0.200.69). < 2 g/day and 2 g/day was inversely associ-
[The strengths of the study were the prospective ated with increased risk of colorectal cancer in
design, the use of a population-based cohort, and unadjusted analyses, while adjusted odds ratios
the use of controls matched for disease severity. were 0.93 (95% CI, 0.223.91) for the group at
The major limitations were the lack of detailed the lower dose and 0.85 (95% CI, 0.322.26) for
exposure information and small size.] the group at higher doses. [The limitations of the
The second nested casecontrol study evalu- study were the potential for selection bias, inad-
ated the risk of colorectal neoplasia (adenocarci- equate matching of the controls (using categories
noma, adenoma, and dysplasia combined) in two of 10 years of age, and not matching on hospital
cohorts of patients with IBD in Denmark and in or year of diagnosis of IBD), limited exposure
Minnesota, USA (Jess et al., 2007). Both cohorts information, limited documentation of covari-
included patients with ulcerative colitis or Crohn ates controlled in the analyses. Odds ratios were
disease. Cases were identified via linkage with adjusted for the use of other 5-ASA drugs, but this
cancer registries, and controls matched for sex, may not be appropriate since these drugs could
vital status, and age at diagnosis, and clinical work via the same mechanisms as sulfasalazine.
factors related to IBD were identified from each The study population may have overlapped with
regional cohort. Exposure and clinical informa- the cohort reported by Moody et al. (1996); both
tion was obtained from medical records. The studies identified patients from the same data-
adjusted relative risk for colorectal neoplasia was base of patients, but the years of study recruit-
close to unity for regular use (>2g/day) or cumu- ment were not reported in the study by Eaden
lative dose of sulfasalazine. Primary sclerosing et al. (2000).]
cholangitis was a strong risk factor for colorectal One study evaluated risk factors for colorectal
neoplasia. [The advantages and limitations of this neoplasia (cancer, adenoma, and dysplasia) among
study were similar to those of the Swedish study. patients with chronic ulcerative colitis who were
An additional limitation of this study was that part of a colonoscopy surveillance programme
there was not a separate analysis for cancer of the (Rutter et al., 2004). Sixty-eight cases with
colorectum alone; of the 43 cases of colorectal neoplasia were identified and 136 controls from
neoplasia, 23 were cancer.] the surveillance population were matched on sex
and clinical characteristics of IBD. Long-term
use (> 10 years) of sulfasalazine was associated
2.3 Casecontrol studies with a non-statistically significant elevated odds
See Table2.2 ratio for colorectal neoplasia (cancer, adenoma, or
Three studies selected cases and controls dysplasia) of 1.58 (95% CI, 0.713.51; 37 exposed
from patients with ulcerative colitis. Eaden et al. cases); use of sulfasalazine for >3months to 10
(2000) evaluated the risk of colorectal cancer years was associated with an odds ratio of 0.97
among patients with ulcerative colitis and (95% CI, 0.412.26; 17 exposed cases). Analysis of
controls in England and Wales. Cases (n=102) the cancer cases only found that both exposure
221
IARC MONOGRAPHS 108
categories were associated with elevated, but very 2.4 Meta-analysis

imprecise odds ratios based on small numbers
(see Table2.1). [A significant association between A meta-analysis of four cohort studies
severity of inflammation and colorectal neoplasia assessed the association between long-term treat-
was noted, and this hindered interpretation of the ment with sulfasalazine and risk of colorectal
association of colorectal cancer and treatment cancer (Diculescu et al., 2003). [The Working
with sulfasalazine. Other concerns included the Group could not interpret this study due to the
small numbers of exposed cancer cases, lack of lack of information on analytical methods and
adjustment for risk factors, and limitations in the the inclusion of studies that were not specific
generalizability of the findings due to the selec- for treatment with sulfasalazine. Furthermore,
tion of subjects from a surveillance programme. summary measures of association were not
In addition, drug use may have been related to calculated.]
duration of ulcerative colitis, and thus matching
by duration of ulcerative colitis may bias the odds
ratio towards the null.] 3. Cancer in Experimental Animals
Terdiman et al. (2007) evaluated use of
sulfasalazine in a population based casecontrol 3.1 Mouse
study consisting of 18440 cases of cancer of the
colorectum and 368 800 controls (matched by See Table3.1
age, sex, and calendar year) identified from two
administrative databases covering all regions in Oral administration
the USA, including 364 cases and 1172 controls
In one study, groups of 50 male and 50 female
with a diagnosis of IBD. Information on claims
B6C3F1 mice (age, 6 weeks) were given sulfasala-
for sulfasalazine prescriptions and other clinical
zine (USP grade) at a dose of 0, 675, 1350, or
variables was obtained from the claim data-
2700 mg/kg bw in corn oil by gavage once per
base. A statistically significant increased risk of
day, 5days per week, for 104weeks. There was
cancer of the colorectum was found among all
a 618% decrease in mean body weight in male
patients using sulfasalazine 1 year before diag-
and female mice at the highest dose compared
nosis (crude OR, 2.33; 95% CI, 1.803.01), while
with controls. The incidence of hepatocellular
among patients with IBD only the adjusted odds
adenoma in males and females, and the inci-
ratio for sulfasalazine treatment was 1.19 (95%
dence of hepatocellular adenoma or carcinoma
CI, 0.831.72); no exposureresponse relation-
(combined) in males and females, were signifi-
ship was observed with number of prescriptions,
cantly greater than those in controls, and the
regardless of adjustment for multiple covariates
incidences increased with a positive trend. The
(P for trend, 0.27). [The strengths of this study
incidence of hepatocellular carcinoma was signif-
were the large size and population-based design,
icantly increased in female mice (Iatropoulos
and information on many (but not all) potential
et al., 1997; NTP, 1997a).
confounders, and some exposureresponse anal-
In a first experiment in a group of related
yses. The major limitation was the short exposure
studies, groups of 5060 male B6C3F1 mice (age,
duration (1 year before diagnosis); in addition,
6 weeks) were given sulfasalazine (USP grade) at
information was not available on several key
a dose of 0 or 2700mg/kg bw by gavage in corn
potential confounders, such as family history of
oil once per day, 5days per week, for 103weeks
colon cancer or IBD and other factors not related
(~2years); in a second experiment, an unexposed
to medications.]
222
Table 3.1 Studies of carcinogenicity in mice given sulfasalazine by gavage
Strain (sex) Dosing regimen, Incidence, (%) and/or multiplicity of tumours Significance Comments
Duration Animals/group at start
Reference
B6C3F1 Sulfasalazine (in corn oil) at a dose of 0, 675, 1350, or Males *P<0.001 (Poly-3 test) Purity, USP
(M,F) 2700mg/kg bw per day, 5 days per wk, for 104wk Hepatocellular adenoma: **P=0.002 (Poly-3 test) grade
104wk 50 M and 50 F/group 13/501, 32/50*, 28/50**, 42/50* ***P=0.004 (Poly-3 test)
NTP (1997a), Hepatocellular carcinoma: ****P=0.005 (Poly-3
Iatropoulos 13/50, 15/50, 23/50, 8/50 test)
et al. (1997) Hepatocellular adenoma or carcinoma (combined): *****P0.05 (Poly-3
24/501, 38/50***, 38/50****, 44/50* test)
Females 1 P<0.001 (Poly-3 trend
Hepatocellular adenoma: test)

12/501, 28/50*, 25/50**, 28/49* 2 P=0.005 (Poly-3 trend
Hepatocellular carcinoma: test)

2/50, 10/50*****, 10/50*****, 9/49*****
Hepatocellular adenoma or carcinoma (combined):
14/502, 32/50*, 28/50*, 29/49*
B6C3F1 (M) Exp. 1: sulfasalazine (in corn oil) at a dose of 0 or Hepatocellular adenoma: *P<0.001 (increase, Purity, USP
Up to 156wk 2700mg/kg bw per day, 5 days per wk, for 103wk Exp. 1: 13/50, 42/50* logistic regression test) grade
NTP (1997b), (~2 yr), fed ad libitum Exp. 2: 8/50, 42/50*
Abdo & Kari Exp. 2: unexposed group fed such that mean body Exp. 3: 13/52, 9/50 (~2 yr)
(1996) weight matched that of the treated group fed ad Exp. 3: 10/48, 14/50 (up to 3 yr)
libitum Hepatocellular carcinoma:
Exp. 3 (dietary restriction): two groups of 110mice Exp. 1: 13/50, 8/50
(one control and one dosed) were given identical Exp. 2: 6/50, 8/50
quantities of feed such that the control group would Exp. 3: 7/52, 1/50 (~2 yr)
attain body weights of approximately 80% that of Exp. 3: 16/48, 6/50 (up to 3 yr)
the ad libitum-fed controls; 60 mice per group were Hepatocellular adenoma or carcinoma (combined):
evaluated at 103wk (~2 yr) and the remaining 50 Exp. 1: 24/50, 44/50*
mice per group at 156wk (3yr), or when survival Exp. 2: 14/50, 44/50*
reached 20% Exp. 3: 18/52, 9/50 (~2 yr)
5060M/group Exp. 3: 21/48, 18/50 (up to 3 yr)
bw; body weight; Exp., experiment; F, female; M, male; wk, week; USP, United States Pharmacopoeia; yr, year
223
Sulfasalazine
IARC MONOGRAPHS 108
group was fed such that the mean body weight of Oral administration
the group matched that of the treated group fed
ad libitum. In a third experiment (dietary restric- In one study, groups of 50 male and 50
tion), two groups of 110 animals, one control female F344/N rats (age, 6 weeks) were given
group and one group given sulfasalazine at a sulfasalazine (USP grade) at a dose of 0, 84, 168,
dose of 2700mg/kg bw in corn oil were offered or 337.5mg/kg bw by gavage in corn oil once per
identical quantities of feed such that the control day, 5days per week, for 105weeks (core study;
group would attain body weights of approxi- continuous exposure). An additional group of
mately 80% those of the control group fed ad male rats (stop-exposure group) was treated
libitum. Sixty mice from each group were eval- with sulfasalazine in corn oil at 337.5mg/kg bw
uated at 103 weeks and the remaining 50 mice for 26weeks, and then with corn oil only for the
from each group were evaluated at 156 weeks remainder of the study (79 weeks). Survival of
(3 years), or at the time when survival reached male rats at the highest dose in the core study was
20%. significantly lower than that of controls, with
The mean body weight at 1year and survival most deaths occurring during the last 8weeks of
at 103weeks (~2years) for the mice treated with the study. Survival of all other treated groups was
sulfasalazine were decreased by 15% and 19%, similar to that of controls.
respectively, relative to controls. The body weight The incidence of transitional cell papilloma
and survival of the weight-matched vehicle-con- of the urinary bladder in the core study was
trol group were similar to those of the treated increased with a positive trend in the groups of
group fed ad libitum. Under the dietary restric- treated male rats; the incidence in the group at
tion protocol, the control and treated groups the highest dose was significantly increased. The
weighed 42g and 34g at 1year and had respective transitional cell neoplasms of the urinary tract
survival rates of 84% and 88% after 103 weeks. observed in the core study were not observed in
Exposure to sulfasalazine under ad-libitum the stop-exposure group. In exposed females,
feeding conditions for 103 weeks (~2 years) there were also low incidences of [rare] transi-
caused significantly increased incidences of tional cell papilloma of the kidney and of the
hepatocellular adenoma, and hepatocellular urinary bladder. All rats with transitional cell
adenoma or carcinoma (combined) in exposed papillomas of the urinary tract also had grossly
mice compared with the controls fed ad libitum visible concretions (calculi) in the kidney and/or
and the weight-matched controls. In contrast to urinary bladder (Iatropoulos et al., 1997; NTP,
the findings of the first two experiments, the inci- 1997a).
dence of hepatocellular tumours in dietary-re- In a first experiment in a group of related
stricted mice was significantly decreased in the studies, groups of 5060 male F344/N rats (age,
exposed group after 103 weeks, and was similar 6 weeks) were given sulfasalazine (USP grade)
to that in non-treated mice after 3years (Abdo & at a dose of 0 or 337.5mg/kg bw in corn oil by
Kari, 1996; NTP, 1997b). gavage once per day, 5days per week, for up to
104weeks; in a second experiment, an unexposed
group was fed such that the mean body weight of
3.2 Rat the group matched that of the treated group fed
ad libitum. In a third experiment (dietary restric-
See Table3.2
tion), two groups of 110rats, one control group
and one group given sulfasalazine at a dose of
337.5mg/kg bw in corn oil were offered identical
224
Table 3.2 Studies of carcinogenicity in rats given sulfasalazine by gavage
Strain (sex) Dosing regimen, For each target organ: Incidence, (%) and/or Significance Comments
Duration Animals/group at start multiplicity of tumours
Reference
F344/N Sulfasalazine (in corn oil) at a dose of 0, 84, 168, or 337.5mg/kg Males 1P<0.001 Purity, USP
(M,F) bw, once per day, 5 days per wk, for 105wk (core study); an Transitional cell papilloma of the urinary (Poly-3 trend grade
105wk additional group of male rats (stop-exposure group) was treated bladder: 0/501, 0/49, 2/50 (4%), 6/50 (12%)*; test)
NTP with sulfasalazine (in corn oil) at 337.5mg/kg bw for 26wk and stop exposure, 0/47 *P=0.011
(1997a), then with corn oil only for the remainder of the study (79wk) Females (Poly-3 test)
Iatropoulos 50 M and 50 F/group Transitional cell papilloma of the urinary
et al. (1997) bladdera: 0/49, 0/50, 2/50 (4%), 0/50
Transitional cell papilloma of the kidneyb:
0/50, 0/50, 0/50, 2/50 (4%)
F344/N (M) Exp. 1: sulfasalazine in corn oil at a dose of 0, or 337.5mg/kg bw, Transitional cell papilloma of the urinary *P=0.011 Purity, USP
Up to once per day, 5 days per wk for up to 104wk, fed ad libitum bladder: (logistic grade
130wk Exp. 2: unexposed group fed such that mean body weight Exp. 1: 0/50, 6/50* regression)
NTP matched that of the treated group fed ad libitum. Exp. 2: 0/50, 6/50*
(1997b), Exp. 3 dietary restriction: two groups of 110rats (one control Exp. 3: 0/51, 0/50 (~2 yr)
Abdo & and one dosed) were given identical quantities of feed such that Exp. 3: 0/49, 1/49 (up to 130 wk)
Kari (1996) the control group would attain body weights of approximately
80% those of the ad libitum-fed controls; 60rats/group were
evaluated at 104wk (~2 yr) and the remaining 50rats/group at
130wk, or when survival reached 20%
5060males/group
a Historical incidence for 2-year studies by the NTP in rats fed corn oil by gavage (vehicle control groups): 3/903 (0.3%0.8%); range, 02%
b Historical incidence for 2-year studies by the NTP in rats fed corn oil by gavage (vehicle control groups): 0/920
bw, body weight; Exp., experiment; F, female; M, male; NTP, National Toxicology Program; wk, week; USP, United States Pharmacopoeia; yr, year
225
Sulfasalazine
IARC MONOGRAPHS 108
quantities of feed such that the control group sulfasalazine there were 141 tumours of the
would attain body weights of approximately 80% intestine (P0.05, ANOVA test) with a tumour
that of the controls fed ad libitum. Sixty rats from multiplicity of 11.82.16 (P<0.05, t-test) (Davis
each group were evaluated at 104weeks and the et al., 1992).
remaining 50 rats from each group were evalu-
ated at 130weeks, or at the time when survival
reached 20%. 4. Mechanistic and Other
After 1 year, mean body weights for the Relevant Data
control and treated rats in the first experiment
were similar. Since there was negligible body 4.1 Absorption, distribution,
weight loss throughout the study, no adjustments metabolism, and excretion
were made to the weight-matched control group,
thereby yielding a redundant control group. 4.1.1 Humans
Survival at 2 years in the first experiment was (a) Absorption, distribution, metabolism, and
70% and 46% for the control and treated rats, excretion
respectively.
The metabolic scheme for sulfasalazine in
The incidence of transitional cell papilloma of
humans is shown in Fig.4.1 (Das & Dubin, 1976;
the urinary bladder was significantly greater in
NTP, 1997a).
exposed rats than in the controls fed ad libitum
Sulfasalazine is not absorbed to any signif-
in the first experiment, or weight-matched
icant extent from the stomach (Das & Dubin,
controls in the second experiment. All rats with
1976). Slow absorption of small amounts (~10
transitional cell papilloma of the urinary bladder
30%) via the small intestine has been reported
also had grossly visible concretions in the kidney
before enterohepatic recycling, and with the
and/or urinary bladder. In the third experiment,
majority of unchanged drug reaching the colon
no significant increase in the incidence of transi-
(Das & Dubin, 1976; Azad Khan et al., 1982).
tional cell papilloma of the urinary bladder was The sulfasalazine molecule comprises 5-ASA
observed (Abdo & Kari, 1996; NTP, 1997b). and sulfapyridine moieties, linked by an azo
bond, which is cleaved by bacterial azoreductases
3.3 Studies of co-carcinogenicity in the colon, releasing 5-ASA and sulfapyridine
(Azad Khan et al., 1982). This cleavage is the
A group of 12 male Wistar rats was given rate-limiting step for clearance of sulfasala-
1,2-dimethylhydrazine at a dose of 40mg/kg bw zine (Das & Dubin, 1976). Most of the 5-ASA
as a single subcutaneous injection each week, is excreted; approximately 50% directly in the
concurrently with sulfasalazine at a dose of faeces, and at least 25% via the kidneys (after
60 mg/kg bw per day by gavage, for 20 weeks. absorption and acetylation in the liver) (Das &
One control group of 11 male Wistar rats was Dubin, 1976; Azad Khan et al., 1982). In contrast,
given 1,2-dimethylhydrazine only. Development sulfapyridine is almost completely absorbed. In
of colon tumours (mainly adenocarcinomas) the liver, sulfapyridine undergoes hydroxylation
was assessed histologically at week 21. All rats and/or N-acetylation to 5-hydroxysulfapyri-
developed tumours of the intestine. In the dine, N4-acetylsulfapyridine, and N4-acetyl-5-
control group receiving 1,2-dimethylhydrazine hydroxysulfapyridine subsequently forming
only, there were 70 tumours of the intestine glucuronic acid conjugates, before excretion
with a tumour multiplicity of 6.4 0.69, while mainly in the urine (Das & Dubin, 1976; Azad
in the group given 1,2-dimethylhydrazine plus Khan et al., 1982).
226
Sulfasalazine
Fig.4.1 Metabolic pathways of sulfasalazine in humans

H OO C
3 2 O 3' 4'
4 1 2'
HO N =N S NH 5'
5 6 O 1' N 6'
Sulfasalazine
H OO C
O
HO N H2 H 2N S NH
O N
5-Aminosalicylic acid (5-ASA) Sulfapyridine
H OO C
HO N HCO CH 3
N-Acetyl-5-aminosalicylic acid
O O
H 2N S NH OH CH3 CO NH S NH
O N O N
5'-Hydroxysulfapyridine N 4-Acetylsulfapyridine
O O
COO H CH3 C O NH S NH OH
H 2N S NH O
OH O N
O N
O
N 4-Acetyl-5'-hydroxysulfapyridine
OH OH
5'-Hydroxysulfapyridine-O-glucuronide
O
COO H
CH3 C O HN S NH O
N OH
O
O
OH OH
N 4-Acetyl-5'-hydroxysulfapyridine-O-glucuronide
From Das & Dubin (1976), NTP (1997a). With kind permission from Springer Science and Business Media.
227
IARC MONOGRAPHS 108
In studies of serum from 10 healthy male (a) Variation in absorption, distribution,

volunteers given single oral doses of 4 g of metabolism, and excretion
sulfasalazine, parent drug was detectable at (i) IBD and rheumatoid arthritis
1.5hours after dosing, and at maximum concen-
trations at 35hours in nine subjects, and after The characteristics of absorption, metabo-
7 hours in one subject (Schrder & Campbell, lism, and excretion of the parent drug in four
1972). Metabolites (sulfapyridine, and acetylated patients with IBD (ulcerative colitis or Crohn
and glucuronidated derivatives) were detected in disease) were similar to those in four healthy
the serum at 35 hours after dosing (Schrder subjects, each given a single oral dose of sulfasala-
& Campbell, 1972). The pharmacokinetics zine (3 or 4g). However, absorption and urinary
of rectally administered sulfasalazine have excretion of the metabolite, sulfapyridine, was
been studied in three healthy male Japanese decreased in patients with IBD. The metabolism
volunteers (Tokui et al., 2002). Sulfasalazine of sulfasalazine was markedly reduced in patients
(6.5mmol), given as a single suppository, reached taking antibiotics and after removal of the large
maximum plasma concentration (2.50.4M) bowel (Azad Khan et al., 1982).
in 5 hours (Tmax), and an area under the curve Pharmacokinetic studies of sulfasalazine
(AUC) of 27.4 4.8 M.h. Parent drug was and its principal metabolites in 13 patients with
almost completely hydrolysed in the colon, and rheumatoid arthritis and 8 patients with IBD
the urinary recovery was only approximately given sulfasalazine as a single oral dose of 2 g
0.2%. Maximum plasma concentration (Cmax) of (Astbury et al., 1990), showed that patients with
the metabolite N-acetyl-5-aminosalicylic acid rheumatoid arthritis had a significantly higher
was 0.50.2 M, reached in 12 hours, while that (and more sustained) plasma concentration
of sulfapyridine was 1.2 0.4 M, reached in of sulfapyridine than did patients with IBD
5hours. 5-ASA was not detected in the serum. (medians of 14.0 g/mL and 7.4 g/mL, respec-
Administration of an enema containing 6.5mmol tively). Two factors may have contributed to high
of 5-ASA, resulted in Cmax 5.82.0 M in 1hour peak plasma concentrations of sulfapyridine in
and an AUC of 29.411.1M.h. In the urine, patients with rheumatoid arthritis: firstly, the
approximately 0.3% was recovered unchanged. metabolism of sulfapyridine may be impaired,
The Cmax for the acetylated metabolite, N-acetyl- and secondly, a larger quantity of sulfasalazine
5-aminosalicylic acid, was 13.3 3.6 M in may reach the lower bowel leading to higher
7hours. More than 10% of 5-ASA was excreted concentrations of the subsequent cleavage
in the urine as acetyl-5-aminosalicylic acid, compounds. The pharmacokinetics of sulfasala-
suggesting that absorption of 5-ASA is favoured zine were variable among patients; maximum
when administered rectally (Tokui et al., 2002). plasma concentrations of sulfapyridine ranged
A study to compare the absorption and from 8 to 22 g/mL in patients with rheumatoid
metabolism of oral preparations of sulfasalazine, arthritis, and from 5 to 18 g/mL in patients
mesalazine (5-ASA) and olsalazine (a dimer of with IBD. The elimination half-life of sulfasala-
5-ASA) used regularly by patients (n = 12, 13, zine ranged from 3 to 8 hours in patients with
and 8, respectively) for treatment of ulcerative rheumatoid arthritis, and from 4 to 9 hours in
colitis, showed considerably greater absorption of patients with IBD (Astbury et al., 1990).
5-ASA and less acetylation, in patients receiving
mesalazine than in those receiving olsalazine or
sulfasalazine (Stretch et al., 1996).
228
Sulfasalazine
(ii) Pregnancy as in maternal serum (Azad Khan & Truelove,

Sulfasalazine and its primary metabolites are 1979). It was estimated that an infant would
able to cross the placenta (Azad Khan & Truelove, receive sulfapyridine at dose of 34 mg/kg bw,
1979; Jrnerot et al., 1981). In five patients with after a maternal dose of 2 g of sulfasalazine per day
ulcerative colitis treated with sulfasalazine (0.5g, (Jrnerot & Into-Malmberg, 1979). Sulfapyridine
four times per day) throughout and after preg- and its acetylated and glucuronidated metabo-
nancy, sulfasalazine was detected in the umbilical lites have been shown to be excreted by babies,
cord blood (mean concentration, 50% of that in 12.5months after maternal dosing (Jrnerot &
maternal serum) and at very low concentrations Into-Malmberg, 1979).
in the amniotic fluid (Azad Khan & Truelove,
(c) Genetic polymorphisms
1979). Analyses of metabolites showed that total
concentrations of sulfapyridine were equal in (i) N-Acetyltransferases
maternal and cord sera, but concentrations The sulfasalazine molecule may be consid-
of free sulfapyridine were significantly lower ered as a slow-release carrier for sulfapyridine,
(P < 0.02) in cord sera. Total concentrations of but there is large inter-individual variation in the
acetylated sulfapyridine were significantly higher rate of metabolism of sulfapyridine, which can
(P<0.025) in cord sera than maternal sera. Total affect steady-state serum concentrations (Das &
concentrations of 5-ASA were very low in all Dubin, 1976). The rate of N-acetylation of 5-ASA
fluids analysed (Azad Khan & Truelove, 1979). to N-acetyl-5-aminosalicylic acid is under genetic
In the study by Jrnerot et al. (1981) of 11 preg- control (Das & Dubin, 1976). The cytosolic
nant patients with IBD treated with sulfasalazine N-acetyltransferase (NAT) family comprises
(1g daily), concentrations of sulfasalazine were NAT1 and NAT2, which catalyse the transfer of
almost identical in cord and maternal serum. acetyl groups from activated acetyl-coenzyme A
Seven of the women were analysed at a later date to the nitrogen of primary amines, hydrazines,
(424 months) after pregnancy; plasma concen- or hydrazides (Kuhn et al., 2010). Single nucleo-
trations of sulfasalazine remained the same, but tide polymorphisms have been detected in NAT2.
concentrations of sulfapyridine had increased, The wildtype has been designated NAT2*4, but
probably reflecting different extents of protein NAT2*5, NAT2*6, and NAT2*7 alleles encode
binding of these compounds, and the different N-acetyltransferase enzymes with amino-acid
distribution volume in pregnancy (Jrnerot et al., changes that cause reduced activity (slow acetyl-
1981). ating function). Patients with a slow acetylator
Small quantities of sulfasalazine and phenotype generally show significantly higher,
sulfapyridine have also been detected in breast and more sustained plasma concentrations of
milk (Azad Khan & Truelove, 1979; Jrnerot & sulfapyridine and its non-acetylated metabo-
Into-Malmberg, 1979). Mean concentrations of lites. The elimination half-life of sulfasalazine in
sulfasalazine and total sulfapyridine in milk, patients with a slow-acetylator phenotype may
compared with concentrations in maternal be approximately 50100% longer than in those
serum, were approximately 30% and 50%, respec- with a fast-acetylator phenotype (Taggart et al.,
tively, as reported by Azad Khan & Truelove 1992).
(1979), and negligible and 40%, respectively as A 13year study of 185 patients with ulcer-
reported by Jrnerot & Into-Malmberg (1979). ative colitis undergoing daily treatment with
The various metabolites of sulfapyridine were sulfasalazine (2 g per day) demonstrated that
present in approximately the same proportions serum concentrations of sulfapyridine (both free
229
IARC MONOGRAPHS 108
and total) were higher in patients with a slow- compared in patients with rheumatoid arthritis
acetylator phenotype (Azad Khan et al., 1983). (8 young and 12 elderly, with equal numbers of
Concentrations of sulfasalazine and of 5-ASA slow and fast acetylators in both age groups), each
were not significantly different in fast and slow receiving sulfasalazine at 2 g per day for 21 days
acetylators. These findings were confirmed by (Taggart et al., 1992). In the elderly, the elimi-
later studies (Taggart et al., 1992). nation half-life of sulfasalazine was increased
The frequency of polymorphism in NAT2 [possibly partly due to slow cleavage of the
varies in different racial or ethnic populations azo bond (Tett, 1993)], and steady-state serum
(Ma et al., 2009). Studies have shown that about concentrations of N-acetyl-5-aminosalicylic acid
60% of patients with IBD, studied in Edinburgh, were higher. The pharmacokinetics of sulfapyri-
Scotland, are slow acetylators (Das et al., 1973), dine were unchanged with age, but were influ-
and a similar proportion was found in healthy enced by acetylation status, in particular, with
volunteers in a study in Liverpool, England increased steady-state serum concentrations in
(Schrder & Evans, 1972). In Germany, a study of slow acetylators. Although age is a determinant
NAT2 genotype and acetylation, using sulfasala- of the steady-state concentration of salicylate
zine as probe drug, showed that 24 out of 44 moieties, the acetylator phenotype seems to play
healthy volunteers (54.5%) were slow acetylators, the greater role in determining serum concentra-
in accordance with the slight prevalence of slow tions of sulfapyridine (Taggart et al., 1992).
acetylators in central European (Caucasian) (ii) Role of transporter proteins
populations which has been reported in several
studies (Kuhn et al., 2010). An efflux ATP-binding cassette (ABC) trans-
The NAT2 genotype has also been investi- porter, the breast cancer resistance BCRP protein
gated in Asian populations. Studies in 21 Japanese (encoded by the ABCG2 gene), has a role in the
subjects (8 healthy subjects and 13 patients with pharmacokinetics of various drugs, including
IBD) given a single oral dose of sulfasalazine sulfasalazine. The BCRP protein is expressed at
at 40 mg/kg bw demonstrated generally good the luminal membrane of cells with key functions
correlation between three NAT2 genotypes in drug transport, namely, placental trophoblast
(rapid, intermediate, slow acetylators) and the cells, hepatocyte bile canaliculi, kidney cells, and
plasma or urinary concentrations of sulfapyri- enterocytes.
dine and N4-acetyl sulfapyridine (Tanigawara The poor bioavailability of sulfasalazine has
et al., 2002). Similar analyses in seven healthy long been attributed to its low solubility and poor
Japanese subjects after 8 days of continued permeability (Das & Dubin, 1976). However,
(multiple dosing) oral doses of 1 g of sulfasala- treatment of human T-cells with sulfasalazine
zine once per day also demonstrated correlation was shown to cause cellular drug resistance that
with genotype (Kita et al., 2001). In 18 healthy was mediated by induction of BCRP (ABCG2),
Chinese men given 1 g of sulfasalazine as a single suggesting that sulfasalazine may be a substrate
oral dose, the NAT2 gene was shown to be an for human BCRP (van der Heijden et al., 2004;
important determinant of metabolite profiles; Urquhart et al., 2008).
the frequency of slow acetylators in the Chinese A subsequent study in vitro, using data on
population was lower than in Caucasians, but expression in human tissue, indicated an asso-
higher than in the Japanese population (Ma et al., ciation between reduced cell surface expression
2009). of the 421C > A variant and reduced BCRP-
The effects of age and acetylator status on mediated transport of sulfasalazine in patients
the pharmacokinetics of sulfasalazine were
230
Sulfasalazine
carrying an A allele at position 421 of the BCRP In male and female B6C3F1 mice given
[ABCG2] gene (Urquhart et al., 2008). sulfasalazine as an intravenous dose at 5 mg/kg
Variation in the ABCG2 gene may impair bw, plasma concentrations of the parent drug
the transport of drugs that are substrates for rapidly declined with a mean elimination half-
the ABC transporter, leading to increased intes- life of 0.5hour in males and 1.2hour in females
tinal absorption, and/or decreased biliary excre- (Zheng et al., 1993). The sex-specific differences
tion, and result in high plasma concentrations, in clearance rate were reflected in AUCsulfasalazine
as demonstrated in 37 healthy Japanese people values (9.21 M.h-1 and 21.39 M.h-1, in males
selected according to ABCG2 and NAT2 genotype and females, respectively).
and given a single oral dose of 2 g of sulfasala- In male and female B6C3F1 mice given
zine (Yamasaki et al., 2008). They showed that sulfasalazine by gavage at various doses (67.5,
in ABCG2-A/A subjects, mean plasma AUC048h 675, 1350, or 2700 mg/kg bw), the bioavailability
and Cmax values for sulfasalazine were signifi- of the parent drug was approximately 17% (range,
cantly higher, and the AUC for sulfapyridine 1618%) at the lowest dose (67.5 mg/kg bw), and
was lower (except in those who also had the was lower (range, 39%) at the higher doses. Both
slow-acetylator NAT2 genotype) than in individ- sulfapyridine and N4-acetylsulfapyridine were
uals without the ABCG2 variant. The increased identified in the plasma. Sulfapyridine was elim-
AUC for sulfasalazine in ABCG2-A/A subjects inated more slowly than the parent compound,
may result from increased oral availability and/ and thus accumulated in all mice given multiple
or decreased hepatic clearance, since BCRP is doses of sulfasalazine. The differential between
expressed on enterocytes and hepatocytes. The plasma concentrations of sulfapyridine and
ratios of AUCacetylsulfapyridine/AUCsulfapyridine were sulfasalazine varied, however, between males and
significantly higher in subjects with the rapid- females: the AUCs for sulfapyridine, at all four
acetylator NAT2 phenotype than in those with doses of sulfasalazine, were higher than those
intermediate or slow genotypes, demonstrating of parent drug by 2132 times in males and by
that inter-individual variability in the pharma- 525 times in females, while maximum plasma
cokinetics of sulfasalazine can be attributed to concentrations were higher than those of the
genetic polymorphism in drug transport and parent drug by 68 times in male mice, and up
metabolism (Yamasaki et al., 2008). to 4 times in females. Plasma concentrations of
N4-acetylsulfapyridine were very low compared
4.1.2 Experimental systems with those of sulfapyridine. [This indicated slow
acetylation of sulfapyridine by B6C3F1 mice,
(a) Absorption, distribution, metabolism, and comparable to that found in humans with the
excretion slow-acetylator phenotype.] The pharmacoki-
Experimental studies in Sprague-Dawley rats netic pattern in B6C3F1 mice given multiple
given diets containing sulfasalazine have shown oral doses of sulfasalazine (i.e. daily doses for
that most of the sulfasalazine is reductively three consecutive days) was similar to that in
cleaved by intestinal bacteria to two compounds, mice given a single oral dose, but accumulation
5-ASA (which is poorly absorbed) and sulfapyri- of sulfapyridine was evident, producing greater
dine (which is well absorbed) (Peppercorn & AUCsulfapyridine values in the multiple-dose study
Goldman, 1972). After metabolism by mamma- than in the single-dose study (Zheng et al., 1993).
lian enzymes, 5-ASA and sulfapyridine are In a study by the NTP (1997a), male F344/N
excreted mainly in the faeces, and in the urine, rats were given sulfasalazine as an intravenous
respectively (Peppercorn & Goldman, 1972). dose at 5 mg/kg bw, and pharmacokinetic
231
IARC MONOGRAPHS 108
parameters were compared with those in the controls. This work thus demonstrated that Bcrp
study in male B6C3F1 mice by Zheng et al. (1993). has a key role in controlling [i.e. maintaining a
The rats retained sulfasalazine longer than mice; low (Dahan & Amidon, 2010)] oral bioavaila-
the AUC for sulfasalazine in rats was double that bility of sulfasalazine (Zaher et al., 2006).
in mice, while values for systemic clearance and In Caco-2 cells, sulfasalazine normally
apparent volume of distribution in mice were exhibits a basolateral-to-apical permeability
double those in rats. The rate of elimination of that is 19 times higher than apical-to-basolateral
sulfasalazine was similar in rats and mice; plasma permeability, indicative of net mucosal secre-
elimination rate constants were 1.47 per hour and tion (Dahan & Amidon, 2009). In this study of
1.28 per hour, respectively, and elimination half- three ATP-dependent efflux transporters (in
lives, 0.53 hour and 0.54 hour, respectively. In Caco-2 cells and rat jejunum), specific inhibitors
F344/N rats given a low oral dose of 67.5 mg/kg of BCRP and of MRP2 were shown to disrupt
bw, sulfasalazine and its metabolites were unde- the normal direction of sulfasalazine perme-
tectable; however, after a higher dose (675 mg/kg ability. The presence of both MRP2 and BCRP
bw), plasma concentrations of parent compound inhibitors produced an efflux ratio of 1, indi-
were detectable within 12 hours (NTP, 1997a). cating no efflux of sulfasalazine. Inhibitors of
P-glycoprotein had no effect on the movement
(b) Role of transporter proteins of sulfasalazine. The results thus suggested that
BCRP is a member of the ATP-dependent efflux efflux transport of sulfasalazine is mediated by
transporters, which includes P-glycoprotein or BCRP and MRP2 (Dahan & Amidon, 2009). A
multi-drug resistance protein 1 (MDR1/ABCB1) more recent study of sulfasalazine absorption
and multidrug resistance-associated protein 2 has shown that curcumin is a potent inhibitor
(MRP2/ABCC2). These transporter proteins of human BCRP. Curcumin not only increased
have significant roles in the processes of drug the plasma AUC08h eightfold in wildtype mice
absorption, distribution, and clearance, and are (but not in Bcrp/ mice), but also increased the
expressed at the apical membrane of cells in the plasma AUC024h by twofold at microdoses of
liver, kidney, brain, placenta, colon, and intes- sulfasalazine and by 3.2-fold at therapeutic doses
tine. From the latter location, on the villus tip of in humans (Kusuhara et al., 2012).
the apical brush-border membrane of intestinal Further studies of the absorption charac-
enterocytes, they actively cause efflux of drugs teristics of sulfasalazine in the isolated mouse
from gut epithelial cells back into the intestinal intestine, have indicated that both influx and
lumen. efflux transporters are involved in the intestinal
In experiments in Bcrp/ [Abcg2/] knockout absorption of sulfasalazine (Tomaru et al., 2013).
mice given sulfasalazine, the AUC for sulfasala- OATP2B1 is a multispecific organic anion influx
zine was greater than in wildtype mice by 13-fold transporter. Like BCRP, it is localized at the
after an intravenous dose (5 mg/kg bw) and brush-border membrane of intestinal epithelial
111-fold after an oral dose (20 mg/kg bw) (Zaher cells, and mediates uptake of many endogenous
et al., 2006). This treatment in the mdr1a [Abcb1a] and xenobiotic substrates from the lumen. Like
knockout mouse did not significantly change the BCRP, sulfasalazine is a substrate for OATP2B1
AUC for sulfasalazine. Furthermore, studies in (Kusuhara et al., 2012; Tomaru et al., 2013). The
wildtype mice treated with an inhibitor of Bcrp study by Kusuhara et al. (2012) was inspired by
(gefitinib) before an oral dose of sulfasalazine the finding that pharmacokinetic data (plasma
resulted in a 13-fold increase in the AUC of AUC for parent drug) from human subjects
plasma sulfasalazine compared with nontreated given sulfasalazine, at either a microdose (100 g
232
Sulfasalazine
suspension) or a therapeutic dose (2 g as tablets), 4.2.1 Humans

demonstrated nonlinearity between doses in the
AUC of plasma sulfasalazine (Kusuhara et al., See Table4.1
2012). Investigations of sulfasalazine transport Mitelman et al. (1982) reported that there was
were performed in three systems in vitro, namely: no clear evidence of chromosomal damage in
(i) ATP-dependent uptake of sulfasalazine by lymphocytes of patients receiving sulfasalazine
membrane vesicles expressing human BCRP; for 1month or 4months at 3 g per day, although
(ii) oral bioavailability of sulfasalazine in vivo, an effect could not be ruled out.
in wildtype and Bcrp/ mice; and (iii) uptake of Increased frequencies of micronucleus forma-
sulfasalazine in HEK293 cells transfected with tion and sister chromatid exchange in patients
the influx transporter OATP2B1 (SLCO2B1). The with IBD receiving sulfasalazine have been
results indicated that sulfasalazine is a substrate reported, but confounding factors were apparent
for OATP2B1, and that saturation of the influx in the study (Erskine et al., 1984).
transporter OATP2B1 at the therapeutic dose is a
possible mechanism underlying nonlinearity in 4.2.2 Experimental systems
the doseexposure relationship for sulfasalazine See Table4.1
(Kusuhara et al., 2012).
The nonsteroidal anti-inflammatory drug (a) Mutagenicity
indomethacin (an inhibitor of the MRP family Sulfasalazine was not mutagenic in assays for
that includes MRP2) has been shown to change, gene mutation in bacteria, including a variety of
in a concentration-dependent manner, the direc- strains of Salmonella typhimurium, Escherichia
tion of membrane permeability to sulfasala- coli, and Klebsiella pneumoniae, in a variety
zine in Caco-2 cells; high concentrations of of protocols, with or without metabolic acti-
indomethacin substantially reduced efflux of vation (Voogd et al., 1980; Zeiger et al., 1988;
sulfasalazine. Efflux was not however abolished, Iatropoulos et al., 1997). In addition, treatment
due to the contribution of BCRP to the control of with sulfasalazine did not result in an increase
absorption. Additionally, an indomethacin-in- in mutations conferring 6-thioguanine resist-
duced increase in sulfasalazine permeability ance in mouse lymphoma L5178Y cells, with or
through the gut wall was also shown in the rat without metabolic activation (Iatropoulos et al.,
jejunal perfusion model. The concomitant intake 1997).
of indomethacin and sulfasalazine may lead to
increased absorption of sulfasalazine in the (b) Chromosomal damage
small intestine, reducing its concentration in the Mackay et al. (1989) reported positive
colon, and potentially altering its therapeutic results in a test for induction of sister chro-
effect (Dahan & Amidon, 2010). matid exchange in cultured human lymphocytes
treated with sulfasalazine at a concentration of
4.2 Genetic and related effects 20160 g/mL in the absence of metabolic acti-
vation. In contrast, Bishop et al. (1990) observed
Several studies, particularly those conducted no increase in the frequency of sister chromatid
in vivo, have demonstrated genotoxicity associ- exchange in Chinese hamster ovary cells treated
ated with sulfasalazine and some of its metab- with sulfasalazine at up to 1000 g/mL, with or
olites. The mutagenicity of sulfasalazine and its without metabolic activation. An increase in the
two major metabolites, sulfapyridine and 5-ASA, formation of micronuclei in cultured human
was reviewed by Iatropoulos et al. (1997).
233
234
Table 4.1 Genetic and related effects of sulfasalazine
Test system Resultsa Doseb Reference

(LED or HID)
Without With
exogenous exogenous
metabolic metabolic
system system
In vitro
Salmonella typhimurium TA100, TA1535, TA97, TA98, reverse mutation c 6666 g/plate Zeiger et al. (1988)
Salmonella typhimurium TA100, TA1535, TA97, TA98, reverse mutation 6250 g/plate Iatropoulos et al. (1997)d
IARC MONOGRAPHS 108
Escherichia coli,WP2 uvrAp, reverse mutation 6250 g/plate Iatropoulos et al. (1997)d
Klebsiella pneumoniae, fluctuation test, base substitution mutation NT 0.5 g/L Voogd et al. (1980)
Mouse lymphoma L5178Y cells, 6-thioguanine resistance 700 g/mL (S9) Iatropoulos et al. (1997)d
500 g/mL (+S9)
Sister chromatid exchange, Chinese hamster ovary cells 160 g/mL (S9) Bishop et al. (1990)
1000 g/mL (+S9)
Sister chromatid exchange, human lymphocytes + NT 20 g/mL Mackay et al. (1989)
Micronucleus formation, human lymphocytes + NT 40 g/mL Mackay et al. (1989)
Chromosomal aberration, Chinese hamster ovary cells 1000 g/mL Bishop et al. (1990)
Chromosomal aberration, human lymphocytes 100 g/mL Iatropoulos et al. (1997)d
In vivo
Micronucleus formation, male B6C3F1 mouse, bone-marrow cells + 1000 mg/kg bw, po 2 Bishop et al. (1990)
Micronucleus formation, male and female B6C3F1 mouse, peripheral + 675 mg/kg bw, po Bishop et al. (1990)
blood erythrocytes 13wk
Micronucleus formation (kinetochore-positive), male B6C3F1 mouse, + 1389 mg/kg bw, po 3 Witt et al. (1992a)
bone-marrow cells
Micronucleus formation (total micronuclei), male B6C3F1 mouse, bone- + 1389 mg/kg bw, po 3 Witt et al. (1992a)
marrow cells
Micronucleus formation (kinetochore-negative), male B6C3F1 mouse, + 5634 mg/kg bw, po 3 Witt et al. (1992a)
bone-marrow cells
Micronucleus formation, male B6C3F1 mouse, bone-marrow cells + 1000 mg/kg bw, po 3 NTP (1997a)
Micronucleus formation, female B6C3F1 mouse, bone-marrow cells + 1000 mg/kg bw, po 3 NTP (1997a)
Micronucleus formation, male B6C3F1 mouse, bone-marrow cells 4000 mg/kg bw, po 1 NTP (1997a)

(LED or HID)
Without With
exogenous exogenous
metabolic metabolic
system system
Micronucleus formation, male F344 rat, bone-marrow cells ? 3000 mg/kg bw, po 3e NTP (1997a)
Chromosomal aberration, male B6C3F1 mouse, bone marrow cells 1000 mg/kg bw, po1 Bishop et al. (1990)
Chromosomal aberration, male B6C3F1 mouse, bone-marrow cells 4000 mg/kg bw, po 3 NTP (1997a)
Chromosomal aberration, male and female Sprague-Dawley rat, bone- 500 mg/kg bw, po 1f Iatropoulos et al. (1997)d
marrow cells
Chromosomal aberration, human lymphocytes 3 g/day for 4months Mitelman et al. (1980)
a +, positive; , negative; ?, inconclusive
b In-vitro test, g/mL; in-vivo test, mg/kg bw per day
c S9 from liver of Aroclor 1254-treated Sprague-Dawley rats and Syrian hamsters
d Studies conducted by Pharmacia, not published; results summarized in Iatropoulos et al. (1997)
e Two trials were conducted; the first trial gave positive results at the highest dose of 2700 mg/kg bw (po 3), and the second gave negative results (HID, 3000 mg/kg bw, po 3). The
overall result was equivocal on the basis of nonreproducibility of the positive response
f Chromosomal aberration was measured at 6, 24, and 48 hours after treatment
HID, highest ineffective dose; LED, lowest effective dose; NT, not tested; po, oral
235
Sulfasalazine
IARC MONOGRAPHS 108
lymphocytes after treatment with sulfasalazine micronucleus formation in bone marrow of male
(effective concentration range, 40160 g/mL) in rats given three doses of sulfasalazine (highest
the absence of metabolic activation was reported dose, 3000 mg/kg bw) were judged to be equiv-
by Mackay et al. (1989). No significant increases ocal (NTP, 1997a); in this assay, an initial trial
in the frequency of chromosomal aberration were gave a positive response at the highest dose of
observed in cultured Chinese hamster ovary 2700 mg/kg bw, but a second trial, with a highest
cells (Bishop et al., 1990), or cultured human dose of 3000 mg/kg bw, gave negative results.
lymphocytes (Iatropoulos et al., 1997), treated This apparent preferential activity in mice was
with sulfasalazine (concentration, up to 1000 or consistent with the observation that mice have a
100 g/mL, respectively). Thus the results of tests greater systemic exposure than rats to sulfapyri-
for chromosomal damage in vitro after treat- dine, the active moiety, after administration of
ment with sulfasalazine were generally negative, similar doses (Zheng et al., 1993).
although sporadic positive results were reported. In one study, no evidence for genotoxicity
In vivo, consistent with results reported in was obtained for sulfasalazine when tested for
assays in vitro, no increases in the frequency the induction of micronuclei in mouse bone
of chromosomal aberration were observed marrow, with or without pretreatment with folate.
in male mice or male and female rats treated Likewise, no evidence for formation of DNA
with sulfasalazine by gavage at doses of up to adducts was detected by 32P-postlabelling in rat
4000 mg/kg bw per day (Bishop et al., 1990; and mouse liver and urinary bladder (Iatropoulos
Iatropoulos et al., 1997; NTP, 1997a). et al., 1997). [The nuclease P1 enrichment proce-
The results of assays for micronucleus forma- dure was used in the 32P-postlabelling method.
tion in male and female mice treated with Assuming that N-hydroxylation of sulfapyri-
sulfasalazine were uniformly positive when dine occurred in vivo, adducts derived from this
multiple treatments (at least two) were employed metabolite would probably be lost.]
(Bishop et al., 1990; Witt et al., 1992a, NTP,
1997a). Further investigation of the nature of the 4.2.3 Genotoxicity of sulfasalazine
induced micronuclei revealed that the majority metabolites
were kinetochore-positive, suggesting that the
micronuclei contained whole chromosomes See Table4.2
rather than fragments, and were primarily due Sulfasalazine has two major metabolites,
to aneuploidy events rather than chromosome sulfapyridine (a carrier molecule that allows
breakage (Witt et al., 1992a). This observation transport of sulfasalazine to the intestine, where
was consistent with the negative results in assays it is activated) and 5-ASA, the therapeutically
for chromosomal aberration with sulfasalazine active moiety.
in vitro and in vivo (Mitelman et al., 1980; Bishop (a) 5-ASA
et al., 1990; Iatropoulos et al., 1997; NTP, 1997a).
The negative results of one test in male mice given 5-ASA has not shown activity in any assay for
a single dose of sulfasalazine at 4000 mg/kg bw genotoxicity in vitro or in vivo. It does not induce
underscored the need for multiple treatments to mutations in any of a variety of Salmonella typh-
induce an observable increase in micronucleus imurium strains, with or without metabolic acti-
formation (NTP, 1997a). vation or in Klebsiella pneumoniae in the absence
In addition to the necessity for multiple of metabolic activation (Voogd et al., 1980). No
treatments, sulfasalazine may also demonstrate induction of sister chromatid exchange, micro-
selective activity in mice; the results of a study on nucleus formation, or chromosomal aberration
236
Table 4.2 Genetic and related effects of metabolites of sulfasalazine

(LED or HID)
Without With exogenous
exogenous metabolic
metabolic system
system
5-Aminosalicylic acid
Salmonella typhimurium TA100, TA1537, TA98, reverse mutation 10 g/L (TA1537, TA98) Voogd et al. (1980)
5 g/L (TA100)
Klebsiella pneumoniae, fluctuation test, base-substitution mutation NT 0.5 g/L Voogd et al. (1980)
Sister chromatid exchange, Chinese hamster ovary cells, in vitro 280 g/mL Witt et al. (1992b)
Sister chromatid exchange, human lymphocytes, in vitro NT 16 g/mL Mackay et al. (1989)
Micronucleus formation, human lymphocytes, in vitro NT 16 g/mL Mackay et al. (1989)
Chromosomal aberration, Chinese hamster ovary cells, in vitro 280 g/mL Witt et al. (1992b)
Micronucleus formation, male B6C3F1 mouse, bone-marrow cells, in vivo 250 mg/kg bw, ip 3 Witt et al. (1992b)
Sulfapyridine
Sister chromatid exchange, Chinese hamster ovary cells, in vitro + 1667 g/mL Witt et al. (1992b)
Chromosomal aberration, Chinese hamster ovary cells, in vitro 5000 g/mL Witt et al. (1992b)
Sister chromatid exchange, human lymphocytes, in vitro + NT 200 g/mL Mackay et al. (1989)
Micronucleus formation, male B6C3F1 mouse, bone-marrow cells, in vivo + 1000 mg/kg bw, ip 3 Witt et al. (1992b)
Micronucleus formation (kinetochore-positive), male B6C3F1 mouse, + 714 mg/kg bw, po 3 Witt et al. (1992a)
bone-marrow cells, in vivo
Micronucleus formation (total micronuclei), male B6C3F1 mouse, bone- + 714 mg/kg bw, po 3 Witt et al. (1992a)
marrow cells, in vivo
Micronucleus formation (kinetochore negative), male B6C3F1 mouse, + 1429 mg/kg bw, po 3 Witt et al. (1992a)
bone-marrow cells, in vivo
N4-Acetylsulfapyridine
Micronucleus formation, human lymphocytes, in vitro + NT 20 g/mL Mackay et al. (1989)
N4-Acetyl-5-hydroxysulfapyridine
237
Sulfasalazine
238

(LED or HID)
Without With exogenous
exogenous metabolic
metabolic system
system
5-Hydroxysulfapyridine
IARC MONOGRAPHS 108

N-Acetyl-5-aminosalicylic acid
a
+, positive; , negative
b
In-vitro test, g/mL; in-vivo test, mg/kg bw per day
HID, highest ineffective dose; ip, intraperitoneal; LED, lowest effective dose; po, oral
Sulfasalazine
has been reported in human lymphocytes or lymphocytes. N4-acetylsulfapyridine was capable

Chinese hamster ovary cells in vitro (Mackay of inducing both sister chromatid exchange and
et al., 1989; Witt et al., 1992b). In vivo, no increase micronucleus formation, while N4-acetyl-5-
in the frequency of micronucleated polychro- hydroxysulfapyridine only induced sister chro-
matic erythrocytes was observed in the bone matid exchange. 5-Hydroxysulfapyridine and
marrow of male mice treated with 5-ASA (dose N4-acetyl-5-aminosalicylic acid did not induce
range, 125250 mg/kg bw per day for 3days) by either sister chromatid exchange or micronu-
intraperitoneal injection (Witt et al., 1992b). cleus formation at the concentrations tested.
(b) Sulfapyridine
4.3 Other mechanistic data relevant
Sulfapyridine has been reported to induce
sister chromatid exchange in Chinese hamster to carcinogenicity
ovary cells and cultured human lymphocytes 4.3.1 Adverse effects
in the absence of metabolic activation (Mackay
et al., 1989; Witt et al., 1992b); no increase in In humans, sulfasalazine is associated with
the frequency of sister chromatid exchange was a wide range of adverse side-effects that include
noted in the presence of metabolic activation in agranulocytosis (Kaufman et al., 1996), hepa-
Chinese hamster ovary cells (Witt et al., 1992b). totoxicity (de Abajo et al., 2004; Jobanputra
Sulfapyridine did not induce chromosomal aber- et al., 2008), nephrotoxicity (Gisbert et al., 2007),
ration in Chinese hamster ovary cells, with or neurotoxicity (Liedorp et al., 2008), and pulmo-
without metabolic activation (Witt et al., 1992b). nary toxicity (Parry et al., 2002). Sulfasalazine is
Mackay et al. (1989) reported that sulfapyri- also associated with reversible infertility in men
dine did not induce micronucleus formation in and in male experimental animals (OMorin
cultured human lymphocytes in the absence et al., 1984). Reactions to sulfasalazine may result
of metabolic activation, at concentrations that from an idiosyncratic delayed-type hypersensi-
reached 400 g/mL. Sulfapyridine induced a tivity reaction that may affect internal organs in
strong, dose-related increase in the frequency variable ways (Jobanputra et al., 2008).
of micronucleated polychromatic erythrocytes Case reports of serious hepatotoxicity asso-
when administered either as multiple intraperi- ciated with sulfasalazine are frequent and occur
toneal injections (Witt et al., 1992b) or by gavage predominantly within the first month of starting
(Witt et al., 1992a). As with sulfasalazine, the therapy; the pattern of liver injury can be hepa-
majority of micronucleated erythrocytes induced tocellular or cholestatic, and may lead to liver
by sulfapyridine in mice were shown to contain failure. Serious hepatotoxicity, which could be a
kinetochores (Witt et al., 1992a), implying that part of the DRESS (drug rash, eosinophilia and
sulfapyridine-induced micronucleus formation systemic symptoms) syndrome is described in
resulted from failure of mitotic chromosomal approximately 0.1% of users, but the estimated
segregation, rather than chromosome breakage. incidence is higher (0.4%) in patients with
inflammatory arthritis (de Abajo et al., 2004;
(c) Metabolites of 5-ASA and sulfapyridine Jobanputra et al., 2008).
Mackay et al. (1989) also tested four Renal toxicity associated with sulfasalazine
acetylated and/or hydroxylated metabolites treatment may be irreversible. Although the
of sulfapyridine and 5-ASA for their ability sulfapyridine moiety is thought to be respon-
to induce sister chromatid exchange and sible for most of the adverse effects of sulfasala-
micronucleus formation in cultured human zine, several case reports in patients with IBD
239
IARC MONOGRAPHS 108
indicate that renal toxicity in humans may (Pounder et al., 1975). Although sulfonamide-in-
occur from treatment with both sulfasalazine duced haemolysis can be severe in patients with
and 5-ASA (Gisbert et al., 2007). Clinically, glucose-6-phosphate dehydrogenase deficiency,
5-ASA-associated nephrotoxicity is typically this study showed that sulfasalazine-induced
expressed as interstitial nephritis, glomerulo- erythrocyte damage also occurred in patients
nephritis, nephritic syndrome, and acute renal with normal levels of this enzyme (Pounder
failure (Barbour & Williams, 1990; Birketvedt et al., 1975).
et al., 2000; Augusto et al., 2009). The incidence of The role of metabolites in sulfasalazine-me-
clinically restrictive renal impairment has been diated toxicity was investigated in vitro, using
estimated at <1 per 500 patients (World et al., human erythrocytes and mononuclear leuko-
1996). The mechanism is unclear, although both cytes as target cells in the presence of human liver
a delayed cell-mediated response, and a dose-de- microsomes; methaemoglobin formation and
pendent effect have been considered (Corrigan cytotoxicity were selected as toxicity end-points.
& Stevens, 2000). 5-ASA-related nephrotoxicity In addition to sulfasalazine, the study included
appeared to be dose-related in female rats given a the metabolites 5-ASA, sulfapyridine, and
single intravenous injection of the sodium salt of 5-hydroxysulfapyridine (Pirmohamed et al.,
5-ASA at doses of up to 5.7 mmol/kg bw (Calder 1991). Bioactivation by human liver microsomes
et al., 1972); however, dose-dependency may that are dependent on reduced nicotinamide
require the administration of doses much higher adenine dinucleotide phosphate (NADPH) to a
than those given to humans. Of note, 5-ASA species that caused methaemoglobinaemia and
combines the structural features of a salicy- cytotoxicity was only observed with sulfapyri-
late and a phenacetin, both of which have well dine. Chromatographic analysis demonstrated
documented nephrotoxic potential (Corrigan & that sulfapyridine was converted to a short-
Stevens, 2000). lived intermediate (t1/2, 8.1 minutes at pH 7.4)
It has been hypothesized that oxidative stress with elution characteristics identical to those
may be a factor in sulfasalazine-induced renal of synthetic sulfapyridine hydroxylamine. This
and hepatic injury. Treatment-related alterations hydroxylamine (10500 M) caused a concen-
in the levels of biomarkers of oxidative stress tration-dependent increase in both methae-
were detected in kidney and liver tissues of male moglobinaemia (2.924.4%) and cytotoxicity
Sprague-Dawley rats given sulfasalazine as daily in the absence of a microsomal system; neither
oral doses at 0, 300, or 600 mg/kg bw for 14 days. At sulfasalazine nor any of the other test metabolites
the highest dose, there were significant decreases had such effects. When the microsomal incuba-
in the activities of renal and hepatic superoxide tions were conducted in the presence of micro-
dismutase, and significant increases in catalase molar concentrations of reducing agents (e.g.
activity, thiobarbituric acid-reactive substances, ascorbic acid, glutathione, or N-acetylcysteine),
and in the oxidized/reduced glutathione ratio sulfapyridine-induced cytotoxity was decreased
(Linares et al., 2009). in mononuclear leukocytes, but there was no
Sulfasalazine can cause haemolytic anaemia effect upon the levels of methaemoglobinaemia.
(Das et al., 1973; Mechanick, 1985) and methae- This suggested that sulfapyridine hydroxy-
moglobinaemia (Miller et al., 1971; Kater, 1974; lamine could readily penetrate erythrocytes,
Azad Khan et al., 1983). In a group of 50 patients where it may undergo redox cycling to nitroso-
receiving sulfasalazine at 2.5 g per day as mainte- sulfapyridine, the species ultimately respon-
nance therapy for ulcerative colitis, approximately sible for the production of methaemoglobin.
40% had elevated levels of methaemoglobin These observations further suggested that
240
Sulfasalazine
N-hydroxylation of sulfapyridine may account acetylated sulfapyridine than fast acetylators

for some of the adverse effects associated with (see also Section 4.1.1), and appear more likely
sulfasalazine (Pirmohamed et al., 1991). [Of note, to experience toxic symptoms when treated with
N-hydroxylation is thought to play a key role in equivalent doses of sulfasalazine (Das & Dubin,
the bioactivation of aromatic amine carcinogens, 1976; Azad Khan et al., 1983; Ricart et al., 2002;
such as 4-aminobiphenyl (IARC, 2012).] Tanaka et al., 2002; Kumagai et al., 2004; Chen
et al., 2007; Soejima et al., 2008).
4.3.2 Effects upon folate pathways
Sulfasalazine has been shown to inhibit the 4.5 Mechanistic considerations
activity of dihydrofolate reductase, methylene-
Sulfasalazine was reported to act as a co-car-
tetrahydrofolate reductase, and serine transhy-
cinogen at a dose of 60 mg/kg bw per day in the
droxymethylase, and also the cellular uptake
1,2-dimethylhydrazine model of colon carcino-
of folate (Selhub et al., 1978; Jansen et al., 2004;
genesis in rats. In the same study, 5-ASA, the
Urquhart et al., 2010).
active pharmacophore unit of sulfasalazine,
acted as a co-carcinogen at a dose of 30 mg/kg bw
4.3.3 Urolithiasis
per day, but not at 60 mg/kg bw per day, which
An increased incidence of transitional cell suggested that 5-ASA exerts a protective effect on
papilloma of the urinary bladder in male rats the colon mucosa, provided a sufficient amount
treated with sulfasalazine has been correlated of the compound reaches the colon (Davis
(P < 0.01) with increased incidences of concre- et al., 1992). On the basis of early proposals that
tions (calculi) in the urinary bladder (NTP, 1997a; localized tissue folate deficiency may account
see also Section 3). In a subsequent study, there for carcinogenesis (Lashner et al., 1989), it was
was decreased incidence of urinary hyperplasia hypothesized that sulfasalazine may be co-car-
in male rats subjected to caloric restriction and cinogenic due to its anti-folate characteristics.
treated with sulfasalazine, and little evidence of Sulfasalazine inhibited dihydrofolate reductase,
urinary bladder concretion, compared with rats methylenetetrahydrofolate reductase, and serine
fed ad libitum (NTP, 1997b; see also Section 3). transhydroxymethylase, and also the cellular
[These data suggested that chronic inflamma- uptake of folate (Selhub et al., 1978; Jansen et al.,
tion associated with urolithiasis may be a factor 2004; Urquhart et al., 2010). Reduced levels of
in sulfasalazine-induced carcinogenesis of the S-adenosylmethionine or 5,10-methylenetet-
bladder in male rats.] rahydrofolate, required for thymidine synthesis,
might account for the effect; however, colonic
4.4 Susceptibility cells may not be completely dependent on blood
stream nutrients (Meenan, 1993). In the rat,
The adverse effects of sulfasalazine have been colonic bacterial folate is incorporated in the
linked to sulfapyridine (Das et al., 1973). This hepatic folate pool (Rong et al., 1991), and this
metabolite, which is well absorbed from the colon, could counteract sulfasalazine-induced folate
is inactivated by NAT2-mediated N-acetylation. depletion (Meenan, 1993). In patients with ulcer-
NAT2 polymorphisms have been associated with ative colitis, folate concentrations measured in
different susceptibilities to the adverse effects of colonic epithelial cells obtained from endoscopic
sulfasalazine. People with the slow-acetylator colon biopsies were not decreased in sulfasala-
genotype have higher serum concentrations of zine-treated patients compared with controls;
free sulfapyridine and lower concentrations of this contrasted with serum concentrations of
241
IARC MONOGRAPHS 108
folate, which were reduced in patients receiving 5. Summary of Data Reported

sulfasalazine (Meenan et al., 1996). These data
suggested that the potentially protective effects 5.1 Exposure data
of folate supplementation against colorectal
carcinogenesis in patients with ulcerative colitis Sulfasalazine is a synthetic aminosalicylate
were not due to correction of localized folate used as an oral anti-inflammatory drug. The
deficiency. most common use of sulfasalazine is for the
Two-year studies in male and female F344/N treatment of autoimmune arthritis. Prescriptions
rats given sulfasalazine by gavage indicated some have been stable over the past decade, with global
evidence for carcinogenic activity on the basis of sales of US$222 million in 2012. Environmental
increased incidences of transitional cell papil- contamination with sulfasalazine in ground-
loma of the urinary bladder, and clear evidence water has been noted, but exposure is likely to
for carcinogenic activity in male and female be predominantly through use as a medication.
B6C3F1 mice on the basis of increased incidences
of hepatocellular adenoma and hepatocellular 5.2 Human carcinogenicity data
carcinoma (NTP, 1997a; see also Section 3).
The data on mutagenicity of sulfasalazine and The available studies of exposure to sulfasala-
its metabolite, sulfapyridine, suggested that the zine included a surveillance study, two cohort
parent drug and the metabolite are predomi- studies, three nested casecontrol studies, and
nantly aneugens (Bishop et al., 1990; Witt et al., three casecontrol studies on cancer of the
1992a, b). Increased frequencies of micronucleus colorectum among patients with inflammatory
formation and sister chromatid exchange in bowel disease or ulcerative colitis. These studies
patients with IBD receiving sulfasalazine have were hampered in their ability to evaluate the
been reported, but confounding factors were association between exposure to sulfasalazine
apparent in the study (Erskine et al., 1984). and risk of cancer of the colorectum by the
Folate deficiency was considered as a possible small numbers of exposed cases, imprecise risk
explanation for the induction of micronucleus estimates, and little information on exposure or
the dose or duration of sulfasalazine use. There
formation by sulfasalazine in vivo. However,
were also concerns about selection bias in some
patients reported to have an elevated frequency
studies based on clinical populations.
of sister chromatid exchange and micronucleus
The best designed studies were a nested
formation had serum folate concentrations that
casecontrol study from a large cohort from
were at the low end of the normal range, and the
the General Practice Research Database in the
observation of reticulocytosis in a 90-day study United Kingdom, and a large population-based
in mice suggested an erythropoietic effect not casecontrol study among patients with inflam-
characteristic of folate deficiency (Bishop et al., matory bowel disease in the USA that evaluated
1990). exposureresponse relationships. Conflicting
The increased incidence of transitional cell findings were reported across studies, with
papilloma of the urinary bladder in male rats four studies reporting relative risks of less than
treated orally with sulfasalazine was correlated unity, two studies reporting estimates close to
with increased incidence of concretions (calculi) unity, and two studies reporting a relative risk
in the urinary bladder (NTP, 1997a). Chronic of greater than unity. Most of these relative risks
inflammation associated with urolithiasis may were not statistically significant. In the studies
be a factor in sulfasalazine-induced carcinogen- that evaluated doseresponse relationships, no
esis of the bladder in male rats. clear patterns were observed.
242
Sulfasalazine
5.3 Animal carcinogenicity data 5.4 Mechanistic and other relevant

In one study in male and female mice given data
sulfasalazine by gavage, there was a signifi- The sulfasalazine molecule contains a
cant increase in the incidence of hepatocellular 5-aminosalicylic acid moiety linked by an azo
adenoma, and of hepatocellular adenoma or bond to a sulfapyridine moiety. Cleavage of
carcinoma (combined) in both sexes; there was the azo bond by bacterial azoreductases in the
also an increase in the incidence of hepatocellular colon releases two pharmacologically active
carcinoma in females. compounds: 5-aminosalicylic acid and sulfapyri-
In a study of dietary restriction in male mice dine. Sulfapyridine is absorbed, and N-acetylated
given diets containing sulfasalazine ad libitum, by the highly polymorphic N-acetyl transferase
there were significant increases in the incidence 2 (NAT2), resulting in considerable inter-in-
of hepatocellular adenoma, and hepatocellular dividual variation in the pharmacokinetics of
adenoma or carcinoma (combined) in exposed sulfasalazine.
mice compared with the controls fed ad libitum Sulfasalazine is not mutagenic in standard
and the weight-matched controls. In contrast, the bacterial assays for gene mutation, with or without
incidence of hepatocellular tumours in dietary-re- exogenous metabolic activation. Tests for chro-
stricted mice was significantly decreased after mosomal damage in vitro after treatment with
2 years in the group exposed to sulfasalazine, sulfasalazine were generally negative, although
and was similar to that in non-treated mice after sporadic positive results have been reported.
3years. Likewise, no increases in the frequency of chro-
In one study in male and female rats given mosomal aberration were observed in male mice
sulfasalazine by gavage, there was a significant or rats treated with sulfasalazine. Positive results
increase in the incidence of transitional cell were consistently obtained in assays for micro-
papilloma of the urinary bladder in males; there nucleus formation in male and female mice in
was also a non-significant increase in the inci- vivo when multiple treatments with sulfasala-
dences of rare transitional cell papilloma of the zine were given; the results suggested that these
kidney and of rare transitional cell papilloma of results were primarily due to aneuploidy events
the urinary bladder in female rats. rather than chromosome breakage.
In a study of dietary restriction in male rats, Sulfasalazine inhibits the activity of dihy-
the incidence of transitional cell papilloma of the drofolate reductase, methylenetetrahydrofolate
urinary bladder was significantly greater in rats reductase, and serine transhydroxymethylase,
receiving sulfasalazine than in controls fed ad and also the cellular uptake of folate. However,
libitum. No significant increase in the incidence folate deficiency does not appear to account for
of transitional cell papilloma of the urinary the effects of sulfasalazine in humans and mice.
bladder was observed in rats subjected to dietary The sulfasalazine metabolite sulfapyridine
restriction and given sulfasalazine. has been shown to undergo N-hydroxylation
In a co-carcinogenicity study in male rats, when incubated with human liver microsomes
sulfasalazine increased the total number and in the presence of NADPH. N-Hydroxylation
multiplicity of 1,2-dimethylhydrazine-induced is known to account for the bioactivation of
intestinal tumours. carcinogenic aromatic amines to DNA-binding
species, and such a pathway would be consistent
with the target organs associated with sulfasala-
zine-induced carcinogenicity in rats. However, no
243
IARC MONOGRAPHS 108
evidence of DNA-adduct formation was detected nephritis. Nephrol Dial Transplant, 24(9):29402.
in rat and mouse liver or urinary bladder. doi:10.1093/ndt/gfp277 PMID:19509026
Azad Khan AK, Nurazzaman M, Truelove SC (1983). The
Male rats treated orally with sulfasalazine effect of the acetylator phenotype on the metabolism
had an increased incidence of transitional cell of sulphasalazine in man. J Med Genet, 20(1):306.
papilloma of the urinary bladder, and this was doi:10.1136/jmg.20.1.30 PMID:6133000
Azad Khan AK, Truelove SC (1979). Placental and
correlated with increased incidences of calculi mammary transfer of sulfasalazine. BMJ, 2(6204):1553
in the urinary bladder. Chronic inflammation doi:10.1136/bmj.2.6204.1553
associated with urolithiasis may be a factor in Azad Khan AK, Truelove SC, Aronson JK (1982). The
sulfasalazine-induced carcinogenesis of the disposition and metabolism of sulphasalazine (salic-
ylazosulphapyridine) in man. Br J Clin Pharmacol,
bladder in male rats. 13(4):5238. PMID:6121576
Barbour VM, Williams PF (1990). Nephrotic syndrome
associated with sulphasalazine. BMJ, 301(6755):818
doi:10.1136/bmj.301.6755.818-b PMID:1977483
6. Evaluation Birketvedt GS, Berg KJ, Fausa O, Florholmen J (2000).
Glomerular and tubular renal functions after long-
term medication of sulphasalazine, olsalazine, and
6.1 Cancer in humans mesalazine in patients with ulcerative colitis. Inflamm
Bowel Dis, 6(4):2759. PMID:11149559
There is inadequate evidence for the carcino- Bishop JB, Witt KL, Gulati DK, MacGregor JT (1990).
genicity of sulfasalazine in humans. Evaluation of the mutagenicity of the anti-in-
flammatory drug salicylazosulfapyridine (SASP).
Mutagenesis, 5(6):54954. doi:10.1093/mutage/5.6.549
6.2 Cancer in experimental animals PMID:1979833
Calder IC, Funder CC, Green CR, Ham KN, Tange JD
There is sufficient evidence for the carcino- (1972). Nephrotoxic lesions from 5-aminosalicylic
genicity of sulfasalazine in experimental animals. Acid. BMJ, 1(5793):1524. doi:10.1136/bmj.1.5793.152
PMID:20791817
Chen M, Xia B, Chen B, Guo Q, Li J, Ye M etal. (2007).
N-acetyltransferase 2 slow acetylator genotype asso-
6.3 Overall evaluation ciated with adverse effects of sulphasalazine in the
treatment of inflammatory bowel disease. Can J
Sulfasalazine is possibly carcinogenic to Gastroenterol, 21(3):1558. PMID:17377643
humans (Group 2B). Coogan PF, Rosenberg L (2007). The use of folic
acid antagonists and the risk of colorectal cancer.
Pharmacoepidemiol Drug Saf, 16(10):11119.
doi:10.1002/pds.1442 PMID:17600846
References Corrigan G, Stevens PE (2000). Review article: intersti-
tial nephritis associated with the use of mesalazine
Abdo KM, Kari FW (1996). The sensitivity of the NTP in inflammatory bowel disease. Aliment Pharmacol
bioassay for carcinogen hazard evaluation can be Ther, 14(1):16. doi:10.1046/j.1365-2036.2000.00683.x
modulated by dietary restriction. Exp Toxicol Pathol, PMID:10632639
48(23):12937. doi:10.1016/S0940-2993(96)80033-9 Dahan A, Amidon GL (2009). Small intestinal efflux
PMID:8672866 mediated by MRP2 and BCRP shifts sulfasalazine
Astbury C, Taggart AJ, Juby L, Zebouni L, Bird HA (1990). intestinal permeability from high to low, enabling
Comparison of the single dose pharmacokinetics of its colonic targeting. Am J Physiol Gastrointest Liver
sulphasalazine in rheumatoid arthritis and inflam- Physiol, 297(2):G3717. doi:10.1152/ajpgi.00102.2009
matory bowel disease. Ann Rheum Dis, 49(8):58790. PMID:19541926
doi:10.1136/ard.49.8.587 PMID:1975737 Dahan A, Amidon GL (2010). MRP2 mediated drug-drug
Augusto JF, Sayegh J, Simon A, Croue A, Chennebault JM, interaction: indomethacin increases sulfasalazine
Cousin M et al. (2009). A case of sulphasalazine-in- absorption in the small intestine, potentially decreasing
duced DRESS syndrome with delayed acute interstitial its colonic targeting. Int J Pharm, 386(12):21620.
doi:10.1016/j.ijpharm.2009.11.021 PMID:19944137
244
Sulfasalazine
Das KM, Dubin R (1976). Clinical pharmacokinetics Fatta D, Achilleos A, Nikolaou A, Meri S (2007).
of sulphasalazine. Clin Pharmacokinet, 1(6):40625. Analytical methods for tracing pharmaceutical resi-
doi:10.2165/00003088-197601060-00002 PMID:15752 dues in water and wastewater. Trends Analyt Chem,
Das KM, Eastwood MA, McManus JPA, Sircus W (1973). 26(6):51533. doi:10.1016/j.trac.2007.02.001
Adverse reactions during salicylazosulfapyridine FDA (2013). Sulfasalazine. Silver Spring (MD): US Food
therapy and the relation with drug metabolism and and Drug Administration, Drugs@FDA. Available
acetylator phenotype. N Engl J Med, 289(10):4915. from: http://www.accessdata.fda.gov/scripts/cder/
doi:10.1056/NEJM197309062891001 PMID:4146729 drugsatfda/index.cfm, accessed 30 June 2014.
Davis AE, Patterson F, Crouch R (1992). The effect of Font G, Juan-Garca A, Pic Y (2007). Pressurized liquid
therapeutic drugs used in inflammatory bowel disease extraction combined with capillary electrophore-
on the incidence and growth of colonic cancer in the sis-mass spectrometry as an improved methodology
dimethylhydrazine rat model. Br J Cancer, 66(5):777 for the determination of sulfonamide residues in
80. doi:10.1038/bjc.1992.359 PMID:1358165 meat. J Chromatogr A, 1159(12):23341. doi:10.1016/j.
de Abajo FJ, Montero D, Madurga M, Garca Rodrguez chroma.2007.03.062 PMID:17433345
LA (2004). Acute and clinically relevant drug-induced Fukino K, Hashimoto M etal. (2007). Analysis between
liver injury: a population based case-control study. clinical efficacy of salazosulfapyridine and plasma
Br J Clin Pharmacol, 58(1):7180. doi:10.1111/j.1365- concentrations of sulfapyridine in patients with rheu-
2125.2004.02133.x PMID:15206996 matoid arthritis. Jpn. J. Ther. Drug Monit, 24:113121.
Diculescu M, Ciocrlan M, Ciocrlan M, Piigoi D, Gisbert JP, Gonzlez-Lama Y, Mat J (2007).
Becheanu G, Croitoru A et al. (2003). Folic acid and 5-Aminosalicylates and renal function in inflammatory
sulfasalazine for colorectal carcinoma chemopre- bowel disease: a systematic review. Inflamm Bowel Dis,
vention in patients with ulcerative colitis: the old 13(5):62938. doi:10.1002/ibd.20099 PMID:17243140
and new evidence. Rom J Gastroenterol, 12(4):2836. Gu G-Z, Xia H-M, Pang ZQ, Liu ZY, Jiang XG, Chen J
PMID:14726972 (2011). Determination of sulphasalazine and its main
Dyson JK, Rutter MD (2012). Colorectal cancer in inflam- metabolite sulphapyridine and 5-aminosalicylic acid in
matory bowel disease: what is the real magnitude of human plasma by liquid chromatography/tandem mass
the risk? World J Gastroenterol, 18(29):383948. spectrometry and its application to a pharmacokinetic
doi:10.3748/wjg.v18.i29.3839 PMID:22876036 study. J Chromatogr B Analyt Technol Biomed Life Sci,
Eaden J, Abrams K, Ekbom A, Jackson E, Mayberry J 879(56):44956. doi:10.1016/j.jchromb.2010.12.034
(2000). Colorectal cancer prevention in ulcerative PMID:21251889
colitis: a case-control study. Aliment Pharmacol Ther, Hanauer SB (2004). Review article: aminosalicylates
14(2):14553. doi:10.1046/j.1365-2036.2000.00698.x in inflammatory bowel disease. Aliment Pharmacol
PMID:10651654 Ther, 20(s4):Suppl 4: 605. doi:10.1111/j.1365-
Elmasry MS, Blagbrough IS, Rowan MG, Saleh HM, Kheir 2036.2004.02048.x PMID:15352896
AA, Rogers PJ (2011). Quantitative HPLC analysis of IARC (2012). Pharmaceuticals. Volume 100 A. A review of
mebeverine, mesalazine, sulphasalazine and dispers- human carcinogens. IARC Monogr Eval Carcinog Risks
ible aspirin stored in a Venalink monitored dosage Hum, 100:Pt A: 1401. PMID:23189749
system with co-prescribed medicines. J Pharm Biomed Iatropoulos MJ, Williams GM, Abdo KM, Kari FW, Hart
Anal, 54(4):64652. doi:10.1016/j.jpba.2010.10.002 RW (1997). Mechanistic studies on genotoxicity and
PMID:21106317 carcinogenicity of salicylazosulfapyridine an anti-in-
eMC; electronic Medicines Compendium (2013). flammatory medicine. Exp Toxicol Pathol, 49(12):15
Sulfasalazine. Available from: http://www.medicines. 28. doi:10.1016/S0940-2993(97)80052-8 PMID:9085070
org.uk, accessed 30 June 2014. IMS Health (2012a). National Disease and Therapeutic
Erskine IA, Mackay JM, Fox DP (1984). Monitoring Index (NDTI). Plymouth Meeting, Pennsylvania: IMS
patients on long-term drug therapy for genotoxic Health
effects. Basic Life Sci, 29:Pt B: 895905. PMID:6152153 IMS Health (2012b) Multinational Integrated Data
Esbjrner E, Jrnerot G, Wranne L (1987). Sulphasalazine Analysis (MIDAS). Plymouth Meeting, Pennsylvania:
and sulphapyridine serum levels in children to IMS Health
mothers treated with sulphasalazine during preg- IMS Health (2012c) National Prescription Audit Plus
nancy and lactation. Acta Paediatr Scand, 76(1):13742. (NPA). Plymouth Meeting, Pennsylvania: IMS Health
doi:10.1111/j.1651-2227.1987.tb10430.x PMID:2882643 Jacoby HI (2000). Gastrointestinal Agents. Kirk-Othmer
European Pharmacopoeia (2008). Sulfasalazine European Encyclopedia of Chemical Technology. John Wiley &
Pharmacopoeia Council of Europe, 2:25142516. Sons, Inc. Available from: http://onlinelibrary.wiley.
Available from: http://www.edqm.eu/en/edqm- com/doi/10.1002/0471238961.0701192010010315.a01.
homepage-628.html, accessed 1 July 2014. pub2/abstract;jsessionid=B9EC963BD4DA13B08CE09
E6F28B43EA7.f03t04, accessed 30 June 2014.
245
IARC MONOGRAPHS 108
Jansen G, van der Heijden J, Oerlemans R, Lems WF, Br J Pharmacol, 166(6):1793803. doi:10.1111/j.1476-
Ifergan I, Scheper RJ et al. (2004). Sulfasalazine is a 5381.2012.01887.x PMID:22300367
potent inhibitor of the reduced folate carrier: implica- Lakatos PL, Lakatos L (2008). Risk for colorectal cancer
tions for combination therapies with methotrexate in in ulcerative colitis: changes, causes and management
rheumatoid arthritis. Arthritis Rheum, 50(7):21309. strategies. World J Gastroenterol, 14(25):393747.
doi:10.1002/art.20375 PMID:15248210 doi:10.3748/wjg.14.3937 PMID:18609676
Jrnerot G, Into-Malmberg MB (1979). Sulphasalazine Lashner BA, Heidenreich PA, Su GL, Kane SV, Hanauer
treatment during breast feeding. Scand J Gastroenterol, SB (1989). Effect of folate supplementation on the
14(7):86971. doi:10.3109/00365527909181418 incidence of dysplasia and cancer in chronic ulcer-
PMID:44005 ative colitis. A case-control study. Gastroenterology,
Jrnerot G, Into-Malmberg MB, Esbjrner E (1981). 97(2):2559. PMID:2568304
Placental transfer of sulphasalazine and sulphapyri- Lee HJ, Zhang H, Orlovich DA, Fawcett JP (2012). The
dine and some of its metabolites. Scand J Gastroenterol, influence of probiotic treatment on sulfasalazine
16(5):6937. doi:10.3109/00365528109182032 metabolism in rat. Xenobiotica, 42(8):7917. doi:10.310
PMID:6119765 9/00498254.2012.660508 PMID:22348441
Jess T, Loftus EV Jr, Velayos FS, Winther KV, Tremaine WJ, Lide DR, editor (2005). CRC Handbook of Chemistry and
Zinsmeister AR etal. (2007). Risk factors for colorectal Physics. 86th ed. Boca Raton (FL), USA: CRC Press.
neoplasia in inflammatory bowel disease: a nested Liedorp M, Voskuyl AE, Van Oosten BW (2008). Axonal
case-control study from Copenhagen county, Denmark neuropathy with prolonged sulphasalazine use. Clin
and Olmsted county, Minnesota. Am J Gastroenterol, Exp Rheumatol, 26(4):6712. PMID:18799104
102(4):82936. doi:10.1111/j.1572-0241.2007.01070.x Linares V, Alonso V, Albina ML, Bells M, Sirvent JJ,
PMID:17222314 Domingo JL etal. (2009). Lipid peroxidation and anti-
Jobanputra P, Amarasena R, Maggs F, Homer D, Bowman oxidant status in kidney and liver of rats treated with
S, Rankin E etal. (2008). Hepatotoxicity associated with sulfasalazine. Toxicology, 256(3):1526. doi:10.1016/j.
sulfasalazine in inflammatory arthritis: A case series tox.2008.11.010 PMID:19071188
from a local surveillance of serious adverse events. Lindberg BU, Broom U, Persson B (2001). Proximal
BMC Musculoskelet Disord, 9(1):48 doi:10.1186/1471- colorectal dysplasia or cancer in ulcerative colitis. The
2474-9-48 PMID:18405372 impact of primary sclerosing cholangitis and sulfasala-
Kater F (1974). Hemolysis, formation of Heinz bodies and zine: results from a 20-year surveillance study. Dis
of methemoglobin under therapy with Azulfidine. Med Colon Rectum, 44(1):7785. doi:10.1007/BF02234825
Klin, 69(11):4668. [German] PMID:4151648 PMID:11805567
Kaufman DW, Kelly JP, Jurgelon JM, Anderson T, Ma JJ, Liu CG, Li JH, Cao XM, Sun SL, Yao X (2009). Effects
Issaragrisil S, Wiholm BE et al. (1996). Drugs in the of NAT2 polymorphism on SASP pharmacokinetics in
aetiology of agranulocytosis and aplastic anaemia. Eur Chinese population. Clin Chim Acta, 407(12):305.
J Haematol Suppl, 60:suppl: 2330. PMID:8987237 doi:10.1016/j.cca.2009.06.025 PMID:19560446
Kita T, Sakaeda T, Adachi S, Sakai T, Aoyama N, Mackay JM, Fox DP, Brunt PW, Hawksworth GM, Brown
Hatanaka H etal. (2001). N-Acetyltransferase 2 geno- JE (1989). In vitro induction of chromosome damage
type correlates with sulfasalazine pharmacokinetics by sulphasalazine in human lymphocytes. Mutat
after multiple dosing in healthy Japanese subjects. Biol Res, 222(1):2736. doi:10.1016/0165-1218(89)90032-3
Pharm Bull, 24(10):117680. doi:10.1248/bpb.24.1176 PMID:2563144
PMID:11642327 McDonnell JP, Klaus F (1976). Sulfasalazine. Analytical
Kuhn UD, Anschtz M, Schmcker K, Schug BS, Hippius Profiles of Drug Substances, 5:51532. doi:10.1016/
M, Blume HH (2010). Phenotyping with sulfasala- S0099-5428(08)60329-9
zine - time dependence and relation to NAT2 phar- McGirt LY, Vasagar K, Gober LM, Saini SS, Beck LA
macogenetics. Int J Clin Pharmacol Ther, 48(1):110. (2006). Successful treatment of recalcitrant chronic
doi:10.5414/CPP48001 PMID:20040334 idiopathic urticaria with sulfasalazine. Arch Dermatol,
Kumagai S, Komada F, Kita T, Morinobu A, Ozaki S, Ishida 142(10):133742. doi:10.1001/archderm.142.10.1337
H etal. (2004). N-acetyltransferase 2 genotype-related PMID:17043190
efficacy of sulfasalazine in patients with rheuma- Mechanick JI (1985). Coombs positive hemolytic anemia
toid arthritis. Pharm Res, 21(2):3249. doi:10.1023/ following sulfasalazine therapy in ulcerative colitis:
B:PHAM.0000016246.84974.ec PMID:15032315 case reports, review, and discussion of pathogenesis.
Kusuhara H, Furuie H, Inano A, Sunagawa A, Yamada Mt Sinai J Med, 52(8):66770. PMID:2867466
S, Wu C et al. (2012). Pharmacokinetic interaction Meenan J (1993). Co-carcinogenic effect of sulphasalazine.
study of sulphasalazine in healthy subjects and the Br J Cancer, 68(5):10434. doi:10.1038/bjc.1993.477
impact of curcumin as an in vivo inhibitor of BCRP. PMID:8105863
246
Sulfasalazine
Meenan J, OHallinan E, Scott J, Weir DG (1996). Influence from: http://oehha.ca.gov/prop65/prop65_list/files/

of sulfasalazine and olsalamine on colonic epithelial P65single052413.pdf, accessed 30 June 2014.
cell folate content in patients with ulcerative colitis. Parry SD, Barbatzas C, Peel ET, Barton JR (2002).
Inflamm Bowel Dis, 2:1637. PMID:23282560 Sulphasalazine and lung toxicity. Eur Respir J,
Miller B, Engelhart A, Goebel KM, Schmidt J, Martini 19(4):75664. doi:10.1183/09031936.02.00267402
GA (1971). [Methemoglobin synthesis during azulfi- PMID:11999006
dine therapy]. Verh Dtsch Ges Inn Med, 77:132630. Pastor-Navarro N, Gallego-Iglesias E, Maquieira A,
PMID:4405361 Puchades R (2007). Immunochemical method for
Mitelman F, Strmbeck B, Ursing B (1980). No cytogenic sulfasalazine determination in human plasma. Anal
effect of metronidazole. Lancet, 1(8180):124950. Chim Acta, 583(2):37783. doi:10.1016/j.aca.2006.10.028
doi:10.1016/S0140-6736(80)91706-7 PMID:6104060 PMID:17386570
Mitelman F, Strmbeck B, Ursing B, Nordle O, Hartley-Asp Peppercorn MA, Goldman P (1972). The role of intestinal
B (1982). Metronidazole exhibits no clastogenic activity bacteria in the metabolism of salicylazosulfapyridine.
in a double-blind cross-over study on Crohns patients. J Pharmacol Exp Ther, 181(3):55562. PMID:4402374
Hereditas, 96(2):27986. doi:10.1111/j.1601-5223.1982. Pinczowski D, Ekbom A, Baron J, Yuen J, Adami HO (1994).
tb00859.x PMID:6125490 Risk factors for colorectal cancer in patients with ulcer-
Mohamed R, Hammel Y-A, LeBreton MH, Tabet JC, Jullien ative colitis: a case-control study. Gastroenterology,
L, Guy PA (2007). Evaluation of atmospheric pressure 107(1):11720. PMID:7912678
ionization interfaces for quantitative measurement of Pirmohamed M, Coleman MD, Hussain F, Breckenridge
sulfonamides in honey using isotope dilution liquid AM, Park BK (1991). Direct and metabolism-dependent
chromatography coupled with tandem mass spectrom- toxicity of sulphasalazine and its principal metabolites
etry techniques. J Chromatogr A, 1160(12):194205. towards human erythrocytes and leucocytes. Br J Clin
doi:10.1016/j.chroma.2007.05.071 PMID:17560585 Pharmacol, 32(3):30310. doi:10.1111/j.1365-2125.1991.
Moody GA, Jayanthi V, Probert CS, Mac Kay H, Mayberry JF tb03903.x PMID:1685664
(1996). Long-term therapy with sulphasalazine protects Porter RS, Kaplan JL (2013). The Merck manuals: Health
against colorectal cancer in ulcerative colitis: a retro- care professionals. Whitehouse Station (NJ): Merck &
spective study of colorectal cancer risk and compliance Co., Inc. Available from: http://www.merckmanuals.
with treatment in Leicestershire. Eur J Gastroenterol com/professional/index.html, accessed 30 June 2014.
Hepatol, 8(12):117983. doi:10.1097/00042737- Pounder RE, Craven ER, Henthorn JS, Bannatyne
199612000-00009 PMID:8980937 JM (1975). Red cell abnormalities associated with
NTP (1997a). NTP Toxicology and Carcinogenesis sulphasalazine maintenance therapy for ulcerative
Studies of Salicylazosulfapyridine (CAS No. 599791) colitis. Gut, 16(3):1815. doi:10.1136/gut.16.3.181
in F344/N Rats and B6C3F1 Mice (savage studies). PMID:235486
Natl Toxicol Program Tech Rep Ser, 457:1327. Ricart E, Taylor WR, Loftus EV, OKane D, Weinshilboum
PMID:12587019 RM, Tremaine WJ etal. (2002). N-acetyltransferase 1
NTP (1997b). Effect of dietary restriction on toxicology and 2 genotypes do not predict response or toxicity
and carcinogenesis studies in F344/N rats and B6C3F1 to treatment with mesalamine and sulfasalazine in
mice. Natl Toxicol Program Tech Rep Ser, 460:1414. patients with ulcerative colitis. Am J Gastroenterol,
PMID:12587016 97(7):17638. doi:10.1111/j.1572-0241.2002.05838.x
Novacek A et al. (1991). Method of producing PMID:12135032
2-hydroxy-5-[[4-[(2-pyridinylamino)sulfonyl]phenyl] Rong N, Selhub J, Goldin BR, Rosenberg IH (1991).
azo]benzoic acid [sulfasalazine Available from: http:// Bacterially synthesized folate in rat large intestine is
www.jazdlifesciences.com/pharmatech/company/ incorporated into host tissue folyl polyglutamates. J
United-States-Biological/Sulfasalazine-2-Hydroxy- Nutr, 121(12):19559. PMID:1941259
5-4-2-pyridinylaminosulfonylphenylazobenzoic- Rosenbaum SE (2011). Basic pharmacokinetics and phar-
AcidSulfasalazine-Azulfidine-Colo-Pleon-Salaz.htm?s macodynamics: An integrated textbook and computer
upplierId=30001720&productId=540654 simulations. Available from: http://www.theannals.
OMorin C, Smethurst P, Dor CJ, Levi AJ (1984). com/content/46/2/305.full, accessed 5 September 2014.
Reversible male infertility due to sulphasalazine: studies Rubin DT, LoSavio A, Yadron N, Huo D, Hanauer SB
in man and rat. Gut, 25(10):107884. doi:10.1136/ (2006). Aminosalicylate therapy in the prevention of
gut.25.10.1078 PMID:6148293 dysplasia and colorectal cancer in ulcerative colitis. Clin
ONeil MJ (2001). The Merck Index. 13th ed. Whitehouse Gastroenterol Hepatol, 4(11):134650. doi:10.1016/j.
Station (NJ): Merck & Co., Inc. cgh.2006.08.014 PMID:17059900
OEHHA; Office of Environmental Health Hazard Rutter M, Saunders B, Wilkinson K, Rumbles S, Schofield
Assessment (2013). Chemicals known to the state G, Kamm M et al. (2004). Severity of inflammation
to cause cancer or reproductive toxicity. Available is a risk factor for colorectal neoplasia in ulcerative
247
IARC MONOGRAPHS 108
colitis. Gastroenterology, 126(2):4519. doi:10.1053/j. in Japanese volunteers. Biol Pharm Bull, 25(2):2647.
gastro.2003.11.010 PMID:14762782 doi:10.1248/bpb.25.264 PMID:11853180
Schrder H, Campbell DE (1972). Absorption, metabo- Tomaru A, Morimoto N, Morishita M, Takayama K,
lism, and excretion of salicylazosulfapyridine in man. Fujita T, Maeda K etal. (2013). Studies on the intestinal
Clin Pharmacol Ther, 13(4):53951. PMID:4402886 absorption characteristics of sulfasalazine, a breast
Schrder H, Evans DA (1972). Acetylator phenotype cancer resistance protein (BCRP) substrate. Drug
and adverse effects of sulphasalazine in healthy Metab Pharmacokinet, 28(1):714. doi:10.2133/dmpk.
subjects. Gut, 13(4):27884. doi:10.1136/gut.13.4.278 DMPK-12-NT-024 PMID:22785334
PMID:4402420 Tuckwell R, Revitt DM, Garelick H, Jones H, Ellis B
SciFinder (2013). SciFinder databases: A division of the (2011). Sources and pathways for pharmaceuticals in
American Chemical Society. Available from: https:// the urban water environment. In: 12th International
www.cas.org/products/scifinder, accessed 30 June 2014. Conference on Urban Drainage, 1116 September 2011,
Selhub J, Dhar GJ, Rosenberg IH (1978). Inhibition of folate Porto Alegre, Brazil.
enzymes by sulfasalazine. J Clin Invest, 61(1):2214. Urquhart BL, Gregor JC, Chande N, Knauer MJ, Tirona
doi:10.1172/JCI108921 PMID:22555 RG, Kim RB (2010). The human proton-coupled
Soejima M, Kawaguchi Y, Hara M, Kamatani N (2008). folate transporter (hPCFT): modulation of intes-
Prospective study of the association between NAT2 tinal expression and function by drugs. Am J Physiol
gene haplotypes and severe adverse events with Gastrointest Liver Physiol, 298(2):G24854. doi:10.1152/
sulfasalazine therapy in patients with rheumatoid ajpgi.00224.2009 PMID:19762432
arthritis. J Rheumatol, 35(4):724 PMID:18398952 Urquhart BL, Ware JA, Tirona RG, Ho RH, Leake BF,
Stretch GL, Campbell BJ, Dwarakanath AD, Yaqoob M, Schwarz UI et al. (2008). Breast cancer resistance
Stevenson A, Morris AI et al. (1996). 5-amino sali- protein (ABCG2) and drug disposition: intestinal
cylic acid absorption and metabolism in ulcerative expression, polymorphisms and sulfasalazine as an in
colitis patients receiving maintenance sulphasalazine, vivo probe. Pharmacogenet Genomics, 18(5):43948.
olsalazine or mesalazine. Aliment Pharmacol Ther, doi:10.1097/FPC.0b013e3282f974dc PMID:18408567
10(6):9417. doi:10.1046/j.1365-2036.1996.85257000.x US Pharmacopeia (2007). Sulfasalazine. Report No.
PMID:8971292 USP30NF25. Rockville (MD), USA: The United States
Taggart AJ, McDermott BJ, Roberts SD (1992). The effect of Pharmacopoieal Convention.
age and acetylator phenotype on the pharmacokinetics of US Pharmacopeia (2013). Sulfasalazine. Report No.
sulfasalazine in patients with rheumatoid arthritis. Clin USP36. Rockville (MD), USA: The United States
Pharmacokinet, 23(4):31120. doi:10.2165/00003088- Pharmacopoieal Convention.
199223040-00006 PMID:1356683 van der Heijden J, de Jong MC, Dijkmans BA, Lems WF,
Tanaka E, Taniguchi A, Urano W, Nakajima H, Matsuda Oerlemans R, Kathmann I etal. (2004). Development
Y, Kitamura Y etal. (2002). Adverse effects of sulfasala- of sulfasalazine resistance in human T cells induces
zine in patients with rheumatoid arthritis are associated expression of the multidrug resistance transporter
with diplotype configuration at the N-acetyltransferase ABCG2 (BCRP) and augmented production of
2 gene. J Rheumatol, 29(12):24929. PMID:12465141 TNFalpha. Ann Rheum Dis, 63(2):13843. doi:10.1136/
Tanigawara Y, Kita T, Aoyama N, Gobara M, Komada ard.2002.005249 PMID:14722201
F, Sakai T et al. (2002). N-Acetyltransferase 2 geno- van Staa TP, Card T, Logan RF, Leufkens HG (2005).
type-related sulfapyridine acetylation and its adverse 5-Aminosalicylate use and colorectal cancer risk in
events. Biol Pharm Bull, 25(8):105862. doi:10.1248/ inflammatory bowel disease: a large epidemiological
bpb.25.1058 PMID:12186410 study. Gut, 54(11):15738. doi:10.1136/gut.2005.070896
Terdiman JP, Steinbuch M, Blumentals WA, Ullman PMID:15994215
TA, Rubin DT (2007). 5-Aminosalicylic acid therapy Voogd CE, Van der Stel JJ, Jacobs JJ (1980). On the muta-
and the risk of colorectal cancer among patients with genic action of some enzyme immunoassay substrates.
inflammatory bowel disease. Inflamm Bowel Dis, J Immunol Methods, 36(1):5561. doi:10.1016/0022-
13(4):36771. doi:10.1002/ibd.20074 PMID:17206695 1759(80)90093-9 PMID:7009752
Tett SE (1993). Clinical pharmacokinetics of slow-acting WHO (2007). Cumulative List No. 14 of INN in Latin,
antirheumatic drugs. Clin Pharmacokinet, 25(5):392 English, French, Spanish, Arabic, Chinese and Russian
407. doi:10.2165/00003088-199325050-00005 (in alphabetical order of the Latin name), with supple-
PMID:7904547 mentary information. Geneva, Switzerland: World
Tokui K, Asai Y, Arakawa T, Matsumoto T, Nabeshima Health Organization. Available from: http://www.vpb.
T (2002). Comparative absorption of 5-aminosalicylic gov.lt/files/539.pdf, accessed 30 June 2014.
acid (5-ASA) after administration of a 5-ASA enema and Witt KL, Bishop JB, McFee AF, Kumaroo V (1992b).
salazosulfapyridine (SASP) after an SASP suppository Induction of chromosomal damage in mammalian cells
in vitro and in vivo by sulfapyridine or 5-aminosalicylic
248
Sulfasalazine
acid. Mutat Res, 283(1):5964. doi:10.1016/0165-

7992(92)90122-X PMID:1380664
Witt KL, Gudi R, Bishop JB (1992a). Induction of kineto-
chore positive and negative micronuclei in mouse bone
marrow cells by salicylazosulfapyridine and sulfapyri-
dine. Mutat Res, 283(1):537. doi:10.1016/0165-
7992(92)90121-W PMID:1380663
World MJ, Stevens PE, Ashton MA, Rainford DJ
(1996). Mesalazine-associated interstitial nephritis.
Nephrol Dial Transplant, 11(4):61421. doi:10.1093/
oxfordjournals.ndt.a027349 PMID:8671848
Yamasaki Y, Ieiri I, Kusuhara H, Sasaki T, Kimura M,
Tabuchi H et al. (2008). Pharmacogenetic character-
ization of sulfasalazine disposition based on NAT2
and ABCG2 (BCRP) gene polymorphisms in humans.
Clin Pharmacol Ther, 84(1):95103. doi:10.1038/
sj.clpt.6100459 PMID:18167504
Zaher H, Khan AA, Palandra J, Brayman TG, Yu L, Ware
JA (2006). Breast cancer resistance protein (Bcrp/abcg2)
is a major determinant of sulfasalazine absorption and
elimination in the mouse. Mol Pharm, 3(1):5561.
doi:10.1021/mp050113v PMID:16686369
Zeiger E, Anderson B, Haworth S, Lawlor T, Mortelmans K
(1988). Salmonella mutagenicity tests: IV. Results from
the testing of 300 chemicals. Environ Mol Mutagen,
11(S12):Suppl 12: 1157. doi:10.1002/em.2850110602
PMID:3277844
Zheng W, Winter SM, Mayersohn M, Bishop JB, Sipes IG
(1993). Toxicokinetics of sulfasalazine (salicylazosul-
fapyridine) and its metabolites in B6C3F1 mice. Drug
Metab Dispos, 21(6):10917. PMID:7905389
249
PENTOSAN POLYSULFATE SODIUM

In this Monograph, pentosan polysulfate
sodium will also be referred to as pentosan.
O
* O *
1.1 Chemical and physical data O
O
1.1.1 Nomenclature O O S
O
S ONa
Chem. Abstr. Serv. Reg. No.: 37319-17-8;
N aO O
140207-93-8
n
Chem. Abstr. Serv. Name: Xylan hydrogen
sulfate, sodium salt (SciFinder, 2010)
(C5H6Na2O10S2)n where n = 612 (US
IUPAC systematic name: [(2R,3R,4S,5R)-2-
Pharmacopeial Convention, 2013)
hydroxy-5-[(2S,3R,4S,5R)-5-hydroxy-3,4-di
Relative molecular mass: [336]n
sulfooxyoxan-2-yl]oxy-3-sulfooxyoxan-4-yl]
hydrogen sulfate, sodium (PubChem, 2013)
Synonyms: Pentosan polysulfate sodium;
pure substance
Pentosan; Xylan hydrogen sulfate sodium salt;
Xylan polysulfate sodium; Sodium pentosan Description: White, odourless powder,
polysulfate; Sodium pentosane polysulfate; slightly hygroscopic (ONeil, 2006)
Xylan sulfate sodium; Xylofuranan sulfate
sodium; (14)--D-Xylan 2,3-bis(hydrogen Density: 1.344 at 20 C, 10% aqueous solution
sulfate) sodium; Sodium xylan polysulfate (ONeil, 2006)
(ONeil, 2006; SciFinder,2010) Spectroscopy data: Index of refraction is
Proprietary names: Elmiron, Cartrophen, 57at 20 C (ONeil, 2006)
Fibrase, Fibrenzym, Hemoclar, SP-54, Solubility: Soluble in water to 50% at pH 6.0
Thrombocid (ONeil, 2006; RxList, 2013)
251
IARC MONOGRAPHS 108
1.1.4 Technical products and impurities to the other, which is irrigation of the bladder
with dimethyl sulfoxide (Hanno et al., 2011).
Pentosan polysulfate sodium is a plant- With its large, heparin-like molecular structure,
derived, semi-synthetic mucopolysaccharide pentosan has anticoagulant and fibrinolytic
with radiochemical purity of 95.6%, with no properties (MicroMedex, 2013), although this
single major impurity (Simon et al., 2005; RxList, use was not observed in recent data from the
2013). USA. Pentosan was approved by the Food and
Drug Administration for the indication of relief
1.2 Analysis of bladder pain or discomfort associated with
interstitial cystitis (FDA, 2013). In the European
Several non-compendial analytical methods Union, other formulations apart from the oral
for the determination of pentosan polysul- form include ointment, rectal suppository, and
fate sodium in pharmaceutical formulations injectable solution. Pentosan is also used in veter-
were available. These included normal-phase inary medicine as an anti-inflammatory drug to
high-performance liquid chromatography with treat arthritis (MicroMedex, 2013).
refractive index detection, capillary electropho-
resis with ultraviolet detection, surface plasmon (b) Dosage
resonance, and gel-permeation chromatography Pentosan polysulfate sodium can be admin-
with refractive index detection. The analyt- istered orally, intramuscularly, rectally, or as a
ical methods are summarized in Table 1.1. No solution instilled into the bladder. For the treat-
compendial analytical methods were available to ment of interstitial cystitis, recommended dosing
the Working Group. is an oral dose of 100 mg, three times per day,
for at least 3 months (IMS Health, 2012b). Use
1.3 Production for deep venous thrombosis prophylaxis involves
intramuscular injection of 50mg every 12hours
1.3.1 Production process (MicroMedex, 2013).
As a semi-synthetic compound, the poly-
saccharide backbone of pentosan polysulfate (c) Trends in use
sodium, xylan, is present in beech tree bark. Pentosan polysulfate sodium is not widely
Extracted xylan from beech bark or other plant used in the USA, with 138 000 drug uses
sources is treated with sulfating agents (e.g. reported in office-based physician visits in
chlorosulfonic acid or sulfuryl chloride). After 2012 (IMS Health, 2012b). Use had declined
sulfation, pentosan polysulfate is treated with by 33% since 2005 (Fig. 1.1). Based on NDTI
sodium hydroxide to form a sodium salt of the data, approximately 50000 patients in the USA
compound (Deshpande et al., 2010). were exposed to pentosan in 2012 (IMS Health,
2012b). About 450000 prescriptions for pentosan
1.3.2 Use were dispensed in the USA in 2012, down slightly
from about 490000 in 2008 (IMS Health, 2012c).
(a) Indications Despite an only modest volume of use, total
Orally administered pentosan polysulfate worldwide sales of pentosan were nonetheless
sodium is used primarily in the treatment of US$276 million in 2012, with 82% occurring in
interstitial cystitis. It is one of only two products the USA. The only other countries with appreci-
approved for treatment of this bladder condi- able use were Spain (US$16 million) and Canada
tion and is often used after inadequate response (US$11 million) (IMS Health, 2012a).
252
Table 1.1 Some non-compendial analytical methods for pentosan polysulfate sodium

Human plasma Immunoassay indirect ELISA 2.6 ng/mL (LLOQ) Abnova (2013)
Human serum Blood collection in tubes Amplified ELISIA using mAb 5-B-10 that recognizes 2,3-, 2,6-, 50 ng/mL (LOD) Kongtawelert &
containing oxalate, separation of and 4,6-disulfate ester ring substitution in pyranose-containing Ghosh (1990)
plasma by centrifugation polysaccharides
Human urine Urine sample, centrifugation, GC-FID 10 g/mL (LOD Lee et al. (1986)
incubation with equal volume Glass column
CPC citrate buffer, centrifugation, Detector: FID
dissolve precipitate in lithium Carrier gas: nitrogen, 30 mL/min
chloride, addition of ethanol
to mixture, centrifugation,
dry precipitate with nitrogen,
addition of 0.5 M HCl, hydrolysis,
dry sample in vacuum oven,
derivatization with TMS reagent
Rabbit serum Collection of blood from rabbit Antiviral bioassay method based on inhibitory activity of PPS on 0.5 g/mL (LOD) Witvrouw et al.
ear, separation of serum by HIV-2 virus in MT-4 cells (human T-lymphoblastoid cell line). (1990)
centrifugation Infection of MT-4 cells with HIV-1 or HIV-2 in culture medium,
transfer to microtitre tray wells containing serum samples, 5-day
incubation at 37C, number of viable cells by MTT
Nanoparticles Drop-wise addition of a solution CZE 0.1 mg/mL (LOQ) Abdel-Haq &
of standard PPS to solution of Buffer: Tris base in water (50200 mM) Boss (2012)
chitosan, magnetic stirring, pH 3.06.5
separation by centrifugation Pressure: 1035 mbar
Running voltage: 10 or 10 to 30 or 30 kV based on separation
mode (normal or reversed)
Injection time: 510 s
Formulation HPLC Muller et al.
Column: silica diol (1984)
Mobile phase: sodium chloride (200 mM)
Flow rate: 1 mL/min
Detector: refractive index
Formulation Dilution of 10 mg of PPS with 10 CZE using indirect detection Degenhardt
mL of purified water Fused silica capillary et al. (1998)
Detector: UV detector (217 nm)
Applied voltage: 20 kV (capillary inlet at cathode)
Current: 14.4 A
Running buffer: BTC buffer, 8.75 mmol/L
253
Pentosan polysulfate sodium
254

Formulation Dilution of 10 mg PPS with 10 mL Reverse polarity CZE using a central composite design 0.25 mg/mL (LOQ) Prochazka et al.
purified water CE-UV diode-array detector (2003)
Polyimide coated fused silica capillary
Detector: 320 nm (reference wavelength: 217 nm)
Running buffer: BTC buffer, 8.75 mmol/L
Formulation SPR technology, biosensor analysis with immobilized enzymes 0.3 g/mL (LOQ) Shen et al.
HNE, HAase or lysozyme (2003)
IARC MONOGRAPHS 108
Immobilization of enzymes: amine coupling

Binding assays in HEPES-containing buffer
Flow rate: 50 L/min
Regeneration: 100 mM acetic acid containing 2.0M sodium
chloride
Formulation Dilution in deionized water HPLC methods NTP (2004)
(1) Column: diol GPC
Mobile phase: 25 mM potassium phosphate, monobasic; 25mM
potassium phosphate, dibasic; 50mM potassium chloride
Mobile phase: acetonitrile in water (5:95); 25 mM potassium
phosphate, monobasic; 25 mM potassium phosphate, dibasic;
50mM potassium chloride
Mobile phase: 0.9% sodium chloride in water
BTC, benzene-1,2,4-tricarboxylic acid; CE, capillary electrophoresis; CPC, cetyl pyridinium chloride; CZE, capillary zone electrophoresis; ELISA, enzyme-linked immunosorbent
assay; ELISIA, enzyme-linked immunosorbent inhibition assay; FID, flame ionization detector; GC, gas chromatography; GPC, gel permeation chromatography; HAase,
hyaluronidase; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; HIV, human immunodeficiency virus; HNE, human neutrophil elastase; HPLC, high-performance liquid
chromatography; LOD, limit of detection; LLOQ, lower limit of quantitation; LOQ, limit of quantitation; MAb, monoclonal antibody; min, minute; MTT, 3-(4,5-dimethylthiazol-2-
yl)-2,5 diphenyltetrazolium bromide; PBS, phosphate buffered saline; PPS, pentosan polysulfate sodium; SPR, surface plasmon resonance; TMB, 3,3,5,5-tetramethylbenzidine; TMS,
trimethylsilyl; UV, ultraviolet
Fig.1.1 Reported use of pentosan by office-based physicians in the USA

250
200
Annual drug uses (thousands)
150
100
50
0
2005 2006 2007 2008 2009 2010 2011 2012
Year
Prepared by the Working Group based on data obtained from IMS Health (2012b)
1.4 Occurrence and exposure 2. Cancer in Humans

Pentosan polysulfate sodium does not occur
in nature. Human exposure is largely limited
to use as a medication. While occupational
exposure in manufacturing is likely to occur,
no specific studies on occupational or environ- 3. Cancer in Experimental Animals
mental exposure to pentosan were identified by
the Working Group. 3.1 Mouse
See Table3.1
1.5 Regulations and guidelines In one study of oral administration, groups
of 50 male and 50 female B6C3F1 mice (age, 6
Pentosan has been approved by drug regula- weeks) were given pentosan polysulfate sodium
tory agencies primarily in the European Union (pharmaceutical grade) at doses of 0 (control), 56,
and USA. In the USA, it was approved by the 168, or 504mg/kg body weight (bw) in deionized
Food and Drug Administration in 1996 (FDA, water by gavage once per day on 5days per week
2013). The Working Group did not identify
for 104 to 105 weeks. A significant decrease in
extraordinary regulatory restrictions on the use body weight was seen in females at the highest
of pentosan as a medication, or regulations on dose, but not in males or any other groups of
environmental exposure. females. Survival of all dosed groups was similar
to that of controls. Liver haemangiosarcomas
255
IARC MONOGRAPHS 108
Table 3.1 Studies of carcinogenicity in mice given pentosan polysulfate sodium by gavage
Strain (sex) Dosing regimen, Incidence,(%) and/or Significance Comments

Duration Animals/group at multiplicity of tumours
Reference start
Abdo et al. Administered at Males *P<0.001 (Poly-3 trend test) Purity,
(2003), NTP doses of 0, 56, 168, Liver haemangiosarcoma: **P0.05 (Poly-3 test) pharmaceutical
(2004) or 504mg/kg bw in 2/50*, 0/50, 4/50, 9/50** ***P=0.031 (Poly-3 trend test) grade
B6C3F1 (M,F) deionized water, 5 Hepatocellular adenoma: ****P=0.003 (Poly-3 trend test)
104105wk days per week, for 19/50, 15/50, 15/50, 20/50 #P=0.010 (Poly-3 trend test)
104105wk Hepatocellular carcinoma: ##P=0.006 (Poly-3 trend test)
50 M and 50 F/group 11/50, 13/50, 15/50, 13/50

Hepatocellular adenoma or
carcinoma (combined):
23/50***, 23/50, 26/50, 31/50**
Female
Liver haemangiosarcomaa:
1/50, 1/49, 1/50, 4/49
Hepatocellular adenoma:
7/50****, 5/49, 4/50, 15/49**
3/50, 3/50, 5/50, 3/50
Hepatocellular adenoma or
carcinoma (combined):
10/50#, 8/49, 9/50, 18/49
Malignant lymphoma:
7/50##, 8/50, 6/50, 16/50**
a Incidence in historical controls receiving NTP-2000 diet in 2-year studies: 24/959 (2.6%1.4%); range, 04%
occurred with a positive trend in males, with 3.2 Rat

the incidence at the highest dose significantly
increased compared with controls. The inci- See Table3.2
dence of liver haemangiosarcoma in females at In one study of oral administration, groups
the highest dose (4 out of 49; 8%) exceeded the of 50 male F344/N rats (age, 6 weeks) received
incidence of this tumour in historical controls pentosan polysulfate sodium (pharmaceutical
(24 out of 959, 2.6%; range, 04%). There was a grade) at oral doses of 0 (control), 14, 42, or
significant increase in the incidence of malignant 126mg/kg bw, and groups of 50 female F344/N
lymphoma in females at the highest dose. rats (age, 6 weeks) received pentosan polysul-
The incidences of hepatocellular adenoma or fate sodium at oral doses of 0 (control), 28, 84,
carcinoma (combined) increased with positive or 252mg/kg bw by gavage in deionized water,
trends in males and in females. The incidences of once per day, 5days per week, for 104105weeks.
hepatocellular adenoma in females at the highest There was no effect on body weight or survival
dose, and of hepatocellular adenoma or carci- in any of the groups of treated rats during the
noma (combined) in males at the highest dose study. There were no significant increases in the
were also significantly increased. The incidences incidence of any neoplasm in treated rats (Abdo
of hepatocellular adenoma in treated males, and et al., 2003; NTP, 2004).
of hepatocellular carcinoma in treated males or
treated females, were not significantly increased
(Abdo et al., 2003; NTP, 2004).
256
Table 3.2 Studies of carcinogenicity in rats given pentosan polysulfate sodium by gavage
Strain (sex) Dosing regimen, Incidence, Significance Comments

Duration Animals/group at start (%) and/or
Reference multiplicity
of tumours
Abdo et al. (2003), Male rats were given doses of 0, 14, 42, See Purity, pharmaceutical grade
NTP (2004) or 126mg/kg bw, and female rats were comments There were no significant
F344/N (M,F) given doses of 0, 28, 84 or 252mg/kg bw increases in the incidence of any
104105wk by gavage in deionized water, 5 days per neoplasm
week, for 104105wk
50M and 50F/group
4. Mechanistic and Other compounds administered intravenously. Orally

Relevant Data administered pentosan did not influence any of
the parameters compared with placebo.
MacGregor et al. (1984) studied the catabolism
4.1 Absorption, distribution, of pentosan polysulfate sodium. Five healthy male
metabolism, and excretion volunteers were given [125I]-labelled pentosan in
conjunction with unlabelled pentosan at a dose
4.1.1 Humans of 0.1, 1, 7, or 50 mg intravenously. The half-
Fellstrom et al. (1987) measured pentosan lives for doses of 0.17 mg ranged from 13 to
polysulfate sodium in plasma and in urine by 18minutes. At a dose of 50mg, the half-life was
radioassay after intravenous and oral adminis- 45 minutes. Tissue distribution studies showed
tration in a group of eight healthy volunteers. that most of the radiolabelled material was local-
After intravenous administration of 40 mg of ized in the liver and spleen. Pentosan was desul-
pentosan, plasma clearance was 49.96.6mL/ fated in the liver and spleen and depolymerized
minute, of which renal clearance constituted in the kidney, and it is likely that desulfation and
4.2 1.2 mL/minute. Only 8% of the intrave- depolymerization of pentosan is saturable.
nous dose was recovered in the urine, suggesting Simon et al. (2005) studied two groups of
that there was extensive metabolism. After daily eight healthy fasted female volunteers who
oral dosing with 400 mg of pentosan, steady- sequentially received a single oral dose of
state trough plasma concentrations were low 200Ci of [3H]-labelled pentosan supplemented
(2050ng/mL), and bioavailability was 0.51%. with 300mg of unlabelled pentosan, or 300Ci
The oral bioavailability of pentosan was of [3H]-labelled pentosan supplemented with
investigated in 18 healthy young male volun- 450mg of unlabelled pentosan. Most (84%) of the
teers who received pentosan as an intravenous administered oral dose was excreted in the faeces
dose of 50mg, or an oral dose of 1500mg, or a as intact pentosan, and a smaller percentage (6%)
placebo (Faaij et al., 1999). Intravenously admin- was excreted in the urine as pentosan of low rela-
istered pentosan significantly increased activated tive molecular mass and desulfated pentosan.
partial thromboplastin time and the activity of Excretion of pentosan was studied in 34
anti-factor Xa, hepatic triglyceride lipase, and female patients with interstitial cystitis who were
lipoprotein lipase compared with placebo in a receiving long-term treatment with pentosan
magnitude comparable to other heparin-like (Erickson et al., 2006). The median concentra-
tion of pentosan in the urine of these patients
257
IARC MONOGRAPHS 108
was 1.2g/mL (range, 0.527.7 g/mL). All the after intravenous administration, with notable
pentosan recovered from the urine of these labelling of connective tissues, and low activity
patients was of low relative molecular mass. in bone and cartilage. There was a high concen-
tration of radiolabel in the urine, and prefer-
4.1.2 Experimental systems ential localization of radiolabel to the lining of
the urinary tract. After oral administration, the
In a pharmacokinetic study of pentosan in tissue distribution of radiolabel was similar, but
New Zealand rabbits, [125I]-labelled pentosan activity was lower (Odlind et al., 1987).
as marker was injected simultaneously with
increasing doses of unlabelled pentosan (Cadroy
et al., 1987). The data indicated that prolonga- 4.2 Genetic and related effects
tion of the half-life of pentosan with increasing
doses resulted from progressive reduction in the 4.2.1 Humans
clearance of the drug, with a constant volume of No data were available to the Working Group.
distribution.
Some studies of distribution were oriented 4.2.2 Experimental systems
towards the pharmacological application of
pentosan in the treatment of interstitial cystitis. (a) Mutagenicity
Kyker et al. (2005) used fluorescently labelled Pentosan polysulfate sodium was not muta-
chondroitin sulfate to track the distribution of genic in Salmonella typhimurium strains TA97,
glycosaminoglycans administered intravesically TA98, TA100, or TA1535, with or without meta-
to C57BL/6NHsd mouse bladder that had been bolic activation, at a concentration range of 100
damaged on the surface. Bladder damaged by to 10000g/plate (NTP, 2004).
trypsin or hydrochloric acid bound the labelled
chondroitin sulfate extensively on the surface, (b) Chromosomal damage
with little penetration into the bladder muscle. No consistent increase in the frequency of
In rabbits given 11.2mg of pentosan by intra- micronucleated polychromatic erythrocytes was
venous administration, median recovery in the seen in bone-marrow cells of male F344/N rats
urine was 47.2% (range, 19.773.2%) for unfrac- or male B6C3F1 mice given pentosan at doses of
tionated pentosan, 74.6% (range, 31.496.3%) 156.252500 mg/kg bw by gavage, three times
for pentosan of low relative molecular mass, at 24-hour intervals. An initial trial had yielded
and 3.3% (range, 2.55.0%) for pentosan of a weakly positive result (P for trend=0.019) in
high relative molecular mass. In rabbits given male rats, but a second trial gave clearly nega-
1.01.2 mg pentosan by oral administration, tive results (NTP, 2004). A subsequent study in
median recovery in the urine was 7.45% (range, male and female B6C3F1 mice given pentosan as
2.146.0%) for pentosan of low relative molecular a daily dose at 63, 125, 250, 500, or 1000mg/kg
mass, and 0.1% (range, 0.00.3%) for pentosan bw by gavage for 3 months also gave negative
of high relative molecular mass (Erickson et al., results. There were no significant differences in
2006). the percentages of polychromatic erythrocytes in
Sprague-Dawley rats were given [3H]-labelled the circulating blood of mice receiving pentosan
pentosan orally or intravenously at a dose of (NTP, 2004).
5mg/kg bw, and killed 1 or 4hours later, respec-
tively. Autoradiography indicated extensive
distribution of radiolabel in the whole animal
258
4.3 Other mechanistic data relevant 4.3.2 Effects on cell proliferation

to carcinogenicity Elliot et al. (2003) observed that pentosan has
marked effects on the growth and extracellular
4.3.1 Effects on cellular physiology matrix of smooth-muscle cells cultured from
Pentosan polysulfate sodium can antagonize human prostate. Pentosan decreased cell prolif-
the binding of fibroblast growth factor-2 (FGF-2) eration and extracellular-matrix production.
to its cell surface receptors, and has been shown This suggested that the drug may have thera-
to modulate the angiogenic activity of FGF-2 in peutic potential in relation to benign prostatic
tumours in humans and mice. Jerebtsova et al. hyperplasia.
(2007) studied the role of FGF-2 and pentosan in The results of treatment of three pros-
the pathogenesis of intestinal bleeding in mice. tate-cancer cell lines (LnCaP, PC3, and DU145)
The results indicated that high steady-state levels with pentosan have been reported (Zaslau et al.,
of circulating FGF-2, plus anticoagulant activity, 2006). In LnCaP cells, there was a mean inhibition
are needed to induce lethal intestinal bleeding in of growth of 12%7% at 24hours (P=0.025),
mice. and 20%15% at 72hours (P<0.001). Similar
Treatment with pentosan prevents the inhibition was observed in the other two cell
progression of nephropathy in streptozoto- lines.
cin-induced diabetes in ageing C57B6 mice Rha et al. (1997) reported that growth of
by decreasing albuminuria, renal macrophage gastric-cancer cell lines expressing midkine, a
infiltration, and expression of tumour necrosis novel heparin-binding growth/differentiation
factor- (Wu et al., 2011). factor, was inhibited by pentosan, which was
Zugmaier et al. (1992) concluded that described as a heparin-binding blocking agent.
pentosan is an in-vitro inhibitor of a variety of Zaslau et al. (2004) reported that pentosan
heparin-binding growth factors released from significantly inhibited the growth of ZR75-1
tumour cells. Of seven tumour cell lines tested, breast-cancer cells; however, a significant increase
six (breast cancer: MDA-MB-231, MDA-MB- in cell proliferation (25% 2%; P < 0.001) was
435, MDA-MB-468; lung cancer: A-549; prostate observed in estrogen-independent MCF-7 breast-
cancer: DU-145; and epidermoid carcinoma: cancer cells.
A-431) were resistant to pentosan in soft-agar The effects of pentosan on tumour growth,
cloning assays, and did not appear to depend on hyperprolactinaemia and angiogenesis in
autocrine stimulation by the heparin-binding diethylstilbestrol-induced anterior pituitary
growth factors. In contrast to this resistance in adenoma in F344 rats was described by Mucha
vitro, subcutaneous growth of tumours from all et al. (2002). Long-term treatment with pentosan
cell lines in athymic nude mice was inhibited in a did not cause any changes in pituitary weight,
dose-dependent fashion by daily intraperitoneal serum prolactin concentration, or density of
injections of pentosan. microvessels. However, there was an increase in
Pentosan inhibits virus adsorption to cells in the number of apoptotic bodies within the ante-
vitro as demonstrated by monitoring the associ- rior pituitary.
ation of radiolabelled HIV-1 virions with MT-4 The mechanism of cell motility inhibition by
cells (Baba et al., 1988). pentosan appears to be independent of cytoskel-
etal structural alterations, including changes
in microfilament and microtubule networks
(Pienta et al., 1992). In vitro, pentosan altered
259
IARC MONOGRAPHS 108
cellular contacts with the extravascular matrix In mice, pentosan caused a significant
and inhibited cell motility. In vivo, pentosan increase in the incidence of liver haemangio-
prolonged survival of male rats injected with sarcoma in males at the highest dose, and an
highly metastatic cells. increase in the incidence of liver haemangio-
sarcoma that occurred with a positive trend in
males. The incidence of liver haemangiosar-
4.4 Susceptibility coma in females at the highest dose exceeded
No data were available to the Working Group. the incidence in historical controls. It caused a
significant increase in the incidence of malig-
nant lymphoma in females at the highest dose.
4.5 Mechanistic considerations Exposure to pentosan increased the trend in the
Most of the experimental studies on pentosan incidences of hepatocellular adenoma or carci-
polysulfate sodium were not directed towards noma (combined) in males and females and also
elucidating a possible mechanism of carcinogen- caused a significant increase in the incidence of
esis. No mechanism of carcinogenesis was indi- hepatocellular adenoma in females and hepato-
cated by the collective findings. cellular adenoma or carcinoma (combined) in
males at the highest dose.
In treated rats, there were no significant
5. Summary of Data Reported increases in the incidence of any neoplasm.
5.1 Exposure data 5.4 Mechanistic and other relevant

data
Pentosan polysulfate sodium is a drug of
high relative molecular mass that is obtained by In humans, pentosan polysulfate sodium is
chemically treating the bark of the beech tree. It desulfated in the liver and spleen and depolym-
is used in oral form to treat bladder conditions erized in the kidney. Intravenous administration
(interstitial cystitis) and in injectable form for of radiolabelled pentosan to rats indicated exten-
the prevention of blood clots. A large proportion sive distribution of radiolabel, particularly in
of the global use of pentosan occurs in the USA connective tissues, a high concentration of radi-
(global sales in 2012, US$276 million, with 82% olabel in the urine, and a preferential localization
occurring in the USA), where prescriptions have of radiolabel to the lining of the urinary tract.
been declining over the past years. Pentosan was not mutagenic when tested
in Salmonella typhimurium, with or without
metabolic activation. Likewise, no evidence of
5.2 Human carcinogenicity data chromosomal damage associated with exposure
No data were available to the Working Group. to pentosan was obtained in studies in rodents.
In vitro, pentosan is an inhibitor of a variety
of heparin-binding growth factors released from
5.3 Animal carcinogenicity data tumour cells.
Pentosan polysulfate sodium was tested for The data did not support any genotoxic mech-
carcinogenicity in one study in male and female anism of carcinogenesis by pentosan.
mice treated by gavage, and in one study in male
and female rats treated by gavage.
260
6. Evaluation Patent No. 20100105889. Available from http://www.

freepatentsonline.com/y2010/0105889.html
Elliot SJ, Zorn BH, McLeod DG et al. (2003). Pentosan
6.1 Cancer in humans polysulfate decreases prostate smooth muscle prolifera-
tion and extracellular matrix turnover. Prostate Cancer
There is inadequate evidence in humans Prostatic Dis, 6:138142. doi:10.1038/sj.pcan.4500632
PMID:12806372
for the carcinogenicity of pentosan polysulfate Erickson DR, Sheykhnazari M, Bhavanandan VP (2006).
sodium. Molecular size affects urine excretion of pentosan
polysulfate. J Urol, 175:11431147. doi:10.1016/S0022-
5347(05)00319-8 PMID:16469641
6.2 Cancer in experimental animals Faaij RA, Srivastava N, van Griensven JM etal. (1999). The
oral bioavailability of pentosan polysulphate sodium in
There is sufficient evidence in experimental healthy volunteers. Eur J Clin Pharmacol, 54:929935.
doi:10.1007/s002280050577 PMID:10192753
animals for the carcinogenicity of pentosan poly- FDA (2013). Pentosan Polysulfate Sodium. US Food and
sulfate sodium. Drug Administration. Available from: http://www.
accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm
Fellstrom B, Bjorklund U, Danielson BG et al. (1987)
6.3 Overall evaluation Pertosan polysulfate (ELMIRON): pharmacokinetics
and effects on the urinary inhibition of crystal growth.
Pentosan polysulfate sodium is possibly Pathogenese und Klinik der Harnsteine (ed. by W.
carcinogenic to humans (Group 2B). Vahlensieck & G. Gassner), pp. 340396. Steinkopff-
Verlag, Darmstadt.
Hanno PM, Burks DA, Clemens JQ, Dmochowski
RR, Erickson D, Fitzgerald MP et al.; Interstitial
References Cystitis Guidelines Panel of the American Urological
Association Education and Research, Inc (2011). AUA
guideline for the diagnosis and treatment of interstitial
Abdel-Haq H, Boss E (2012). Capillary electrophoresis cystitis/bladder pain syndrome. J Urol, 185(6):216270.
as a tool for the characterization of pentosan nano- doi:10.1016/j.juro.2011.03.064 PMID:21497847
particles. J Chromatogr A, 1257:12530. doi:10.1016/j. IMS Health (2012a). Multinational Integrated Data
chroma.2012.07.096 PMID:22921360 Analysis (MIDAS). Plymouth Meeting, Pennsylvania:
Abdo KM, Johnson JD, Nyska A (2003). Toxicity and IMS Health. Available from: http://www.imshealth.
carcinogenicity of Elmiron in F344/N rats and B6C3F1 com/portal/site/imshealth
mice following 2 years of gavage administration. Arch IMS Health (2012b). National Disease and Therapeutic
Toxicol, 77:702711. doi:10.1007/s00204-003-0472-9 Index (NDTI). Plymouth Meeting (Pennsylvania):
PMID:14508637 IMS Health, 20052012. Available from http://www.
Abnova (2013). PPS (Human) ELISA Kit. Available from: imshealth.com/portal/site/imshealth
http://www.abnova.com/protocol_pdf/KA2273.pdf IMS Health (2012c). National Prescription Audit Plus
Baba M, Nakajima M, Schols D et al. (1988). Pentosan (NPA). Plymouth Meeting, Pennsylvania: IMS Health.
polysulfate, a sulfated oligosaccharide, is a potent and 20082012.
selective anti-HIV agent in vitro. Antiviral Res, 9:335 Jerebtsova M, Wong E, Przygodzki R etal. (2007). A novel
343. doi:10.1016/0166-3542(88)90035-6 PMID:2465736 role of fibroblast growth factor-2 and pentosan poly-
Cadroy Y, Dol F, Caranobe C etal. (1987). Pharmacokinetics sulfate in the pathogenesis of intestinal bleeding in
of 125I-pentosan polysulfate in the rabbit. Thromb mice. Am J Physiol Heart Circ Physiol, 292:H743H750.
Res, 48:373378. doi:10.1016/0049-3848(87)90449-X doi:10.1152/ajpheart.00969.2006 PMID:17071728
PMID:2448892 Kongtawelert P, Ghosh P (1990). A monoclonal antibody
Degenhardt M, Benend H, Wtzig H (1998). Quality that recognizes 2,3-, 2,6-, and 4,6-disulphate ester ring
control of pentosane polysulfate by capillary zone elec- substitution in pyranose-containing polysaccharides.
trophoresis using indirect detection. J Chromatogr A, Its production, characterization and application for
817(12):297306. doi:10.1016/S0021-9673(98)00421-X the quantitation of pentosan polysulphate, dextran
PMID:9764502 sulphate, glycosaminoglycan polysulphate and chon-
Deshpande PB, Luthra P, Pandey AK, Paghdar DJ, droitin sulphate E. J Immunol Methods, 126(1):3949.
Govardhana PSGV (2010). Process for the preparation doi:10.1016/0022-1759(90)90009-K PMID:1689359
of pentosan polysulfate or salts thereof. United States
261
IARC MONOGRAPHS 108
Kyker KD, Coffman J, Hurst RE (2005). Exogenous Lett, 118:3746. doi:10.1016/S0304-3835(97)00215-2

glycosaminoglycans coat damaged bladder surfaces in PMID:9310258
experimentally damaged mouse bladder. BMC Urol, 5:4 RxList (2013). RxList - The internet drug index. Available
doi:10.1186/1471-2490-5-4 PMID:15788101 from: http://www.rxlist.com/elmiron-drug-htm
Lee CE, Edwards HE, Navaratnam S, Phillips GO (1986). SciFinder (2010). SciFinder Databases: Registry, Chemcats.
Quantitative analysis of a polysulfated xylan (SP54) in American Chemical Society.
urine using gas-liquid chromatography. Anal Biochem, Shen B, Shimmon S, Smith MM, Ghosh P (2003). Biosensor
152(1):528. doi:10.1016/0003-2697(86)90118-1 analysis of the molecular interactions of pentosan
PMID:2420230 polysulfate and of sulfated glycosaminoglycans with
MacGregor IR, Dawes J, Paton L etal. (1984). Metabolism immobilized elastase, hyaluronidase and lysozyme
of sodium pentosan polysulphate in mancatabolism using surface plasmon resonance (SPR) technology. J
of iodinated derivatives. Thromb Haemost, 51:321325. Pharm Biomed Anal, 31(1):8393. doi:10.1016/S0731-
PMID:6208629 7085(02)00606-4 PMID:12560052
MicroMedex (2013) DRUGDEXEvaluations. MicroMedex Simon M, McClanahan RH, Shah JF et al. (2005).
2.0 database. Ann Arbor, Michigan: Truven Health Metabolism of [3H]pentosan polysulfate sodium (PPS)
Analytics Inc. Available from: http://micromedex. in healthy human volunteers. Xenobiotica, 35:775784.
com/pharmaceutical. doi:10.1080/00498250500230586 PMID:16278190
Mucha S, Melen-Mucha G, Stepien T etal. (2002). Effects US Pharmacopeial Convention (2013). Pentosan
of pentosan polysulfate sodium on the estrogen-in- Polysulfate Sodium 2013. The United States
duced pituitary prolactinoma in Fischer 344 rats. Oncol Pharmacopeial Convention. No. CAS-140207938;
Rep, 9:13851389. PMID:12375053 CAS-116001968 [replaced]. Available from: www.
Muller D, Ndoume-Nze M, Jozefonvicz J (1984). High- uspusan.com.
pressure size-exclusion chromatography of anticoag- Witvrouw M, Baba M, Balzarini J, Pauwels R, De Clercq E
ulant materials. J Chromatogr, 297:3518. doi:10.1016/ (1990). Establishment of a bioassay to determine serum
S0021-9673(01)89055-5 PMID:6490767 levels of dextran sulfate and pentosan polysulfate, two
NTP (2004). NTP technical report on the toxicology potent inhibitors of human immunodeficiency virus. J
and carcinogenesis studies of Elmiron (CAS No. Acquir Immune Defic Syndr, 3(4):3437. PMID:1690285
37319178) in F344/N rats and B6C3F1 mice (gavage Wu J, Guan TJ, Zheng S etal. (2011). Inhibition of inflam-
studies). Natl Toxicol Program Tech Rep Ser, 512:7289. mation by pentosan polysulfate impedes the develop-
PMID:15213766 ment and progression of severe diabetic nephropathy
ONeil MJ (2006). The Merck Index - An encyclopedia of in aging C57B6 mice. Lab Invest, 91:14591471.
chemicals, drugs, and biologicals. 14th Ed. Whitehouse doi:10.1038/labinvest.2011.93 PMID:21808238
Station, NJ: Merck & Co., Inc. Zaslau S, Riggs DR, Jackson BJ et al. (2004). In vitro
Odlind B, Dencker L, Tengblad A (1987). Preferential effects of pentosan polysulfate against malignant
localization of 3H-pentosanpolysulphate to the breast cells. Am J Surg, 188:589592. doi:10.1016/j.
urinary tract in rats. Pharmacol Toxicol, 61:162166. amjsurg.2004.07.007 PMID:15546576
doi:10.1111/j.1600-0773.1987.tb01796.x PMID:2446305 Zaslau S, Sparks S, Riggs D etal. (2006). Pentosan poly-
Pienta KJ, Murphy BC, Isaacs WB et al. (1992). Effect sulfate (Elmiron): in vitro effects on prostate cancer
of pentosan, a novel cancer chemotherapeutic cells regarding cell growth and vascular endothelial
agent, on prostate cancer cell growth and motility. growth factor production. Am J Surg, 192:640643.
Prostate, 20:233241. doi:10.1002/pros.2990200308 doi:10.1016/j.amjsurg.2006.08.008 PMID:17071199
PMID:1374181 Zugmaier G, Lippman ME, Wellstein A (1992). Inhibition
Prochazka S, Mulholland M, Lloyd-Jones A (2003). by pentosan polysulfate (PPS) of heparin-binding
Optimisation for the separation of the oligosaccharide, growth factors released from tumor cells and blockage
sodium pentosan polysulfate by reverse polarity capil- by PPS of tumor growth in animals. J Natl Cancer
lary zone electrophoresis using a central composite Inst, 84:17161724. doi:10.1093/jnci/84.22.1716
design. J Pharm Biomed Anal, 31(1):13341. doi:10.1016/ PMID:1279186
S0731-7085(02)00569-1 PMID:12560057
PubChem (2013). Pentosan Sulfuric Polyester. Available
from: http://pubchem.ncbi.nlm.nih.gov/summary/
summary.cgi?cid=37720&loc=ecrcs, accessed 26 June
2014.
Rha SY, Noh SH, Kwak HJ etal. (1997). Comparison of
biological phenotypes according to midkine expres-
sion in gastric cancer cells and their autocrine activities
could be modulated by pentosan polysulfate. Cancer
262
TRIAMTERENE
1. Exposure Data 1.1.3 Chemical and physical properties of the

pure substance
1.1 Chemical and physical data Description: Odourless yellow powder or
1.1.1 Nomenclature crystalline solid, almost tasteless at first and
with a slightly bitter aftertaste, acidified solu-
Chem. Abstr. Serv. Reg. No.: 396-01-0 tions give a blue fluorescence (ChemicalBook,
Chem. Abstr. Serv. Name: 2,4,7-Pteridine 2013)
triamine, 6-phenyl Melting point: 316C
IUPAC systematic name: 6-Phenyl- Density: 1.502g/cm3 (ChemSpider, 2013)
pteridine-2,4,7-triamine Solubility: Very slightly soluble in water and
Synonyms: 6-Phenyl-2,4,7-triaminopteridine; in ethanol (96%).
2,4,7-triamino-6-phenyl-pteridin Stability data: Stable in formulation: acid,
neutral and alkali. Slowly oxidized upon
exposure to air (Bakshi & Singh, 2002;
Chemical Book, 2013).
Dissociation constants: pKa (strongest
acidic) = 15.88; pKa (strongest basic) = 3.11
H 2N N N NH2
(DrugBank, 2013)
N
N 1.1.4 Technical products and impurities
NH2 (a) Trade names
These are some trade names for medica-
C12H11N7 tions with triamterene as the sole active agent:
Relative molecular mass: 253.26 (ONeil, Ademine; Diren; Ditak; Diucelpin; Diurene;
2006). Diuterene; Dyazide; Dyren; Dyrenium; Dytac;
Jatropur; Riyazine; Teriam; Triteren; Urinis;
Urocaudal (ONeil, 2006; DrugBank, 2013).
263
IARC MONOGRAPHS 108
(b) Impurities of appreciable amounts of potassium in the urine.

The following impurities are listed in the The most commonly reported clinical indica-
British Pharmacopoeia (2009): tions for triamterene in the USA in 20112012
are listed in Table1.2.
5-Nitrosopyrimidine-2,4,6-triamine Triamterene is recommended as a first-line
(nitrosotriaminopyrimidine) antihypertensive in the USA (Chobanian et al.,
2,7-Diamino-6-phenylpteridin-4-ol 2003), and its use is recommended in combina-
2,4-Diamino-6-phenylpteridin-7-ol tion with another class of antihypertension drug
Phenylacetonitrile (benzyl cyanide). in Europe (Mansia et al., 2007, Mancia et al.,
2009). Off-label use for Mnire disease has been
reported in the medical literature (van Deelen
1.2 Analysis & Huizing, 1986), and still occurs in the USA
(Table1.2; IMS Health, 2012a).
Triamterene can be identified by infrared In the USA, triamterene is approved by the
absorption or potentiometric titration assays that Food and Drug Administration (FDA, 2013) for
use its property of producing an intense blueish the management of oedema and as an adjunc-
fluorescence in a 1/1000 solution of formic acid. tive diuretic where its potassium-sparing effect
The selective estimation of triamterene in the is desired. In 201112, 99.5% of its uses were as a
presence of its degradation products or other combination product with hydrochlorothiazide.
compounds in biological fluids is mainly carried The combination of triamterene and hydro-
out by high-performance liquid chromatography chlorothiazide is approved for the treatment of
(HPLC). Compendial and non-compendial hypertension or oedema in patients who develop
analytical methods are summarized in Table1.1. hypokalaemia when receiving hydrochlorothi-
azide alone, or for whom hypokalaemia cannot
be risked. Triamterene may also be used alone or
as an adjunct to other antihypertension drugs.
1.3.1 Production process The European Union (eMC, 2013) lists
three available formulations: triamterene alone,
Triamterene is a synthetic compound that triamterene combined with another diuretic
was first synthesized by Spickett & Timmis (1954) (hydrochlorothiazide, bemetizide, epitizide,
by the reaction of 4-amino-5-nitrosopyrimidine trichlormethiazide, xipamide, or furosemide),
with phenylacetonitrile (NTP, 1993). and triamterene combined with two other anti-
hypertension agents (propranolol/hydrochlo-
1.3.2 Use rothiazide, reserpine/hydrochlorothiazide, or
(a) Indications verapamil/hydrochlorothiazide). These drugs
are indicated in the European Union for oedema
Triamterene has been used since 1961 as a and hypertension.
potassium-sparing diuretic. It is still chiefly used Given its use in chronic conditions,
as an antihypertension agent for the control of triamterene therapy would be expected to be
elevated blood pressure, as well as for the treat- life-long in the absence of adverse effects for the
ment of interstitial fluid accumulation (oedema), patient.
particularly when this co-exists with hyperten- Triamterene is a weak antagonist of folic acid,
sion. Potassium-sparing diuretics, unlike other and a photosensitizing drug (NTP, 1993; Vargas
classes of diuretics, produce diuresis without loss et al., 1998).
264
Table 1.1 Some analytical methods for triamterene

Compendial methods
Identity Infrared spectroscopy British
Assay Potentiometric titration Pharmacopoeia (2009)
Human serum Extraction with methyl tert- HPLC-ECD 5ng/mL (LOQ) Richter et al. (1996)
butyl ether, centrifugation, Column: C18
evaporation, reconstitution in Mobile phase: phosphate buffer:acetonitrile
mobile phase (90:10)
Human plasma Mixing, centrifugation, gentle HPLC-UV 20ng/mL (LOD) Yakatan & Cruz
agitation, evaporated to Column: C18 (1981)
dryness under N2, extract was Flow rate: 1mL/min
reconstituted with 500L of Detector: fluorescence detector
mobile phase Mobile phase: acetonitrile:distilled water:acetic
acid
Wavelength: 365nm
Human blood, plasma, and Deproteination by adding HPLC-UV 20ng/mL (plasma) Srgel et al. (1984)
urine acetonitrile, mixing, Column: C18 0.5g/mL (urine)
centrifugation Mobile phase: acetonitrile in 0.02% phosphoric acid (LOQ)
solvent system
Flow rate: 2mL/min
Excitation wavelength: 365nm
Emission wavelength: 440nm
Human plasma and urine Protein precipitation HPLC 1ng/mL (LOQ) Swart & Botha (1987)
(plasma), centrifugation, Column: C18
collection of supernatant Fluorescence detector
Mobile phase: phosphoric acid and trimethylamine
buffer : acetonitrile : methanol (70:14:8)
Human urine Centrifugation, collection of Matrix isopotential fluorometry 2.4ng/mL (LOD) Pulgarn et al. (2001)
supernatant
Human urine Liquidliquid extraction with LC-ESI-MS/MS 20ng/mL Deventer et al. (2002)
ethyl acetate, centrifugation, Column: C18
evaporation, reconstitution in Mobile phase: 1% acetic acid and acetonitrile
mobile phase Flow rate: 0.3mL/min
Human urine Filtration through a folded CE-LIF 50ng/mL (LOQ) Horstktter et al.
filter Silica-fused capillary (2002)
Phosphate buffer
Wavelength: 353nm
Triamterene
265
266

Human urine Extraction, preconcentration GC-MS 130g/L (LOD) Amendola et al.
and derivatization with Column: C18 (2003)
acetone : methyl iodide (1:10) Selected ion monitoring
Human urine and formulation Solid-phase extraction Spectrofluorimetric method 0.8ng/mL (LOD) Ibaez et al. (2005)
Column: C18 2.3ng/mL (LOQ)
Mobile phase: phosphoric acid and triethylamine
buffer : acetonitrile : methanol (70:17:3)
IARC MONOGRAPHS 108
Bovine milk Extraction with acetonitrile, UPLC-ESI-MS/MS 0.2g/kg (LOD) Shao et al. (2008)
centrifugation, decantation of Column: C18 0.5g/kg (LOQ)
supernatant, evaporation to Flow rate: 0.45mL/min
dryness, reconstitution with Multiple reaction monitoring
water
Rat plasma Tandem solid-phase HPLC-UV 1.4ng/mL (LOD) Li et al. (2011)
extraction method by Mobile phase: acetonitrile : 0.2% acetic acid 4.8ng/mL (LOQ)
connecting two different (20 : 80)
cartridges (C18 and MCX) Flow rate: 0.8mL/min
Detection wavelength: 265nm
CE-LIF, capillary electrophoresis-laser-induced fluorescence; GC-MS, gas chromatography-mass spectrometry; HPLC, high-performance liquid chromatography; HPLC-ECD, high
performance liquid chromatography-electrochemical detection; LC-ESI-MS/MS, liquid chromatography-electrospray ionization-tandem mass spectrometry; LOD, limit of detection;
LOQ, limit of quantitation; min, minute; NR, not reported; UPLC-ESI-MS/MS, ultra performance liquid chromatography-electrospray ionization- tandem mass spectrometry; UV,
ultraviolet
Triamterene
Table 1.2 Most commonly reported clinical indications for triamterene in the USA, 20112012
Diagnosisa ICD-9 code Drug uses (thousands) Percentage of total

Essential hypertension, NOS 401.90 4824 84.6
Oedema, NOS 782.30 141 2.5
Hypertensive heart disease, other 402.90 136 2.4
Hypertension, benign 401.10 60 1.0
Mnire disease 386.00 56 1.0
Chronic ischaemic disease, unspecified, with hypertension 414.50 44 0.8
Hypertensive renal disease 403.90 43 0.8
Metabolic/insulin resistance syndrome 277.70 29 0.5
Swelling of foot 729.80 24 0.4
Arteriosclerotic heart disease with hypertension 414.20 22 0.4
a No diagnosis was stated for 0.5% of drug uses
ICD-9, International Classification of Diseases Ninth Revision; NOS, not otherwise specified
(b) Dosage triamterene in several important respects (Wang

In the USA, triamterene alone is available et al., 2007; Gu et al., 2012).
in doses of 50 mg and 100 mg. In combina- In the USA, triamterene is used moderately,
tion with hydrochlorothiazide, triamterene with 2.8million drug uses in 2012 according to
doses of 37.5 mg, 50 mg and 75 mg are used. IMS Health NDTI data (IMS Health, 2012a). Its
In 201112, IMS Health National Disease and use has declined by 47% since 2005 (Fig. 1.1).
Therapeutics Index (NDTI) data showed that Approximately 1.2 million patients in the USA
the most commonly used form was triamterene received triamterene in 2012 (IMS Health,
37.5 mg/hydrochlorothiazide 25 mg. Once- 2012a). According to the IMS Health National
per-day dosing predominates (94%). The mean Prescription Audit Plus, there was a total of
daily dosage among patients taking triamterene 5.2million prescriptions containing triamterene
is 37mg per day (IMS Health, 2012a). dispensed in the USA in 2012, a decrease of 32%
In the European Union, triamterene is avail- from 7.6 million prescriptions in 2008 (IMS
able as a single agent (50 mg), in combination Health, 2012b). In 2012, nearly all triamterene
with hydrochlorothiazide (50 mg with 25 mg (99.6%) was dispensed in the form of combina-
hydrochlorothiazide), and in combination with tion products containing hydrochlorothiazide
furosemide (50mg triamterene with 40mg furo- (IMS Health, 2012b).
semide) (eMC, 2013). Total worldwide sales of triamterene were
US$141million in 2012 according to IMS Health
(c) Trends in use MIDAS data, with 80% occurring in the USA.
The only other nation with sales of greater than
Triamterene is a less commonly used
US$ 5 million was Germany (US$ 11 million)
antihypertension agent in the USA, accounting
(IMS Health, 2012c).
for 3% of the medications prescribed for high
blood pressure. Other members of the potas-
sium-sparing diuretic class, spironolactone
and amiloride, are chemically distinct from
267
IARC MONOGRAPHS 108
Fig.1.1 Use of triamterene reported by office-based physicians, USA

1600
1400
Quarterly drug uses (thousands)
1200
1000
800
600
400
200
0
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
Prepared by the Working Group based on data obtained from IMS Health (2012a)
1.4 Occurrence and exposure 2. Cancer in Humans

Triamterene does not occur in nature.
Human exposure is predominantly from 2.1 Background
use as a medication. Occupational exposure in Five casecontrol studies, including two
manufacturing is also likely to occur. nested casecontrol studies, assessed the associ-
ation between triamterene and cancer. Cancer of
1.5 Regulations and guidelines the breast was investigated in three studies, and
cancers of the lip and colon were each assessed
Triamterene has been widely approved by in one study; however, only one study on cancer
drug regulatory agencies around the world. of the breast reported risk estimates specifically
In the USA, it was approved by the Food and for triamterene. The studies are reported below,
Drug Administration in 1964 (FDA, 2013). [The organized by relevance, and in Table2.1.
Working Group did not identify extraordinary
regulatory restrictions on the use of triamterene
as a medication, or regulations on environmental
exposure.]
268
Table 2.1 Casecontrol studies of cancer and triamterene
Reference Total Control Exposure Organ site Exposure categories Exposed Relative risk Covariates
Study location No. source assessment (ICD code) cases (95% CI) Comments
and period cases (hospital,
Total population)
No.
controls
Williams et al. 481 Population Mailed Breast Triamterene (ever- 4 0.4 Age, race, centre. Other
(1978) 1268 questionnaires use) >1yr P>0.05 risk factors did not have
USA to patients Recent use NR 0.4 confounding effects
Period, NR confirmed P>0.05 Response rate: patient
via physician 5 yr 0 NR questionnaire, 88%;
questionnaire physician questionnaire, 73%.
Triamterene usually used in
combination with thiazide.
Only four exposed cases
treated with triamterene for
>1 yr
Friedman et al. 712 Nested case Pharmacy Lip HCTZ/triamterene, 71 1.98 (1.522.58) Smoking
(2012) 22904 controls; database (squamous prescriptions before White, non-Hispanic without
San Francisco, cohort of (prescriptions cell cancer diagnosis, HIV or organ transplant; cases
USA, 1994 health-care dispensed) carcinoma, 3 identified by linking to cancer
2008 subscribers 97%) <1yr supply NR 0.91 (0.601.39) registry; controls randomly
1 to <5yr supply NR 1.87 (1.372.57) selected and matched on age,
sex, and year of cohort entry;
5 yr supply NR 2.82 (1.744.55)
analysis, 2-yr lag
Mack et al. 99 Nested Medical records; Breast Hypertensive NR 0.9 [CI cannot be Women, age 5389 yr; controls
(1975) 396 casecontrols not blinded drugs (triamterene calculated] matched by age, community
Los Angeles, (retirement alone, 19%) ever- entry; cases identified from
USA, 19715 community, use. Never used community records or
enrolled from rauwolfia class of surrounding hospital and
1968 to 1973) drugs surveillance programmes
Ever-used rauwolfia 2.9 [CI cannot
class of drugs be calculated]
Triamterene
269
270
Study location No. source assessment (ICD code) cases (95% CI) Comments
and period cases (hospital,
Total population)
No.
controls
Coogan et al. 5989 Hospital Nurse Breast Regular use (4/wk 21 1.22 (0.612.46) Race, education, menopausal
(2009) 5504 medical administered for at least 3mo) status, parity, BMI,
Boston, records questionnaire/ potassium-sparing female-hormone use, oral
IARC MONOGRAPHS 108
New York, self-reported diuretics that did contraceptives, and alcohol

Philadelphia, not contain thiazide Women, aged 2279 yr;
Baltimore, Duration (yr): controls matched to cases
USA, 1976 <5 10 1.50 (0.573.96) on age, interview year, study
2007 centre
5 11 1.05 (0.382.88)
P for trend 0.64
Coogan & 1229 Population Telephone Colorectum Dihydrofolate Age, sex, NSAID use, number
Rosenberg 1165 interview reductase inhibitors of doctor visits, alcohol
(2007) (triamterene, consumption, education,
Massachusetts, methotrexate, and vitamin use, colonoscopy,
USA, 20014 sulfasalazine) dietary factors
Regular use 34 1.6 (0.92.8) Age, 5074 yr. Factors
<5yr 16 1.3 (0.63.0) considered but had little
effect on odds ratio: race,
5yr 18 1.9 (0.84.2)
exercise, BMI, family history,
P for trend 0.6 cholecystectomy, smoking
status, hormonal replacement
therapy. Most common folate
antagonist: triamterene (30%)
BMI, body mass index; CI, confidence interval; HCTZ, hydrochlorothiazide; mo, month; NR, not reported; NSAID, nonsteroidal anti-inflammatory drug; wk, week; yr, year
Triamterene
2.2 Triamterene and cancer of the triamterene was not assessed separately. Cases
breast of cancer of the lip (squamous cell carcinoma,
97%) were identified via the programmes cancer
Odds ratios (OR) specific for triamterene registry, and drug use was determined from phar-
use were reported in a casecontrol study of 481 macy database. Analyses were based on 712 cases
women with cancer of the breast, 421 women and 22 904 matched controls and were lagged
with benign breast lesions, and 1268 controls by 2 years. The smoking-adjusted odds ratio for
identified from a mammography screening cancer of the lip and at least three prescriptions
project in the USA (Williams et al., 1978). A for hydrochlorothiazide-triamterene was 1.98
small proportion of subjects were treated with (95% CI, 1.522.58); and the risk increased with
triamterene, and only four women used it for increasing duration of use. For prescriptions of
more than 1year. Most women also took other 1 to <5years, the adjusted odds ratio was 1.87
drugs. Triamterene use, both ever and recent, (95% CI, 1.372.57), and for prescriptions of
was associated with a decreased risk of cancer of 5years the adjusted odds ratio was 2.82 (95% CI,
the breast. Odds ratios adjusted for age, race, and 1.744.55). Odds ratios for hydrochlorothiazide
centre were 0.4 (95% CI, not reported; P>0.05) were higher than those for the hydrochlorothi-
for both exposure periods. Triamterene use (ever azide-triamterene combination. [This study was
or for 5years) was associated with a non-signif- relatively informative for evaluating the effects of
icantly increased risk of benign breast lesions; treatment with hydrochlorothiazide-triamterene
however, triamterene was generally given in combined and risk of cancer of the lip because of
combination with thiazide, which was an inde- its large size, nested design, use of a pharmacy
pendent risk factor for benign breast lesions database for drug usage, and consideration of
in this study. [The strengths of this study were potential confounders by study-selection criteria
the verification of exposure information and and multivariable analysis. While the study did
detailed information on multiple-drug use and not adjust for exposure to sunlight, it seems
potential confounders. However, the study had unlikely that exposure to sunlight was sufficiently
a limited ability to evaluate the risks specific for greater in cases than controls to account for the
triamterene use because only a small number of increase in risk by up to threefold. Nevertheless,
subjects were treated with triamterene and most the study was not informative for evaluating
were exposed to multiple drugs.] specific effects of triamterene and risk of cancer.]
2.3 Combined use of triamterene 2.4 The drug class including

and hydrochlorothiazide, and triamterene, and cancer of the
cancer of the lip breast or colorectum
Friedman et al. (2012) performed casecontrol Mack et al. (1975) evaluated the association of
analyses of the association between cancer of the cancer of the breast with antihypertension drugs
lip and use of common antihypertension drugs in a nested casecontrol study among residents
in a cohort of medical-insurance subscribers in of a retirement community in Los Angeles, USA.
San Francisco, USA. Triamterene combined with The study included 99 cases of cancer of the breast
hydrochlorothiazide (see Monograph on hydro- and 396 controls matched for age and entry date,
chlorothiazide in this volume) was among the and information about medication was abstracted
most common prescriptions for diuretics, but from medical records of the community health
271
IARC MONOGRAPHS 108
centre. The rauwolfia class of drugs, including The potential for recall bias was reduced by the
reserpine, were the primary focus of the study, use of hospital controls.]
but some analyses were conducted for other A casecontrol study of colorectal cancer
antihypertension drugs (methyldopa alone or grouped triamterene with antagonists of folic
combined with other drugs, 35%; triamterene acid rather than other antihypertension drugs
alone, 19%; chlorthalidone alone, 15%; hydrala- (Coogan & Rosenberg, 2007). The study included
zine alone, 15%; spironolactone alone, 7%; guan- 1229 cases of adenocarcinoma of the colon and
ethidine alone, 3%; and combined drugs not rectum identified from cancer registries and
including methyldopa, 6%). The odds ratio for participating hospitals in Massachusetts, USA,
ever-use of other antihypertension drugs was and 1165 population-based controls matched for
0.9 (95% CI, not reported) among women who age, sex, and geographical location; cases and
never used the rauwolfia class of drugs, and 2.9 controls were free from Crohn disease or ulcera-
(95% CI, not reported) among ever-users of the tive colitis. Triamterene was the most commonly
rauwolfia class of drugs. [The major limitations of used folic-acid antagonist (30% of all antagonists).
the study were the lack of information specific for Elevated odds ratios were observed for regular
triamterene, and the potential for confounding use of dihydrofolate-reductase inhibitors, which
by other antihypertension drugs. Other limita- included methotrexate and sulfasalazine in addi-
tions were the inadequate exposure information, tion to triamterene (adjusted OR, 1.6; 95% CI,
low prevalence of exposure (14% in controls), 0.92.8) and risks were somewhat higher among
short follow-up (in part due to the advanced age those who had used these drugs for 5 years or
of the subjects), and the potential lack of general- more (adjusted OR, 1.9; 95% CI, 0.84.2). [The
izability to the general population because of the study was not specific for triamterene use, and
restricted subject population.] although the population was large, only a small
Another study of cancer of the breast eval- percentage had taken dihydrofolate-reductase
uated risks associated with several different inhibitors including triamterene. There was also
classes of diuretics, including thiazides, potas- potential for misclassification of exposure due to
sium-sparing diuretics that do not contain self-reporting.]
thiazide (including triamterene), and loop
diuretics, using medical records of several hospi-
tals for 19762007 (Coogan et al., 2009). The 3. Cancer in Experimental Animals
study included 5989 cases of invasive cancer of
the breast, and 5504 matched hospital controls See Table3.1
with diagnoses unrelated to use of diuretics. The
adjusted odd ratio for cancer of the breast and
regular use (defined as four times per week for 3.1 Mouse
at least 3months) of potassium-sparing diuretics In one study of carcinogenicity with oral
was 1.22 (95% CI, 0.612.46). The odds ratio was administration, referred to hereafter as the first
elevated for use for < 5years (adjusted OR, 1.50; study, groups of 5060male and 5060female
95% CI, 0.573.96), but not for use for 5years. B6C3F1 mice (age, 6 weeks) were given feed
[The major limitations of the study were the lack containing triamterene (purity, > 99%) at a
of information specific for triamterene, and the concentration of 0 (control), 100, 200, or 400ppm
low percentage of the population using potassi- for 2 years. These concentrations were equivalent
um-sparing diuretics (21 out of 5504 controls). to average daily doses of approximately 0, 10, 25,
272
Table 3.1 Studies of carcinogenicity with triamterene in mice and rats
Species, strain Dosing regimen Incidence of tumours Significance Comments

Duration
Reference
Mouse, B6C3F1 First study: First study *P<0.001 Purity, >99%
(M,F) 0 (control), 100, 200 or 400ppm Hepatocellular adenoma: **P<0.05 Second study performed
104wk in feed: 0, 10, 25, 45mg/kg bw 17/50 (34%), 22/50 (44%), 19/50 (38%), 20/60 (33%) (M); ***P<0.01 because of a dosing error in
NTP (1993) (M); 0, 15, 30, 60mg/kg bw (F) 10/50 (20%)*, 22/50 (44%)**, 23/50 (46%)***, 36/60 (60%)* (F) mice at the highest dose in
Second study: Hepatocellular carcinoma: the first study
0 (control) or 400ppm in 5/50 (10%)**, 7/50 (14%), 3/50 (6%), 13/60 (22%)** (M)a; [The Working Group
feed: 0, 40mg/kg bw (M); 0, 4/50 (8%), 4/50 (8%), 3/50 (6%), 8/60 (13%) (F) considered that the results
60mg/kg bw (F) Hepatocellular adenoma or carcinoma (combined): for the group at the highest
5060M and 5060F/group 20/50 (40%), 26/50 (52%), 19/50 (38%), 29/60 (48%)** (M)b; dose could not be evaluated
(age, 6 wk), for both studies 13/50 (26%)*, 26/50 (52%)**, 25/50 (50%)**, 37/60 (62%)* (F)c because of overdosing]
Second study *P<0.001
Hepatocellular adenoma: **P<0.005
21/50 (42%), 36/50 (72%)* (M);
7/50 (14%), 28/51 (55%)* (F)
9/50 (18%), 11/50 (22%) (M);
5/50 (10%), 11/51 (22%) (F)d
Hepatocellular adenoma or carcinoma (combined):
25/50 (50%), 38/50 (76%)** (M)b;
10/50 (20%), 31/50 (61%)* (F)c
Rat, F344/N 0 (control), 150, 300, 600ppm Hepatocellular adenoma: *P<0.05 Purity, >99%
(M,F) in feed: 0, 5, 10, 25mg/kg bw 0/50 (0%), 6/50 (12%)*, 4/50 (8%), 3/49 (6%) (M)e No significant increase in
104wk (M); 0, 5, 15, 30mg/kg bw (F) tumour incidence in females
NTP (1993) 50M and 50F/group (age, 6
wk)
a Historical rates of hepatocellular carcinoma in feed studies in control male mice (meanSD): 122/865 (14.1%7.2%); range, 327%
b Historical rates of hepatocellular adenoma or carcinoma (combined) in feed studies in control male mice (meanSD): 249/865 (28.8%10.9%); range, 1758%
c Historical rates of hepatocellular adenoma and carcinoma (combined) in feed studies in control female mice (meanSD): 98/863 (11.4%7.6%); range, 334%
d Historical rates of hepatocellular carcinoma in feed studies in control female mice (meanSD): 28/863 (3.2%2.9%); range, 010%
e Historical rates of hepatocellular adenoma in feed studies in control male rats (meanSD): 19/799 (2.4%2.9%); range, 08%
bw, body weight; F, female; M, male; SD, standard deviation; wk, week
Triamterene
273
IARC MONOGRAPHS 108
or 45mg/kg body weight (bw) for males, and 0, These concentrations were equivalent to average
15, 30, or 60mg/kg bw for females (NTP, 1993). daily doses of approximately 0, 5, 10, or 25mg/kg
Survival of exposed groups was similar to that bw for males, and 0, 5, 15, or 30mg/kg bw for
of controls except for the group of male mice at females (NTP, 1993). Survival of exposed groups
400 ppm. Because of a dosing error, male and was similar to that of controls. Triamterene
female mice at the highest dietary concentration caused a significant increase in the incidence of
(400ppm) actually received approximately four hepatocellular adenoma in male rats at the lowest
times the targeted concentration (approximately dose (6 out of 50; 12%), which exceeded the
1600ppm) of triamterene for 7days at week40. range for historical controls (08%;19 out of 799,
During week40, 12 male and 4 female mice died. 2.4%). Hepatocellular adenoma was present in
The surviving mice in the group receiving the all three dosed groups of males and not in males
highest dose were kept in this study, but because in the control group. There was no significant
of uncertainty regarding the effect of this 1 week increase in the incidence of tumours in female
of increased exposure on the outcome of the rats. [Hepatocellular adenoma is a tumour that
study, a second study was conducted with groups is known to progress to malignancy.]
of 5060 male and 5060 female B6C3F1 mice
(age, 6 weeks) given feed containing triamterene
at 0 (control) or 400ppm (equivalent to average 4. Mechanistic and Other
daily doses of approximately 40 mg/kg bw for Relevant Data
males, and 60mg/kg bw for females) for 2 years.
In the first study, triamterene caused signifi-
cant increases in the incidences of hepatocellular 4.1 Absorption, distribution,
adenoma in females at the lowest and interme- metabolism, and excretion
diate doses. [The Working Group considered
that the results for the group receiving the 4.1.1 Humans
highest dose could not be evaluated because of General pharmacokinetic parameters of
the overdosing.] In the second study, survival of triamterene and its major metabolite, 4-hydrox-
exposed groups was similar to that of controls. ytriamterene sulfate (see Fig. 4.1), were investi-
There were significant increases in the incidence gated in a randomized crossover trial with six
of hepatocellular adenoma in males and females, healthy volunteers who were given triamterene
and of hepatocellular adenoma or carcinoma by intravenous infusion (10 mg over 10 minutes),
(combined) in females. The incidence of liver or orally (50 mg tablet) (Gilfrich et al., 1983).
foci was increased in some groups of treated mice After intravenous infusion, terminal half-lives for
in both the first and second studies. Treatment triamterene and 4-hydroxytriamterene sulfate
with triamterene also caused treatment-related were 255 42 and 188 70 minutes, respec-
thyroid follicular cell hyperplasia. tively. Total triamterene plasma clearance was
4.41.41L/minute, which is indicative of rapid
metabolism of this compound. After oral admin-
3.2 Rat
istration, triamterene was rapidly absorbed from
In one study of carcinogenicity with oral the gastrointestinal tract. The parent drug and
administration, groups of 50male and 50female 4-hydroxytriamterene sulfate were detectable in
F344/N rats (age, 6 weeks) were given feed plasma after 15 minutes, and maximum concen-
containing triamterene (purity, > 99%) at a trations of 26.4 17.7 and 779 310 ng/mL,
concentration of 0 (control), 150, 300, or 600ppm. respectively, were reached after 1.5 hours. The
274
Triamterene
Fig.4.1 Chemical structures of triamterene and its metabolites, 4-hydroxytriamterene and

4-hydroxytriamterene sulfate
H 2N N N NH2
N Triamterene
N
NH2
H 2N N N NH2
N 4'-Hydroxytriamterene
N
NH2
HO
H 2N N N NH2
O N 4'-Hydroxytriamterene sulfate
N
S NH2
HO O
O
half-lives of unchanged triamterene and the investigated in a crossover study of 10 healthy

metabolite were not significantly different from volunteers given oral triamterene (doses, 0, 12.5,
those observed after intravenous administration. 25, 50, or 100 mg), and of 12 healthy volunteers
The bioavailability of triamterene after intra- given triamterene (doses, 0, 25, or 50 mg) with
venous or oral administration was calculated hydrochlorothiazide (doses, 12.5, or 25 mg).
from plasma and urine concentrations, and Coadministration did not affect the renal excre-
found to be 5222%, demonstrating consider- tion of either parent drug, but significantly
able inter-subject variability. By inclusion of data reduced renal excretion of 4-hydroxytriamterene
for total excretion of triamterene and 4-hydroxy sulfate (Mhrke et al., 1997).
triamterene sulfate, the absorption was calcu- 4-Hydroxytriamterene sulfate shows more
lated to be 83.225.9% (Gilfrich et al., 1983). The protein binding (91% protein-bound) than
high absorption of triamterene is indicative of its triamterene (55%) (Knauf et al., 1983), possibly
ability, as a weak lipid-soluble base (pKa=6.2) to because of ionic bonding between sulfate and
cross lipid membranes by non-ionic diffusion protein, in addition to hydrophobic bonding
(Kau, 1978). Triamterene is more highly concen- (Gilfrich et al., 1983). 4-Hydroxytriamterene
trated in erythrocytes than in plasma (Gilfrich sulfate is almost completely eliminated by
et al., 1983). tubular secretion, while triamterene, although
The pharmacokinetic interaction of partly eliminated via this route, can also undergo
triamterene and hydrochlorothiazide was glomerular filtration, due to the fact that a
275
IARC MONOGRAPHS 108
substantial proportion (45%) is not bound to 4.1.2 Experimental systems

protein (Knauf et al., 1983).
Triamterene undergoes significant first- Intestinal absorption of triamterene in the
pass metabolism with rapid hydroxylation of colon and the whole small intestine of the rat was
the phenyl ring at the 4-position, yielding the shown to occur via a carrier-mediated mecha-
phase-I metabolite 4-hydroxytriamterene (see nism (Montalar et al., 2003) likely to comprise
Fig. 4.1). This intermediate metabolite is tran- two carriers, and also via an efflux process (Kau
sient and is detected at most in trace amounts & Sastry, 1977).
(<1ng/mL) in urine or plasma (Gilfrich et al., The tissue distribution in male Sprague-
1983). Hydroxylation seems to be mediated virtu- Dawley rats given [14C]triamterene intravenously
ally exclusively by cytochrome P450 1A2, and showed extensive accumulation of radiolabel
inhibition or induction of this isoenzyme will (Kau & Sastry, 1977). High concentrations of the
change the time-course of both triamterene and parent drug were found in most tissues (except
its pharmacologically active phase-II metabolite the brain, fat, and testes), and blood concentra-
(Fuhr et al., 2005). tions were low. No metabolites were detected.
4-Hydroxytriamterene is rapidly conju- The highest concentrations of triamterene were
gated via cytosolic sulfotransferases to yield reached within the first 20 minutes in highly
the principal phase-II metabolite 4-hydroxytri- perfused tissues such as kidneys, liver, heart,
amterene sulfate (Gilfrich et al., 1983; NTP, 1993; lungs, and skeletal muscle. The kidneys contained
Horstktter et al., 2002). In addition, phase-II the highest concentrations of triamterene at all
metabolism produces very small quantities timed intervals and doses, but the largest dose
of other metabolites, such as N-glucuronides deposition was in skeletal muscle (part of the
(Lehmann, 1965). peripheral compartment). Elimination was
The parent drug and its metabolite slow (estimated plasma half-life, 2.8 hours),
4-hydroxytriamterene sulfate are excreted in the possibly due to triamterene binding to tissue
urine and faeces (Kau & Sastry, 1977; NTP, 1993). in the central compartment (e.g., kidneys,
Renal clearance of triamterene administered by liver). [This study demonstrated the ability of
intravenous infusion (10 mg over 10 minutes) triamterene to bind to tissue, which influences
was 0.220.1L/minute, and that of 4-hydroxy its rate of distribution and elimination.]
triamterene sulfate was 0.17 0.061 L/minute. Studies of tissue distribution of triamterene
Total urinary recovery of triamterene and of in guinea-pigs and baboons reported triamterene
the sulfate was 4.5% and approximately 50%, concentrations in muscle and heart that were
respectively (Gilfrich et al., 1983). Renal clear- much higher than in plasma, low concentrations
ance of orally administered (50 mg) triamterene in the brain, active transport of triamterene in
and of the sulfate was 0.180.05L/minute and the kidney, and transfer of triamterene from
0.15 0.03 L/minute, respectively (Gilfrich mother to fetus (Pruitt et al., 1975).
et al., 1983). 4-Hydroxytriamterene sulfate and A quantitative study of the renal handling
N-glucuronides were also excreted into the bile of [3H]triamterene in an isolated, perfused
(Mutschler et al., 1983; NTP, 1993). rat-kidney model showed that triamterene
undergoes glomerular filtration, active tubular
secretion and passive re-absorption by a pH-de-
pendent mechanism (Kau, 1978).
When [14C]triamterene (2 mg/kg bw) was
administered subcutaneously to Sprague-Dawley
276
Triamterene
rats, 45% of the total radiolabel was excreted in the 4.3 Other mechanistic data relevant
urine, and 50% in the faeces, over 72 hours (Kau to carcinogenicity
et al., 1975). In the urine and faeces, 7279% of
the administered dose was excreted as unchanged 4.3.1 Effects on cell physiology
triamterene, 1015% as free 4-hydroxytri-
amterene, 15% as its sulfate, and 2% as a minor Triamterene is a potassium-sparing diuretic
unidentified metabolite. [The low level of sulfate that blocks the epithelial sodium channel on
conjugation may have been a consequence of the the luminal side of the collecting tubule in the
route of administration.] The liver was identi- kidney. Its major physiological action is to inhibit
fied as the major site of triamterene metabolism; transport of sodium ions and reduce blood
triamterene was not metabolized in the kidney. volume. The action of triamterene is different
No metabolites were detected in the plasma. from that of thiazide-based drugs, but similar to
that of amiloride (Busch et al., 1996). [There is no
known link between sodium-channel inhibition
4.2 Genetic and related effects and carcinogenesis.]
4.2.1 Humans
4.3.2 Effects on cell function
In cultures of the BJAB-B958 human
lymphoma cell line, there was a dose-dependent
4.2.2 Experimental systems inhibition of cell growth, with cell death
See Table4.1 following extended incubation with triamterene
at 80M. The activity of dihydrofolate reductase
(a) Mutagenicity was blocked by triamterene, and its metabolites
Triamterene (up to 10 000 g/plate) was 4-hydroxytriamterene and 4-hydroxytriam-
not mutagenic in Salmonella typhimurium terene sulfate are less effective inhibitors of this
strains TA98, TA1535, TA1537 or TA100, with enzyme (Schalhorn et al., 1981).
or without metabolic activation (S9 from rat or Photosensitivity, a known clinical side-effect
hamster liver) (NTP, 1993). Triamterene did not of triamterene, was demonstrated in vitro in
induce dominant lethal mutations in germ cells cultures of human peripheral blood lymphocytes
of male CD-1 mice when given at doses of up and neutrophils co-exposed to triamterene and
to 100 mg/kg bw per day by gavage for 5 days ultraviolet A (UV-A) irradiation. At a concen-
(Manson et al., 1986). tration of 60g/mL, triamterene decreased cell
viability to 4050%, indicating photosensitiza-
(b) Chromosomal damage tion of twofold. Other experiments also showed
No increase in the frequency of chromo- that triamterene produced singlet oxygen species
somal aberration was induced by triamterene at when irradiated with UV-A light in the presence
concentrations of up to 600 g/mL in Chinese of molecular oxygen. (Vargas et al., 1998).
hamster ovary cells in vitro in the presence or
absence of metabolic activation. Triamterene did
induce sister chromatid exchange in Chinese
hamster ovary cells in vitro in the presence or
absence of metabolic activation (NTP, 1993).
277
IARC MONOGRAPHS 108
Table 4.1 Genetic and related effects of triamterene
Test system Resultsa Concentration or dose Reference

(LED or HID)
Without With
exogenous exogenous
metabolic metabolic
system system
In vitro
Salmonella typhimurium TA100, TA98, TA1535, TA1537, 10000 g/plate NTP (1993)
reverse mutation
Chromosomal aberrations, Chinese hamster ovary cells 600g/mL NTP (1993)
Sister chromatid exchange, Chinese hamster ovary cells + + 10g/mL, S9 NTP (1993)
160g/mL, + S9
In vivo
Dominant lethal mutations, germ cells of male CD-1 mice NT 100mg/kg bw per day Manson et al.
for 5days (1986)
bw, body weight; HID, highest ineffective dose; LED, lowest effective dose; S9, 9000 g supernatant
4.4 Susceptibility 4.4.2 Renal impairment

4.4.1 Liver disease Although renal elimination is only a minor
route of excretion for triamterene, it is the main
Patients with liver cirrhosis have reduced route of elimination of 4-hydroxytriamterene
ability to hydroxylate triamterene, as evidenced sulfate. Thus, in individuals with renal impair-
by high plasma concentrations of triamterene ment, accumulation of the sulfate is substantial
and low concentrations of 4-hydroxytri- and progressive, but negligible for triamterene.
amterene sulfate. After administration of The kinetics of triamterene were observed in 32
200 mg of triamterene, peak plasma concen- patients with widely varying degrees of creati-
trations in eight patients without liver disease nine clearance (10135mL/minute), an indicator
were 55948ng/mL and 2956320ng/mL for of renal function. In patients with reduced renal
triamterene and 4-hydroxytriamterene sulfate, function, significant accumulation in plasma
respectively. In the seven patients with alco- and reduced renal clearance of the sulfate were
holic cirrhosis, peak plasma concentrations of reported. Plasma concentrations of the parent
triamterene were increased to 1434184ng/mL, drug were not increased (Knauf et al., 1983).
while the concentrations of the sulfate were
reduced to 469 84 ng/mL. Renal clearance 4.4.3 Age
was also reduced in patients with cirrhosis: the
clearance of triamterene and the sulfate were Early reports concluded that the mean peak
2.80.7 and 38.06.6mL/minute, respectively, concentration of triamterene after an oral dose of
compared with 14.41.5 and 116.711.6mL/ 50mg is higher in older than in younger patients
minute, respectively, in patients without liver (84 and 41 ng/mL, respectively), that the time
disease (Villeneuve et al., 1984). to reach peak concentrations of 4-hydroxytri-
amterene sulfate is prolonged [suggesting that
hydroxylation decreases with age], and that the
systemic clearance of the parent drug and its
278
Triamterene
metabolite declines significantly with age (NTP, 5. Summary of Data Reported

1993).
In contrast, a more recent study in which
an oral dose of triamterene (50 mg) was given
5.1 Exposure data
to 11 healthy elderly individuals (mean age, Triamterene is a synthetic potassium-sparing
68 5 years), and to 10 healthy young indi- diuretic; unlike other diuretics, it causes limited
viduals (mean age, 252years) did not report excretion of potassium in the urine. Triamterene
a significant reduction in hydroxylation of is most often prescribed for hypertension as a
triamterene (Fliser et al., 1999). Renal clearance combination tablet that includes hydrochlorothi-
was similar in elderly and young individuals. azide. In 2012, while global sales were greatest in
This study excluded those with conditions that the USA (US$141 million), the volume of annual
may adversely affect renal or hepatic function prescriptions was modest and had declined over
(e.g. hypertension, malnutrition, and cardiac time.
failure) and all subjects were on a standardized
diet to control intake of protein and electrolytes,
particularly sodium. It was assumed by Fliser 5.2 Human carcinogenicity data
et al. (1999) that the subjects in previous studies Five casecontrol studies were available to
may have had various conditions that affected assess the association between triamterene and
renal and/or hepatic function. cancer.
In two of these studies, all subjects in the
4.5 Mechanistic consideration treatment group were exposed to triamterene.
One of the studies reported a risk estimate that
Triamterene was not mutagenic in S. typhi- was specific for triamterene and cancer of the
murium, with or without exogenous metabolic breast; however, most of the women took other
activation, did not induce chromosomal aber- drugs and few subjects had used triamterene for
rations in Chinese hamster ovary cells, with or more than 1 year. The other study reported a rela-
without exogenous metabolic activation (NTP, tive risk for cancer of the lip and treatment with
1993), and did not induce the dominant lethal combined triamterene and hydrochlorothiazide,
mutation in the germ cells of male CD-1 mice and thus any effect attributable to triamterene
in vivo (Manson et al., 1986). However, positive could not be separated from the effects attribut-
results were obtained for induction of sister chro- able to hydrochlorothiazide.
matid exchange in Chinese hamster ovary cells, The three remaining casecontrol studies
both with and without metabolic activation (NTP, reported findings on cancers of the breast or
1993). Both inhibition of dihydrofolate reductase colorectum for triamterene in combination
and photosensitization are possible mechanisms with other drugs (hydrochlorothiazide, other
for the induction of DNA damage by triamterene antihyper tension drugs, potassium-sparing
(Schalhorn et al., 1981; Vargas et al., 1998). diuretics, and other inhibitors of dihydrofolate
reductase).
The available studies were not informative
for evaluation of the association between risk of
cancer and exposure specifically to triamterene.
279
IARC MONOGRAPHS 108
5.3 Animal carcinogenicity data germ cells of male CD-1 mice in vivo. Triamterene
induced sister chromatid exchange in Chinese
Triamterene was tested for carcinogenicity by hamster ovary cells, in the presence or absence
oral administration in two studies in mice and of exogenous metabolic activation. Therefore,
one study in rats. triamterene may produce genetic toxicity directly
In a first feeding study in male and female at the chromosomal level without metabolic
mice, triamterene caused a significant increase activation.
in the incidence of hepatocellular adenoma in Triamterene is an inhibitor of dihydrofolate
females. In a second feeding study, triamterene reductase in vitro; its metabolites 4-hydroxy-
caused significant increases in the incidences of triamterene and 4-hydroxytriamterene sulfate
hepatocellular adenoma in males and females, are less effective inhibitors of the enzyme. When
and of hepatocellular adenoma or carcinoma irradiated with ultraviolet A light, triamterene
(combined) in females. produced singlet oxygen species in the pres-
In a feeding study in male and female rats, ence of molecular oxygen. In-vitro coexposure
triamterene caused an increase in the incidence of human peripheral blood lymphocytes and
of hepatocellular adenoma (a tumour that is neutrophils to triamterene and ultraviolet A
known to progress to malignancy) in males. resulted in decreased cell viability.
Hepatocellular adenoma was reported in all dose Inhibition of dihydrofolate reductase and
groups, but not in rats in the control group. There photosensitization are possible mechanisms for
was no significant increase in the incidence of the induction of DNA damage by triamterene.
tumours in female rats.
5.4 Mechanistic and other relevant 6. Evaluation

data
In humans, triamterene is primarily metab-
olized to 4-hydroxytriamterene sulfate via There is inadequate evidence in humans for
sulfotransferase-mediated conjugation of the the carcinogenicity of triamterene.
phase-I metabolite, 4-hydroxytriamterene.
4-Hydroxytriamterene sulfate binds strongly to 6.2 Cancer in experimental animals
proteins, to a greater extent than the parent drug.
Intravenous administration of radiolabelled There is sufficient evidence in experimental
triamterene in rats resulted in extensive accu- animals for carcinogenicity of triamterene.
mulation of radiolabel, with the highest concen-
trations in highly perfused tissues, particularly
the kidneys.
6.3 Overall evaluation
Triamterene was not mutagenic when tested Triamterene is possibly carcinogenic to
in Salmonella typhimurium, in the presence humans (Group 2B).
or absence of exogenous metabolic activation.
Triamterene also gave negative results in assays
for the induction of chromosomal aberration in
Chinese hamster ovary cells, in the presence or
absence of exogenous metabolic activation, and
did not induce dominant lethal mutation in the
280
Triamterene
References ht t p : // w w w. a c c e s s d a t a . fd a . g ov/s c r ip t s /c d e r/
drugsatfda/index.cfm, accessed 8 September 2014.
Fliser D, Bischoff I, Hanses A, Block S, Joest M, Ritz E etal.
Amendola L, Colamonici C, Mazzarino M, Botr F (2003). (1999). Renal handling of drugs in the healthy elderly.
Rapid determination of diuretics in human urine by Creatinine clearance underestimates renal function
gas chromatography-mass spectrometry following and pharmacokinetics remain virtually unchanged.
microwave assisted derivatization. Anal Chim Acta, Eur J Clin Pharmacol, 55(3):20511. doi:10.1007/
475(1-2):12536. doi:10.1016/S0003-2670(02)01223-0 s002280050619 PMID:10379636
Bakshi M, Singh S (2002). Development of validated Friedman GD, Asgari MM, Warton EM, Chan J, Habel
stability-indicating assay methodscritical review. J LA (2012). Antihypertensive drugs and lip cancer in
Pharm Biomed Anal, 28(6):101140. doi:10.1016/S0731- non-Hispanic whites. Arch Intern Med, 172(16):124651.
7085(02)00047-X PMID:12049968 doi:10.1001/archinternmed.2012.2754 PMID:22869299
British Pharmacopoeia (2009). Triamterene. London, UK, Fuhr U, Kober S, Zaigler M, Mutschler E, Spahn-
Medicines and healthcare products regulatory agency. Langguth H (2005). Rate-limiting biotransformation
Busch AE, Suessbrich H, Kunzelmann K, Hipper A, of triamterene is mediated by CYP1A2. Int J Clin
Greger R, Waldegger S etal. (1996). Blockade of epithe- Pharmacol Ther, 43(7):32734. doi:10.5414/CPP43327
lial Na+ channels by triamterenes - underlying mecha- PMID:16035375
nisms and molecular basis. Pflugers Arch, 432(5):7606. Gilfrich HJ, Kremer G, Mhrke W, Mutschler E, Vlger
doi:10.1007/s004240050196 PMID:8772124 KD (1983). Pharmacokinetics of triamterene after i.v.
ChemicalBook (2013). ChemicalBook: Chemical Search administration to man: determination of bioavaila-
Engine. Available from: http://www.chemicalbook. bility. Eur J Clin Pharmacol, 25(2):23741. doi:10.1007/
com. BF00543797 PMID:6628507
ChemSpider (2013). ChemSpider: The free chemical Gu Q, Burt VL, Dillon CF, Yoon S (2012). Trends in antihy-
database. Royal Society of Chemistry. Available from: pertensive medication use and blood pressure control
http://www.chemspider.com. among United States adults with hypertension: the
Chobanian AV, Bakris GL, Black HR, Cushman WC, National Health and Nutrition Examination Survey,
Green LA, Izzo JL Jr etal. ; Joint National Committee 2001 to 2010. Circulation, 126(17):210514. doi:10.1161/
on Prevention, Detection, Evaluation, and Treatment of CIRCULATIONAHA.112.096156 PMID:23091084
High Blood Pressure. National Heart, Lung, and Blood Horstktter C, Kober S, Spahn-Langguth H, Mutschler
Institute; ; National High Blood Pressure Education E, Blaschke G (2002). Determination of triamterene
Program Coordinating Committee (2003). Seventh and its main metabolite hydroxytriamterene sulfate
report of the Joint National Committee on Prevention, in human urine by capillary electrophoresis using
Detection, Evaluation, and Treatment of High Blood ultraviolet absorbance and laser-induced fluorescence
Pressure. Hypertension, 42(6):120652. doi:10.1161/01. detection. J Chromatogr B Analyt Technol Biomed Life
HYP.0000107251.49515.c2 PMID:14656957 Sci, 769(1):10717. doi:10.1016/S1570-0232(02)00002-8
Coogan PF, Rosenberg L (2007). The use of folic PMID:11936683
acid antagonists and the risk of colorectal cancer. Ibaez GA, Escandar GM, Espinosa Mansilla A, Muoz
Pharmacoepidemiol Drug Saf, 16(10):11119. de la Pea A (2005). Determination of triamterene in
doi:10.1002/pds.1442 PMID:17600846 pharmaceutical formulations and of triamterene and
Coogan PF, Strom BL, Rosenberg L (2009). Diuretic its main metabolite hydroxytriamterene sulfate in
use and the risk of breast cancer. J Hum Hypertens, urine using solid-phase and aqueous solution lumines-
23(3):2168. doi:10.1038/jhh.2008.131 PMID:18971940 cence. Anal Chim Acta, 538(1-2):7784. doi:10.1016/j.
Deventer K, Delbeke FT, Roels K, Van Eenoo P (2002). aca.2005.02.001
Screening for 18 diuretics and probenecid in doping IMS Health (2012a). National Disease and Therapeutic
analysis by liquid chromatography-tandem mass Index (NDTI). Plymouth Meeting (Pennsylvania):
spectrometry. Biomed Chromatogr, 16(8):52935. IMS Health, 20052012. Available from: http://www.
doi:10.1002/bmc.201 PMID:12474217 imshealth.com/portal/site/imshealth, accessed 8
DrugBank (2013). DrugBank: Open Data Drug & Drug September 2014.
Target Database. Version 4.2. Available from: http:// IMS Health (2012b). National Prescription Audit Plus
www.drugbank.ca/ (NPA). Plymouth Meeting, Pennsylvania: IMS Health.
eMC (2013). electronic Medicines Compendium (eMC). 20082012.
Available from: http://www.medicines.org.uk, accessed IMS Health (2012c). Multinational Integrated Data
8 September 2014. Analysis (MIDAS). Plymouth Meeting, Pennsylvania:
FDA (2013). Triamterene. US Food and IMS Health. Available from: http://www.imshealth.
Drug Administration. Available from: com/portal/site/imshealth, accessed 8 September 2014.
281
IARC MONOGRAPHS 108
Kau ST (1978). Handling of triamterene by the isolated methotrexate. J Drug Target, 11(4):21523. doi:10.1080/
perfused rat kidney. J Pharmacol Exp Ther, 206(3):7019. 10611860310001615947 PMID:14578108
PMID:702330 Mutschler E, Gilfrich HJ, Knauf H, Mhrke
Kau ST, Sastry BV (1977). Distribution and pharmacoki- W, Vlger KD (1983). Pharmacokinetics of
netics of triamterene in rats. J Pharm Sci, 66(1):536. triamterene. Clin Exp Hypertens A, 5(2):24969.
doi:10.1002/jps.2600660112 PMID:833742 doi:10.3109/10641968309048825 PMID:6831748
Kau ST, Sastry BV, Alvin JD, Bush MT (1975). Metabolism NTP 1993). NTP Toxicology and carcinogenesis studies
of triamterene in the rat. Drug Metab Dispos, 3(5):345 of triamterene (CAS No. 396010) in F344/N rats and
51. PMID:241615 B6C3F1 mice (feed studies). Natl Toxicol Program Tech
Knauf H, Mhrke W, Mutschler E (1983). Delayed elimina- Rep Ser, 420:1367. PMID:12616291
tion of triamterene and its active metabolite in chronic ONeil MJ (2006). The Merck Index - An Encyclopedia of
renal failure. Eur J Clin Pharmacol, 24(4):4536. Chemicals, Drugs, and Biologicals. 14th ed. Whitehouse
doi:10.1007/BF00609885 PMID:6861860 Station (NJ): Merck & Co., Inc.
Lehmann K (1965). [Separation, isolation and iden- Pruitt AW, McNay JL, Dayton PG (1975). Transfer char-
tification of metabolic products of triamterene] acteristics of triamterene and its analogs. Central
Arzneimittelforschung, 15(7):8126. PMID:5898685 nervous system, placenta, and kidney. Drug Metab
Li H, He J, Liu Q, Huo Z, Liang S, Liang Y (2011). Dispos, 3(1):3041. PMID:234832
Simultaneous analysis of hydrochlorothiazide, Pulgarn JA, Molina AA, Lpez PF (2001). Simultaneous
triamterene and reserpine in rat plasma by high perfor- direct determination of amiloride and triamterene
mance liquid chromatography and tandem solid- in urine using isopotential fluorometry. Anal
phase extraction. J Sep Sci, 34(5):5427. doi:10.1002/ Biochem, 292(1):5968. doi:10.1006/abio.2001.5064
jssc.201000754 PMID:21344645 PMID:11319818
Mack TM, Henderson BE, Gerkins VR, Arthur M, Baptista Richter K, Oertel R, Kirch W (1996). New sensitive
J, Pike MC (1975). Reserpine and breast cancer in a method for the determination of hydrochlorothiazide
retirement community. N Engl J Med, 292(26):136671. in human serum by high-performance liquid chroma-
doi:10.1056/NEJM197506262922603 PMID:1138164 tography with electrochemical detection. J Chromatogr
Mancia G, Laurent S, Agabiti-Rosei E, Ambrosioni E, A, 729(1-2):2936. doi:10.1016/0021-9673(95)00900-0
Burnier M, Caulfield MJ et al. ; European Society of PMID:9004952
Hypertension(2009). Reappraisal of European guide- Schalhorn A, Siegert W, Sauer HJ (1981). Antifolate effect
lines on hypertension management: a European Society of triamterene on human leucocytes and on a human
of Hypertension Task Force document. J Hypertens, lymphoma cell line. Eur J Clin Pharmacol, 20(3):219
27(11):212158. doi:10.1097/HJH.0b013e328333146d 24. doi:10.1007/BF00544601 PMID:7286039
PMID:19838131 Shao B, Zhang J, Yang Y, Meng J, Wu Y, Duan H (2008).
Mansia G, De Backer G, Dominiczak A, Cifkova R, Simultaneous analysis of thirteen diuretics residues
Fagard R, Germano G et al. European Society of in bovine milk by ultra-performance liquid chroma-
Cardiology(2007). 2007 ESH-ESC Guidelines for tography/tandem mass spectrometry. Rapid Commun
the management of arterial hypertension: the task Mass Spectrom, 22(21):342733. doi:10.1002/rcm.3740
force for the management of arterial hypertension PMID:18837072
of the European Society of Hypertension (ESH) and Srgel F, Lin ET, Hasegawa J, Benet LZ (1984). Liquid
of the European Society of Cardiology (ESC). Blood chromatographic analysis of triamterene and its
Press, 16(3):135232. doi:10.1080/08037050701461084 major metabolite, hydroxytriamterene sulfate, in
PMID:17846925 blood, plasma, and urine. J Pharm Sci, 73(6):8313.
Manson JM, Guerriero FJ, Brown T, San Sebastian J (1986). doi:10.1002/jps.2600730633 PMID:6737274
Lack of in vivo mutagenicity and testicular toxicity of Spickett RGW, Timmis GM (1954). The synthesis of
triamterene in mice. Fundam Appl Toxicol, 7(4):533 compounds with potential anti-folic acid activity. Part
46. doi:10.1016/0272-0590(86)90104-1 PMID:3803749 I. 7-Amino- and 7-hydroxy-pteridines. J Chem Soc,
Mhrke W, Knauf H, Mutschler E (1997). Pharmacokinetics 288795. doi:10.1039/jr9540002887
and pharmacodynamics of triamterene and hydro- Swart KJ, Botha H (1987). Rapid method for the determi-
chlorothiazide and their combination in healthy nation of the diuretic triamterene and its metabolites in
volunteers. Int J Clin Pharmacol Ther, 35(10):44752. plasma and urine by high-performance liquid chroma-
PMID:9352394 tography. J Chromatogr A, 413:3159. doi:10.1016/0378-
Montalar M, Nalda-Molina R, Rodrguez-Ibez M, 4347(87)80246-3 PMID:3558684
Garca-Valcrcel I, Garrigues TM, Merino V et al. van Deelen GW, Huizing EH (1986). Use of a diuretic
(2003). Kinetic modeling of triamterene intes- (Dyazide) in the treatment of Menires disease. A
tinal absorption and its inhibition by folic acid and double-blind cross-over placebo-controlled study.
282
Triamterene
ORL J Otorhinolaryngol Relat Spec, 48(5):28792.

doi:10.1159/000275884 PMID:3537899
Vargas F, Fuentes A, Sequera J, Mndez H, Fraile G,
Velsquez M et al. (1998). In Vitro Approach to
investigating the phototoxicity of the diuretic drug
triamterene. Toxicol In Vitro, 12(6):6617. doi:10.1016/
S0887-2333(98)00057-5 PMID:20654456
Villeneuve JP, Rocheleau F, Raymond G (1984).
Triamterene kinetics and dynamics in cirrhosis. Clin
Pharmacol Ther, 35(6):8317. doi:10.1038/clpt.1984.121
PMID:6734036
Wang YR, Alexander GC, Stafford RS (2007). Outpatient
hypertension treatment, treatment intensification, and
control in Western Europe and the United States. Arch
Intern Med, 167(2):1417. doi:10.1001/archinte.167.2.141
PMID:17242314
Williams RR, Feinleib M, Connor RJ, Stegens NL (1978).
Case-control study of antihypertensive and diuretic use
by women with malignant and benign breast lesions
detected in a mammography screening program. J Natl
Cancer Inst, 61(2):32735.http://www.ncbi.nlm.nih.
gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&l
ist_uids=277719&dopt=Abstract PMID:277719
Yakatan GJ, Cruz JE (1981). High-performance liquid
chromatographic analysis of triamterene and p-hy-
droxytriamterene in plasma. J Pharm Sci, 70(8):94951.
doi:10.1002/jps.2600700832 PMID:7310671
283
HYDROCHLOROTHIAZIDE

1.1 Identification of the agent O O O O
1.1.1 Nomenclature S S
N H
Chem. Abstr. Serv. Reg. No.: 58-93-5 (SciFinder, H2 N
2013) Cl N
Chem. Abstr. Serv. Name: 2H-1,2,4-Benzo H
thiadiazine-7-sulfonamide, 6-chloro-3,4-di-
hydro-1,1-dioxide (SciFinder, 2013)
IUPAC systematic name: 6-Chloro-1,1-dioxo- C7H8ClN3O4S2
3,4-dihydro-2H-1,2,4-benzothiadiazine-7- Relative molecular mass: 297.74 (ONeil, 2001)
sulfonamide
Synonyms: Dihydrochlorothiazid; Dihydro 1.1.3 Chemical and physical properties of the
chlorothiazide; Dihydrochlorothiazidum; pure substance
Dihydrochlorurit; Dihydrochlorurite;
Dihydrox ychlorothiazidum; HCTZ; HCZ; Description: A white or almost white, crys-
Hydrochlorothiazid; Hydrochlorthiazide talline powder; odourless or almost odourless
(DrugBank, 2013); Hidroclorotiazida; (WHO, 2006)
Hydrochlorothiazidum; Chlorsulfonamido Density: 1.693g/cm3 (calculated) (Lookchem,
dihydrobenzothiadiazine dioxide; Chloro 2013)
sulthiadil; 3,4-dihydrochlorothiazide (IPCS,
2013) Melting point: 273275 C (ONeil, 2001)
WHO International Nonproprietary Name Spectroscopy data: Infrared, Raman, ultra-
(INN): Hydrochlorothiazidum (WHO, 2006) violet, proton nuclear magnetic resonance
(1H-NMR) and 13C-NMR spectral data have
been reported.
Solubility: Very slightly soluble in water
(722mg/L at 25C; Deppeler, 1981); soluble
in ethanol at ~750g/L and in acetone (WHO,
2006); soluble in dilute ammonia; freely
soluble in sodium hydroxide solution, in
285
IARC MONOGRAPHS 108
n-butylamine and in dimethylformamide; 6-Chloro-N-[(6-chloro-7-sulfamoyl-2,3-

sparingly soluble in alcohol; insoluble in dihyd ro-4H-1,2,4-benzothiadiazin-4-yl
ether, chloroform and in dilute mineral acids 1,1-d iox ide)met hyl]-3,4 -d i hyd ro -2 H-
(HSDB, 2013). 1,2,4-benzot hiadiazine-7-su lfonamide
Octanol/water partition coefficient (log P): 1,1-dioxide.
0.07 (Chemspider, 2013)
Dissociation constant: pKa = 7.9 ; pKa2 = 9.2
(ONeil, 2001)
1.2 Analysis
An overview of selected analytical methods is
1.1.4 Technical products and impurities presented in Table1.1.
(a) Trade names

Acuren; Adelphan; Apo-Hydro; Aqua
zide; Clorana; Colidur; Colonraitai; Cotrazid; 1.3.1 Production process
Decazon; Dehydratin; Dehydrazid; Depress;
Hydrochlorothiazide is synthesized by
Dichlotride; Dichlozid; Di-Eudrin; Dihydrochl
either the reaction of para-formaldehyde with
Ozide; Dihydrochlorothiazide; Dihydrodiazid;
5-chloro-2,4-disulfamoylaniline in nonaqueous
Disalunil; Disothiazide; Dithiazide; Dithiazide;
media, or the reaction of formaldehyde with
Diunorm; Diunorm; Diurace; Diural; Diuren;
6-chloro-7-sulfamoyl-2H-1,2,4-benzothiadi-
Diurex; Diurezin; Diuzid; Do-Hydro; Drenol;
azine-1,1-dioxide in aqueous alkaline solution
Duberzide; Edepress; Esidrex; Esidrix; H.C.T.;
(Deppeler, 1981).
HCT [manufacturer]; Hexazide; Hidro
clorotiazida; Hidrochlorotiazid Alkaloid;
Hidromed; Hidroronol; Hidrosaluretil; Hidro 1.3.2 Use
tiadol; HTZ; Hybozide; Hychlozide; Hydrex; (a) Indications
Hydride; Hydrochlorothiazide; Hydrochloro
Hydrochlorothiazide is a thiazide-type
thiazidum Polpharma; Hydroklortiazid
diuretic chiefly used as an antihypertension
Evolan; Hydromed; Hydrozide; Hypodehydra;
agent for the control of elevated blood pres-
Hypothiazid; Hytaz; Hyzide; Keshiau; Klorzide;
sure (Table1.2). It is often combined with other
Koliside; Locoid; Lonpra; Microzide; Monozid;
agents in the treatment of hypertension, either
Nefrix; Newtolide; Nisidrex; Nor-Tiazida; Oretic;
through separate prescriptions for hydrochloro-
Ridaq; Rofucal; Tandiur; Tiazid; Urilzid; Xenia
thiazide and the other agents, or through the use
(MicroMedex, 2013).
of combination products in which a single tablet
(b) Impurities contains hydrochlorothiazide plus one other
antihypertensive medication (more rarely, two
Some impurities are described in the other agents).
European Pharmacopoeia (2005): In the USA, hydrochlorothiazide is indicated
Chlorothiazide (active drug rarely used as an for the management of hypertension either as
alternative to hydrochlorothiazide) the sole therapeutic agent or in combination
4-Amino-6-chlorobenzene-1,3-disulfona with other antihypertensives and is recom-
mide (salamide) mended as first-line medication (Chobanian
et al., 2003). The Food and Drug Administration
286
Table 1.1 Analytical methods for hydrochlorothiazide
Sample matrix Sample preparation Analytical method Detection limit Reference

Compendial
methods
Assay Potentiometry Indian
Titrant: 0.1 M TBAH Pharmacopoeia
Reference electrode: calomel or silversilver chloride (2010)
Indicator electrode: glass electrode
1 mL of 0.1 M TBAH is equivalent to 0.01488g of HCTZ
Assay for UV-visible spectroscopy wavelength: 273 nm Indian
HCTZ tablets Pharmacopoeia
(2010)
Assay for HPLC column: C18 US
HCTZ tablets Mobile phase: monobasic sodium phosphate and acetonitrile Pharmacopeia
(9:1) (2007)
pH 3.00.1
Flow rate: 2 mL/min
Human plasma Addition of internal standard (clortalidone) HPLCtandem mass spectrometry 5 ng/mL (LLOQ) Sousa et al.
and MTBE, centrifugation, addition of Column: C18 Linearity: (2009)
solvent to organic phase, evaporation and Mobile phase: acetonitrile and water (80:20, v/v) SRM 5400 ng/mL
reconstitution in acetonitrile and water transition: 296.10 m/z, 204.85 m/z
(1:1, v/v)
Human urine Mix urine samples with deionized water, HPLCnarrow bore chromatography 1 g/mL (LOD) Farthing et al.
centrifuge Column: C18 Linearity: (1998)
Mobile phase: acetic acid and acetonitrile (93:7) pH 3 250 g/mL
Wavelength: 272 nm
Human serum Gel filtration of the sera on Sephadex G-15, HPLC 50 ng/mL (LOD) Christophersen
extraction of the protein-free fraction of the Column: C18 (Spherisorb ODS) et al. (1977)
effluent with ethyl acetate Mobile phase: 15% methanol in water
Rat plasma To plasma add internal standard (HFTZ), LCESIMS Linearity: Takubo et al.
extraction with MTBE Column: C8 41000ng/mL (2004)
Mobile phase: distilled water and acetonitrile (85:15) Accuracy:
Negative ionization mode 100.8113.1%
Precision:
0.2816.4%
Hydrochlorothiazide
287
288
Sample matrix Sample preparation Analytical method Detection limit Reference

Human plasma, Derivatization with Reverse phase HPLC 3.3 ng/mL Huang et al.
simultaneous 2,4-dibromoacetophenone (pBPB) to form Column: C18 (LOQ) (2006)
determination captopril-pBPB adduct. Extraction of Mobile phase: acetonitrile, trifluoroacetic acid and water
of HCTZ and HCTZ and derivatized captopril with ether (gradient elution)
captopril and dichloromethane Flow rate:1.2mL/min
Human plasma, Liquidliquid extraction with diethyl ether LCMS Linearity: Yan et al.
simultaneous and dichloromethane (60:40) Column: C8 1.00600ng/mL (2008)
IARC MONOGRAPHS 108
quantitation Mobile phase: acetonitrile, 10 mM ammonium acetate and

of HCTZ and formic acid (gradient elution)
telmisartan Flow rate: 1.2 mL/min
SRM transition: 295.9 m/z 268.9 m/z
Negative ionization mode
Human blood Addition of the internal standard, benzene GLC 0.05g/mL Vandenheuvel
and plasma extraction, extraction with ethyl acetate, Column: glass U-tube (sensitivity) et al. (1975)
back-extraction into NH4OH, adjustment Carrier gas: argon:methane (95:5)
to pH3.7 and extraction with ethyl acetate, Flow rate: 60 mL/min
evaporation, dissolution of the residue in Detector: electron capture detector
trimethylanilinium hydroxide in methanol
In tablets Dissolve drug in 0.02 M NaOH, dilute with Electrochemical study 5.0 ng/mL Abdel Razak
BrittonRobinson buffer pH 3.3 Electrode: glass carbon electrode (LOD) (2004)
pH 3.3
Oxidation potential: + 1040 mV
In urine Centrifugation at 4000 g/mL spiked with Electrochemical study 14 ng/mL Abdel Razak
HCTZ and diluted with BrittonRobinson Electrode: glass carbon electrode (2004)
buffer pH 3.3 pH 3.3
Oxidation potential: + 1040 mV
In Solution of HCTZ in acetone Diffuse reflectance spectroscopy: Whatman 42 filter paper as 1.32 102mol/L Gotardo et al.
pharmaceutical the solid support (LOD) (2005)
formulations Solvents: acetone and methanol (HPLC grade); PDAC used Linearity:
for spot reaction with HCTZ 3.36102 to
1.01101mol/L
GLC, gas-liquid chromatography; HCTZ, hydrochlorothiazide; HFTZ, hydrofluorothiazide; HPLC, high-performance liquid chromatography; LC-MS, liquid chromatography
mass spectroscopy; LCESIMS, liquid chromatographyelectrospray ionizationmass spectrometry; LOD, limit of detection; LLOQ, lower limit of quantification; LOQ, limit of
quantification; MTBE, methyl tert-butyl ether; PDAC, para-dimethylamino cinnamaldehyde; SRM, single reaction monitoring; TBAH, tetrabutylammoniumhydroxide; UV, ultraviolet;
v/v, volume per volume
Hydrochlorothiazide
Table 1.2 Most commonly reported clinical indications for hydrochlorothiazide in the USA, 2012
Diagnosisa ICD-9 code Drug uses (in thousands) Percentage of total

Hypertension, benign 401.10 187 0.8
Surgery after heart disease V67.03 153 0.6
Hypertensive renal disease 403.90 136 0.6
Oedema, NOS 782.30 122 0.5
Cerebrovascular accident 436.00 86 0.4
Metabolic/insulin resistance syndrome 277.70 82 0.3
Total with reported diagnoses 24383 100
No diagnosis was stated for 0.2% of drug uses.
a

has approved hydrochlorothiazide alone, and (b) Dosage

thirty-four combinations containing two agents, Hydrochlorothiazide alone (as sold in the
and four combinations containing three agents, USA) is available in doses of 12.5 mg, 25 mg,
although not all of these forms were currently 50mg, and 100mg. In combination with other
available (FDA, 2013). The most common pharmaceuticals, the dose of hydrochloro-
drugs combined with hydrochlorothiazide are thiazide is generally 12.5 mg or 25 mg (eMC,
triamterene (see Monograph on triamterene in 2013; MicroMedex, 2013).
this volume), lisinopril, losartan, and valsartan When used as a single agent tablet in the
(IMS Health, 2012a). USA in 201112, hydrochlorothiazide was most
The European Medicines Agency indica- frequently used at a dose of 25mg (69%), followed
tion for hydrochlorothiazide is for treatment by 12.5mg (25%) with higher doses used infre-
of hypertension. Labelling includes use for quently (5%). In combination products, a hydro-
hypertension and oedema for combination chlorothiazide dosage of 12.5mg is most common
drugs containing hydrochlorothiazide and (52%), followed by 25 mg (31%) then 37.5 mg
another diuretic agent. Hydrochlorothiazide is (15%). Both alone and in combination products,
a recommended drug in Europe (Mansia et al., once-per-day dosing predominates (>90%) with
2007, 2009). Although hydrochlorothiazide is a twice-per-day and less-than-daily (e.g. every
registered product, it is generally available only second day) dosing being less common. Overall,
in combination products. The European Union the mean daily dosage among patients reported
listed nineteen combination products containing to be taking hydrochlorothiazide is 22 mg per
hydrochlorothiazide and one other drug, and day (IMS Health, 2012a).
two containing two other agents, although not
all are currently in use (eMC, 2013). (c) Trends in use
Thiazide diuretics, including hydrochloro
thiazide, are among the most frequently used
antihypertension agents in the USA and
western Europe, accounting for roughly 30%
289
IARC MONOGRAPHS 108
of medications prescribed to patients with high in Germany were treated with hydrochloro-
blood pressure (Wang et al., 2007). Other data thiazide. This suggested that hydrochloro-
confirmed that 2530% of USA patients with thiazide represented about 25% of all drugs used
hypertension were taking thiazide diuretics in for hypertension in this country. These results
20092010 (Gu et al., 2012). The use of hydrochlo- were very consistent with previous reports that
rothiazide increased modestly in the mid-2000s diuretics represented 39% of all drugs mentioned
(Stafford et al., 2006; Gu et al., 2012) after publi- for German patients with hypertension, and
cation of the results of the Antihypertensive compendium information suggesting that
and Lipid-Lowering Treatment to Prevent Heart hydrochlorothiazide was not generally available
Attack Trial (ALLHAT, 2002; Cushman et al., on its own, but rather as a combination product.]
2002), which showed reasonable equivalence
between chlorthalidone (a thiazide-related
diuretic), amlodipine, and lisinopril. Reported
1.4 Occurrence and exposure
uses of hydrochlorothiazide (alone or in combi- Human exposure to hydrochlorothiazide
nation products) at visits to physicians in the is largely limited to use as a medication. While
USA then decreased, from 28.9 million uses in occupational exposure in manufacturing is
2008 to 24.3 million in 2012. Approximately 10 likely to occur, no specific studies on hydro-
million patients in the USA were reported to be chlorothiazide as an agent in occupational or
exposed to hydrochlorothiazide in 2012 (IMS environmental exposure were identified by the
Health, 2012a). Prescriptions containing hydro- Working Group.
chlorothiazide dispensed in the USA in 2012
amounted to 126.5 million, a slight decrease
from 136.5 million prescriptions in 2008 (IMS 1.5 Regulations and guidelines
Health, 2012b; of these, 48.0 million prescrip- Hydrochlorothiazide has been widely
tions were dispensed for hydrochlorothiazide approved by drug regulatory agencies around the
as a single-agent product (38% of the total for world. In the USA, it was approved by the Food
hydrochlorothiazide). Overall, these two data and Drug Administration (FDA, 2013) in 1959.
sources suggested modest declines in use over The Working Group did not identify extraor-
the past decade. dinary regulatory restrictions on use of hydro-
Total worldwide sales of hydrochlorothiazide chlorothiazide as a medication, or regulations on
in 2012 were US$10.1 billion. Sales were highest in environmental exposure.
the USA at US$3.4 billion, followed by Germany
(US$ 0.9 billion), Japan (US$ 0.7 billion), Italy
(US$0.6 billion), France (US$0.5 billion), Brazil 2. Cancer in Humans
(US$0.4 billion), Spain (US$0.4 billion), India
(US$ 0.15 billion), and China (US$ 0.1 billion).
Hydrochlorothiazide is a diuretic in a class
The generally lower sales outside of the USA
of thiazide compounds primarily used to treat
(even when adjusted for population size) reflect
hypertension, but also oedema and congestive
the greater use of other thiazide and thiazide-
heart failure. In addition to diuretic effects on
type diuretics in other countries (IMS Health,
the kidney, hydrochlorothiazide has photosensi-
2012c).
tizing properties, enhancing skin sensitivity to
[Based on data from Wang et al. (2007) and
sunlight exposure.
Kuehlein et al. (2011), the Working Group calcu-
lated that 44% of patients treated for hypertension
290
Hydrochlorothiazide
Epidemiological studies have investigated one prescription. Limitations further included

associations with use of hydrochlorothiazide the heterogeneity of other skin cancers and
using pharmacy information in pre-paid health inability to examine basal cell and squamous
plans and data from national databases linking cell skin cancers. Also, information was lacking
with physician and/or cancer registry informa- to evaluate the potential confounding or effect
tion. Some studies evaluated thiazides as a class modification by factors related to sun exposure.
of drugs either through prescription records or For cancers of the lip and other sites, the results
self-reported use. The types of cancers investi- referred to hydrochlorothiazide in combination
gated or observed in these studies included those with other drugs. Since computerized pharmacy
related to exposure to sunlight (e.g. lip or cuta- records only began in 1994, only a limited latency
neous malignancies), and those of the kidney period could be observed.]
(e.g. renal cell carcinoma), with a few reports of On the basis of the findings from Friedman
other malignancies (e.g. cancers of the prostate, et al. (2009), a similarly designed nested case
colon, breast, and endometrium). See Table 2.1 control study on cancer of the lip was carried out
and Table2.2. from 1994 to 2008; the study included 712 cases (of
which 97.2% were squamous cell carcinomas of
the lip) and 22904 non-Hispanic, white controls
2.1 Cancers of the lip and skin (Friedman et al., 2012). An increased odds ratio
Using data from 1994 to 2006 from the Kaiser for cancer of the lip was observed for people
Permanente Medical Care Program in northern having three or more prescriptions of any medi-
California, USA, Friedman et al. (2009) screened cine containing hydrochlorothiazide (OR, 2.19;
for drugs potentially related to cancer occur- 95% CI, 1.742.76) or exclusively hydrochloro-
rence by analysing cancer cases and controls thiazide (OR, 2.03; 95% CI, 1.233.36) at least
matched on age (same year), sex, year of starting 2years before the reference date. Odds ratios were
drug coverage, and index date (matched to the higher with longer duration of prescriptions. [A
cases date of diagnosis). The analysis considered strength of this study was the large sample size
the potentially confounding effects of HIV posi- that allowed them to look at hydrochlorothiazide
tivity, and other medical conditions. Elevated alone; however, there were no data on potentially
risks were observed for cancers of the lip (odds modifying or confounding factors, including
ratio, OR, 2.29; 95% CI, 1.842.86) and all other factors related to sun exposure. The definition of
types of cancer of the skin combined, including cancer of the lip presumably used the definition
Merkel cell, malignant fibrous histiocytoma, provided by the Surveillance, Epidemiology, and
dermatofibrosarcoma, skin appendage carci- End Results Program (SEER) of the National
noma and other rarer types (OR, 1.56; 95% CI, Cancer Institute, USA.]
1.202.01) in relation to three or more prescrip- In a population-based nested casecontrol
tions of hydrochlorothiazide at least 2 years study from North Jutland, Denmark, diagnoses
before diagnosis. No association was observed of melanoma, squamous cell carcinoma, and
with cutaneous melanoma or other cancers. basal cell carcinoma from 1989 to 2003 were
[Many comparisons were made. Risk estimates identified through the Danish cancer registry,
were only provided if they indicated an associa- which includes non-melanoma as well as mela-
tion with an odds ratio >1.50 with three or more noma of the skin (Jensen et al., 2008). Four
prescription with a 2-year lag, P<0.01 for differ- controls per case were selected from the Danish
ence from odds ratio 1.00, or a higher odds ratio Civil Registration System, matched on age
for three or more prescriptions compared with (exact), sex, and area of residence. Prescriptions
291
292
Table 2.1 Casecontrol studies of thiazides (including hydrochlorothiazide) and cancer
Reference Total cases Control Exposure Organ site Exposure categories Exposed Relative risk Covariates
Location; Total source assessment (ICD code) cases (95% CI) Comments
period controls (hospital,
population)
Friedman NR KPMCP KPMCP HCTZ use, including Controls matched on sex, year
et al. (2009) subscribers computerized combinations, 3 of birth, and year of starting
Northern prescription dispensings, 2-yr laga drug coverage
California, NR records Kidney (renal No 1.00 (ref.)
USA; pelvis)
IARC MONOGRAPHS 108
Yes 537 1.71 (1.541.91) Nested casecontrol study

August of 10 controls to each case
1994 identified by the KPMCP
December cancer registry
2006
Lip No 1.00 (ref.) Uncertainty as to whether
the association with
kidney cancer was due to
hypertension or the drug
Yes 147 2.29 (1.842.86) Lung cancer only weakly
related to lip cancer, making
confounding by smoking less
likely to be an explanation
Skin (other No 1.00 (ref.) Other types of cancer of the
than lip) skin included: 35 Merkel cell
carcinoma, 14 malignant
fibrous histiocytoma, 8
dermatofibrosarcoma, 7 skin
appendage carcinoma, 5 or
fewer of 14 other rarer types
Yes 95 1.56 (1.202.01)
population)
Friedman 712 Subscribers KPMCP Lip HCTZ use, including Matched on age, sex, year of
et al. (2012) 22904 to KPMCP computerized combinations, 3 cohort entry. Adjusted for
Northern controls prescription dispensings, 2-yr laga cigarette smoking
California, records HCTZ prescriptions b 103 Non-Hispanic whites
USA; No HCTZ 1.00 (ref.) only; age, 30 yr; excludes
19942008 prescriptions transplant recipients
and HIV-positive; 3:1,
men:women
3 HCTZ 2.19 (1.742.76) No stratum-specific numbers
prescriptions provided, just the overall
HCTZ prescriptions 19 number of cases prescribed
onlya HCTZ alone, only and with
No HCTZ 1.0 (ref.) triamterene
prescriptions
3 HCTZ 2.03 (1.233.36)
prescriptions
Years of supply:
HCTZ:b 103
<1yr 0.98 (0.661.45)
1 to <5yr 2.03 (1.542.68)
5yr 4.22 (2.826.31)
HCTZ-triamterene:b 71
<1yr 0.91 (0.601.39)
1 to <5yr 1.87 (1.372.57)
5yr 2.82 (1.744.55)
Hydrochlorothiazide
293
294
population)
Jensen et al. 5964 BCC, Population North Jutland Skin (BCC, HCTZ prescriptions Chronic medical conditions,
(2008) 1129 SCC, Countys SCC, MM) before diagnosis previous use of oral
North 1151 MM prescription MM: glucocorticoids, prescriptions
Jutland 32412 database None 1.00 (ref.) for other photosensitizing
County, controls diuretics
IARC MONOGRAPHS 108
Any 98 1.32 (1.031.70)

Denmark; Sparse data for HCTZ only
19892003 Prescriptions >1 yr 1.30 (0.991.71) (n=13 cases); no association
since diagnosis with MM in this limited
Prescriptions >5 yr 1.24 (0.861.78) group
since diagnosis
Linear per 10000 mg 0.99 (0.951.03)
SCC:
None 1.00 (ref.) Sparse data for HCTZ
Any 159 1.58 (1.291.93) only (n=5 cases); no clear
Prescriptions >1 yr 1.67 (1.362.07) association with SCC in this
since diagnosis limited group
Prescriptions >5 yr 1.92 (1.462.54)
since diagnosis
Linear per 10000 mg 1.03 (1.011.06)
BCC:
None 1.00 (ref.) Wider CIs for HCTZ only
Any 542 1.05 (0.951.16) (n=84 cases); no clear
Prescriptions >1 yr 1.05 (0.941.17) association with BCC
since diagnosis
Prescriptions >5 yr 1.10 (0.951.26)
since diagnosis
Linear per 10000 mg 1.02 (1.001.03)
population)
Robinson 1637 SCC, Population Personal Skin (BCC, Thiazides (diuretics) 239 Age, sex, number of previous
et al. (2013) 1605 BCC interview SCC) episodes of painful sunburn
New 1906 OR for photosensitizing
Hampshire, controls cardiovascular drugs (mainly
USA; thiazides), 1.3 (95% CI,
19932009 1.01.6)
SCC 1.0 (ref.) OR restricted to HCTZ was
1.3 (0.72.4) similar
BCC No association
de Vries 1371 (602 Hospital Questionnaire Skin (BCC, Thiazide Age, sex, phototype, and
et al. (2012) BCC; 409 (same SCC, CMM) (bendroflumethiazide) country
Multicentre SCC; 360 dermatology CMM 33 1.22 (0.771.93) Study period not reported.
(Finland, CMM) outpatient BCC 94 1.27 (0.921.75) Age- and sex-matched
Germany, 1550 clinics and controls
SCC 99 1.66 (1.162.37)
Greece, controls hospital
Italy, Malta, departments
Poland, as the cases)
Scotland
and Spain)
Hiatt et al. 257 (167 Subscribers KPMCP Kidney Thiazide, ever-use History of smoking, BMI,
(1994) men, 90 to KPMCP prescription (renal cell Men: hypertension, history of
Northern women) records and carcinoma) No 1.0 (ref.) kidney infection at check-up
California, 257 chart review ORs were highest for the
Yes 1.2 (0.62.1)
USA; controls category of longest time since
196489 (167 men, Women: first use, duration of use,
90 women) No 1.0 (ref.) number of mentions, and
Yes 4.0 (1.510.8) grams, but trends were not
statistically significant
Stanford 1362 Population Home Breast Thiazide, ever-use 167 1.22 (0.91.6) Age at diagnosis
et al. (1986) interview Exposure information for
USA antihypertensive and oedema
(BCDDP); medications was truncated at
19737 the time of diagnosis for cases
or at an equivalent time for
controls
Hydrochlorothiazide
295
296
population)
Li et al. 975 Population In-person Breast Thiazides Age
(2003) 1007 (CMS list) interview Ever-use 246 1.4 (1.11.8) Of 1210 eligible cases
Washington controls 6 mo to 5 yr 81 1.5 (1.12.2) identified, 975 women (80.6%)
state, USA; were interviewed (14%
> 5 yr 159 1.3 (1.01.7)
April 1997 refused, 4% died, 1% moved
IARC MONOGRAPHS 108
May 1999 away, 1% refused contact with

patients). Of the 1365 eligible
women who were selected as
controls, 1007 women (73.8%)
were interviewed (22%
refused, 2% died, 2% moved
away, 1% not located)
Fortuny 469 Population Personal Endometrium Thiazides Adjusted for age (linear
et al. (2009) 467 interview No 369 1.0 (ref.) spline), BMI (four categories),
New Jersey, controls Yes 100 1.8 (1.13.0) demographic factors
USA; (education, race), other
Duration of use:
20015 estrogen-related variables
<3yr 28 1.6 (0.83.5) (menarche, hormone therapy,
36yr 23 1.2 (0.62.8) oral contraceptives, age at
>6yr 48 2.4 (1.24.8) menopause, parity), smoking,
P (trend test) 0.39 family history of endometrial
cancer, plus all other variables
included in table
Three methods were used to
locate controls (random-digit
dialling for women aged <65
yr, CMS lists for women aged
65 yr, area sampling for
women aged 55 yr)
population)
Hallas et al. 149417 Population Prescription Any cancer Thiazides Prior discharge diagnosis of
(2012) 597668 (Danish records All malignancies 11509 1.25 (1.221.28) COPD or inflammatory bowel
Denmark; controls cancer (database of disease, modified Charlson
20005 registry, the Danish index that contains 19
Danish Medicines categories of comorbidity
national Agency) Four controls matched by age
registry of and sex were selected for each
patients, case by a risk-set sampling.
prescription For all drug classes, those
database of exposed who had taken at
the Danish least 1000 defined daily doses
Medicines during the past 5yr before the
Agency, index date were considered
and Danish
person
registry)
a
Users of other drugs excluded
b
All use of drug, regardless of other drugs dispensed
BCDDP, Breast Cancer Detection Demonstration Project; BCC, basal cell carcinoma; BMI, body mass index; CI, confidence interval; COPD, chronic obstructive pulmonary disease;
CMM, cutaneous malignant melanoma; CMS, Centers for Medicare & Medicaid Services; HCTZ, hydrochlorothiazide; KPMCP, Kaiser Permanente Medical Care Program; mo, month;
MM, malignant melanoma; NR, not reported; ref., reference; RR, relative risk; SCC, squamous cell carcinoma; yr, year
Hydrochlorothiazide
297
298
Table 2.2 Cohort studies of thiazides (including hydrochlorothiazide) and cancer
Reference Total No. of Exposure Organ site (ICD Exposure No. of Relative risk Covariates
Location; subjects assessment code) categories cases/ (95% CI) Comments
period deaths
cited in NR KPMCP Kidney Thiazide use SMRa Limited details on study design.
Friedman et al. computerized Any use 55 1.36 (1.031.77) Limited data on potential
(2009) prescription confounding factors
USA; 19692002 records
Ruiter et al. 10692 with no Computerized Skin (BCC) Thiazide use Hazard ratio Age, sex, smoking, tendency to
(2010) prescriptions for pharmacy sunburn, residence in a country
IARC MONOGRAPHS 108
Never used 385 1.0 (ref.)

The diuretics before records Ever used 137 1.0 (0.951.05) with high ambient UV radiation,
Netherlands; April 1991 hair and eye colour, other
19862007 Days of use: photosensitizing drugs
<94 34 1.02 (0.721.45) Cases identified from general
94524 35 0.98 (0.691.39) practitioners and linkage with
5241646 34 0.86 (0.601.22) national cancer registry. Unclear
>1646 34 1.10 (0.771.58) how many BCC cases may have
been missed
Flaherty et al. 118191 women Self-report in Kidney (renal Thiazide use Limited exposure information,
(2005) and 48953 men 1980, then every cell carcinoma) hypertension an independent risk
United States 2yr factor
Nurses Health Current use in 22 1.5 (1.02.4) Age-adjusted
Study, USA; 1980, women 22 1.4 (0.92.3) Age, BMI, hypertension
19802000; and
Current use in 20 1.5 (0.92.5) Age-adjusted
HPFS, USA;
1986, men 20 0.8 (0.51.5) Age, BMI, hypertension
1998
Friedman & Ury 143574 KPMCP All cancers, Thiazides SMR Age-, sex-standardized
(1980) outpatients computerized prostate Prostate cancer 53 1.4 SMR based on expected rates
Northern with at least one prescription P for trend <0.05 from the 3rd National Cancer
California, USA; prescription records Survey in the San Francisco Bay
All cancers 585 1.1
196976 filled in 196973 Area. Drugs not specifically
P for trend NS HCTZ
van den Eeden & 143574 KPMCP All cancers (56 Thiazides 1464 SMR 1.07 (1984)
Friedman (1995) prescription cancer sites) SMR 1.02 (1988)
San Francisco, records Gall bladder 16 SMR 1.8
USA; 196988 (P<0.05)
Reference Total No. of Exposure Organ site (ICD Exposure No. of Relative risk Covariates
Location; subjects assessment code) categories cases/ (95% CI) Comments
period deaths
Tenenbaum et 14166 patients Derived All cancers, All cancers: Hazard ratio Significant covariates were
al. (2001) with previous from intake colon No diuretics 622 1.00 (ref.) included: age, sex, smoking, and
Tel Aviv, Israel; myocardial examination HCTZ 29 1.41 (0.972.05) triglycerides
19926 infarction and/ [unclear from No information available on
HCTZ/ 16 1.45 (0.882.38)
or stable angina paper] duration or doses of drugs
Amiloride
syndrome, administered
screened for Colon cancer: Hazard ratio
participation in No diuretics 73 1.00 (ref.)
BIP study (2153 HCTZ 5 2.12 (0.855.26)
on diuretics; Amiloride/ 4 3.15 (1.158.65)
375 on HCTZ HCTZ
alone; 199 on
amiloride/
HCTZ
combination)
a
HCTZ not distinguished from other thiazides in previous screenings in a smaller cohort, but elevated risk of kidney cancer detected for thiazides as a group.
BCC, basal cell carcinoma; BIP, Bezafibrate Infarction Prevention; HCTZ, hydrochlorothiazide; HPFS, Health Professionals Follow-up Study; KPMCP, Kaiser Permanente Medical Care
Program; NR, not reported; NS, not significant; ref., reference; SMR, standardized morbidity ratio; UV, ultraviolet; yr, year
Hydrochlorothiazide
299
IARC MONOGRAPHS 108
for diuretics were obtained from a database of uncertain whether complete ascertainment of
ambulatory patient prescriptions that began in basal cell carcinoma was achieved.]
1989 and had full coverage of the population A population-based casecontrol study of
in 1991. Taking into account history of chronic cases of basal cell carcinoma (n = 1605) and
disease, prescriptions for oral glucocorticoids squamous cell carcinoma (n=1637) and controls
and other photosensitizing diuretics, elevated that were frequency-matched on age and sex
odds ratios were found for squamous cell carci- (n=1906) from New Hampshire, USA, included
noma (OR, 1.58; 95% CI, 1.291.93) and mela- self-reported use of photosensitizing drugs such
noma (OR, 1.32; 95% CI, 1.031.70) in relation as hydrochlorothiazide (Robinson et al., 2013).
to prescriptions for hydrochlorothiazide. A weak Elevated odds ratios for squamous cell carci-
association was observed for basal cell carcinoma noma of the skin were observed for reported use
(OR, 1.05; 95% CI, 0.951.16). For squamous cell of photosensitizing cardiovascular drugs (the
carcinoma, the odds ratios increased linearly majority of which were thiazides) (OR, 1.3; 95%
with increasing total dose of prescriptions; the CI, 1.01.6), and specifically for thiazides (OR,
dose trend was weak for basal cell carcinoma, 1.3; 95% CI, 0.72.4); the manuscript indicated
and was not present for melanoma. Additionally, that the odds ratio for hydrochlorothiazide
for squamous cell carcinoma, the association was similar [which was not surprising since
was stronger, with a longer lag period from hydrochlorothiazide comprised the majority of
time of prescription to diagnosis. A sensitivity the thiazides sold in the USA]. Multiple poten-
analysis indicated that underascertainment of tially confounding factors were considered in
skin-cancer diagnosis could have led to an under- the analysis, including risk factors related to sun
estimate of risk. [A limitation of this study was exposure, and history of cigarette smoking. No
that it relied on data from medical records and associations were observed with use of thiazides
on prescriptions, and thus was not able to assess and basal cell carcinoma. [Limitations of this
the potential confounding or modifying effects study were the reliance on self-reported drug
of factors related to sun exposure. Additionally, use information, lack of statistical power to
hydrochlorothiazide was frequently given with evaluate effect modification by sunlight-related
amiloride, and there were too few subjects to factors, and failure to report the risk estimate for
evaluate the effects of therapy with hydrochlo- hydrochlorothiazide.]
rothiazide only.] A multicentre, hospital-based study of cancer
In a prospective cohort study from the of the skin carried out in Europe (Finland,
Netherlands, Ruiter et al. (2010) used comput- Germany, Greece, Italy, Malta, Poland, Scotland,
erized outpatient pharmacy records to assess and Spain) examined thiazide use determined
the relation between thiazide use and basal cell through questionnaires completed partly by
carcinoma. The analysis included 10 692 indi- the participant and partly by a dermatology
viduals, largely Caucasian, with no diuretic department. The study included 409 cases of
prescriptions before April 1991. Basal cell carci- squamous cell carcinoma, 602 cases of basal
nomas were identified by general practitioners cell carcinoma, 360 cases of cutaneous malig-
and linkage with the national cancer registry nant melanoma, and 1550 controls. Cases were
for 1986 to 2007. After adjustment for multiple consecutive patients, recently diagnosed (within
potentially confounding factors, no excess risk 3months of study entry), aged 18 years or older,
of basal cell carcinoma was observed with either from one of the participating dermatology prac-
cumulative duration of use or average grams of tices. Controls were patients visiting the hospital
thiazide prescription (Ruiter et al., 2010). [It was clinics for reasons other than cancer of the skin
300
Hydrochlorothiazide
and were frequency-matched to cases on age (as could be attributed to hydrochlorothiazide use
far as possible in 5-year strata) and sex. Cases or to hypertension, since use of other cardio-
and controls were excluded if they were unable vascular drugs, e.g. Clonidine, Diltiazem, and
to complete the questionnaire, did not agree to Gemfibrozil, were also related to risk of renal cell
take part, had Fitzpatrick skin types VVI, had cancer.]
ever received phototherapy, or had photoaller- Also using records from the Kaiser
gies or lupus erythematosus. A minimum of Permanente Medical Care Program, Hiatt et al.
3months, daily intake of self-reported thiazide (1994) conducted a nested casecontrol study of
(bendroflumethiazide) was associated with an 257 cases of renal cell carcinoma and one-to-one
odds ratio of 1.66 (95% CI, 1.162.37) for squa- matched controls. Cases included members
mous cell carcinoma, an odds ratio of 1.27 (95% enrolled in the programme with documented
CI, 0.921.75) for basal cell carcinoma, and an renal cell carcinoma diagnosed between 1964
odds ratio of 1.22 (95% CI, 0.771.93) for cuta- and 1989 and who had received a standardized
neous melanoma (de Vries et al., 2012). multiphasic health check-up (offered from 1964
to 1988) before diagnosis. Controls who had also
undergone the multiphasic health check-up were
2.2 Renal cell carcinoma matched to cases on the age ( 1 year) when
A standardized morbidity ratio (SMR) anal- they had the check-up, and were required to be
ysis was conducted of outpatients enrolled in enrolled in the Kaiser Permanente Medical Care
the Kaiser Permanente Medical Care Program Program when their case was diagnosed. Data
with at least one prescription for thiazide filled on thiazide use up to 6months before diagnosis
between 19691973 who were followed for cancer (and a matched date for controls), were abstracted
until 2002 A total of 55 observed versus 40.34 from the medical records. An association was
expected cancers of the kidney were observed found between thiazide use and renal cell carci-
among those with thiazide prescriptions (SMR, noma among women (OR, 4.0; 95% CI, 1.510.8)
1.36; 95% CI, 1.031.77) (Friedman et al., 2009). but not men (OR, 1.2; 95% CI, 0.62.1) adjusted
[Limited details on the study design were for multiple potentially confounding factors,
provided, and there were limited data avail- including hypertension (Hiatt et al., 1994). Odds
able on potentially confounding or modifying ratios did not increase with estimated number of
factors.] grams used (based on time since first use, dura-
In the nested casecontrol analyses of tion, and number of mentions of use in the chart).
data from the Kaiser Permanente Medical [Data on potentially confounding or modifying
Care Program from 19942006 conducted by factors were available through the multiphasic
Friedman et al. (2009), an increased risk of cancer check-up.]
of the renal pelvis (OR, 1.71; 95% CI, 1.541.91) A cohort analysis of renal cell carcinoma was
was observed in association with three or more conducted in the United States Nurses Health
prescriptions of hydrochlorothiazide at least Study and Health Professionals Follow-up
2 years before diagnosis. [This observation was Study (HPFS) of 118 191 women and 48 953
one of many comparisons. The study reported men without a history of cancer (Flaherty et al.,
limited data on potentially confounding factors, 2005). Renal cell carcinomas self-reported (up to
i.e. those not available in medical records, e.g. 2000 in the Nurses Health Study, and 1998 in
cigarette smoking history. While hypertension the HPFS) were confirmed by medical record in
was evaluated in the analysis, it was uncertain more than 80% of cases, and the 156 women and
whether findings for cancer of the renal pelvis 110 men with histologically verified (via biopsy,
301
IARC MONOGRAPHS 108
nephrectomy or autopsy) renal cell carcinoma were related to elevated standardized morbidity
were included in the analysis. Thiazide use was ratio for cancer of the gall bladder, raising the
based on self-report (beginning in 1980 for the possibility the association was due to the indica-
Nurses Health Study, and 1986 for the HPFS) tion rather the specific drug.]
and was updated every 24years. For women, the Two casecontrol studies of cancer of the
age-adjusted relative risk estimate was 1.5 (95% breast evaluated self-reported use of thiazides. A
CI, 1.02.4), and after adjustment for history of casecontrol analysis of thiazides was conducted
hypertension, and updated body mass index was in the USA within the Breast Cancer Detection
1.4 (95% CI, 0.92.3); among men, the relative Demonstration Project, a multicentre breast-
risks were 1.5 (95% CI, 0.92.5), and 0.8 (95% CI, screening trial involving in-person interviews
0.51.5), respectively. The result was not altered with women with cancer of the breast (diag-
when using updated information on thiazide use. nosed between 1973 and 1977) and controls
[There was limited information on thiazide use (neither recommended for a biopsy nor had a
derived from a postal questionnaire. As in other biopsy during participation in the programme)
studies, it was not possible to exclude the possi- (Stanford et al., 1986). Response rates were 86%
bility of confounding by hypertension, since for cases and 74% for controls. Self-reported use
hypertension is a major indication for hydro- of thiazides for at least 6months compared with
chlorothiazide use. Hypertension was reported women without a history of hypertension was
to be an independent risk factor for renal cell associated with an age-adjusted odds ratio of
carcinoma in this study.] 1.22 (95% CI, 0.91.6). Odds ratios by duration
of use were 1.28 for <5years, 1.50 for 59years,
and 1.33 for 10 years (P for trend, 0.06). For
2.3 Other cancers years since first use were 1.27 for <5years, 1.70
Other cancers were assessed in analyses of for 59years, and 1.06 for 10 years (P for trend,
standardized morbidity ratio using the Kaiser 0.12).
Permanente Medical Care Program prescription A more recent population-based case
pharmacy database and the Kaiser Permanente control study from western Washington State,
cancer registry with the northern California USA, included 975 cases of cancer of the breast
(USA) cancer registry as the referent population. in women aged 6579 years diagnosed between
Among 143 574 outpatients, with at least one 1997 and 1999, and identified through the cancer
prescription filled between 19691973 who were registry for the region (Li et al., 2003). Controls
followed for cancer until 1976 using hospital-dis- (n = 1007) were identified through Center for
charge records for the programme and the cancer Medicare and Medicaid services, and cases were
registry, Friedman & Ury (1980) found an elevated limited to those who were registered in this
age- and sex-standardized morbidity ratio for system. Response rates were 81% of cases and
cancer of the prostate (SMR, 1.4; P<0.05). In a 74% of controls. In-person interviews encom-
subsequent analysis, with follow-up data until passed a detailed history of cardiovascular
1988, van den Eeden & Friedman (1995) reported medications used, and included duration and
a greater than expected incidence of tumours of dose, using a life-events calendar and photo-
the gall bladder with thiazide use (16 observed, graphs of medicines to enhance recall. Ever-use
8.9 expected; SMR, 1.8; P < 0.05). [The limita- of hydrochlorothiazide was reported by 19% of
tions of these hypothesis-generating analyses are controls. The odds ratio for cancer of the breast
mentioned in Sections 2.1 and 2.2. Additionally, among women who reported use of thiazides for
in the 1995 study, other hypertension drugs also 6 months or more compared with women who
302
Hydrochlorothiazide
had never used any antihypertension medica- of thiazides for at least 6 months was 1.8 (95%
tion was 1.4 (95% CI, 1.11.8), and 1.5 (95% CI, CI, 1.13.0) [compared with those with no use or
1.1.2.2) for 6months to 5years of use, and 1.3 use for less than 6months]. Odds ratios were 1.6
(95% CI, 1.01.7) for > 5 years of use. Multiple (95% CI, 0.83.5), 1.2 (95% CI, 0.62.8), 2.4 (95%
potentially confounding factors were taken into CI, 1.24.8) for < 3, 36, and > 6 years of use,
consideration, including race, income, marital respectively (P for trend, 0.39). [Low response
status, education, age at menarche, parity, age at rates could have introduced selection bias.]
first birth, type of menopause, age at menopause, A cohort analysis was conducted of 14 166
duration of oral contraceptive use, ever-use individuals recruited between 1990 and 1992
of hormone replacement therapy, first-degree to participate in the Bezafibrate Infarction
family history of breast carcinoma, smoking Prevention study in Israel (Tenenbaum et al.,
status, average daily intake of alcohol, and body 2001). Participants were followed for incidence
mass index. However, none changed the estimate and mortality from cancer until 1996 using the
by more than 10%, and thus estimates were only Israel population registry and national cancer
adjusted for age. [This study was large and had registry. Participants included those with a
relatively extensive information on exposure. history of heart disease (myocardial infarction,
The use of recall aids to prompt reporting of stable angina syndrome) but without a perma-
drug use was an additional strength. The lack of nent pacemaker implantation, cerebrovas-
clear trends with duration could be due to poorer cular disease, chronic hepatic or renal disease,
recall for use further in the past, or to lack of a peripheral vascular disease, malignant disease,
true association.] estrogen therapy, type 1 diabetes mellitus, or
A population-based casecontrol study of use of lipid-modifying drugs. Incidence of all
invasive epithelial endometrial cancer evaluated cancers was increased among those who used
thiazide use in New Jersey, USA (Fortuny et al., hydrochlorothiazide (hazard ratio, HR, 1.41; 95%
2009). Cases were derived from The Estrogen, CI, 0.972.05) or a combination hydrochloro-
Diet, Genetics, and Endometrial Cancer (EDGE) thiazide/amiloride therapy (HR, 1.45; 95% CI,
study a population-based casecontrol study 0.882.38). An elevated incidence of cancer of
conducted in six counties in northern New the colon was observed among those who used
Jersey. Cases diagnosed with endometrial cancer hydrochlorothiazide (n=5 cases; HR, 2.12; 95%
between 1 July 2001 and 30 June 2005 were CI, 0.855.26) or a combination hydrochloro-
identified by the cancer registries for the region thiazide/amiloride therapy (n=4 cases; HR, 3.15;
and state. To be eligible, cases were required to 95% CI, 1.158.65). Age, sex, smoking status, and
be aged 21 years and over and residing in one triglycerides were included in the models if they
of the six counties (Bergen, Essex, Hudson, were statistically significant using stepwise Cox
Middlesex, Morris, and Union). Controls, models. [No other specific cancer types were
without a history of hysterectomy, were iden- mentioned in relation to hydrochlorothiazide.
tified through random-digit dialling (age, < 65 The limitations of this study included the small
years) and Centers for Medicare and Medicaid number of cancers of the colon. There was no
records. Response rates were 30% for the cases information about use, i.e. dose or duration. The
and 39% for the controls. After adjustment precise method of ascertaining medication use
for multiple potentially confounding factors, was not explained and was assumed to be derived
including age, sex, level of education, smoking, from the intake examination. Given that the
body mass index, hypertension, diabetes, and cohort was part of a clinical trial, the results may
use of other drugs, the overall odds ratio for use not be generalizable.]
303
IARC MONOGRAPHS 108
A population-based casecontrol study was adenoma or carcinoma (combined) were seen in

conducted on cancers from 2000 to 2005 using males. No significant increases in the incidence
the Danish cancer registry, and drug use was of any neoplasms were seen in females. [Although
estimated from the Prescription Database of the significant increases in the incidences of hepato-
Danish Medicines Agency (Hallas et al., 2012). cellular adenoma, and hepatocellular adenoma
A thiazide user was defined as a person having or carcinoma (combined) were observed in male
taken 1000 defined daily doses of the drug, and mice exposed to hydrochlorothiazide, these
the comparison group was never-users. The odds findings may have been the result of an unusu-
ratio for all cancers combined was 1.25 (95% CI, ally low incidence of hepatocellular neoplasms
1.221.28). [This was a large study with no details in the dietary control group in this study (15%)
on the types of cancer among subjects exposed compared with those seen in control groups
to thiazides. The association with all cancers (30%) from the National Toxicology Program
combined could be due to confounding by indi- (NTP) historical database. The Working Group
cation as other antihypertension drugs also were noted that no data on historical controls were
related to overall cancer incidence.] available from the laboratory where this agent
was tested.]
3. Cancer in Experimental Animals 3.1.2 Rat

In a feeding study, groups of 24 male and
3.1 Oral administration 24female rats (age, 68weeks) were given diets
containing hydrochlorothiazide [USP grade] at
See Table3.1
a dietary concentration of 0 (control), or 0.1%
(1000 ppm) for 104 weeks (Lijinsky & Reuber,
3.1.1 Mouse 1987). The rats were observed until natural death
In a feeding study, groups of 50 male and or moribundity occurred, or the end of the study
50female mice (age, 78weeks) were given diets at 130 weeks. Although the total observation
containing hydrochlorothiazide [USP grade] at period was not specified, median times of death
dietary concentrations of 0 (control), 2500, or in male and female rats exposed to hydrochloro-
5000ppm and held until death or completion of thiazide were 111 and 114 weeks, respectively.
the 103104-week exposure period (NTP, 1989; Median times of death in untreated controls were
Bucher et al., 1990). No changes in survival, 107 weeks in males and 122 weeks in females.
changes in group mean body weight, or gross Dietary administration of hydrochlorothiazide
clinical evidence of toxicity were identified in at 1000ppm was well tolerated. Although body
male or females exposed to hydrochlorothiazide. weights were not reported, a high incidence of
A significant increase in the incidences of hepa- chronic progressive nephropathy was seen in
tocellular adenoma, and hepatocellular adenoma male and female rats exposed to hydrochloro-
or carcinoma (combined) was seen in males at thiazide; this finding suggested that the dose
the highest dose; an increase in the incidence of of 1000 ppm used in the study approached the
hepatocellular carcinoma in mice at the highest maximum tolerated dose for this agent. No
dose (9 out of 50 versus 4 out of 48) was not statis- statistical analyses of the incidences of preneo-
tically significant (P=0.161). Statistically signif- plastic or neoplastic lesions were reported. The
icant, dose-related increases in the incidence incidence of adrenal pheochromocytoma was
of hepatocellular adenoma, and hepatocellular increased [P<0.005] from 0 out of 24 in control
304
Table 3.1 Studies of carcinogenicity in mice and rats given diets containing hydrochlorothiazide
Species, strain Dosing regimen, Incidence of tumours Significance Comments

Duration
Reference
Mouse, B6C3F1 Diets containing hydrochlorothiazide Hepatocellular adenoma: *P<0.01 Purity, USP grade
(M,F) at 0 (control), 2500, or 5000ppm 3/48 (6%)*, 8/49 (16%), (trend) No historical control data from the laboratory where
103104wk 50M and 50F/group (age, 78-wk) 14/50 (28%)** (M) **P0.012 this agent was tested
NTP (1989), Hepatocellular carcinoma: Unusually low incidence of hepatocellular neoplasms
Bucher et al. 4/48 (8%), 4/49 (8%), 9/50 in the male control group (15%) compared with control
(1990) (18%) (M) groups from the NTP historical database (30%)
Hepatocellular adenoma or No significant increases in the incidence of any
carcinoma (combined): neoplasm were reported in females
7/48 (15%)*, 10/49 (20%),
21/50 (42%)** (M)
Rat, F344 (M,F) Diets containing hydrochlorothiazide Adrenal *[P<0.005] Purity, USP grade
130wk at 0 (control), or 1000ppm for 104wk. pheochromocytoma: No statistical analyses were reported
Lijinsky & Reuber Rats were held untreated for up to an 6/24, 9/24 (M); 0/24, 9/24*
(1987) additional 26wk (F)
24M and 24F/group (age, 68 wk)
Rat, F344 (M,F) Diets containing hydrochlorothiazide Purity, USP grade
105106wk at 0 (control), 250, 500, or 2000ppm Decreased body weight in all exposed groups may have
NTP (1989), 50M and 50F/group (age, 78-wk) reduced sensitivity to neoplastic development
Bucher et al. No significant increases in the incidence of any
(1990) neoplasm were reported in either sex
F, female, M, male; USP, United States Pharmacopeia; wk, week
Hydrochlorothiazide
305
IARC MONOGRAPHS 108
females to 9 out of 24 in exposed females; the 3.2 Coexposure with modifying

incidences of adrenal pheochromocytoma in agents
males were 6 out of 24 and 9 out of 24 in control
and treated animals, respectively. [The Working In experimental groups that were compo-
Group noted the unusually high incidence of nents of the study described above, Lijinsky &
adrenal pheochromocytoma in males in the Reuber (1987) exposed groups of 24 male and
control group.] Exposure to hydrochlorothiazide 24 female F344 rats to diets containing hydro-
also induced significant increases [P < 0.001] chlorothiazide at 0 ppm (control) or 1000 ppm
in the incidence of parathyroid hyperplasia in in combination with sodium nitrite at 2000ppm
males and females. for 104weeks, and maintained until 130weeks.
In a feeding study, groups of 50 male and The study was designed to determine whether
50female rats (age, 78weeks) were given diets exposure to hydrochlorothiazide associated with
containing hydrochlorothiazide [USP grade] at sodium nitrite could result in the formation
dietary concentrations of 0 (control), 250, 500, of carcinogenic N-nitroso compounds in the
or 2000 ppm, and maintained until death or stomach. No significant increases in neoplastic
completion of the 105106-week exposure period or preneoplastic lesions were reported. [No statis-
(NTP, 1989; Bucher et al. 1990). No gross clinical tical analyses of the incidence of preneoplastic or
evidence of agent toxicity was identified in either neoplastic lesions were reported.]
male or female rats exposed to hydrochloro-
thiazide. Survival curves were very similar in
all groups of males. Although an apparent trend 4. Mechanistic and Other
towards early mortality was seen in females Relevant Data
exposed to hydrochlorothiazide, the differences
in survival curves were not statistically signif-
icant. Mean body weights in all groups of rats 4.1 Absorption, distribution,
exposed to hydrochlorothiazide were below metabolism, and excretion
those in the control groups; at study termina-
tion, body weight suppression in both sexes was 4.1.1 Humans
greater than 15%. This reduction in mean body (a) Pharmacokinetics of single doses
weight was apparently secondary to suppression
Pharmacokinetic studies using [14C]hydro-
of food intake, but was not clearly dose-related.
chlorothiazide given orally (n = 4) or intrave-
No significant increases in the incidence of any
nously (n=2) to healthy subjects have shown that
neoplasms were reported in all groups of exposed
hydrochlorothiazide is absorbed mainly (~70%)
rats. [The Working Group noted that decreased
via the duodenum and upper jejunum, and only
body weight in all groups exposed to hydro-
to a minor extent via the stomach (Beermann
chlorothiazide may have reduced the sensitivity
et al., 1976).
of these animals to neoplastic development.]
High-performance liquid chromatography
Confirming the results of Lijinsky & Reuber
(HPLC) analysis of plasma and urine samples
(1987), significant increases in the incidence of
from 12 healthy volunteers given single oral
hyperplasia of the parathyroid gland were seen in
doses of hydrochlorothiazide (25, 50, 100, or
all groups of exposed rats of males and females.
200 mg as tablets or suspensions) showed that
plasma profiles (i.e. mean peak plasma concentra-
tions, times of peak concentrations, areas under
306
Hydrochlorothiazide
plasma curves) and recovery of unchanged drug after (rather than before) food (Barbhaiya et al.,
in the urine were linearly related to dose (Patel 1982). [The inconsistency between these studies
et al., 1984). This was consistent with a previous was due to procedural differences, for example,
report from Beermann & Groschinsky-Grind differing doses used, and variation in times
(1977), who used gas-liquid chromatography when food was permitted after dosing. Fasting
(GLC) analyses. Absorption from all doses was and non-fasting subjects were permitted food
rapid; peak plasma concentrations were achieved 4hours after dosing in the later study (Barbhaiya
at approximately 2 hours. These findings were et al., 1982) but not until 10 hours after dosing,
common to both tablet and suspension formula- in the fasted subjects only, in the earlier study
tions (Patel et al., 1984). (Beermann & Groschinsky-Grind, 1978b).] It was
A study in healthy volunteers to evaluate the suggested that prolonged abstinence from food
urinary excretion of hydrochlorothiazide from in the fasted group had altered gastrointestinal
25 mg and 50 mg tablets sourced from seven secretion and motility, affecting drug absorption
different distributors and at least six manufac- (Barbhaiya et al., 1982).
turers demonstrated equivalent bioavailability Hydrochlorothiazide, in all therapeutic
for all products (Meyer et al., 1975). Thus, the doses, is approximately 40% bound to plasma
bioavailability of hydrochlorothiazide did not protein, and accumulates in erythrocytes. The
appear to differ between different oral prepara- ratio of uptake between erythrocytes and plasma
tions (Beermann, 1984). is approximately 3.5:1 (Beermann et al., 1976).
Absorption of hydrochlorothiazide is gener- Equilibrium of hydrochlorothiazide between
ally considered to follow first-order kinetics plasma and erythrocytes is reached 4 hours
(Barbhaiya et al., 1982); however, one study after an oral dose (Beermann et al., 1976). HPLC
of plasma concentrations over the 48 hours analyses in seven healthy volunteers showed
following an oral dose of 100 mg in four healthy that 24 hours after a single oral dose of hydro-
volunteers (Redalieu et al., 1985) suggested zero- chlorothiazide of 100 mg, the concentration of
order absorption. [Zero-order kinetics implies hydrochlorothiazide bound to erythrocytes was
accumulation of excess dissolved drug at the approximately ninefold the concentration in
absorption site, or saturable absorption, neither plasma (Yamazaki et al., 1989).
of which have so far been demonstrated with The study by Beermann et al. (1976) of [14C]
hydrochlorothiazide.] hydrochlorothiazide given as oral (n = 4) or
Reports on the influence of ingested food intravenous (n = 2) doses to healthy subjects
on the absorption of hydrochlorothiazide were demonstrated negligible recovery in the faeces
conflicting. In a study reporting increased and duodenal bile, and showed that the main
absorption in the presence of food (Beermann excretory route of hydrochlorothiazide was via
& Groschinsky-Grind, 1978b), the mean urinary the kidneys. Mean renal clearance was approxi-
recovery of an oral dose of 75 mg of hydrochloro- mately 300 mL/minute. There was no reabsorp-
thiazide in eight subjects, under both fasting and tion (Beermann et al., 1976). Clearance was by
non-fasting conditions, was found by GLC to be glomerular filtration and active secretion via the
47% and 55%, respectively. In contrast, a study organic anion transport system at the proximal
of eight healthy volunteers each given 50 mg tubules (Beermann et al., 1976; Barbhaiya et al.,
of hydrochlorothiazide reported reduction in 1982; Kim et al., 2003).
plasma absorption (modestly affected by varying Elimination of hydrochlorothiazide from
accompanying fluid volumes) and reduction in plasma was biphasic over 2427 hours; plasma
urinary recovery of hydrochlorothiazide taken concentrations fell rapidly over the initial 12
307
IARC MONOGRAPHS 108
hours, and then more slowly (Beermann et al., (ACBS), a hydrolysis product of hydrochloro-
1976; Barbhaiya et al., 1982; Patel et al., 1984). thiazide. Concentrations of this metabolite were
The urinary excretion rate closely resembled this higher (4.3% of hydrochlorothiazide excreted)
time course (Barbhaiya et al. 1982). The plasma in patients urine 24 hours after taking hydro-
elimination half-life was about 6hours initially, chlorothiazide than in the same batch of bulk
but up to 15 hours terminally (Barbhaiya et al., tablets (0.4%), so it was unlikely to be a tablet
1982; Patel et al., 1984). Approximately 70% and contaminant.
90% of the oral and intravenous administered Okuda et al., (1987) also demonstrated that,
doses, respectively, were recovered unchanged in while concentrations of hydrochlorothiazide
the urine of healthy volunteers (Beermann et al., in erythrocytes and plasma peaked at 6 hours
1976), and thus reflected gastrointestinal absorp- after dosing and then slowly declined, the
tion of hydrochlorothiazide. concentrations of ACBS were continuing to rise
at 24 hours. Thus it seems ACBS is formed by
(b) Pharmacokinetics of repeated doses hydrolysis after administration of hydrochloro-
Three patients who had been receiving thiazide, and is excreted more slowly than it is
hydrochlorothiazide only for a minimum of produced. Furthermore, ACBS appears to have
3months to treat hypertension (without cardiac a stronger affinity to erythrocytes than does
failure), were given an oral dose of 50 mg of [14C] hydrochlorothiazide; concentrations of ACBS
hydrochlorothiazide after an overnight fast. Peak and hydrochlorothiazide in erythrocytes were
concentrations occurred at 34hours in plasma approximately equal, but ACBS concentrations
and at 4 to 5 hours in blood cells, and urinary in plasma were approximately 10 times lower
recovery for two of the patients was of the same than those of hydrochlorothiazide (Okuda et al.,
magnitude as that in healthy subjects, but lower 1987).
in one patient (who had decreased renal func- Metabolites of hydrochlorothiazide were
tion) (Beermann et al., 1976). also investigated in urine of six healthy volun-
teers after oral administration of a single tablet
(c) Metabolism containing 25 mg of hydrochlorothiazide and
It is generally considered that since hydro- 25 mg of spironolactone. Hydrochlorothiazide
chlorothiazide is excreted in urine almost and ACBS (at a concentration at least 10 times
entirely as unchanged drug, it is not metabolized higher than the parent drug) and the minor
in humans (Beermann et al., 1976). Radiographic metabolite, chlorothiazide, were detected in the
analysis of urine extracts (n=110) collected from urine by liquid chromatography-mass spectrom-
five healthy subjects and three patients given etry 120 hours after administration (Deventer
[14C]hydrochlorothiazide orally revealed a single et al., 2009).
spot with the same chromatographic properties
(d) Variation in absorption, distribution, and
as hydrochlorothiazide and representing >95%
excretion
of the radiolabel. However, two samples from
one subject collected on the second and third (i) Pregnancy
days after dosing revealed some radiolabelled A study of 10 pregnant women given a daily
material (< 0.5% of total radiolabel excreted) dose of hydrochlorothiazide of 50 mg for at
that did not correspond to hydrochlorothiazide least 2 weeks (to treat oedema and/or hyperten-
(Beermann et al., 1976). The nature of this sion) demonstrated that the diuretic crossed the
material was found by Okuda et al. (1987) to be human placenta, resulting in concentrations in
2-amino-4-chloro-1,3-benzenedisulfonamide
308
Hydrochlorothiazide
the umbilical cord plasma that were similar to administered dose (i.e. approximately half the
those in maternal plasma. In amniotic fluid, the amount normally recovered in the urine).
concentration of hydrochlorothiazide was higher (iv) Cardiac conditions
than that in maternal or umbilical cord plasma
by up to 5 and 19 times, respectively (Beermann The pharmacokinetics of hydrochloro-
et al., 1980). thiazide have been shown to be altered in
A subsequent case report confirmed these patients with congestive heart failure (Beermann
findings, and also reported very low concentra- & Groschinsky-Grind, 1979). GLC analysis of
tions of hydrochlorothiazide in breast milk (rela- plasma and urine of seven patients given oral
tive to maternal blood). Hydrochlorothiazide hydrochlorothiazide (5075 mg) indicated
was not however detectable (detection limit, substantial reduction in the extent and rate of
20 ng/mL) in the blood of the nursing infant absorption of hydrochlorothiazide (recovery of
(Miller et al., 1982). only 2137% of the administered dose in three
patients, and delayed plasma peak in another).
(ii) Impaired renal function The total urinary excretion of hydrochloro-
A study in 23 patients with varying degrees thiazide in two other patients was approximately
of impaired renal function showed reduction 50% of the administered dose. Since the reduced
in the extent and rate of elimination of hydro- intestinal mobility shown in cardiac failure
chlorothiazide; only about 10% of an oral dose would be expected to promote the uptake of
was recovered, and the elimination half-life hydrochlorothiazide, the observed decrease in
was increased from a mean value of 6.4 hours absorption was suggested to be due to changes
in healthy individuals to 11.5hours (in patients in the intestinal wall and/or in blood.
with a mean creatinine clearance of 60 mL/ This study also highlighted a substantial
minute) and to 21 hours (in patients with a mean reduction in renal clearance (range, 10187 mL/
creatinine clearance of 19 mL/minute). Intestinal minute) in these cardiac patients. These patients
absorption was considered not to be reduced in were older than those studied previously (age,
these patients, since the area-under-the-curve 4060 years), and it was considered that reduced
values were greater in those with low creatinine renal function and age may have been factors
clearance than in healthy subjects. In patients in the reduction of absorption (Beermann &
with severe renal impairment, the elimination Groschinsky-Grind, 1979).
half-life was prolonged further to approximately Absorption of hydrochlorothiazide was
34 hours and recovery of hydrochlorothiazide reported to be unchanged in patients with hyper-
was greatly reduced even after extending the tension (Beermann et al., 1976).
collection period (Niemeyer et al., 1983). (v) Racial differences
(iii) Gastrointestinal surgery Racial differences in the pharmacokinetics
Absorption of hydrochlorothiazide has been of hydrochlorothiazide have been investigated in
shown to be impaired in patients who have a matched group of (nine black and nine white)
undergone intestinal-shunt surgery for obesity. hypertensive patients given a single dose of
GLC urine analysis (Backman et al., 1979) for 25mg of hydrochlorothiazide. Analyses of serial
five patients who received an oral dose of hydro- samples of blood and urine collected over 36
chlorothiazide of 775 mg (at times ranging from hours demonstrated that the pharmacokinetics
1.5 to 6 years after surgery) showed that the of hydrochlorothiazide did not differ according
recovery of unchanged drug was only 31% of the to race (Ripley et al., 2000).
309
IARC MONOGRAPHS 108
(vi) Pharmacokinetic and drug interactions (>100 g/mL), the renal excretion changed from
Absorption of an oral dose of 75 mg of net secretion to net reabsorption due to satu-
hydrochlorothiazide given to six healthy volun- ration of the secretory system and substantial
teers was shown to be increased by concomitant passive tubular reabsorption.
administration of propantheline. This effect was
attributed to the reduction in gastrointestinal 4.2 Genetic and related effects
motility caused by this anticholinergic drug
(Beermann & Groschinsky-Grind, 1978a). The 4.2.1 Humans
centrally-acting -adrenergic agonist guanabenz
and the excipient polyvinylpyrrolidone 10 000,
increased the absorption of hydrochlorothiazide.
Cholestyramine and colestipol reduced mean 4.2.2 Experimental systems
peak plasma concentrations of hydrochloro- See Table4.1
thiazide (Welling, 1986).
No significant pharmacokinetic interactions (a) Mutagenicity
have been noted between hydrochlorothiazide The Working Group did not identify any new
and propranolol, metoprolol, sotalol, or acebut- data on the mutagenicity of hydrochlorothiazide
olol. A similar lack of significant interactions has published since the previous IARC Working
been noted between hydrochlorothiazide and Group review (IARC, 1990). In two studies, hydro-
spironolactone and indomethacin, allopurinol chlorothiazide was not mutagenic to Salmonella
and its metabolite, oxipurino1, and phenytoin typhimurium in the presence or absence of an
(Welling, 1986). exogenous metabolic system (Waskell, 1978;
Early studies on hydrochlorothiazide and Andrews et al., 1984). Hydrochlorothiazide was
triamterene as components of a fixed drug not mutagenic in bacterial screening systems,
combination revealed differences in bioavaila- with or without enzymatic activation, but can
bility from combination tablets and capsules, but form chemically reactive mutagenic products
subsequent work suggested that these differences after reaction with nitrite, a product of nitric
may have been due to the effects of formulation oxide (Andrews et al., 1984). [The Working Group
rather than drug interaction (Welling, 1986). noted that only one concentration was used in
both studies.] In a third study, in strain TA98,
4.1.2 Experimental systems but not in TA1535, TA1537, or TA100, a small,
reproducible, concentration-dependent increase
A study of intrahepatic distribution in rats
in the mean number of revertants was observed
demonstrated that, after steady-state intravenous
in the absence but not in the presence of an exoge-
infusion, hydrochlorothiazide distributes homo-
nous metabolic system (Mortelmans et al., 1986).
geneously but not instantaneously throughout
Hydrochlorothiazide did not induce reversion in
the liver. This was proposed to be because hydro-
an arg strain of Escherichia coli (Hs30R) (Fujita,
chlorothiazide does not undergo metabolism in
1985). At concentrations greater than 500 g/mL,
the liver (AbdelHameed et al., 1993).
hydrochlorothiazide produced cytotoxic effects
In a study in isolated perfused rat kidney,
and induced mutations resulting in resistance to
Masereeuw et al. (1997) demonstrated that renal
trifluorothymidine in L5178Y mouse lymphoma
clearance of hydrochlorothiazide exceeded clear-
cells in the absence of an exogenous metabolic
ance by glomerular filtration only at low perfu-
system (NTP, 1989).
sate concentrations. At higher concentrations
310
Hydrochlorothiazide
Table 4.1 Genetic and related effects of hydrochlorothiazide
Test system Resultsa Concentration or dose Reference

(LED or HID)
Without With
exogenous exogenous
metabolic metabolic
system system
In vitro
Salmonella typhimurium TA1535, reverse 10000g/plate Mortelmans et al. (1986)
mutation assay
Salmonella typhimurium TA1535, reverse 1000g/plate Andrews et al. (1984)
mutation assay
mutation assay
Salmonella typhimurium TA100, reverse 5000g/plate Waskell (1978)
mutation assay
mutation assay
mutation assay
mutation assay
Salmonella typhimurium TA98, reverse + 3333g/plate Mortelmans et al. (1986)
mutation assay
Salmonella typhimurium TA98, reverse 5000g/plate Waskell (1978)
mutation assay
mutation assay
Escherichia coli (Hs30R) reversion assay in an NT 0.26mM Fujita (1985)
arg strain
Nondisjunction in Aspergillus nidulans spot NT NR Bignami et al. (1974)
test
Chromosome aberration, Chinese hamster 1250g/mL Galloway et al. (1987)
ovary cells
Chromosome aberration, Chinese hamster (+) 0.5mg/mL Ishidate et al. (1981)
lung cell line
Sister chromatid exchange, Chinese hamster + + 900g/mL Galloway et al. (1987)
ovary cells
Mutations resulting in trifluorothymidine + NT 500g/mL NTP (1989)
resistance, L5178Y mouse lymphoma cells
Cytokinesis blocked micronucleus assay, + NT 40 g/mL Andrianopoulos et al.
cultured human lymphocytes (2006)
In vivo
Drosophila melanogaster, sex-linked recessive 10000ppm in diet Valencia et al. (1985)
lethal mutation
Drosophila melanogaster, sex-linked recessive 10000g/kg, injection Valencia et al. (1985)
lethal mutation
+, positive; (+), weakly positive; , negative
a
HID, highest ineffective dose; LED, lowest effective dose; NR, not reported; NT, not tested
In the presence of ultraviolet A irradiation, production of DNA cyclobutane pyrimidine

hydrochlorothiazide significantly enhanced the dimers (thyminethymine dimers, detected by
311
IARC MONOGRAPHS 108
HPLC) in isolated DNA and in the skin of mice the chloride site on this transporter, thus inhib-
defective in DNA repair (Kunisada et al., 2013). iting Na+ transport and reducing blood volume.
Clinical evidence also indicates that hydro-
(b) Chromosomal damage chlorothiazide can act as a photosensitizer in
Chromosomal damage caused by hydro- the presence of irradiation by ultraviolet A or
chlorothiazide was reviewed by a previous ultraviolet B (Addo et al., 1987). The incidence of
IARC Working Group (IARC, 1990). In a phototoxic exanthem is occasional (>1/1000 to
spot test, hydrochlorothiazide did not induce <1/100) (Rote Liste Service GmbH, 2012)
nondisjunction and mitotic crossing-over in
Aspergillus nidulans (Bignami et al., 1974). 4.4 Susceptibility
Hydrochlorothiazide did not induce sex-linked
recessive lethal mutation in Drosophila melano- No data were available to the Working Group.
gaster fed or injected with hydrochlorothiazide
solutions at 10 mg/mL (Valencia et al., 1985). 4.5 Mechanistic considerations
Significant increases in the frequency of sister
chromatid exchange were observed in Chinese The possible association between exposure
hamster ovary cells in the presence and absence of to hydrochlorothiazide and cancer of the skin
an exogenous metabolic system (Galloway et al., may result from drug-related photosensitization,
1987). Although the results of tests for chromo- which would cause DNA damage (production of
somal aberration were considered to be negative, dimers by hydrochlorothiazide in the presence of
very high frequencies of chromatid gaps were sunlight) and may also lead to a chronic inflam-
noted at 900 and 1000 g/mL (Galloway et al., matory reaction in the skin.
1987). Chromosomal aberrations were not found
in Chinese hamster lung cells, but polyploidy
was observed after treatment with hydrochloro-
5. Summary of Data Reported
thiazide for 48 hours (Ishidate et al., 1981).
One new study was identified since 5.1 Exposure data
the previous IARC (1990) evaluation. Hydrochlorothiazide is a thiazide-based
Hydrochlorothiazide was found to induce diuretic that is recommended as a first-line
micronucleus formation and chromosome therapy for hypertension. Most frequently,
breakage in cultured human lymphocytes via hydrochlorothiazide is used with other drugs that
chromosome delay (Andrianopoulos et al., lower blood pressure, including in combination
2006). products. In the USA, use of hydrochlorothiazide
has declined slightly over the past decade.
4.3 Other mechanistic data relevant
to carcinogenicity 5.2 Human carcinogenicity data
Effects on cell physiology The occurrence of cancer among patients
Hydrochlorothiazide is a thiazide-based using hydrochlorothiazide has been examined in
diuretic that acts on an electroneutral Na+/Cl cohort and casecontrol studies from the USA and
cotransporter in the distal convoluted tubules Denmark, and in an observational cohort within
of kidney nephrons. The major physiological an intervention trial of heart disease patients
action of hydrochlorothiazide is to compete for from Israel. Other cohort and casecontrol
312
Hydrochlorothiazide
studies have examined use of thiazides (but not was minimal, since positive associations were
specifically hydrochlorothiazide) in multiple observed in studies that could account for this
regions of Europe and in the USA. effect. Effect modification by sun exposure is
potentially important, but had not been thor-
5.2.1 Cancers of the skin and lip oughly examined.
In contrast to the results for squamous cell
Associations between use of hydrochlorothi- carcinoma, results from casecontrol and cohort
azide and squamous cell carcinoma of the skin or studies that examined basal cell carcinoma and
lip were assessed in two casecontrol studies in malignant melanoma of the skin were weaker,
Denmark, and California, USA. The casecontrol lacked doseresponse relationships, or gave
study from Denmark reported an excess risk of results that were close to unity.
squamous cell carcinoma of the skin associated
with hydrochlorothiazide use, and risk increased
5.2.2 Other cancer sites
with increasing dose; cancer of the lip was not eval-
uated. A casecontrol analysis of a cohort study An increased risk of cancer of the kidney
from California detected an excess risk of cancer associated with use of hydrochlorothiazide (one
of the lip and other skin cancers (not including study) or thiazide was reported in three studies in
squamous cell carcinoma, basal cell carcinoma, two independent study populations in the USA;
or melanoma) among users of hydrochlorothi- an earlier study in one of those populations had
azide. This was followed by a nested casecontrol also found an increased risk of renal cell carci-
study of cancer of the lip in the same population, noma among women with unspecified thiazide
which reported a statistically significant twofold use. This association was difficult to interpret
increase in risk for three or more prescriptions, owing to potential confounding by hypertension,
and increasing odds ratios with duration of use. an independent risk factor for renal disease.
This was the only study with adequate statistical Two casecontrol studies on cancer of the
power to assess use of hydrochlorothiazide alone. breast assessed thiazide use: the odds ratios
Two other casecontrol studies of cancer of the found were 1.2, and 1.4, respectively. Increased
skin in Europe and the USA reported increased risks of cancer of the gall bladder, colon, prostate,
odds ratios for squamous cell carcinoma of the and endometrium associated with thiazide use
skin associated with use of thiazide or photosen- were reported each in a single study.
sitizing cardiovascular drugs (mainly thiazides); In conclusion, there were few studies on the
these findings supported the results of studies risk of other cancers in relation to use of hydro-
reporting on hydrochlorothiazide specifically. chlorothiazide, and results had the potential to
While the available data on hydrochloro- be confounded by drug indication.
thiazide and skin cancer generally suggested
associations with squamous cell carcinoma for
sites potentially exposed to sunlight (i.e. skin and
lip), only a few studies have evaluated the asso- Hydrochlorothiazide was tested for carcino-
ciation between hydrochlorothiazide exposure genicity in one feeding study in male and female
and cancer of the skin and lip, and even fewer mice, in two feeding studies in male and female
studies have examined dose or duration effects. rats, and in one feeding study with coexposure
The Working Group considered that the poten- to sodium nitrite in male and female rats. In the
tial confounding effect of sunlight exposure, a first study, hydrochlorothiazide caused a signif-
major risk factor for squamous cell carcinoma, icant increase in the incidence of hepatocellular
313
IARC MONOGRAPHS 108
adenoma, and of hepatocellular adenoma or 6. Evaluation

carcinoma (combined) in male mice; there were
no significant increases in the incidence of any
neoplasm in female mice. In the second study,
there was an increased incidence of adrenal There is limited evidence in humans for the
pheochromocytoma in female rats. No signifi- carcinogenicity of hydrochlorothiazide. Positive
cant increase in the incidence of any neoplasm associations were observed for squamous cell
was observed in male rats in the second study, carcinoma of the skin and lip.
or in male and female rats in the third study. The
study of coexposure also gave negative results.
6.2 Cancer in experimental animals
5.4 Mechanistic and other relevant There is limited evidence in experi-
data mental animals for the carcinogenicity of
hydrochlorothiazide.
Hydrochlorothiazide is excreted essentially
unchanged in humans.
Hydrochlorothiazide was not mutagenic in
standard bacterial screening assays, but produced Hydrochlorothiazide is possibly carcinogenic
cytotoxic effects and induced mutation in L5178Y to humans (Group 2B).
mouse lymphoma cells in the absence of exoge-
nous metabolic activation. Hydrochlorothiazide
increased the frequency of sister chromatid References
exchange, but not chromosomal aberration, in
Chinese hamster ovary cells, both in the presence Abdel Razak O (2004). Electrochemical study of hydro-
and absence of an exogenous metabolic system. chlorothiazide and its determination in urine and
Hydrochlorothiazide induced polyploidy, but not tablets. J Pharm Biomed Anal, 34(2):43340. doi:10.1016/
S0731-7085(03)00497-7 PMID:15013158
chromosomal aberration, in Chinese hamster AbdelHameed MH, Chen TM, Chiou WL (1993).
lung cells. In-vitro induction of micronucleus Intrahepatic distribution of hydrochlorothiazide and
formation and chromosome breakage via chro- quinidine in rats: implications in pharmacokinetics. J
Pharm Sci, 82(10):9926. doi:10.1002/jps.2600821004
mosome delay were observed in human lympho- PMID:8254499
cytes exposed to hydrochlorothiazide. Addo HA, Ferguson J, Frain-Bell W (1987). Thiazide-
In the presence of ultraviolet A irradiation, induced photosensitivity: a study of 33 subjects. Br J
hydrochlorothiazide enhanced the production of Dermatol, 116(6):74960. doi:10.1111/j.1365-2133.1987.
tb04893.x PMID:3620339
DNA cyclobutanepyrimidine dimers, both in ALLHAT (2002). Major outcomes in high-risk hyperten-
isolated DNA and in the skin of DNA repair-de- sive patients randomized to angiotensin-converting
ficient mice. enzyme inhibitor or calcium channel blocker vs
The possible association between exposure diuretic: The Antihypertensive and Lipid-Lowering
Treatment to Prevent Heart Attack Trial (ALLHAT).
to hydrochlorothiazide and cancer of the skin JAMA, 288(23):298197. doi:10.1001/jama.288.23.2981
may result from drug-related photosensitization, PMID:12479763
which would cause DNA damage (production of Andrews AW, Lijinsky W, Snyder SW (1984). Mutagenicity
of amine drugs and their products of nitrosation. Mutat
dimers by hydrochlorothiazide in the presence of Res, 135(2):1058. doi:10.1016/0165-1218(84)90162-9
sunlight) and may also lead to a chronic inflam- PMID:6363913
matory reaction in the skin. Andrianopoulos C, Stephanou G, Demopoulos NA
(2006). Genotoxicity of hydrochlorothiazide in
314
Hydrochlorothiazide
cultured human lymphocytes. I. Evaluation of chro- Chemspider (2013). ChemSpider: The free chemical
mosome delay and chromosome breakage. Environ database. Royal Society of Chemistry. http://www.
Mol Mutagen, 47(3):16978. doi:10.1002/em.20180 chemspider.com/chemical-structure.3513.html
PMID:16304670 Chobanian AV, Bakris GL, Black HR, Cushman WC,
Backman L, Beerman B, Groschinsky-Grind M, Hallberg Green LA, Izzo JL Jr etal.; Joint National Committee on
D (1979). Malabsorption of hydrochlorothiazide Prevention, Detection, Evaluation, and Treatment of
following intestinal shunt surgery. Clin Pharmacokinet, High Blood Pressure. National Heart, Lung, and Blood
4(1):638. doi:10.2165/00003088-197904010-00006 Institute; National High Blood Pressure Education
PMID:421412 Program Coordinating Committee (2003). Seventh
Barbhaiya RH, Craig WA, Corrick-West HP, Welling report of the Joint National Committee on Prevention,
PG (1982). Pharmacokinetics of hydrochlorothiazide Detection, Evaluation, and Treatment of High Blood
in fasted and nonfasted subjects: a comparison Pressure. Hypertension, 42(6):120652. doi:10.1161/01.
of plasma level and urinary excretion methods. J HYP.0000107251.49515.c2 PMID:14656957
Pharm Sci, 71(2):2458. doi:10.1002/jps.2600710226 Christophersen AS, Rasmussen KE, Salvesen B (1977).
PMID:7062255 Determination of hydrochlorothiazide in serum by
Beermann B (1984). Aspects on pharmacokinetics of some high-pressure liquid chromatography. J Chromatogr,
diuretics. Acta Pharmacol Toxicol (Copenh), 54:Suppl 132(1):917. doi:10.1016/S0021-9673(00)93774-9
1: 1729. doi:10.1111/j.1600-0773.1984.tb03626.x PMID:833234
PMID:6369882 Cushman WC, Ford CE, Cutler JA, Margolis KL, Davis
Beermann B, Fhraeus L, Groschinsky-Grind M, BR, Grimm RH et al. ; ALLHAT Collaborative
Lindstrm B (1980). Placental transfer of hydro- Research Group(2002). Success and predictors of blood
chlorothiazide. Gynecol Obstet Invest, 11(1):458. pressure control in diverse North American settings:
doi:10.1159/000299817 PMID:7390276 the antihypertensive and lipid-lowering treatment to
Beermann B, Groschinsky-Grind M (1977). prevent heart attack trial (ALLHAT). J Clin Hypertens
Pharmacokinetics of hydrochlorothiazide in man. (Greenwich), 4(6):393404. doi:10.1111/j.1524-
Eur J Clin Pharmacol, 12(4):297303. doi:10.1007/ 6175.2002.02045.x PMID:12461301
BF00607430 PMID:590315 de Vries E, Trakatelli M, Kalabalikis D, Ferrandiz L,
Beermann B, Groschinsky-Grind M (1978a). Enhancement Ruiz-de-Casas A, Moreno-Ramirez D etal.; EPIDERM
of the gastrointestinal absorption of hydrochloro- Group (2012). Known and potential new risk factors
thiazide by propantheline. Eur J Clin Pharmacol, for skin cancer in European populations: a multicentre
13(5):3857. doi:10.1007/BF00644613 PMID:668798 case-control study. Br J Dermatol, 167:Suppl 2: 113.
Beermann B, Groschinsky-Grind M (1978b). doi:10.1111/j.1365-2133.2012.11081.x PMID:22881582
Gastrointestinal absorption of hydrochlorothiazide Deppeler HP (1981). Hydrochlorothiazide. In: Analytical
enhanced by concomitant intake of food. Eur J Clin Profiles of Drug Substances, Volume 10:405441.
Pharmacol, 13(2):1258. doi:10.1007/BF00609756 Deventer K, Pozo OJ, Van Eenoo P, Delbeke FT (2009).
PMID:658108 Detection of urinary markers for thiazide diuretics
Beermann B, Groschinsky-Grind M (1979). after oral administration of hydrochlorothiazide
Pharmacokinetics of hydrochlorothiazide in patients and altizide-relevance to doping control analysis.
with congestive heart failure. Br J Clin Pharmacol, J Chromatogr A, 1216(12):246673. doi:10.1016/j.
7(6):57983. doi:10.1111/j.1365-2125.1979.tb04646.x chroma.2009.01.032 PMID:19187939
PMID:465280 DrugBank (2013). Open Data Drug & Drug Target
Beermann B, Groschinsky-Grind M, Rosn A (1976). Database. Version 4.1. Available from: http://www.
Absorption, metabolism, and excretion of hydro- drugbank.ca
chlorothiazide. Clin Pharmacol Ther, 19(5 Pt 1):5317. eMC (2013). Electronic Medicines Compendium.
PMID:1277708 Available from: http://www.medicines.org.uk
Bignami M, Morpurgo G, Pagliani R, Carere A, Conti G, Di European Pharmacopoeia (2005). Ed. 6.0. Strasbourg:
Giuseppe G (1974). Non-disjunction and crossing-over Europe Directorate for the Quality of Medicines of the
induced by pharmaceutical drugs in Aspergillus Council of Europe (EDQM).
nidulans. Mutat Res, 26(3):15970. doi:10.1016/S0027- Farthing D, Fakhry I, Ripley EB, Sica D (1998). Simple
5107(74)80070-9 PMID:4604824 method for determination of hydrochlorothiazide
Bucher JR, Huff J, Haseman JK, Eustis SL, Elwell MR, Davis in human urine by high performance liquid chro-
WE Jr et al. (1990). Toxicology and carcinogenicity matography utilizing narrowbore chromatography. J
studies of diuretics in F344 rats and B6C3F1 mice. 1. Pharm Biomed Anal, 17(8):14559. doi:10.1016/S0731-
Hydrochlorothiazide. J Appl Toxicol, 10(5):35967. 7085(98)00021-1 PMID:9800665
doi:10.1002/jat.2550100509 PMID:2254588
315
IARC MONOGRAPHS 108
FDA (2013). Drugs@FDA. Available from: http://www. study (California, USA). Cancer Causes Control,
accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm, 5(4):31925. doi:10.1007/BF01804982 PMID:8080943
accessed 8 September 2014. HSDB (2013) Hazardous Substances Data Bank.
Flaherty KT, Fuchs CS, Colditz GA, Stampfer MJ, Speizer Hydrochlorothiazide. United States National Library
FE, Willett WC et al. (2005). A prospective study of of Medicine. Available from: http://www.toxnet.nlm.
body mass index, hypertension, and smoking and the nih.gov, accessed 8 September 2014.
risk of renal cell carcinoma (United States). Cancer Huang T, He Z, Yang B, Shao L, Zheng X, Duan G (2006).
Causes Control, 16(9):1099106. doi:10.1007/s10552- Simultaneous determination of captopril and hydro-
005-0349-8 PMID:16184476 chlorothiazide in human plasma by reverse-phase
Fortuny J, Sima C, Bayuga S, Wilcox H, Pulick K, HPLC from linear gradient elution. J Pharm Biomed
Faulkner S et al. (2009). Risk of endometrial cancer Anal, 41(2):6448. doi:10.1016/j.jpba.2005.12.007
in relation to medical conditions and medication use. PMID:16413728
Cancer Epidemiol Biomarkers Prev, 18(5):144856. IARC (1990). Hydrochlorothiazide. IARC Monogr Eval
doi:10.1158/1055-9965.EPI-08-0936 PMID:19383893 Carcinog Risks Hum, 50:293305. PMID:2292804
Friedman GD, Asgari MM, Warton EM, Chan J, Habel IMS Health (2012a). National Disease and Therapeutic
LA (2012). Antihypertensive drugs and lip cancer in Index (NDTI). Plymouth Meeting, Pennsylvania: IMS
non-Hispanic whites. Arch Intern Med, 172(16):124651. Health; 200512.
doi:10.1001/archinternmed.2012.2754 PMID:22869299 IMS Health (2012b). National Prescription Audit Plus
Friedman GD, Udaltsova N, Chan J, Quesenberry CP (NPA). Plymouth Meeting, Pennsylvania: IMS Health;
Jr, Habel LA (2009). Screening pharmaceuticals for 200812.
possible carcinogenic effects: initial positive results for IMS Health (2012c). Multinational Integrated Data
drugs not previously screened. Cancer Causes Control, Analysis (MIDAS). Plymouth Meeting, Pennsylvania:
20(10):182135. doi:10.1007/s10552-009-9375-2 IMS Health. Available from: http://www.imshealth.
PMID:19582585 com/portal/site/imshealth
Friedman GD, Ury HK (1980). Initial screening for Indian Pharmacopoeia (2010). The Indian Pharmacopoeia
carcinogenicity of commonly used drugs. J Natl Cancer Commission. Ghaziabad, India.
Inst, 65(4):72333. PMID:6932525 IPCS (2013). Hydrochlorothiazide. Geneva: International
Fujita H (1985). Arginine reversion and lambda induction Programme on Chemical Safety, World Health
in E. coli with benzothiadiazine diuretics irradiated Organization. Available from: http://www.inchem.org/
with near-ultraviolet light. Mutat Res, 158(3):1359. documents/pims/pharm/hydrochl.htm.
doi:10.1016/0165-1218(85)90076-X PMID:2934628 Ishidate M, Sofuni T, Yoshikawa K (1981). Chromosomal
Galloway SM, Armstrong MJ, Reuben C, Colman S, Brown aberration tests in vitro as a primary screening tool for
B, Cannon C et al. (1987). Chromosome aberrations environmental mutagens and/or carcinogens. GANN
and sister chromatid exchanges in Chinese hamster Monograph on Cancer Research, 27:95108.
ovary cells: evaluations of 108 chemicals. Environ Jensen AO, Thomsen HF, Engebjerg MC, Olesen AB,
Mol Mutagen, 10(S10):Suppl 10: 1175. doi:10.1002/ Srensen HT, Karagas MR (2008). Use of photo -
em.2850100502 PMID:3319609 sensitising diuretics and risk of skin cancer: a popula-
Gotardo MA, Pezza L, Pezza HR (2005). Determination of tion-based case-control study. Br J Cancer, 99(9):15228.
hydrochlorothiazide in pharmaceutical formulations doi:10.1038/sj.bjc.6604686 PMID:18813314
by diffuse reflectance spectroscopy. Ecletica Quimica, Kim GH, Na KY, Kim SY, Joo KW, Oh YK, Chae SW
30(2):1724. doi:10.1590/S0100-46702005000200002 et al. (2003). Up-regulation of organic anion trans-
Gu Q, Burt VL, Dillon CF, Yoon S (2012). Trends in antihy- porter 1 protein is induced by chronic furosemide or
pertensive medication use and blood pressure control hydrochlorothiazide infusion in rat kidney. Nephrol
among United States adults with hypertension: the Dial Transplant, 18(8):150511. doi:10.1093/ndt/gfg186
National Health And Nutrition Examination Survey, PMID:12897087
2001 to 2010. Circulation, 126(17):210514. doi:10.1161/ Kuehlein T, Laux G, Gutscher A, Goetz K, Szecsenyi J,
CIRCULATIONAHA.112.096156 PMID:23091084 Campbell S etal. (2011). Diuretics for hypertensionan
Hallas J, Christensen R, Andersen M, Friis S, Bjerrum L inconsistency in primary care prescribing behaviour.
(2012). Long term use of drugs affecting the renin-an- Curr Med Res Opin, 27(3):497502. doi:10.1185/030079
giotensin system and the risk of cancer: a popula- 95.2010.547932 PMID:21208153
tion-based case-control study. Br J Clin Pharmacol, Kunisada M, Masaki T, Ono R, Morinaga H, Nakano E,
74(1):1808. doi:10.1111/j.1365-2125.2012.04170.x Yogianti F etal. (2013). Hydrochlorothiazide enhances
PMID:22243442 UVA-induced DNA damage. Photochem Photobiol,
Hiatt RA, Tolan K, Quesenberry CP Jr (1994). Renal cell 89(3):64954. doi:10.1111/php.12048 PMID:23331297
carcinoma and thiazide use: a historical, case-control Li CI, Malone KE, Weiss NS, Boudreau DM, Cushing-
Haugen KL, Daling JR (2003). Relation between use
316
Hydrochlorothiazide
of antihypertensive medications and risk of breast ONeil MJ (2001). The Merck Index - An Encyclopedia of
carcinoma among women ages 6579 years. Cancer, Chemicals, Drugs, and Biologicals, 13th ed. Whitehouse
98(7):150413. doi:10.1002/cncr.11663 PMID:14508839 Station (NJ): Merck & Co., Inc.
Lijinsky W, Reuber MD (1987). Pathologic effects of Okuda T, Itoh S, Yamazaki M, Nakahama H, Fukuhara
chronic administration of hydrochlorothiazide, with Y, Orita Y (1987). Biopharmaceutical studies
and without sodium nitrite, to F344 rats. Toxicol Ind of thiazide diuretics. III. In vivo formation of
Health, 3(3):41322. doi:10.1177/074823378700300313 2-amino-4-chloro-m-benzenedisulfonamide as
PMID:3686543 a metabolite of hydrochlorothiazide in a patient.
Lookchem (2013). Lookchem database. Chem Pharm Bull (Tokyo), 35(8):35168. doi:10.1248/
Hydrochlorothiazide. Available from: http://www. cpb.35.3516 PMID:3427729
lookchem.com, accessed 8 September 2014. Patel RB, Patel UR, Rogge MC, Shah VP, Prasad VK, Selen
Mansia G, De Backer G, Dominiczak A, Cifkova R, A etal. (1984). Bioavailability of hydrochlorothiazide
Fagard R, Germano G et al.; European Society of from tablets and suspensions. J Pharm Sci, 73(3):359
Hypertension; European Society of Cardiology (2007). 61. doi:10.1002/jps.2600730317 PMID:6716243
2007 ESH-ESC Guidelines for the management of Redalieu E, Chan KK, Tipnis V, Zak SB, Gilleran TG,
arterial hypertension: the task force for the manage- Wagner WE Jr etal. (1985). Kinetics of hydrochloro-
ment of arterial hypertension of the European Society thiazide absorption in humans. J Pharm Sci, 74(7):7657.
of Hypertension (ESH) and of the European Society doi:10.1002/jps.2600740714 PMID:4032251
of Cardiology (ESC). Blood Press, 16(3):135232. Ripley E, King K, Sica DA (2000). Racial differences in
doi:10.1080/08037050701461084 PMID:17846925 response to acute dosing with hydrochlorothiazide.
Mansia G, Laurent S, Agabiti-Rosei E, Ambrosioni E, Am J Hypertens, 13(2):15764. doi:10.1016/S0895-
Burnier M, Caulfield MJ et al.; European Society of 7061(99)00168-5 PMID:10701815
Hypertension (2009). Reappraisal of European guide- Robinson SN, Zens MS, Perry AE, Spencer SK, Duell EJ,
lines on hypertension management: a European Society Karagas MR (2013). Photosensitizing agents and the
of Hypertension Task Force document. J Hypertens, risk of non-melanoma skin cancer: a population-based
27(11):212158. doi:10.1097/HJH.0b013e328333146d case-control study. J Invest Dermatol, 133(8):19505.
PMID:19838131 doi:10.1038/jid.2013.33 PMID:23344461
Masereeuw R, Moons WM, Russel FG (1997). Saturable Rote Liste Service GmbH (2012). HCT-ratiopharm 12,5 mg
accumulation and diuretic activity of hydro- Tabletten.
chlorothiazide in the isolated perfused rat kidney. Ruiter R, Visser LE, Eijgelsheim M, Rodenburg EM,
Pharmacology, 54(1):3342. doi:10.1159/000139467 Hofman A, Coebergh JW et al. (2010). High-ceiling
PMID:9065959 diuretics are associated with an increased risk of
Meyer MC, Melikian AP, Whyatt PL, Slywka GW (1975). basal cell carcinoma in a population-based follow-up
Hydrochlorothiazide bioavailability: an evaluation of study. Eur J Cancer, 46(13):246772. doi:10.1016/j.
thirteen products. Curr Ther Res Clin Exp, 17(6):5707. ejca.2010.04.024 PMID:20605443
PMID:808380 SciFinder (2013). SciFinder Databases: Registry, Chemcats.
MicroMedex (2013). 2.0. Ann Arbor, Michigan: Truven American Chemical Society. Available from: http://
Health Analytics Inc. www.cas.org/products/scifinder
Miller ME, Cohn RD, Burghart PH (1982). Sousa CEM, Bedor DCG, Gonalves TM et al. (2009).
Hydrochlorothiazide disposition in a mother and her Rapid determination of hydrochlorothiazide in
breast-fed infant. J Pediatr, 101(5):78991. doi:10.1016/ human plasma by high performance liquid chroma-
S0022-3476(82)80323-5 PMID:7131161 tography-tandem mass spectrometry. Lat Am J Pharm,
Mortelmans K, Haworth S, Lawlor T, Speck W, Tainer 28:7937.
B, Zeiger E (1986). Salmonella mutagenicity tests: Stafford RS, Monti V, Furberg CD, Ma J (2006). Long-
II. Results from the testing of 270 chemicals. term and short-term changes in antihypertensive
Environ Mutagen, 8(S7):Suppl 7: 1119. doi:10.1002/ prescribing by office-based physicians in the United
em.2860080802 PMID:3516675 States. Hypertension, 48(2):2138. doi:10.1161/01.
Niemeyer C, Hasenfuss G, Wais U, Knauf H, Schfer- HYP.0000229653.73128.b6 PMID:16785334
Korting M, Mutschler E (1983). Pharmacokinetics of Stanford JL, Martin EJ, Brinton LA, Hoover RN (1986).
hydrochlorothiazide in relation to renal function. Eur Rauwolfia use and breast cancer: a case-control study. J
J Clin Pharmacol, 24(5):6615. doi:10.1007/BF00542218 Natl Cancer Inst, 76(5):81722. PMID:3457968
PMID:6873147 Takubo T, Okada H, Ishii M, Hara K, Ishii Y (2004).
NTP (1989). Toxicology and carcinogenesis studies of Sensitive and selective liquid chromatography-electro-
hydrochlorothiazide (CAS No. 58935) in F344/N rats spray ionization tandem mass spectrometry analysis
and B6C3F1 mice (feed studies). Natl Toxicol Program of hydrochlorothiazide in rat plasma. J Chromatogr
Tech Rep Ser, 357:1194. PMID:12695784
317
IARC MONOGRAPHS 108
B Analyt Technol Biomed Life Sci, 806(2):199203.

doi:10.1016/j.jchromb.2004.03.060 PMID:15171930
Tenenbaum A, Grossman E, Fisman EZ, Adler Y, Boyko
V, Jonas M et al. (2001). Long-term diuretic therapy
in patients with coronary disease: increased colon
cancer-related mortality over a 5-year follow-up. J Hum
Hypertens, 15(6):3739. doi:10.1038/sj.jhh.1001192
PMID:11439311
US Pharmacopeia (2007). Hydrochlorothiazide,
USP28-NF23. Rockville (MD), USA: The United States
Pharmacopoeial Convention.
Valencia R, Mason JM, Woodruff RC, Zimmering S (1985).
Chemical mutagenesis testing in Drosophila. III.
Results of 48 coded compounds tested for the National
Toxicology Program. Environ Mutagen, 7(3):32548.
doi:10.1002/em.2860070309 PMID:3930234
van den Eeden SK, Friedman GD (1995). Prescription
drug screening for subsequent carcinogenicity.
Pharmacoepidemiol Drug Saf, 4(5):27587. doi:10.1002/
pds.2630040504
Vandenheuvel WJA, Gruber VF, Walker RW, Wolf FJ
(1975). GLC analysis of hydrochlorothiazide in blood
and plasma. J Pharm Sci, 64(8):130912. doi:10.1002/
jps.2600640810 PMID:1151702
Wang YR, Alexander GC, Stafford RS (2007). Outpatient
hypertension treatment, treatment intensification, and
control in Western Europe and the United States. Arch
Intern Med, 167(2):1417. doi:10.1001/archinte.167.2.141
PMID:17242314
Waskell L (1978). A study of the mutagenicity of anes-
thetics and their metabolites. Mutat Res, 57(2):14153.
doi:10.1016/0027-5107(78)90261-0 PMID:351387
Welling PG (1986). Pharmacokinetics of the thiazide
diuretics. Biopharm Drug Dispos, 7(6):50135.
doi:10.1002/bdd.2510070602 PMID:3828483
WHO(2006). The International Pharmacopoeia, . 4th ed.
Geneva, Switzerland: World Health Organization.
Yamazaki M, Itoh S, Okuda T, Tanabe K, Nakahama H,
Fukuhara Y etal. (1989). Binding of hydrochlorothiazide
to erythrocytes. J Pharmacobiodyn, 12(7):4238.
doi:10.1248/bpb1978.12.423 PMID:2593083
Yan T, Li H, Deng L, Guo Y, Yu W, Fawcett JP etal. (2008).
Liquid chromatographic-tandem mass spectrometric
method for the simultaneous quantitation of telmis-
artan and hydrochlorothiazide in human plasma.
J Pharm Biomed Anal, 48(4):12259. doi:10.1016/j.
jpba.2008.08.021 PMID:18838240
318
PIOGLITAZONE AND ROSIGLITAZONE
1. Exposure Data Chem. Abstr. Serv. Name: 2,4-Thiazolidine

dione, 5-[[4-[2-(5-ethyl-2-pyridinyl)ethoxy]
Thiazolidinediones are a class of synthetic phenyl]methyl]- (SciFinder, 2013).
compounds that exert direct effects on the IUPAC systematic Name: 5-[[4-[2-(5-
mechanisms of insulin resistance, and result Ethylpyridin-2-yl)ethoxy]phenyl]methyl]-
in improved insulin action and reduced hyper- 1,3-thiazolidine-2,4-dione (Drugbank, 2013;
insulinaemia. In the present Monograph, the Pubchem, 2013).
Working Group evaluated pioglitazone and WHO International Nonproprietary Name
rosiglitazone, two thiazolidinediones that (INN): Pioglitazonum (WHO, 2007).
initially showed great promise as receptor-me-
diated oral therapy for type 2 diabetes mellitus. (b) Pioglitazone hydrochloride
Rosiglitazone and pioglitazone were introduced
to the market at about the same time (1999 in the Chem. Abstr. Serv. Reg. No.: 112529-15-4
USA, and 20012002 in Taiwan, China; Tseng, (SciFinder, 2013).
2012d). Some patients may therefore have been Chem. Abstr. Serv. Name: 2,4-Thiazolidine
exposed to both drugs, which were sometimes dione,5-[[4-[2-(5-ethyl-2-pyridinyl)ethoxy]
prescribed sequentially. The Working Group did phenyl]methyl]-, hydrochloride (1:1)
not consider other thiazolidinediones, such as (SciFinder, 2013).
troglitazone, which was marketed for only a short Proprietary Names: Actos; Glustin; Zactose
period (19972000), before being withdrawn (Drugbank, 2013; Pubchem, 2013).
from the world market subsequent to reports of
United States Nonproprietary Name (USAN):
fatal hepatotoxicity (Julie et al., 2008).
Pioglitazone hydrochloride
1.1 Chemical and physical data on

pioglitazone
1.1.1 Nomenclature
(a) Pioglitazone

(SciFinder, 2013).
319
IARC MONOGRAPHS 108
1.1.2 Structural and molecular formulae and The solubility is highly dependent on pH, and
relative molecular mass is greater at lower pH. Solubility according
to pH: 52.60g/mL (pH1.83); 38.63g/mL
(a) Pioglitazone (pH2.57); 4.55g/mL (pH3.92); 6.35g/mL
O (pH7.39); 19.19g/mL (pH8.82); 49.96g/mL
N O
S (pH9.52) (Seedher & Kanojia, 2009); 100g/mL
H 3C
N H in 1:1 dimethyl sulfoxide:phosphate-buffered
saline (pH 7.2); 2.5 mg/mL in dimethylfor-
O mamide and dimethyl sulfoxide (Cayman
C19H20N2O3S (ONeil, 2006) SDS, 2013)
Relative molecular mass: 356.44
Stability data: Exposure to heat (105 C)
(b) Pioglitazone hydrochloride results in a change of appearance; exposure
to heat and UV light results in a slight drop
O
N O (1.52%) in assay; exposure to 0.1N sodium
S
N H . HCl
hydroxide results in degradation; exposure to
H 3C
heat (105C) and peroxide results in a slight
O increase in total impurities (EMA CHMP,
C19H20N2O3S. HCl 2012)
Relative molecular mass: 392.90 Octanol/water partition coefficient: Log
P=2.723.73 (Giaginis et al., 2007)
1.1.3 Chemical and physical properties of the Vapour pressure: 2.881014 mmHg at 25C,
pure substance estimated (EMA CHMP, 2012)
(a) Pioglitazone
(b) Pioglitazone hydrochloride
Description: Colourless needles from
dimethylformamide and water (ONeil, 2006) Description: White crystalline powder,
odourless (Physicians Desk Reference, 2012);
Density: 1.260 g/cm3 at 20 C (Langchem,
colourless prisms from ethanol (ONeil, 2006)
2013)
Density: 1.26 g/cm3 (ChemicalBook, 2013)
Melting point: 183184 C (ONeil, 2006;
Milne, 2000) Melting point: 193194 C (ONeil, 2006;
Milne, 2000)
Spectroscopy data: Ultraviolet (UV)
(Venkatesh et al., 2006), proton nuclear Solubility: Practically insoluble in water,
magnetic resonance (1H NMR) (Madivada insoluble in ether, slightly soluble in ethanol,
et al., 2009), 13C NMR (Madivada et al., 2009), very slightly soluble in acetone and acetoni-
infrared (IR) (Madivada et al., 2009), and trile (ONeil, 2006); very soluble in dimethyl-
mass spectrometry (MS) (Wang & Miksa, formamide (Physicians Desk Reference, 2012)
2007; Thevis et al., 2005) have been reported Vapour pressure: 3.01013 mmHg at 25C
Solubility: 14.05g/mL (water); 25.07g/mL (ChemicalBook, 2013)
(0.15M NaCl); 10.61g/mL (0.1M phosphate
buffer) (Seedher & Kanojia, 2008); 46.85mg/L
at 25C (EMA CHMP, 2012)
320
Pioglitazone and rosiglitazone
1.1.4 Technical products and impurities 1.2 Analysis of pioglitazone

Pioglitazone hydrochloride is used to formu- Physical properties used for the identification
late the finished dosage forms described below. of the substance, e.g. IR and melting point, are
presented in Section 1.1.3. Selected non-compen-
(a) Trade names
dial methods are presented in Table1.1.
Actos; Glustin; Glizone; Pioz; Zactose (Rx There are numerous methods including
List, 2013). HPLC with UV or MS detection for the anal-
ysis of pioglitazone in different matrices, such
(b) Impurities
as formulations, plasma, and serum. The limit of
Three impurities were detected up to concen- quantitation (LOQ) in serum using the method
tration of 0.1% by reversed-phase high-perfor- by Palem et al. (2011) is 1 ng/mL; Lin et al.
mance liquid chromatography (HPLC) and were (2003) reported a LOQ of 0.5ng/mL in human
characterized by 1H NMR, 13C NMR, MS, and IR plasma. Other reported techniques for analysing
spectral data (Kumar et al., 2004): formulations include capillary electrophoresis
5-(4-Hydroxybenzyl)-1,3-thiazolidine-2,4- (Radhakrishna et al., 2002a) and potentiometric
dione sensors (Mostafa & Al-Majed, 2008).
5-(4-Fluorobenzyl)-1,3-thiazolidine-2,4-
dione 1.3 Production and use of
2-[2-(4-Bromophenoxy) ethyl-5-ethyl] pioglitazone
pyridine.
1.3.1 Production
Four impurities in pioglitazone were prepared
and characterized by NMR spectroscopy (Richter Pioglitazone exists in two polymorphic
et al., 2007): crystal forms. Polymorph 1 is used in the manu-
facture of the finished drug product. The manu-
5-{4-[2-(5-Ethyl-6-{4-[2-(5-ethylpyridin- facture of pioglitazone consists of six steps, the
2-yl)-ethoxy]phenyl}pyrid-2-yl)ethoxy] last in combination with hydrochloride to yield
benzyl}-1,3-thiazolidine-2,4-dione pioglitazone hydrochloride (EMEA, 2012).
5-{4-[2-(5-Ethyl-4-{4-[2-(5-ethylpyridin-
2-yl)et hox y]phenyl}py rid-2-yl)et hox y] 1.3.2 Use
benzyl}-1,3-thiazolidine-2,4-dione
5-{6,4-bis-[2-(5-Ethyl-pyridin-2-yl)ethoxy] (a) Indications
biphenyl-3-ylmethyl}-1,3-thiazolidine-2,4- Pioglitazone acts as an insulin sensi-
dione tizer, providing a means to improve glycaemic
5-{4 -[2-(5-Et hyl-py r id i n-2-yl)et hox y] control by reducing insulin resistance and thus
b e n z y l} -3 -[2 - (5 - e t h y l - p y r i d i n -2 -y l) decreasing hyperglycaemia in patients with type
ethyl]-1,3-thiazolidine-2,4-dione. 2 diabetes mellitus (Sweetman, 2011; FDA, 2013a).
Pioglitazone acts only in the presence of endoge-
nous insulin. It is indicated particularly for over-
weight patients as an adjunct to diet and exercise
to improve glycaemic control. It is contraindi-
cated in patients with type 1 diabetes mellitus,
and for the treatment of diabetic ketoacidosis. It
321
322
Table 1.1 Analytical methods for pioglitazone and rosiglitazone

Human plasma Pre-activation of SPE column with LC-UV 50 ng/mL (LLOQ) Sripalakit et al.
acetonitrile and KH2PO4, addition of IS Column: C18 (2006)
and KH2PO4 to pioglitazone solution in Mobile phase: methanol, acetonitrile and mixed
plasma, extraction using SPE column, phosphate buffer (pH2.6, 10mM) (40:12:48, v/v/v)
elution using acetonitrile and water, Flow rate: 1.2 mL/min
filtration, and analysis of filtrate Wavelength: 269nm
IARC MONOGRAPHS 108
Human plasma Addition of pioglitazone standard LC-UV 25 ng/mL (LLOQ) Souri et al. (2008)
solutions, IS, diethylether, mixing, Column: C18
centrifugation, addition of NaOH to Mobile phase: acetonitrile and 140mM KH2PO4 (pH
organic layer, mixing, centrifugation, and 4.45) (40:60, v/v)
injection of aqueous layer in HPLC Flow rate:1.4 mL/min
Wavelength: 269 nm
Human plasma Addition of IS, 0.1 M ammonium acetate, LC-ESI-MS 0.5 ng/mL (LLOQ) Lin et al. (2003)
pH adjustment, extraction with methyl Column: C18
tert-butyl ether and butyl chloride, Mobile phase: acetonitrile:water (60:40) with 10mM
centrifugation, evaporation, residues ammonium acetate and 0.02% TFA
dissolved in mobile phase, centrifugation Flow rate: 0.2mL/min
SRM transition: 357 m/z 134 m/z (positive mode)
Human serum Samples diluted 1:1 (v/v) with acetonitrile LC-ESI-MS 9 ng/mL (LLOQ) Xue et al. (2003)
containing IS Column: C18
Mobile phase: 50% of 10 mM ammonium acetate
in acetonitrile: water (10: 90) and 50% of
water:acetonitrile (10:90)
Flow rate: 1/35 mL/min
SRM transition: 357m/z 134 m/z (positive mode)
Pig serum Addition of IS, NaOH solution (1 M), LC-UV 1 ng/mL (LLOQ) Palem et al. (2011)
dichloromethane, centrifugation, Column: C18
separation of organic layer, evaporation, Mobile phase: acetonitrile:50 mM ammonium acetate
reconstitution with methanol and analysis buffer (pH5) (67:33, v/v)
Flow rate: 1mL/min
Wavelength: 240 nm
Dog serum Serum samples loaded on the column, LC-UV 25 ng/mL (LLOQ) Zhong & Lakings
elution with acetonitrile, eluate mixed with Column: C18 (1989)
purified water, and analysis Mobile phase: Acetonitrile:water (41:59, v/v)
containing 1.2mL/L acetic acid (pH6.00.05)
Wavelength: 229 nm

Rat serum Addition of IS solution (rosiglitazone), LC-UV 15 ng/mL (LOD) Ravikanth et al.
precipitation by addition of ethylacetate, Column: C18 50 ng/mL (LOQ) (2011)
centrifugation, and analysis Mobile phase: Methanol:ammonium acetate (30mM,
pH5) (60:40, v/v)
Wavelength: 269 nm
Human serum Serum sample: activation of SPE column, LC-UV Serum: Yamashita et al.
and urine addition of phosphate buffer, elution with Column: C18 0.010.05g/mL (1996)
methanol and 0.02M sodium acetate, Mobile phase: 0.05Mphosphate buffer Urine:
addition of acetic acid, evaporation, (pH6.0):methanol (9:1, v/v) and 0.05M phosphate 0.10.5g/mL
dissolve residue in 0.1M KH2PO4, buffer (pH6.0): methanol:acetonitrile (4:2:4, v/v)
extraction in diethylether, evaporation and Flow rate: 1.0mL/min
addition of IS Wavelength: 269nm
Urine sample: addition of 0.1 M KH2PO4,
extraction with mixture of diethylether
and dichloromethane (4:1, v/v),
evaporation, dissolution in IS solution, and
analysis
Bulk and Bulk sample: sample in mixture of aqueous HPLC Radhakrishna
pharmaceutical 0.1% ortho-phosphoric acid and acetonitrile Column: C18 et al. (2002a)
formulation at 1:1 (v/v) Mobile phase: 10 mM KH2PO4:acetonitrile (pH6.0)
Formulation: 20 weighed tablets ground Flow rate: 1mL/min
to a fine powder, extraction with diluting Wavelength: 225nm
solution, centrifugation
Tablet Tablets ground to fine powder, dissolve in HPLC Jain et al. (2008)
formulation methanol, sonication, filtration, dilution, Column: C18
and analysis Mobile phase: methanol:phosphate buffer (pH4.3)
(75:25, v/v)
Wavelength: 258nm
Tablet Finely powdered tablets, addition of HPLC 42 ng/mL (LOD) Jedlicka et al.
methanol, sonication, centrifugation, Column: C18 (2004)
supernatant diluted with 60% methanol, Mobile phase: ammonium formate buffer (0.05M,
injection on column pH4.1):acetonitrile (45:55, v/v)
Flow rate 1.0mL/min
Wavelength: 266nm
323
324

Formulation Powder transferred to volumetric flask, HPLC Venkatesh et al.
volume adjustment with acetonitrile and Column: C18 (2006)
methanol (1:1), sonication, filtration, Mobile phase: formic acid, (0.05M, pH3.0),
addition of IS and injection onto HPLC water:acetonitrile (5:95, v/v) and water:methanol
column (10:90,v/v)
Flow rate:1.0mL/min
Wavelength: 260nm
IARC MONOGRAPHS 108
Environmental Addition of acetonitrile to stabilize HPLC-TOF-MS LOQ Martn et al.

sample sample, store at 4 C, filtration, addition of Column: C18 Waste water: (2012)
rosiglitazone as IS, and analysis Mobile phase: acetonitrile (containing formic acid 0.1%, 1.1ng/L
v/v) and an aqueous 10 mM ammonium formate solution River water:
(containing formic acid 0.1%, v/v) 1.2ng/L
Flow rate: 0.7mL/min Tap water: 3.1ng/L
Formulation Addition of mobile phase to flask RP-UPLC 0.01 g/mL (LOD) Xavier &
containing tablet powder, sonication, Column: C18 0.05 g/mL (LOQ) Basavaiah (2012)
filtration, and analysis Mobile phase: acetonitrile:buer (pH3.2) (20:80, v/v)
Wavelength: 220nm
Formulation Crushing of tablet, transfer of powder to Potentiometric sensors Mostafa & Al-
beaker, dissolve in acidic water solution, The sensing membranes incorporate ion association Majed (2008)
sonication, pH adjustment to 3.0 using complexes of pioglitazone cation and sodium
phosphate buffer tetraphenylborate or phosphomolybdic acid or
phosphotungstic acid as electroactive material.
Marketed Tablet finely powdered, extraction with Chiral normal-phase HPLC 100 ng/mL (LOD) Gowramma et al.
formulation methanol, further dilutions to achieve a Column: chiral 400 ng/mL (LOQ) (2012a, b)
concentration of 100g/mL of pioglitazone Mobile phase: hexane and n-propyl alcohol (80:20, v/v)
using methanol Flow rate: 1.0mL/min
Wavelength: 233nm
Bulk and Twenty weighed tablets ground to a fine Capillary electrophoresis Pioglitazone Radhakrishna
pharmaceutical powder, extraction with diluting solution, Separation mode: micellar electrokinetic unsaturated et al. (2002a)
formulation centrifugation chromatographic fused-silica capillary impurity:
Background electrolyte: 80 parts of 20mM sodium 0.29g/mL (LOD)
borate (pH9.3) containing 50mM SDS and 20 parts of 0.74g/mL (LOQ)
acetonitrile.
Wavelength: 210nm
Voltage: 25kV
HPLC, high-performance liquid chromatography; HPLC-TOF-MS, high-performance liquid chromatography time of flight mass spectrometry; IS, internal standard; KH 2PO4,
monopotassium phosphate; LC-UV, liquid chromatography ultraviolet spectroscopy; LC-ESI-MS, liquid chromatography electrospray ionization mass spectrometry; LOD, limit of
detection; LLOQ, lower limit of quantification; LOQ, limit of quantification; NaOH, sodium hydroxide; RP-UPLC, reverse phase-ultra high pressure liquid chromatography; SDS,
sodium dodecyl sulfate; SPE, solid-phase extraction; SRM, selected reaction monitoring; TFA, trifluoroacetic acid; UV, ultraviolet
Table 1.2 Most commonly reported clinical indications for pioglitazone and rosiglitazone in the
USA, 20112012
Drug uses (in thousands) Percentage of total

Diagnosisa ICD-9 codeb
Pioglitazone Rosiglitazone Pioglitazone Rosiglitazone
Diabetes mellitus NOS 250.001 3500 90 54.0 28.4
Diabetes type II, non-insulin dependent 250.003 2585 126 39.9 39.8
Diabetic nephropathy 250.302 66 7 1.0 2.1
Diabetic neuropathy 250.503 60 0.9
Diabetic kidney disease 250.301 57 13 0.9 4.1
Diabetes type I, insulin dependent 250.002 45 0.7
Metabolic/insulin resistant syndrome 277.701 35 60 0.5 18.9
Elevated glucose 790.201 27 0.4
Polycystic ovary syndrome 256.401 7 2.1
All other diagnoses 110 14 1.7
Total with reported diagnoses 6484 316 100.0 100.0
No diagnosis was stated for 0.3% of drug uses.
a
b The ICD-9 codes given are a more detailed, proprietary version developed by IMS Health.
From IMS Health (2012b)
is also contraindicated in patients with advanced to congestive heart failure (Singh et al., 2007a),
congestive heart failure (FDA, 2013a). and fractures (Loke et al., 2009), and subse-
Apart from its approved indication for quent warnings about the risk of cancer of the
treatment of type 2 diabetes mellitus, pioglita- bladder (FDA, 2013a). Prescription trends from
zone is also used for other off-label indications the Netherlands also declined after regulatory
(Table1.2). warnings concerning fractures and development
of cancer of the bladder (Ruiter et al., 2012).
(b) Dosage Total worldwide sales of pioglitazone were
Pioglitazone is available as tablets of 15, 30 US$3.34 billion in 2012, with 71% occurring in
and 45mg dosage titrated on adequacy of ther- the USA (US$2.37 billion). Other nations with
apeutic response (Sweetman, 2011). Pioglitazone significant sales of pioglitazone included Japan
is also available in combination products, (US$ 322 million), India (US$ 106 million),
including pioglitazone and metformin (Actoplus United Kingdom (US$ 71 million) and Italy
Met), pioglitazone and glimeperide (Duetact), (US$66 million) (IMS Health, 2012a).
and pioglitazone and alogliptin (Oseni) (FDA, Pioglitazone was reported in 2.4 million
2013a; ChemSpider, 2013). drug uses in the USA in 2012, a decline from 6.7
million reported uses in 2006, according to IMS
(c) Trends in use Health National Disease and Therapeutic Index
Pioglitazone was one of the most widely data. Based on these same data, approximately
used drugs for the treatment of type 2 diabetes 600000 patients in the USA were taking piogl-
among adults in 20002005. However, prescrip- itazone in 2012 (IMS Health, 2012b). See also
tion sales of thiazolidinediones in general, and Fig.1.1.
of pioglitazone in particular, have declined
following several studies that suggested links
325
IARC MONOGRAPHS 108
Fig.1.1 Trends in use of pioglitazone by office-based physicians in the USA

2000
1600
1200
800
400
0
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
Prepared by the Working Group from data obtained from IMS Health, National Disease and Therapeutic Index (IMS Health, 2012b).
1.4 Occurrence and exposure to or used (Pubchem, 2013). No information was

pioglitazone available to the Working Group on the potential
number of workers exposed.
1.4.1 Natural occurrence
Pioglitazone is not reported to occur natu- 1.5 Regulations and guidelines for
rally. The production and medicinal use of pioglitazone
pioglitazone may contaminate the environment
through various waste streams (Pubchem, 2013). Pioglitazone was first approved for use in the
If released into the air, pioglitazone is removed USA on 15 July 1999 (FDA, 2013b). The Food
by wet or dry deposition, since it exists solely and Drug Administration (FDA) approved a risk
in the particulate phase in the atmosphere. evaluation and mitigation strategy (REMS) for
Pioglitazone does not volatilize from dry soil pioglitazone to ensure that the benefits of this
surfaces based upon its vapour pressure. If drug outweighed the risks; however, this REMS
released into the water, pioglitazone is expected was later rescinded.
to adsorb to suspended solids and sediment. The French Agency for the Safety of Health
Products (AFFSAPS) suspended the use of medi-
1.4.2 Occupational exposure cations containing pioglitazone in 2011, on the
basis of a French study linking pioglitazone to
Occupational exposure to pioglitazone may cancer of the bladder (AFFSAPS, 2013). Following
occur through inhalation and dermal contact this study, the Federal Institute for Drugs and
at workplaces where pioglitazone is produced Medical Devices (BrFAM) in Germany also
326
recommended the suspension of sales of piogl- 1.6.2 Structural and molecular formulae and
itazone (BrFAM, 2011). relative molecular mass
(a) Rosiglitazone
1.6 Chemical and physical data on CH 3 O
N
rosiglitazone N
O
S
N H
1.6.1 Nomenclature
(a) Rosiglitazone O
C18H19N3O3S
Relative molecular mass: 357.43
(SciFinder, 2013)
Chem. Abstr. Serv. name: 2,4- (b) Rosiglitazone maleate
Thiazolidinedione, 5-[[4-[2-(methyl-2-pyrid- CH 3 O
N
inylamino)ethoxy]phenyl]methyl] (SciFinder, N
O
S
2013) N H . HO OH
O O
IUPAC Systematic Name: 5-[[4-[2- O
[Me t hy l(py r id i n-2-y l)a m i no]e t hox y] C18H19N3O3S. C4H4O4

phenyl]methyl]-1,3-thiazolidine-2,4-dione Relative molecular mass: 473.50
(Pubchem, 2013) From ONeil (2006), SciFinder (2013)
WHO INN: Rosiglitazone (WHO, 2007)
(b) Rosiglitazone maleate pure substance
(a) Rosiglitazone
(SciFinder, 2013) Description: Solid, colourless crystals from
Chem. Abstr. Serv. Name: 2,4- methanol (ONeil, 2006)
Thiazolidinedione, 5-[[4-[2-(methyl-2-pyrid- Melting point: Rosiglitazone: 151155 C
inylamino)ethoxy]phenyl]methyl]-, (2Z)-2- (ONeil, 2006)
butenedioate (1:1) (SciFinder, 2013) Density: 1.3150.06g/cm3 at 20 C and pres-
IUPAC Systematic Name: (Z)-But-2-enedioic sure 760 Torr (SciFinder, 2013)
acid;5-[[4-[2-[methyl(pyridin-2-yl)amino] Spectroscopy data: UV (Venkatesh et al.,
ethoxy]phenyl]methyl]-1,3-thiazolidine-2,4- 2006), 1H NMR, 13C NMR, IR (potassium
dione (Pubchem, 2013) bromide), and MS have been reported. (Pang
Proprietary Names: Gaudil (Pubchem, 2013); et al., 2009; Wang & Miksa, 2007)
Rezult (SciFinder, 2013); Avandia (GSK, 2012) Solubility: 30.67 g/mL in water at 25 C;
3.79g/mL in 0.1M phosphate buffer at 25C
(Seedher & Kanojia, 2008; Seedher & Kanojia,
2009); 35.09g/mL in 0.15M NaCl at 25C
(Seedher & Kanojia, 2008); soluble in the mg/
mL range in ethanol, dimethyl sulfoxide, and
dimethylformamide (Cayman SDS, 2013);
10.45mg/L in water at 25C (NLM, 2013)
327
IARC MONOGRAPHS 108
Octanol/water partition coefficient: Log 1.7 Analysis of rosiglitazone

P=2.783.02 (Giaginis et al., 2007)
Vapour pressure: 1.141013 mmHg at 25C Physical properties used for the identification
(SciFinder, 2013) of the substance, e.g. IR and melting point, are
presented in Section 1.6.3.
Selected non-compendial methods are
(b) Rosiglitazone maleate
presented in Table 1.1. Rosiglitazone can be
Description: White to off-white solid (ONeil, analysed in different matrices such as plasma,
2006; GSK, 2012) serum, urine and formulations, by HPLC and
with detection by UV or MS. Detection and
Melting point: 122123C (ONeil, 2006)
quantification limits for determination of rosigl-
Solubility: Readily soluble in ethanol and in itazone in human serum by HPLC method with
buffered aqueous solution at pH2.3; solubility UV detection are 0.033g/mL and 0.102g/mL,
decreases with increasing pH in the physio- respectively (Sultana et al., 2011). Rosiglitazone
logical range (ONeil, 2006; GSK, 2012) can be analysed in human plasma with lower
Stability data: Stable for 2years when stored limit of quantification of 1.00 ng/mL using
at 4 C. Stock solutions are stable for up to liquid chromatography electrospray ionization
3months when stored at 20C (Enzo PDS, mass spectrometry (LC-ESI-MS) (OMaille et al.,
2012) 2008). Other analytical methods for detection in
human urine include square-wave adsorptive
1.6.4 Technical products and impurities stripping voltammetry method (Al-Ghamdi &
Hefnawy, 2012); analysis of formulation can be
Rosiglitazone maleate is used to formulate also achieved with capillary electrophoresis with
the finished dosage forms described below. UV detection (Yardmc et al., 2007) or by UV
spectroscopy (Sireesha et al., 2011).
(a) Trade names
Avandia (GSK, 2012); Roglit 4; Romerol;
Rosit-2; Sensulin; Tazone-4 (BDdrugs, 2013)
1.8 Production and use of
rosiglitazone
(b) Impurities
1.8.1 Production
Desmethyl impurity or 5-(4-(2-(pyridin-2-yl-
amino)ethoxy)benzyl)thiazolidine-2,4-dione Rosiglitazone is produced by dissolving
(Krishna et al., 2008) 2-N-methyl-2-pyridylaminoethanol in dim-
ethylformamide and adding sodium hydroxide
Dimer impurity or 5-((2,4-dioxothiazo-
lid in-5-yl)(4 -(2-(met hyl(py rid in-2-yl) under atmospheric nitrogen (Cantello
amino)ethox y)phenyl)methyl)-5-(4-(2- et al., 1994; Pubchem Substance, 2013).
(methyl(pyridin-2-yl)amino)ethoxy)benzyl) 4-Fluorobenzaldehyde is added. The resulting
thiazolidine-2,4-dione (Krishna et al., 2008) mixture is then dissolved in toluene and
Succinate impurity or 2-(5-(4-(2-(methyl- piperidine is added. The compound obtained
(p y r id i n-2 -y l) a m i no) e t hox y)b e n z y l) is dissolved in dioxane and hydrogenated at
-2,4-dioxothiazolidin-3-yl)succinic acid room temperature and atmospheric pressure.
(Krishna et al., 2008) Rosiglitazone is separated by filtration, vacuum
4-(2-(Methyl(pyridin-2-yl)amino)ethoxy) concentration and recrystallization.
benzaldehyde (Radhakrishna et al., 2002b)
328
1.8.2 Use medication in September 2010 (EMA, 2010).

Countries with appreciable continuing sales
(a) Indications included China (US$12 million), Canada (US$11
Rosiglitazone maleate is an antidiabetic agent million), Mexico (US$ 7 million), Australia
indicated for the improvement of glycaemic (US$ 3 million), USA (US$ 3 million), and
control in adults with type 2 diabetes mellitus, Argentina (US$3 million). The use of rosiglita-
as an adjunct to diet and exercise. Rosiglitazone zone rapidly declined in the USA after increased
maleate decreases hyperglycaemia by reducing risk of cardiovascular disease associated with
insulin resistance in the presence of endogenous use of rosiglitazone was reported in May 2007
insulin (Avandia, 2010; Sweetman, 2011). (Nissen & Wolski, 2007; Singh et al., 2007b).
In the USA, rosiglitazone maleate is only Rosiglitazone was reported as 105000 drug
indicated for patients already using rosiglitazone, uses in the USA in 2012, down from 6.7 million
or for those not taking rosiglitazone and who uses in 2005 (IMS Health, 2012b; see Fig. 1.2).
have been unable to achieve adequate glycaemic Approximately 10000 patients in the USA were
control using other diabetes medications, or who taking rosiglitazone in 2012, down from about
have decided not to take pioglitazone (Avandia, 40000 in 2011 (IMS Health, 2012b).
2010). In the USA, the most commonly reported Prescribing trends from the Netherlands
indication for rosiglitazone is type 2 diabetes show a decline in prescriptions for rosiglita-
mellitus (see Table1.2). zone after regulatory warnings (Ruiter et al.,
Rosiglitazone is also used for off-label indi- 2012). Since the marketing authorization for
cations such as polycystic ovarian syndrome and rosiglitazone was suspended, rosiglitazone is no
insulin resistance syndrome. longer available for use in Europe (EMA, 2010).
Side-effects include fluid retention, conges- Prescriptions for rosiglitazone have also declined
tive heart failure, and liver disease (Pubchem in several other countries across Asia, such as
Substance, 2013). Taiwan, China (Lu & Li, 2013).
(b) Dosage
Rosiglitazone is available as tablets of 2mg,
1.9 Occurrence and exposure to
4 mg, and 8 mg. Rosiglitazone is also avail- rosiglitazone
able in combination with glimeperide or with 1.9.1 Natural occurrence and environmental
metformin. Rosiglitazone is started at a dose of
fate
4mg and can be titrated up to a dose of 8mg, if
inadequate response is obtained in combination Rosiglitazone is not reported to occur natu-
with metformin or sulfonylureas (Sweetman, rally. The production and use of rosiglitazone
2011). may result in its release to the environment
through various waste streams. If released to air,
(c) Trends in use it will exist solely in the particulate phase and
Total worldwide sales of rosiglitazone in will be removed from the atmosphere by wet or
2012 were US$43 million, a decline from much dry deposition (Pubchem Substance, 2013).
higher levels in the previous decade (IMS Health,
2012a). In 2012, there were limited, if any, sales
in most countries of the European Union as a
consequence of the decision of the European
Medicines Agency to suspend marketing of this
329
IARC MONOGRAPHS 108
Fig.1.2 Trends in use of rosiglitazone by office-based physicians in the USA

2000
1600
1200
800
400
0
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
Prepared by the Working Group from data obtained from IMS Health, National Disease and Therapeutic Index (IMS Health, 2012b).
1.9.2 Occupational exposure Europe (EMA, 2010), where rosiglitazone is no

longer in use after being linked in several studies
Occupational exposure to rosiglitazone may to an increase in the risk of myocardial infarc-
occur through inhalation of dust and via dermal tion (Nissen & Wolski, 2007; Singh et al., 2007b).
contact at workplaces where rosiglitazone is In June 2013, an FDA advisory committee
produced or used (Pubchem Substance, 2013). No meeting was held to discuss the re-adjudication
information was available to the Working Group of data on cardiovascular events associated with
on the potential number of workers exposed. rosiglitazone from a large randomized controlled
trial (FDA, 2013c). The advisory committee
1.10 Regulations and guidelines for recommended that restrictions on rosiglitazone
rosiglitazone be lessened, since re-adjudication did not reveal
a statistically significant increase in the risk of
In the USA, rosiglitazone is only available cardiovascular events.
under a REMS, and approved by FDA on the
basis of safety and effectiveness (DHHS/FDA,
2007; Woodcock et al., 2010). The marketing
authorization for rosiglitazone was withdrawn in
330
2. Cancer in Humans The French health insurance databases

SNIIRAM (Systme national dinformation
Thiazolidinediones (rosiglitazone, pioglita- inter-rgimes de lAssurance maladie) and PMSI
zone and troglitazone) have been used as orally (Programme de mdicalisation des systmes din-
administered glucose-lowering drugs in patients formation) cover all employees and represent
with type 2 diabetes mellitus. Rosiglitazone and approximately 75% of the French population.
pioglitazone were introduced to the market at These databases contain all reimbursement data
about the same time (1999 in the USA, and 2001 for the patients health expenditure, including
and 2002 for rosiglitazone and pioglitazone, medication and outpatient medical and nursing
respectively, in Taiwan, China; Tseng, 2012d). care prescribed or performed by health-care
Some patients may therefore have been exposed professionals. International Classification of
to both drugs, which were sometimes prescribed Diseases 10th Revision (ICD-10) codes are
sequentially. Troglitazone was marketed for only applied in the databases and hospital discharge
a short period (19972000), before being with- information can be linked. To evaluate the asso-
drawn from the world market subsequent to ciation between the use of pioglitazone or rosigl-
reports of fatal hepatotoxicity (Julie et al., 2008). itazone and risk of various cancers, a cohort of
There was a concern regarding the poten- 1491060 diabetic patients (aged 4079 years on
tial for ascertainment bias in the observational 31 December 2006) from this national health
studies, since differences in the intensity and insurance scheme was created. Patients included
frequency of ascertainment between the piogli- had filled at least one prescription for an anti-
tazone and control groups were unknown. Since diabetic drug (i.e. metformin, sulfonylurea,
pioglitazone is associated with an increased risk pioglitazone, rosiglitazone, other oral antidi-
of oedema and congestive heart failure, patients abetic drugs and/or insulin) in 2006. Patients
taking pioglitazone were potentially more likely were excluded if they had cancer of the bladder
to undergo more frequent urine analysis, which diagnosed before study entry or within the first
could lead to detection of microscopic haema- 6months after study entry. Diagnosis of cancer
turia, more frequent cystoscopies, and eventually of the bladder or other cancers was followed up
a diagnosis of cancer of the bladder. until 31 December 2009 (Neumann et al., 2012).
Pioglitazone and rosiglitazone may have [The Working Group noted that the period of
different effects on the risk of cancer and the follow-up was only 3years. It was unclear which
Working Group therefore evaluated these drugs patients may have used in the past, before
compounds separately, whenever data were avail- enrollment into the cohort.]
able. Studies that reported results for non-specific In the United Kingdom, The Health
thiazolidinediones were considered uninforma- Improvement Network (THIN) database (since
tive by the Working Group and are not cited in 2003), managed by the Medicines and Healthcare
this Monograph. Products Regulatory Agency, MHRA) is similar
Several of the studies on these agents were in structure and content to the General Practice
based on analyses of large databases from France, Research Database (GPRD, 19942002), which
the United Kingdom, the USA, and Taiwan, provides electronic medical records of approxi-
China, which are briefly described below. mately 10 million patients living in the United
Associations of multiple cancers with specific Kingdom [the Working Group estimated a 50%
thiazolidinediones were reported (see Table2.1) overlap in the two databases]. Data available
include demographic information, medical diag-
noses (using Read codes, a standard classification
331
332
Table 2.1 Cohort studies of cancer and exposure to pioglitazone or rosiglitazone
Reference Total No. of Exposure Organ site (ICD Exposure categories Exposed Relative risk Covariates
Location, subjects assessment code) cases (95% CI) Comments
follow-up
period
Bladder cancer
Lewis et al. 193099 Prescription Bladder (linkage Pioglitazone Included diabetic patients aged
(2011)a records with KPNC Never use 791 1.00 (ref.) 40 yr between 1997 and 2002.
KPNC, USA, cancer registry) Ever use 90 1.2 (0.91.5) Numbers of cases reported
19972008 as median incidence rate per
IARC MONOGRAPHS 108
Time since starting (mo):

100000 person-years.
<18 NR 1.2 (0.81.7)
1836 NR 1.4 (0.92.1)
>36 NR 1.3 (0.91.8)
P for trend 0.07
Duration of therapy
(mo):
<12 NR 0.8 (0.61.3)
1224 NR 1.4 (0.92.1)
>24 NR 1.4 (1.032.0)
P for trend 0.03
Cumulative dose (mg):
110500 NR 1.0 (0.71.5)
1050128000 NR 1.2 (0.81.8)
>28000 NR 1.4 (0.962.1)
P for trend 0.08
Neumann 1491060 Prescription Bladder Pioglitazone (+) vs (): Age, sex (when applicable), and
et al. (2012) (pioglitazone records (discharge Both sexes 2016 1.22 (1.051.43) exposure to glucose-lowering
France, exposed: diagnosis with Men 1790 1.28 (1.091.51) drugs.
20069 n=155535) ICD-10 C67, Patients with diabetes in two
Women 226 0.78 (0.441.37)
combined large national linked databases:
with specific Rosiglitazone (+) vs (): health insurance system
aggressive Both sexes 2016 1.08 (0.921.26) (SNIIRAM) and hospitalization
treatment) Men 1790 1.10 (0.931.30) (PMSI), 20062009. Age range:
Women 226 0.89 (0.531.49) 4079 yr; lack of consideration
of potential confounders like
smoking and comorbidities.
follow-up
period
Neumann et Pioglitazone (), both 1.00 (ref.) Age, sex (when applicable),
al. (2012) sexes level of pioglitazone use (i.e.
France, Cumulative dose (mg): cumulative dose and duration
20069 <10500 NR 1.12 (0.891.40) of exposure, respectively) and
(cont.) exposure to other glucose-
1050027999 NR 1.20 (0.931.53)
lowering drugs
28000 NR 1.75 (1.222.50)
Duration of exposure (d):
<360 NR 1.05 (0.821.36)
360719 NR 1.34 (1.021.75)
720 NR 1.36 (1.041.79)
Pioglitazone (), men 1.00 (ref.)
<10500 NR 1.17 (0.921.48)
1050027999 NR 1.24 (0.961.60)
28000 NR 1.88 (1.302.71)
<360 NR 1.10 (0.841.43)
360719 NR 1.39 (1.061.84)
720 NR 1.44 (1.091.91)
Pioglitazone (), women 1.00 (ref.)
<10500 NR 0.77 (0.361.65)
1050027999 NR 0.84 (0.352.06)
28000 NR 0.57 (0.084.11)
<360 NR 0.76 (0.341.72)
360719 NR 0.87 (0.322.35)
720 NR 0.71 (0.222.23)
333
334
follow-up
period
Wei et al. 207714 Prescription Bladder Pioglitazone (yes vs no) 869 1.16 (0.831.62) Type 2 diabetes patients aged
(2013)b (pioglitazone records (database 40 yr. HR, 1.22 (95% CI,
General exposed, records) 0.801.84) in a propensity-
Practice 23548; matched analysis done in a group
Research unexposed, of patients without missing data
IARC MONOGRAPHS 108
Database, 184166) on baseline characteristics

United
Kingdom,
200110
Tseng (2012a)c 54928 Medical Bladder Pioglitazone
Taiwan, reimbursement (ICD-9 188) Never-users 155 1.00 (ref.)
China, records in the Ever-users 10 1.31 (0.662.58)
20069 Taiwan, China,
Time since starting (mo):
National Health
Insurance <18 4 1.38 (0.493.83)
database 1836 5 1.54 (0.623.84)
>36 1 0.65 (0.094.77)
P for trend 0.6352
Duration of therapy
(mo):
<12 8 1.54 (0.733.26)
12 2 0.82 (0.203.35)
P for trend 0.6919
110500 8 1.45 (0.693.06)
>10500 2 0.94 (0.233.84)
P for trend 0.7125
follow-up
period
Tseng (2013a)d 547584 Medical Bladder (ICD-9 Pioglitazone (yes vs no) 1869 1.02 (0.751.39) Benign prostatic hyperplasia
Taiwan, diabetic men reimbursement 188) Rosiglitazone (yes vs no) 1869 1.12 (0.921.37) is a significant risk factor for
China, records in the bladder cancer in diabetic men.
20069 Taiwan, China, The hazard ratios for pioglitazone
National Health and rosiglitazone are estimated
Insurance in diabetic men with benign
database prostatic hyperplasia.
Fujimoto et al. 21335 Hospital Bladder Pioglitazone (yes vs no) 170 1.75 (0.893.45) NR, probably not adjusted.
(2013) records Single centre, lack of adjustment
Japan, for confounders.
200011
Colorectal cancer
Neumann 1485146 Prescription Colorectum Pioglitazone (+) vs () 10618 0.97 (0.901.05) Age, sex (when applicable), and
et al. (2012)e records (ICD-10 C18 to Rosiglitazone (+) vs () 10618 0.88 (0.820.95) exposure to glucose-lowering
France, C21) drugs
20069
Tseng (2012b)e 995843 Reimbursement Colon (ICD-9 Pioglitazone (yes vs no) 3 versus 0.78 (0.252.49)
Taiwan, databases 153) 2386
China, 20035 Rosiglitazone (yes vs no) 29 vs 1.22 (0.811.84)
2360
Ferrara et al. 252467 Prescription Colon Never-use of other TZD 1.00 (ref.)
(2011) records Ever-use of other TZD 1260 1.1 (0.81.5)
KPNC, USA, Never-use of pioglitazone 1.00 (ref.)
19972005
Ever-use of pioglitazone 1260 0.9 (0.71.1)
Lung cancer
Ferrara et al. 252467 Prescription Lung/bronchus Never-use of other TZD 1.00 (ref) Ten categories of cancer sites,
(2011)f records (linkage with Ever-use of other TZD 1637 0.9 (0.61.3) diabetic patients aged 40 yr
KPNC, USA, KPNC cancer Never-use of pioglitazone 1.00 (ref.)
19972005 registry)
Neumann 1493472 Prescription Lung (ICD-10 Pioglitazone (+) vs () 9298 0.94 (0.871.02) Age, sex, and exposure to
et al. (2012) records C33 and C34) Rosiglitazone (+) vs () 9298 0.91 (0.840.99) glucose-lowering drugs
France,
20069
335
336
follow-up
period
Prostate cancer
Ferrara et al. 252467 Prescription Prostate (linkage Never-use of other TZD 1.00 (ref.) Ten categories of cancer sites.
(2011)f records with KPNC Ever-use of other TZD 2105 1.0 (0.71.3) Diabetic patients aged 40 yr
KPNC, USA, cancer registry) Never-use of pioglitazone 1.00 (ref.)
IARC MONOGRAPHS 108
19972005
Tseng (2011)g 494630 Medical Prostate Pioglitazone (yes vs no) 889 0.77 (0.105.75)
Taiwan, reimbursement (ICD-9 185) Rosiglitazone (yes vs no) 889 0.88 (0.431.80)
China, 20035 records in the
Taiwan, China,
National Health
Insurance
database
Breast cancer
Ferrara et al. 252467 Prescription Female breast Never-use of other TZD 1.0 (ref.)
(2011) records (linkage with Ever-use of other TZD 1561 0.9 (0.71.2)
KPNC, USA, KPNC cancer Never-use of pioglitazone 1.0 (ref.)
19972005 registry)
Neumann 671510 Prescription Female breast Pioglitazone (+) vs () 6820 0.91 (0.831.00) Age, sex, and exposure to
et al. (2012) records (ICD-10 C50) Rosiglitazone (+) vs () 6820 0.80 (0.730.88) glucose-lowering drugs
France,
20069
follow-up
period
Other cancers
Ferrara et al. 252467 Prescription Linkage with Never-use of other TZD 1.00 (ref.)
(2011) records KPNC cancer
KPNC, USA, registry
19972005 NHL Ever-use of other TZD 569 0.7 (0.41.2)
Corpus uterus Ever-use of other TZD 552 1.2 (0.81.9)
Pancreas Ever-use of other TZD 431 1.0 (0.61.8)
Kidney/renal Ever-use of other TZD 430 1.3 (0.72.3)
pelvis
Rectum Ever-use of other TZD 390 0.7 (0.41.5)
Melanoma Ever-use of other TZD 373 1.0 (0.51.8)
Never-use of pioglitazone 1.00 (ref.)
NHL Ever-use of pioglitazone 569 1.3 (1.01.8)
Corpus uterus Ever-use of pioglitazone 552 1.1 (0.81.5)
Pancreas Ever-use of pioglitazone 431 1.2 (0.81.7)
Kidney/renal Ever-use of pioglitazone 430 0.7 (0.41.1)
pelvis
Rectum Ever-use of pioglitazone 390 1.2 (0.81.8)
Melanoma Ever-use of pioglitazone 373 1.3 (0.92.0)
Neumann 1495787 Prescription Kidney (ICD-10 Pioglitazone (+) vs () 2861 0.91 (0.791.06) Age, sex, and exposure to
et al. (2012) records C64) Rosiglitazone (+) vs () 2861 0.98 (0.861.13) glucose-lowering drugs
France, 1495411 Head and neck Pioglitazone (+) vs () 2868 0.85 (0.730.99) Age, sex, and exposure to
20069 (ICD-10 C00 to glucose-lowering drugs
Rosiglitazone (+) vs () 2868 0.79 (0.670.92)
C14)
Tseng (2012c)h 999730 Reimbursement Thyroid (ICD-9 Pioglitazone (yes vs no) 943 0.52 (0.073.93)
Taiwan, databases 193) Rosiglitazone (yes vs no) 0.67 (0.231.95)
China,
19962005
337
338
follow-up
period
Tseng (2013c)i 998540 Reimbursement Oral cavity, lip, Pioglitazone (yes vs no) 766
Taiwan, databases and pharynx Men 1.70 (0.2213.20)
China, 20035 (ICD-9 140, 141, Women no incident cases
143, 144, 145, of oral cancer
146, 148, and
IARC MONOGRAPHS 108
Rosiglitazone (yes vs no)

149)
Men 1.15 (0.443.04)
Women 0.90 (0.203.98)
a Age, sex, race/ethnicity, current smoking, renal function, bladder condition, congestive heart failure, income, baseline A1C, newly diagnosed with diabetes at start of follow-up,
duration of diabetes, other cancer before baseline, other diabetic medications (other TZDs, metformin, sulfonylureas, other oral hypoglycaemic drugs, insulin).
b Age, sex, duration of diabetes, smoking status and BMI before entry into the study, and insulin treatment and number and type of different oral hypoglycaemic drug classes used
during the follow-up period.

c Age, sex, diabetes duration, nephropathy, urinary-tract disease, hypertension, COPD, cerebrovascular disease, IHD, peripheral arterial disease, eye disease, dyslipidaemia, heart
failure, rosiglitazone, sulfonylurea, meglitinide, metformin, acarbose, insulin, statin, fibrate, angiotensin-converting enzyme inhibitor/angiotensin-receptor blocker, Ca-channel
blocker, region of residence, occupation, and other cancer before baseline.
d Age, diabetes duration, nephropathy, urinary-tract diseases, hypertension, COPD, cerebrovascular disease, IHD, peripheral arterial disease, eye disease, dyslipidaemia, heart failure,
obesity, alcohol-related diagnosis, non-alcohol-related chronic liver disease, rosiglitazone/pioglitazone, sulfonylurea, meglitinide, metformin, acarbose, insulin, statin, fibrate, ACEI/
ARB, Ca-channel blockers, -blockers, 5- reductase inhibitors, clopidogrel, ticlopidine, dipyridamole, cyclophosphamide, diuretics, other cancer before baseline and potential
detection examinations.
e Age, sex, diabetes, hypertension, COPD, asthma, stroke, nephropathy, IHD, peripheral arterial disease, eye disease, dyslipidaemia, obesity, statin, fibrate, ACEI/ARB, Ca-channel
blocker, aspirin, dipyridamole, clopidogrel/ticlopidine, NSAIDs, sulfonylurea, metformin, insulin, acarbose, rosiglitazone, region of residence, occupation, and colon-cancer detection
examinations.
f Age, ever use of other diabetes medications, year of cohort entry, sex, race/ethnicity, income, current smoking, baseline HbA1c, diabetes duration, new diabetes diagnosis, creatinine,
and congestive heart failure.

g Age, diabetes duration, hypertension, COPD, stroke, nephropathy, IHD, peripheral arterial disease, eye disease, obesity, dyslipidaemia, statin, fibrate, ACEI/ARB, Ca-channel blocker,
sulfonylurea, metformin, insulin, acarbose, rosiglitazone, region of residence, and occupation.

h Age, sex, diabetes, living region, occupation, detection examination, hypertension, COPD, stroke, nephropathy, IHD, peripheral arterial disease, eye disease, obesity, dyslipidaemia,
benign thyroid disease, other cancer, sulfonylurea, metformin, insulin, acarbose, pioglitazone/rosiglitazone, statin, fibrate, ACEI/ARB, Ca-channel blocker, aspirin, ticlopidine,
clopidogrel, NSAIDs.
i Age, diabetes, obesity, hypertension, COPD, alcohol-related diagnoses, stroke, nephropathy, IHD, peripheral arterial disease, eye disease, dyslipidaemia, statin, fibrate, ACEI/ARB,
Ca-channel blockers, sulfonylurea, metformin, insulin, acarbose, pioglitazone/rosiglitazone, living region, occupation, potential detection.
ACEI/ARB, angiotensin-converting enzyme inhibitors/angiotensin receptor blockers; BMI, body mass index; Ca, calcium; COPD, chronic obstructive pulmonary disease; d, day;
HbA1c, glycated haemoglobin; HR, hazard ratio; IHD, ischaemic heart disease; KPNC, Kaiser Permanente Northern California; mo, month; NHL, non-Hodgkin lymphoma; NR, not
reported; NSAIDs, nonsteroidal anti-inflammatory drugs; PMSI, Programme de mdicalisation des systmes dinformation; ref., reference; SNIIRAM, Systme national dinformation
inter-rgimes de lAssurance maladie; TZD, thiazolidinediones; vs, versus; yr; year
system used in the United Kingdom primary Insurance of Taiwan, China, for evaluating the
care settings), lifestyle, and measures taken association between use of thiazolidinediones
during clinical practice. The database is regu- and risk of various cancers. Since March 1995,
larly updated and practitioners contributing a compulsory and universal system of health
data receive training for consistency in data insurance (National Health Insurance) has been
recording. Studies conducted by Wei et al. (2013) implemented in Taiwan, China. All contracted
and Azoulay et al. (2012) used this database. medical institutes must submit computerized
In the USA, the Kaiser Permanente Northern and standard claim documents for reimburse-
California (KPNC) pharmacy database covers ment. More than 99% of the population of 23
3.2 million members of this health plan (approx- million people were enrolled in this insurance
imately 30% of the population of the area) and system, and > 98% of the hospitals nationwide
includes outpatient prescriptions dispensed were under contract with the insurance. The
at a KPNC pharmacy. Approximately 95% of average number of annual physician visits in
members fill their prescriptions at KPNC phar- Taiwan, China, is one of the highest around the
macies. The KPNC diabetes registry gathers world, at approximately 15 visits per year per
longitudinal electronic medical records and capita in 2009. The National Health Research
clinically related data for patients with diabetes Institute is the only institute approved, as per
from the following sources: primary hospi- local regulations, for handling these reimburse-
tal-discharge diagnoses of diabetes, two or ment databases for academic research. The data-
more outpatient-visit diagnoses of diabetes, any bases contain detailed records on every visit for
prescription of a diabetes-related medication, each patient, including outpatient visits, emer-
or any record of glycated haemoglobin (HbA1c) gency-department visits, and hospital admis-
> 6.7%. The database includes information of sion. The databases also include principal and
cancer registries, pharmacy records, laboratory secondary diagnostic codes, prescription orders,
records, and inpatient and outpatient medical and claimed expenses. Certain computerized
diagnoses. Patients who met any of the following databases, including a database from the national
criteria were eligible for forming a cohort for cancer registry (with a high level of complete-
the analysis of the association between pioglita- ness), are also available for data linkage. Most
zone use and risk of cancer referred to in this studies used the International Classification of
Monograph: (i) as of 1 January 1997, diagnosed Diseases Ninth Revision, Clinical Modification
with diabetes, aged 40 years, and members (ICD-9-CM) codes for disease or cancer diag-
of KPNC; (ii) diagnosed with diabetes, reached nosis, with or without data linkage with the
age 40 years between 1 January 1997 and 31 cancer registry. [The Working Group noted that
December 2002, and were KPNC members on there may have been a detection bias associated
their 40th birthday; or (iii) having diabetes and with this database because of the high number of
aged 40 years when joining KPNC between 1 patient visits per capita.]
January 1997 and 31 December 2002. A total of
193099 patients (30173 ever-users and 162926
never-users of pioglitazone) were followed and a
2.1 Cancer of the bladder
mid-point interim analysis was published in 2011 See Fig.2.1
(Lewis et al., 2011).
Several independent groups (Tseng, 2011,
2012a, 2012b, 2013a; Chang et al., 2012) used the
reimbursement databases of the National Health
339
IARC MONOGRAPHS 108
Fig.2.1 Risk of cancer of the bladder associated with ever-use or never-use of pioglitazone or
rosiglitazone, by study designa
aWeights are from random-effects analysis

2.1.1 Clinical trial were diagnosed within 1 year of randomization,

which might have precluded a causeeffect rela-
The potential risk of cancer of the bladder tionship.] After excluding one case of previously
associated with exposure to pioglitazone in diagnosed cancer of the bladder in the placebo
humans was first raised by the large randomized group (found later to show benign histology),
PROspective pioglitAzone Clinical Trial In Hillaire-Buys et al. (2011) recalculated the crude
macroVascular Events (PROactive), which showed relative risk of cancer of the bladder for pioglita-
an imbalance of incident cases of cancer of the zone versus placebo in PROactive as 2.83 (95% CI,
bladder, with greater number in patients rand- 1.027.85; P value for Fisher exact test=0.04). [In
omized to pioglitazone than to placebo (14 cases a randomized controlled trial, the confounders
versus 6 cases, or 0.5% versus 0.2%; P = 0.069) are balanced at baseline and any differences in
(Dormandy et al., 2005). [Insufficient data were ascertainment are likely to be non-differential.
provided to calculate a risk estimate. The aim of The Working Group noted these findings in light
the PROactive clinical trial was primarily to eval- of the very low incidence of cancer of the bladder.
uate macrovascular events associated with use of After excluding the eleven cases of cancer of the
pioglitazone. The Working Group also noted that bladder diagnosed within 1 year of randomiza-
11 out of the 20 cases of cancer of the bladder tion, one with benign histology in the placebo
340
group, and six with known risk factors for cancer excluded if they had an occupationally related
of the bladder, only three cases remained two in cancer of the bladder, or if they were diagnosed
the group receiving pioglitazone and one in the before entry or within the first 6 months after
placebo group.] study entry. Patients were followed up for a mean
of 39.9 months (27.4 months due to exposure),
2.1.2 Cohort studies starting 6months after study entry. Ten percent
of the patients (155 535 out of 1 491 060) took
See Table 2.1 a minimum of two prescriptions for pioglita-
Lewis et al. (2011) studied the incidence of zone over 6 consecutive months. Compared
cancer of the bladder among 193099 members of with non-use, the estimated hazard rate ratio
the KPNC health plan who were enrolled in the for cancer of the bladder associated with use of
plans diabetes registry and were aged 40 years pioglitazone was 1.22 (95% CI, 1.051.43) and
between 1997 and 2002. The registry routinely for rosiglitazone was 1.08 (95% CI, 0.921.26),
compiled electronic medical records data from after adjustment for age, sex, and exposure to
various sources, including cancer registries, phar- glucose-lowering drugs. Doseresponse analyses
macy records, laboratory records, and medical were available only for pioglitazone and showed
diagnoses. Incident cases of cancer of the bladder increasing hazard ratios with increasing dura-
were identified from 2002 to 2008 via the KPNC tion and cumulative dose. The hazard ratios for
cancer registry and ever-use of specific diabetes cumulative doses of <10500, 1050027999 and
medications (defined as two or more prescrip- 28000 mg compared with never-users of piogl-
tions within 6 months) was determined from the itazone was 1.12 (95% CI, 0.891.40), 1.20 (95% CI,
pharmacy database. In the mid-point interim 0.931.53) and 1.75 (95% CI, 1.222.50), respec-
analysis of a 10-year longitudinal KPNC cohort tively; and were 1.05 (95% CI, 0.821.36), 1.34
study in the USA (Lewis et al., 2011), the overall (95% CI, 1.021.75) and 1.36 (95% CI, 1.041.79),
hazard ratio was 1.2 (95% CI, 0.91.5) for cancer of respectively, for duration of exposure < 360,
the bladder for ever-users of pioglitazone versus 360719 and 720 days compared with never-
never-users; patients who used pioglitazone for users. [Smoking was not accounted for in the
>24 months showed a risk with adjusted hazard analyses and therefore may have confounded the
ratio of 1.4 (95% CI, 1.032.0). [Although adjust- reported results. Sex-specific analyses suggested
ment for smoking was a strength of this study, an association observed only in men, but not in
only current smoking was considered, which women. Data on smoking were not available for
may not have fully controlled for confounding.] adjustment. Since pioglitazone is usually used
Neumann et al. (2012) analysed the risk of as a second- or third-line antidiabetic drug,
cancer of the bladder associated with exposure users of pioglitazone may have had longer dura-
to pioglitazone and rosiglitazone in a cohort tion of diabetes, poorer glycaemic control, and
of 1491060 diabetic patients aged 4079 years higher rates of chronic diabetic complications
who had been prescribed at least one dose of and comorbidities. All these characteristics may
glucose-lowering drugs in 2006. Subjects were affect the risk of cancer of the bladder (Perez,
followed between 2006 and 2009 using the French 2013; Tseng, 2012d). The length of follow-up
national health insurance information system limited evaluation of the long-term impact of
(SNIIRAM) linked with the French hospital treatment.]
discharge database (PMSI). Overall, 2016 cases In a matched cohort study by Wei et al.
of cancer of the bladder were identified (men, (2013) that used a propensity score approach
1790 cases; and women, 226 cases). Patients were (derived from baseline characteristics of age, sex,
341
IARC MONOGRAPHS 108
smoking, body mass index, and diabetes dura- cancer of the bladder associated with the use of
tion), the association between use of pioglitazone thiazolidinediones (Tseng, 2012a, 2013a, b).
and risk of cancer of the bladder was assessed Tseng (2012a) followed a random sample
in patients with type 2 diabetes using the of 54 928 patients with type 2 diabetes in the
General Practice Research Database. Between reimbursement databases of the National Health
2001 and 2010, 207714 patients aged 40 years Insurance for 4years from 1 January 2006 to 31
were studied: 23 548 users of pioglitazone and December 2009. Among 165 incident cases of
184 166 patients receiving other antidiabetic cancer of the bladder, 10 (0.39%) were ever-users
medications. Follow-up started at the date of and 155 (0.30%) were never-users of pioglitazone,
first prescription for pioglitazone or other oral and were not necessarily using other antidiabetic
antidiabetic drugs during the study period and drugs. The hazard ratio for ever-users versus
ended in December 2010. Patients with a cancer never-users of pioglitazone was 1.31 (95% CI,
diagnosis before the entry date or less than 90 0.662.58) after adjustment for age, sex, diabetes
days of follow-up time were excluded. Incident duration, various comorbidities, and medica-
cases of cancer of the bladder were obtained from tions. Doseresponse relationships were also
general practitioner records during follow-up. evaluated, but no trend was observed. [Smoking
Hazard ratios were computed, comparing the and body mass index were not available for anal-
risk of developing cancer of the bladder in the yses from the databases.]
group receiving pioglitazone and in the group In a second study drawn from the entire data-
receiving treatment with other oral antidiabetic base of the National Health Insurance of Taiwan,
drugs. A propensity score matched analysis was China, Tseng (2013a) evaluated the risk of cancer
used in patients without missing data on base-
of the bladder associated with use of pioglitazone
line characteristics to minimize confounding by
and rosiglitazone in a subgroup of 85 152 men
indication (n=34498). The following potential
with type 2 diabetes and benign prostatic hyper-
confounders were included: smoking status, age,
plasia. The hazard ratios (HR) for cancer of the
sex, duration of diabetes from first diagnosis to
bladder among the diabetic patients with benign
the first treatment with oral antidiabetic drug
during the study period, body mass index before prostatic hyperplasia for ever-users of pioglita-
entry into the study, and insulin treatment and zone (HR, 1.02; 95% CI, 0.751.39) and rosigli-
number and type of different oral antidiabetic tazone (HR, 1.12; 95% CI, 0.921.37) were close
drug classes used during follow-up. During to 1.0. [The study was not primarily aimed at
the study period, 66 new cases of cancer of the analysing the risk of cancer of the bladder asso-
bladder (mean follow-up time, 3.5years) occurred ciated with use of pioglitazone or rosiglitazone,
in the pioglitazone group, and 803 cases in the and therefore no doseresponse relationship was
group receiving other oral antidiabetic drugs assessed. Smoking and body mass index could
(mean follow-up time, 5.3 years) (adjusted HR, not be adjusted for because of lack of such infor-
1.16; 95% CI, 0.831.62). [The use of a propensity mation in the databases. There was a concern
score to control for confounding by adjustment over overlapping of the study population with
was a strength of this study. There was a poten- that of Tseng (2012a).]
tial overlap in the studied population because the In a third study drawn from the entire data-
authors used a similar database to that used by base of the National Health Insurance of Taiwan,
Azoulay et al. (2012).] China, Tseng (2013b) evaluated the association
The National Health Insurance of Taiwan, between use of rosiglitazone and risk of cancer of
China, was used to conduct several analyses of the bladder after excluding patients who had ever
been exposed to pioglitazone. A total of 885236
342
patients with type 2 diabetes and receiving oral 2.1.3 Nested casecontrol studies
antidiabetic agents (except pioglitazone) and/or
insulin were studied for incidence of cancer of See Table 2.2
the bladder from 1 January 2006 to 31 December To determine whether the use of pioglita-
2009. Among these patients, 102926 were ever- zone was associated with an increased risk of
users and 782310 were never-users of rosiglita- incident cancer of the bladder in people with
zone, with 356 and 2753 incident cases of cancer type 2 diabetes, Azoulay et al. (2012) conducted
of the bladder, respectively. The hazard ratio a nested casecontrol analysis within a cohort of
for cancer of the bladder for ever-users versus 115727 people with type 2 diabetes in the United
never-users of rosiglitazone was 0.98 (95% CI, Kingdom General Practice Research Database.
0.871.104) after adjustment for age, sex, diabetes Participants were newly treated with oral hypo-
duration, various comorbidities, and medica- glycaemic agents between 1 January 1988 and 31
tions. Doseresponse relationships were also December 2009. All incident cases of cancer of
evaluated, but neither the P values for the hazard the bladder occurring during follow-up (n=470)
ratios of the categories nor the P values for trends were identified and 376 cases were matched to up
were significant. [This study evaluated use of to 20 controls (n=6699) on year of birth, year
rosiglitazone and risk of cancer of the bladder of cohort entry, sex, and duration of follow-up.
after excluding potential residual confounding Exposure was defined as ever-use of pioglitazone
from pioglitazone. The study used databases and/or rosiglitazone (defined by the presence of
covering the whole nation and spanning the at least one prescription between cohort entry
whole period since the start of rosiglitazone use and the year before the index date), along with
in Taiwan, China. However, data on smoking and measures of duration and cumulative dosage.
body mass index were not available for analyses. Analyses were adjusted for smoking status,
There was a concern regarding overlapping of the excessive alcohol use, obesity, HbA1c, previous
study population with that of Tseng (2012a). The bladder conditions, previous cancer (other than
follow-up duration of 4years may also have been non-melanoma skin cancer), Charlson comor-
too short.] bidity score, and ever-use of other antidiabetic
Fujimoto et al. (2013) identified nine cases of agents (metformin, sulfonylureas, insulin, and
cancer of the bladder in a cohort of 663 patients other oral hypoglycaemic agents). Overall,
who had taken pioglitazone, in a database of ever-use of pioglitazone was associated with an
21335 patients with type 2 diabetes from a single increased rate of cancer of the bladder (rate ratio,
Japanese hospital between 2000 and 2011. They 1.83; 95% CI, 1.103.05), with a positive expo-
reported a hazard ratio of 1.75 (95% CI, 0.893.45) sureresponse trend (P=0.030). The highest risk
for cancer of the bladder among pioglitazone was observed in patients exposed for > 24 months
users compared with all patients with diabetes. (RR, 1.99; 95% CI, 1.143.45) and in those with
[The Working Group noted that incident cases a cumulative dosage > 28 000 mg (RR, 2.54;
were defined as any bladder cancers diagnosed 95% CI, 1.056.14). [Enrolment of new users of
after onset of drug therapy. No information was diabetes medications, who may have had less
given about total follow-up time. Duration and severe disease, and adjustment for smoking were
dose of drug were only given for identified cases, potential strengths of this study. However, there
as were data about smoking status. No other was a potential overlap in the study population
details were given about confounders.] with that of Wei et al. (2013).]
343
344
Table 2.2 Casecontrol studies of cancer and exposure to pioglitazone or rosiglitazone
Study No. source assessment (ICD code) cases (95% CI) Comments
location cases (hospital,
and No. of population)
period controls
Bladder cancer
Azoulay 376 General Prescription Bladder (Read Never-use of any TZD 319 1.00 (ref.) Up to 20 controls
et al. 6699 Practice records codes) Exclusive ever-use of 19 1.83 (1.103.05) per case, matched on
(2012)a Research year of birth, year of
IARC MONOGRAPHS 108
pioglitazone
United Database Exclusive ever-use of 36 1.14 (0.781.68) cohort entry, sex, and
Kingdom, rosiglitazone duration of follow-up
1988
Ever-use of both pioglitazone 2 0.78 (0.183.29)
2009
and rosiglitazone
Pioglitazone
Cumulative duration (mo):
12 1 0.56 (0.074.42)
1324 2 3.03 (0.6314.52)
>24 16 1.99 (1.143.45)
P for trend 0.050
Cumulative dosage (mg):
10500 7 1.58 (0.693.62)
1050128000 6 1.66 (0.703.94)
>28000 6 2.54 (1.056.14)
P for trend 0.030
Rosiglitazone
Cumulative duration (d):
519 9 0.80 (0.401.62)
>5191022 13 1.33 (0.732.40)
>1022 14 1.34 (0.752.40)
P for trend 0.32
Cumulative dosage (mg):
2464 8 0.71 (0.341.49)
>24645152 15 1.50 (0.862.62)
>5152 13 1.27 (0.692.32)
P for trend 0.49
period controls
Chang 1583 Population Reimbursement Bladder Pioglitazone
et al. 6308 databases (linkage Never-user NR 1.00 (ref)
(2012)b through Ever-user NR 0.95 (0.701.29)
Taiwan, national
China, cancer Cumulative dosage (DDD):
20007 registry) <120 NR 0.83 (0.541.27)
120 NR 1.07 (0.721.57)
Cumulative duration (yr):
1 NR 0.87 (0.611.25)
12 NR 1.20 (0.652.22)
23 NR 0.84 (0.302.36)
3 NR 1.56 (0.514.74)
Cumulative dose duration:
High NR 1.13 (0.691.83)
Intermediate NR 1.06 (0.651.71)
Low NR 0.77 (0.481.25)
Rosiglitazone
Never-user NR 1.00 (ref.)
Ever-user NR 1.05 (0.811.36)
Cumulative dosage (DDD):
<120 NR 1.05 (0.771.45)
120 NR 1.05 (0.781.40)
1 NR 1.07 (0.801.42)
12 NR 1.26 (0.861.86)
23 NR 0.60 (0.331.08)
3 NR 1.14 (0.651.99)
High NR 1.09 (0.771.54)
Low NR 1.14 (0.821.58)
345
346
period controls
Song 329 Hospital Electronic Bladder Pioglitazone () 308 1.00 (ref.) Alcohol, smoking,
et al. (2012) 658 medical records (confirmed by Pioglitazone (+) 21 2.09 (0.2616.81) coexisting cancer,
Republic cytology) haemoglobin and
of Korea, albumin
IARC MONOGRAPHS 108
200511 Single centre

(severance hospital),
1:2 age-sex-matched
cases:controls; age,
>20 yr. No significant
differences in diabetes
duration, BMI,
and renal function
between cases and
controls
Hassan 122 Hospital Interviewed Liver TZD () 116 1.00 (ref.) Age, sex, race,
et al. (2010) 86 with a (pathologically TZD (+) 6 0.3 (0.10.7) education, cigarette
USA, structured confirmed) smoking, alcohol
20008 and validated drinking, HCV, HBV,
questionnaire and family history of
cancer
Chang 10741 Population Liver (linkage Pioglitazone Age- and sex-
et al. 41847 through Never-user 10267 1.00 (ref.) matched controls (
(2012)c national (non-use) 4 controls per case).
Taiwan, cancer Ever-user 474 0.83 (0.720.95) A significantly lower
China, registry) risk of liver cancer was
20007 Cumulative dosage (DDD): mainly observed in
<120 225 0.87 (0.721.05) diabetic patients with
chronic liver disease
120 249 0.80 (0.670.95)
and with higher
Cumulative duration (yr): cumulative dosage or
1 352 0.87 (0.741.02) longer duration of use
12 79 0.80 (0.591.07)
23 30 0.71 (0.451.14)
3 13 0.44 (0.230.86)
period controls
Chang et Rosiglitazone Age- and sex-
al. (2012)c Never-user of rosiglitazone 9154 (non- 1.00 (ref.) matched controls (
Taiwan, use) 4 controls per case).
China, Ever-user of rosiglitazone 1587 0.73 (0.650.81) A significantly lower
20007 risk of liver cancer was
(cont.) mainly observed in
<120 725 0.86 (0.750.98) diabetic patients with
120 862 0.64 (0.560.72) chronic liver disease
Cumulative duration (yr): and with higher
1 1034 0.78 (0.690.88) cumulative dosage or
12 330 0.66 (0.560.79) longer duration of use
23 135 0.59 (0.460.74)
3 88 0.64 (0.490.85)
Colorectal cancer
Chang 7200 Population Reimbursement Colorectum Pioglitazone Age- and sex-matched
et al. 28712 databases (linkage Never-user 6822 (non- 1.00 (ref.) controls ( 4 controls
(2012)d through use) per case)
Taiwan, national Ever-user 378 1.04 (0.911.20)
China, cancer
20007 registry)
<120 170 1.15 (0.951.39)
120 280 0.97 (0.821.16)
1 278 1.15 (0.981.34)
12 60 0.82 (0.611.11)
23 26 0.86 (0.551.33)
3 14 0.77 (0.431.39)
Rosiglitazone
Never-user 6127 1.00 (ref.)
Ever-user 1073 0.86 (0.760.96)
<120 434 0.89 (0.771.03)
120 639 0.83 (0.730.95)
347
348
period controls
Chang et Cumulative duration (yr):
al. (2012)d 1 678 0.91 (0.801.04)
Taiwan, 12 220 0.78 (0.650.94)
China,
IARC MONOGRAPHS 108
23 99 0.69 (0.540.88)
20007
(cont.) 3 76 0.83 (0.631.10)
Lung cancer
Chang 5361 Population Reimbursement Lung (linkage Pioglitazone
et al. 21313 databases through Never-user NR 1.00 (ref.)
(2012)e national Ever-user NR 1.14 (0.951.37)
Taiwan, cancer
China, registry)
20007 Cumulative dosage (DDD):
<120 NR 1.13 (0.891.44)
120 NR 1.20 (0.961.50)
Cumulative duration (yr)
1 NR 1.25 (1.011.53)
12 NR 1.09 (0.771.53)
23 NR 0.91 (0.511.61)
3 NR 0.20 (0.050.86)
High NR 1.00 (0.751.33)
Low NR 1.16 (0.881.51)
Rosiglitazone
Never-user NR 1.00 (ref)
Ever-user NR 1.12 (0.901.39)
<120 NR 1.37 (1.081.74)
120 NR 1.00 (0.801.25)
period controls
Chang et Cumulative duration (yr):
al. (2012)e 1 NR 1.26 (1.011.58)
Taiwan, 12 NR 0.83 (0.631.09)
China,
23 NR 0.96 (0.701.33)
20007
(cont.) 3years NR 1.17 (0.821.67)
Cumulative dose duration
High NR 0.95 (0.741.22)
Low NR 1.34 (1.051.72)
a Alcohol, obesity, smoking, HbA1c, previous bladder conditions, previous cancer (other than non-melanoma skin cancer), Charlson comorbidity score, and ever-use of other
antidiabetic agents (metformin, sulfonylureas, insulin, and other oral hypoglycaemic agents).
b Multivariable model with stepwise selection of covariates, including pioglitazone, rosiglitazone, short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea
(mean daily dosage in quartiles), number of oral antidiabetic agents, nephropathy, glinides, ACE inhibitors, chronic kidney disease, Ca-channel blockers, neuropathy.
c Multivariate model with stepwise selection of covariates, including pioglitazone, rosiglitazone, short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea
(mean daily dosage in quartiles), number of oral antidiabetic agents, chronic liver disease, statins, aspirin, -blockers, chronic kidney disease, glinides (oral antidiabetic agent),
nephropathy, cerebrovascular disease, Ca-channel blockers, cardiovascular disease, chronic lung disease.
d Multivariate model with stepwise selection of covariates, including pioglitazone, rosiglitazone, short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea
(mean daily dosage in quartiles), number of oral antidiabetic agents, glinides, nephropathy, neuropathy, chronic liver disease, statins, retinopathy, Ca-channel blockers, ACE inhibitors,
peripheral vascular disease, depression, -blockers, aspirin, chronic kidney disease, chronic lung disease, cerebrovascular disease.
e Multivariable model with stepwise selection of covariates, including pioglitazone, rosiglitazone, short-acting human insulin, metformin (mean daily dosage in quartiles), sulfonylurea
(mean daily dosage in quartiles), number of oral antidiabetic agents, chronic lung disease, glinides, retinopathy, Ca-channel blockers, chronic kidney disease, statins, angiotensin
receptor blockers, chronic liver disease, -glucosidase inhibitors.
ACE, angiotensin-converting enzyme; BMI, body mass index; Ca, calcium; d, day; DDD, defined daily doses; HbA1c, glycated haemoglobin; HBV, hepatitis B virus; HCV, hepatitis C
virus; mo, month; NR, not reported; ref., reference; TZD, thiazolidinediones; yr, year
349
IARC MONOGRAPHS 108
2.1.4 Casecontrol studies models. The important risk factor of tobacco

smoking could not be adjusted for, but chronic
See Table 2.2 lung disease, a proxy indicator of smoking, was
Chang et al. (2012) conducted a nationwide included as a covariate. The study population
casecontrol study to evaluate the risk of several potentially overlapped with that of Tseng (2012a,
malignancies in diabetic patients who received 2013a, b).]
thiazolidinediones (pioglitazone or rosiglita- In the Republic of Korea, Song et al. (2012)
zone). A total of 606 583 patients with type 2 conducted a casecontrol study in diabetic
diabetes, aged 30 years, without a history of patients with cancer of the bladder (n=329) who
cancer, were identified from the National Health presented at one hospital between November
Insurance claims database, Taiwan, China, 2005 and June 2011. Cases exposed to pioglita-
between 1 January 2000 and 31 December 2000. zone were matched by sex and age to 658 control
As of 31 December 2007, patients with incident patients without cancer of the bladder who were
cancer of the liver, colorectum, lung, or urinary listed on the hospital diabetes registry. The odds
bladder were included as cases, and up to four ratio for cancer of the bladder associated with
age- and sex-matched controls were selected by a history of pioglitazone use was 2.09 (95% CI,
risk-set sampling. Information was collected 0.2616.81) [but it was unclear whether this model
on prescribed drug types (according to the was adjusted for potential confounders]. [The
Anatomic Therapeutic Chemical classification Working Group considered that the description
system, A10BG02 for rosiglitazone and A10BG03 of the methods and statistical analysis was inad-
for pioglitazone), dosage, date of prescription, equate, with conflicting information and some
supply days, and total number of pills dispensed elements of the design (i.e. casecontrol study in
from the outpatient pharmacy-prescription data- a single tertiary hospital, history of pioglitazone
base. Approximately 26.1% of patients had ever or other medications prescribed before visiting
received rosiglitazone, and 14.1% had received this tertiary centre was not available, recall bias
pioglitazone. The mean cumulative duration in information on confounders from the retro-
was 522 days, and the mean daily dosage was spective nature of the study, opposite associa-
0.14 defined daily doses per day for rosiglitazone, tion between pioglitazone use and cancer of the
compared with 375 days and 0.11 defined daily bladder in univariate and multivariate analyses)
doses per day for pioglitazone. gave cause for concern.]
With 1583 cases of cancer of the bladder
and 70 559 diabetic controls, an increased risk
of cancer of the bladder was associated with
2.1.5 Other study designs
only 3 years of pioglitazone use (1.56; 95% CI, A study was conducted by Piccinni et al.
0.514.74). The hazard ratio for the highest cate- (2011) using the FDA Adverse Event Reporting
gory of cumulative dose duration was 1.13 (95% System, in the USA. Cases of cancer of the
CI, 0.691.83). [This study had a longer follow-up bladder were compared with all other reports
period than several others evaluating rosiglita- of adverse effects within the system; a reporting
zone and pioglitazone, but these drugs did not odds ratio of 4.30 (95% CI, 2.826.52) was esti-
become available in Taiwan, China, until after mated for cancer of the bladder associated with
follow-up began. The methods were not clearly pioglitazone compared with other antidiabetic
described and it was not clear how cumulative drugs, and was elevated in each sex separately.
dose was estimated. The investigators included [The Working Group noted that interpretation
chronic kidney disease and various drugs in the of these results was challenging because there
350
was no information about the population at risk: Longer duration of treatment (> 24 months)
adverse event reports for drugs may not be a (1.42, 1.171.72) and higher cumulative dose
representative sample of the population. Analyses (> 28 000 mg) of pioglitazone (1.64, 1.282.12)
using the Adverse Events Reporting System may were associated with a significantly higher risk.
suffer from notoriety bias, but this study was A meta-analysis by Ferwana et al. (2013)
conducted before concerns about pioglitazone included six studies and reported a hazard ratio
were well known.] of 1.23 (95% CI, 1.091.39) associated with use of
pioglitazone.
2.1.6 Meta-analyses
Several meta-analyses have evaluated the 2.2 Cancer of the liver
association between the use of thiazolidine-
There were no cohort studies evaluating
diones and cancer of the bladder (Zhu et al.,
the association between cancer of the liver and
2012; Colmers et al., 2012b; Bosetti et al., 2013;
specific drugs of the thiazolidinedione class.
Ferwana et al., 2013). [The Working Group found
In a casecontrol study, Chang et al. (2012)
it difficult to compare meta-analyses since each
(see Section 2.1.4 for description of study; see
included different, but partially overlapping,
Table 2.2), reported a lower risk of cancer of
studies, some of which were unpublished.]
the liver associated with both pioglitazone and
Zhu et al. (2012) conducted a meta-analysis
rosiglitazone. The adjusted odds ratio (OR) for
on the association between pioglitazone and
pioglitazone was 0.83 (95% CI, 0.720.95), and
cancer of the bladder from five studies, including
for rosiglitazone was 0.73 (95% CI, 0.650.81).
one prospective, randomized, controlled study
Odds ratios decreased with increasing categories
(Dormandy et al., 2005), three cohort studies
of duration of pioglitazone use (OR, 0.44; 95% CI,
(Lewis et al., 2011; Neumann et al., 2012; Tseng,
0.230.86 for 3 years). [Although several risk
2012a), and one casecontrol study (Chang et al.,
factors for cancer of the liver were accounted for
2012). There was no evidence for the presence of
in the analysis, important potential confounders
significant heterogeneity between the five studies
such as smoking, alcohol use, and hepatitis
(Q=2.68, P=0.61; I2=0.0%). The meta-relative
status, were not accounted for.]
risk was 1.17 (95% CI, 1.031.32) for all studies.
In patients with cumulative treatment exposure
to pioglitazone for > 24 months, the meta-rel- 2.3 Cancer of the colorectum
ative risk was 1.38 (95% CI, 1.121.70), and in
those with a cumulative dose of >28000mg, the 2.3.1 Casecontrol studies
meta-relative risk was 1.58 (95% CI, 1.122.06). See Table 2.2
A meta-analysis by Colmers et al. (2012b) In a multivariate analysis for risk of cancer
included unpublished results for cancer of the of the colorectum associated with use of piogli-
bladder associated with specific thiazolidinedi- tazone or rosiglitazone, Chang et al. (2012) (see
ones. The meta-relative risk for pioglitazone was Section 2.1.4 for description of study) reported
1.22 (95% CI, 1.071.39) and for rosiglitazone was odds ratios of 1.04 (95% CI, 0.911.20) associ-
0.87 (95% CI, 0.342.23). ated with pioglitazone use, and 0.86 (95% CI,
In a meta-analysis by Bosetti et al. (2013), the 0.760.96) associated with rosiglitazone use. The
meta-relative risk was 1.20 (95% CI, 1.071.34) magnitude of the odds ratio for the highest expo-
from six studies on pioglitazone, and 1.08 (95% sure duration of 3 years was very similar for
CI, 0.951.23) from three studies on rosiglitazone. rosiglitazone (OR, 0.83; 95% CI, 0.631.10) and
351
IARC MONOGRAPHS 108
for pioglitazone (OR, 0.77; 95% CI, 0.431.39). a multivariable analysis. Hazard ratios of 0.78
Furthermore, a trend of decreasing odds ratios (95% CI, 0.252.49) for pioglitazone users, and
with increasing cumulative duration of exposure 1.22 (95% CI, 0.811.84) for rosiglitazone users
was observed for both rosiglitazone and piogli- were reported.
tazone. [See Section 2.1.4 for the strengths and
limitations of this study.] 2.3.3 Meta-analyses
A meta-analysis derived from three observa-
2.3.2 Cohort studies
tional studies (Colmers et al., 2012a) calculated
See Table 2.1 a meta-risk ratio of 0.97 (95% CI, 0.901.04)
Ferrara et al. (2011) evaluated risk of cancer of for cancer of the colorectum associated with
the colorectum associated with pioglitazone use pioglitazone use (see Section 2.1.6 for further
in a cohort of 252467 male and female patients description).
aged 40 years in the KPNC diabetes registry (see
Section 2, introduction, for description). Data on
use of diabetes medications were obtained from
2.4 Cancer of the lung
the pharmacy clinical database, and the filling 2.4.1 Casecontrol studies
of two prescriptions of pioglitazone within 6
months was defined as ever used. Information See Table 2.2
was collected from electronic medical records In a multivariable analysis for cancer of the
about all confounders, except smoking, which lung, Chang et al. (2012) (see Section 2.1.4 for
was supplemented by postal-survey data. The description of study) reported an odds ratio for
hazard ratio for cancer of the colorectum asso- pioglitazone use of 1.14 (95% CI, 0.951.37) and
ciated with ever-use versus never-use of piogli- 1.12 (95% CI, 0.901.39) associated with rosigl-
tazone was 0.9 (95% CI, 0.71.1), after adjusting itazone use. A higher risk was reported with a
for a large number of potential confounders cumulative duration of 1 year for use of either
including current smoking, age, and ever-use pioglitazone or rosiglitazone, with adjusted odds
of other diabetes medications. [This study was ratios of 1.25 (95% CI, 1.011.53) and 1.26 (95%
based on the same population as Lewis et al. CI, 1.011.58), respectively. [The investigators
(2012). The authors were only able to examine included chronic kidney disease and various
recently initiated therapy and short-term use drugs in the models. The important risk factor
(median, 1.6years) of pioglitazone, although the of tobacco smoking could not be adjusted for,
latency period until development of cancer of the but chronic lung disease, a proxy indicator of
bladder may be longer.] smoking, was included as a covariate.]
Neumann et al. (2012) (see Section 2.1.2 for
description of study) investigated risk of cancer 2.4.2 Cohort studies
of the colorectum in users of pioglitazone or See Table 2.1
rosiglitazone compared with non-users, and In a multivariable analysis in the study by
reported hazard ratios of 0.97 (95% CI, 0.901.05) Ferrara et al. (2011) (see Section 2.3.2 for descrip-
for pioglitazone, and 0.88 (95% CI, 0.820.95) for tion of study), no effect was found on the incidence
rosiglitazone. of cancer of the lung or bronchus in pioglitazone
Tseng (2012b) (see Section 2.1.2 for descrip- users when compared with never-users (adjusted
tion of study) assessed risk of cancer of the colon HR, 1.0; 95% CI, 0.81.3). [Adjustment for
in thiazolidinedione users versus non-users in smoking was a strength of this study, but only
352
current smoking was considered, which may not 2.6 Cancer of the breast
have fully controlled for confounding.]
Similarly, in a cohort study of diabetic In the PROactive clinical trial (Dormandy
patients in France Neumann et al. (2012) (see et al., 2005; see Section 2.1.1 for description of
Section 2.1.2 for description of study) reported study), an imbalance in the number of cases of
no significant difference in the risk of cancer of cancer of the breast was noted, with three cases
the lung in users compared with non-exposed in the pioglitazone group and eleven in the
controls for pioglitazone (adjusted HR, 0.94; 95% placebo group. [Insufficient data were provided
CI, 0.871.02), or for rosiglitazone (adjusted HR, to calculate a risk estimate. The PROactive trial
0.91; 95% CI, 0.840.99). Hazard ratios were not was primarily aimed at evaluating macrovas-
adjusted for smoking. cular events associated with pioglitazone use.]
In a multivariable analysis in the study by
2.4.3 Meta-analyses Ferrara et al. (2011) (see Section 2.3.2 and 2.4.2
for comments and description of study; see
A meta-analysis of pulmonary malignan- Table 2.1), no effect was found in the incidence
cies by Monami et al. (2008) showed that the of cancer of the breast in pioglitazone users when
meta-odds ratio for rosiglitazone versus compar- compared with never-users (adjusted HR, 1.0;
ators in clinical trials was 0.67 (95% CI, 0.301.51). 95% CI, 0.81.3).
Colmers et al. (2012a) reported a meta-rel- Similarly, a cohort study of diabetic patients
ative risk for ever-users of pioglitazone versus in France (Neumann et al., 2012; see Section
never-users of 0.95 (95% CI, 0.881.02) from two 2.3.2 for comments and description of study; see
observational studies. Table 2.1) reported no significant difference in
the risk of cancer of the breast in pioglitazone
2.5 Cancer of the prostate users when compared with non-exposed controls
(adjusted HR, 0.91; 95% CI, 0.831.00). A signif-
See Table 2.1 icantly reduced risk of cancer of the breast was
No casecontrol studies evaluated the asso- found in rosiglitazone users when compared
ciation between use of pioglitazone or rosiglita- with non-exposed controls (adjusted HR, 0.80;
zone and cancer of the prostate. 95% CI, 0.730.88).
A cohort study by Ferrara et al. (2011) (see The meta-relative risk estimated by Colmers
Section 2.3.2 for description of study) found no et al. (2012a) for cancer of the breast in ever-
association when comparing ever-use of piogli- users of pioglitazone versus never-users from
tazone versus never-use (adjusted HR, 1.0; 95% two observational studies was 0.93 (95% CI,
CI, 0.81.2). Tseng (2011) (see Section 2 introduc- 0.851.01).
tion for description of study) reported an inverse
association for ever-use of pioglitazone versus
never-use (adjusted HR, 0.77; 95% CI, 0.105.75)
2.7 Other site-specific cancers
and for rosiglitazone (adjusted HR, 0.88; 95% CI, See Table 2.1
0.431.80). Several of the studies described above also
Two meta-analyses (Monami et al., 2008; reported on other site-specific cancers. The study
Colmers et al., 2012a) reported meta-relative risks by Ferrara et al. (2011) reported relative risks of
of around unity for cancer of the prostate associ- >1 for some other cancers.
ated with the use of pioglitazone or rosiglitazone. Neumann et al. (2012) reported a hazard ratio
of near unity for cancer of the kidney among
353
IARC MONOGRAPHS 108
users of pioglitazone or rosiglitazone, and hazard There was a significant positive trend towards
ratios of approximately 0.8 for both agents for increased mortality with increasing dose in
cancer of the head and neck. male mice. Increased incidences of benign pheo-
By using the National Health Insurance data- chromocytoma of the adrenal gland were seen
base of Taiwan, China, Tseng evaluated the asso- in exposed male mice, and increased incidences
ciation of pioglitazone and rosiglitazone with the of leiomyosarcoma of the uterine cervix were
risk of cancer of the thyroid (Tseng, 2012c), and seen in exposed female mice when compared
cancer of the oral cavity, lip, and pharynx (Tseng, with controls. [Although it was noted that the
2013c), respectively. The hazard ratio for cancer FDA identified these differences as being statis-
of the thyroid was 0.52 (95% CI, 0.073.93) for tically significant, the Working Group could not
pioglitazone and 0.67 (95% CI, 0.231.95) for confirm the statistical analyses because original
rosiglitazone (Tseng, 2012c); for cancer of the study data (e.g. mortality) were not available
oral cavity, lip, and pharynx, the hazard ratio (FDA, 1999a).]
was 1.70 (95% CI, 0.2213.20) for pioglitazone
and 1.15 (95% CI, 0.443.04) for rosiglitazone (b) Genetically engineered mouse
(Tseng, 2013c). Pino et al. (2004) reported increased inci-
In a meta-analysis by Colmers et al. (2012a), dence and multiplicity of adenoma of the large
the meta-risk ratio for cancer of the kidney was intestine in ApcMin/+ mice (a genetically engi-
0.89 (95% CI, 0.761.04) for ever-users of piogl- neered mouse model that overexpresses the Apc
itazone versus never-users from two observa- gene, leading to rapid development of intes-
tional studies. tinal neoplasms) with dietary exposure to any
of several peroxisome proliferator-activated
receptor (PPAR) agonists, including pioglita-
3. Cancer in Experimental Animals zone. In this study, male C57BL/6J-ApcMin/+ mice
(age, 67 weeks) were fed diets containing piogl-
3.1 Pioglitazone itazone at a concentration selected to achieve a
dose of 150 mg/kg bw per day for 8 weeks. All
3.1.1 Oral administration mice exposed to pioglitazone (15 out of 15, 100%
[P<0.01]) developed adenoma of the large intes-
(a) Mouse
tine, compared with 60% (9 out of 15 mice) in
See Table3.1 the dietary control group. Pioglitazone increased
As part of its pharmacology review of the the multiplicity of tumours of the large intestine,
New Drug Application package (NDA 21073) but not of the small intestine [data and statistics
for pioglitazone submitted by the Takeda were provided in graphical form]. [The Working
America Research and Development Center, the Group noted that, although used extensively as
FDA summarized the results of a 2-year study a model system for cancer chemoprevention,
that was performed to evaluate the potential the predictive value of the ApcMin/+ mouse in
carcinogenicity of pioglitazone in mice (FDA, the identification of agents that may promote or
1999a). In this study, groups of 60 male and 60 otherwise stimulate carcinogenesis in the human
female CD-1 mice [age not reported] received colon was unknown.]
pioglitazone by gavage at doses of 0 (vehicle),
0 (placebo suspension), 3, 10, 30, or 100mg/kg (c) Rat
body weight (bw) per day for 104 weeks. [Vehicle See Table3.2
and placebo suspension were not specified.]
354
Table 3.1 Studies of carcinogenicity in mice given pioglitazone orally
Strain (sex) Dosing regimen, Incidence of tumours Significance Comments

Reference
CD-1 (M,F) 0(vehicle control), 0(placebo Benign pheochromocytoma of the *P<0.05 (Peto test) The Working Group was unable
104wk suspension), 3, 10, 30, 100mg/kg bw adrenal gland: to confirm the statistical
FDA (1999a) per day by gavage in an unspecified 0/60, 0/60, 0/60, 0/60, 2/60*, 1/60* analyses.
vehicle or suspension. (M) Increase in mortality in exposed
60M and 60F/group Leiomyosarcoma of the uterine males.
cervix:
0/60, 0/60, 0/60, 0/60, 1/60*, 1/60* (F)
C57BL6J-ApcMin/+ Pioglitazone mixed in feed and Large intestine adenoma : *[P<0.01] Genetically engineered
(M) given to achieve daily doses of 0 or 9/15, 15/15* P0.05 for increased mouse sensitive to intestinal
8wk ~150mg/kg bw multiplicity of large carcinogenesis.
Pino et al. (2004) 15/group intestine tumour [data No increases in the multiplicity
and statistics read from of small intestine tumours.
graph]
bw, body weight; F, female, M, male; wk, week
355
356
Table 3.2 Studies of carcinogenicity in rats given pioglitazone orally

Reference
Sprague-Dawley 0(vehicle control), 0(placebo Transitional cell carcinoma of urinary bladder: P<0.025 (trend) (M) Increase in
(M,F) suspension), 1, 4, 8 (M only), 16, or 0/60, 0/60, 0/60, 2/60, 3/60, 5/60, 4/60 (M) mortality in
104wk 63mg/kg bw per day by gavage in an 0/60, 0/60, 0/60, 0/60, , 0/60, 0/60 (F) exposed M and F
FDA (1999a) unspecified vehicle or suspension Transitional cell papilloma of urinary bladder: NS
60M and 60F/group 0/60, 0/60, 0/60, 0/60, 4/60, 2/60, 2/60 (M);
0/60, 0/60, 1/60, 1/60, , 1/60, 0/60 (F)
IARC MONOGRAPHS 108
Fibrosarcoma of subcutis: P<0.025 (trend)

0/60, 0/60, 0/60, 0/60, 0/60, 2/60, 2/60 (M)
Subcutaneous lipoma: P<0.025 (trend)
0/60, 0/60, 1/60, 0/60, , 1/60, 3/60 (F)
CD (M) 0(control) or 16mg/kg bw by gavage in Urinary bladder *P0.05 (Peto test)
104wk citric acid granules for 85wk Papilloma: 0/78, 7/82*
Sato et al. (2011) 90/group Carcinoma: 0/78, 1/82
bw, body weight; F, female, M, male; NS, not significant; wk, week
As part of its pharmacology review of the New 3.1.2 Coexposure with modifying agents
Drug Application package (NDA 21073) for
pioglitazone submitted by the Takeda America See Table3.3
Research and Development Center, the FDA A group of 34 male and 35 female strain H
summarized the results of a 2-year study that Swiss mice was exposed by whole-body inhala-
was performed to evaluate the potential carcino- tion to mainstream cigarette smoke for 4months,
genicity of pioglitazone in rats (FDA, 1999a). starting 12 hours after birth, and then kept in
In this study, groups of 60 male and 60 female filtered air until the experiment was terminated
Sprague-Dawley rats [age not reported] received at age 7 months. After weaning (at age 45
pioglitazone by gavage at doses of 0 (vehicle), weeks), the mice also received diets containing
0 (placebo suspension), 1, 4, 8 (males only), 16, pioglitazone at a concentration of 120 mg/kg. A
or 63mg/kg bw per day for 104 weeks. [Vehicle control group of 34 male and 38 female mice was
and placebo suspension were not specified.] exposed to mainstream cigarette smoke only. In
There was a significant positive trend towards females, 5 out of 35 (P<0.01) mice exposed to
increased mortality with increasing dose in male pioglitazone developed adenoma of the kidney
and female rats. versus 0 out of 38 controls. In males, 3 out of 32
Treatment with pioglitazone caused a signif- mice developed kidney adenoma versus 0 out of
icant positive trend in the incidence of transi- 34 controls (La Maestra et al., 2013).
tional cell carcinoma of the urinary bladder in
male rats. Although female rats did not demon- 3.2 Rosiglitazone
strate an increased incidence of transitional
cell tumours of the urinary bladder, urothelial 3.2.1 Oral administration
hyperplastic lesions were identified in male and (a) Mouse
female rats exposed to pioglitazone. In addition,
pioglitazone induced a small but significant posi- See Table3.4
tive trend in the incidence of fibrosarcoma of the As part of its pharmacology review of
subcutis in male rats, and a significant positive the New Drug Application package (NDA
trend in the incidence of subcutaneous lipoma in 21071) for rosiglitazone that was submitted by
female rats (FDA, 1999a). SmithKline Beecham Pharmaceuticals, the FDA
Sato et al. (2011) reported a study in which summarized the results of a 2-year feeding study
two groups of 90 male CD rats (age, 6 weeks) that was performed to evaluate the potential
received pioglitazone (in citric acid granules) by carcinogenicity of rosiglitazone in mice (FDA,
gavage at a dose of 0 (control), or 16mg/kg bw per 1999b). In this study, groups of 60 male and 60
day for 85 weeks, followed by a 19-week observa- female CD-1 mice [age not reported] received
tion period. There was a significant increase in diet supplemented with rosiglitazone at concen-
the incidence of papilloma of the urinary bladder trations that were selected to provide doses of 0
in the exposed group (7 out of 82 rats) compared (control), 0.4, 1.5, or 6.0mg/kg bw for 105 weeks.
with the controls (0 out of 78 rats). There was A significant positive trend towards increased
also one carcinoma of the urinary bladder in mortality with increasing dose was seen in male
the exposed group compared with none in the and female mice. The reduction in survival
controls. of male mice in the group at the highest dose
necessitated early termination of this dose group
at week 95, instead of week 105. The only signif-
icant increase in the incidence of any neoplasm
357
358
Table 3.3 Studies of carcinogenicity involving exposure to pioglitazone or rosiglitazone with modifying agents
Species, strain (sex) Modifying agent Dosing regimen, Incidence of tumours Significance
Reference
Mouse, H Swiss Mainstream cigarette smoke by whole- After weaning (at age, 45 wk), Kidney adenoma: *P<0.01
(M,F) body inhalation for 4mo, starting 12 mice were also fed diets containing 0/34, 3/32 (M)
7mo hours after birth, followed by filtered air
pioglitazone at a concentration of 0 0/38, 5/35* (F)
La Maestra et al. for 3mo (control) or 120 mg/kg.
(2013) 34, 34/group (M)
38, 35/group (F)
IARC MONOGRAPHS 108
Rat, F344 (F) N-Butyl-N-(4-hydroxybutyl)nitrosamine After 2 wk, rats were given Urinary bladder carcinoma: *P<0.01
7mo (150mg) by gavage in thanol : water, rosiglitazone by gavage at 0 (control), 20/35, 34/34*
Lubet et al. (2008) 2/wk for 8wk 50mg/kg bw per day
35, 35/group
Rat, F344 (F) N-Butyl-N-(4-hydroxybutyl)nitrosamine 2 wk later, rats were given rosiglitazone Urinary bladder carcinoma: *P<0.01
8mo (150mg) by gavage in ethanol : water, by gavage at 0 (control), 10mg/kg bw 8/29, 28/30*
Lubet et al. (2008) 2/wk for 8wk per day
29, 30/group
Rat, F344 (F) N-Butyl-N-(4-hydroxybutyl)nitrosamine 2 wk later, rats were given rosiglitazone Urinary bladder carcinoma : *P<0.05
10mo (150mg) by gavage in ethanol : water, by gavage at 0 (control), 0.4, or 2mg/kg 12/25, 19/29, 24/30*
Lubet et al. (2008) 2/wk for 8wk bw per day
25, 29, 30/group
bw, body weight; F, female, mo, month; M, male; wk, week
Table 3.4 Studies of carcinogenicity in mice given diets containing rosiglitazone

Reference
CD-1 (M,F) Rosiglitazone mixed in feed to Liver haemangiosarcoma: *P=0.013 There was a significant trend for increasing
105wk achieve doses of 0 (control), 0.4, 1.5, 0/60, 4/60*, 0/60, 0/60 (M) mortality with increasing dose in M and F.
FDA (1999b) or 6.0mg/kg bw per day High-dose males were killed at wk95.
60M and 60F/group No increase in the incidence of any type of
neoplasm in females
C57BL6J-ApcMin/+ Rosiglitazone mixed in feed to Large intestine adenoma : *[P<0.01] Genetically engineered mouse sensitive to
(M) achieve doses of 0 or ~20mg/kg bw 9/15, 14/14* P0.05 for increased intestinal carcinogenesis. No increases in
8wk per day multiplicity of large multiplicity of small intestine tumours.
Pino et al. (2004) 15/group intestine tumours
[data and statistics
read from graph]
bw, body weight; F, female, M, male; wk, week; NR, not reported
359
IARC MONOGRAPHS 108
in any dose group was an increased incidence [age not reported] received rosiglitazone by
of liver haemangiosarcoma in male mice at the gavage at doses of 0 (control), 0.05, 0.3, or
lowest dose (FDA, 1999b). [Because no signifi- 2.0mg/kg bw per day in 1% methylcellulose for 2
cant evidence of a doseresponse relationship years. In comparison to vehicle controls, a signif-
was seen, the Working Group concluded that icant increase in mortality was seen in males at
there was no treatment-related positive trend in the highest dose. Significant increases in the
the incidence of liver haemangiosarcoma, or of incidence of subcutaneous lipoma were seen in
any other tumour type in either sex.] males at the intermediate dose, and in females at
the highest dose. [Increases in the incidence of
(b) Genetically engineered mouse subcutaneous lipoma and adipocyte hyperplasia
Pino et al. (2004) reported increased incidence (a putative preneoplastic lesion that is linked to
and multiplicity of adenoma of the large intestine lipoma) in males and females, were considered to
in ApcMin/+ mice (a genetically engineered mouse be treatment-related.]
model that overexpresses the Apc gene, leading to
rapid development of intestinal neoplasms) with 3.2.2 Coexposure with modifying agents
dietary exposure to rosiglitazone. In this study,
See Table3.3
male C57BL6J-ApcMin/+ mice (age, 67 weeks)
In two studies evaluating the activity of
were fed rosiglitazone at a dietary concentration
rosiglitazone as a chemopreventive agent for
selected to achieve a dose of 20 mg/kg bw per
cancer of the urinary bladder, groups of female
day, for 8 weeks. All mice exposed to rosiglita-
F344 rats received N-butyl-N-(4-hydroxybutyl)
zone (14 out of 14, 100% [P < 0.01]) developed
nitrosamine (BBN) by gavage in 0.1 mL
adenoma of the large intestine, compared with
ethanol:water (25:75, v/v), twice per week, for
60% (9 out of 15 mice) in the dietary controls
8 weeks. Beginning 2 weeks after the last dose of
group. Rosiglitazone increased the multiplicity
BBN, parallel groups of rats were given rosiglita-
of tumours of the large intestine, but not of the
zone at a dose of 0.4, 2, 10, or 50mg/kg bw per
small intestine [data and statistics were provided
day by gavage, or the vehicle (carboxymethylcel-
in in graphical form]. [The Working Group
lulose:polyethylene glycol 400; 50:50, v/v) only,
noted that, although used extensively as a model
for 710 months, followed by necropsy and histo-
system for cancer chemoprevention, the predic-
pathological examination of the urinary bladder
tive value of the ApcMin/+ mouse in the identifi-
(Lubet et al., 2008).
cation of agents that may promote or otherwise
In the first study, the incidence of carcinoma
stimulate carcinogenesis in the human colon was
of the urinary bladder in rats exposed to BBN plus
unknown.]
rosiglitazone (50mg/kg bw per day) for 7months
(c) Rat was 100% (34 out of 34; P<0.01), compared with
57% (20 out of 35) in the group exposed to BBN
See Table3.5
only. In the follow-up study with lower doses of
As part of its pharmacology review of the New
rosiglitazone, the incidence of carcinoma of the
Drug Application package (NDA 21071) for
urinary bladder at 8months in the group receiving
rosiglitazone submitted by SmithKline Beecham
BBN plus rosiglitazone (10mg/kg bw per day) was
Pharmaceuticals, the FDA summarized the
93% (28 out of 30; P<0.01), compared with 28%
results of a 2-year study that was performed to
(8 out of 29) in the group exposed to BBN only. In
evaluate the potential carcinogenicity of rosigli-
the same study, the incidence of carcinoma of the
tazone in rats (FDA, 1999b). In this study, groups
urinary bladder at 10 months in groups treated
of 60 male and 60 female Sprague-Dawley rats
360
Table 3.5 Studies of carcinogenicity in rats given rosiglitazone by gavage

Reference
Sprague- Rosiglitazone given at Subcutaneous lipoma: *P=0.001 Significantly increased mortality in high-
Dawley (M,F) doses of 0, 0 (second 3/60, 4/60, 5/59, 13/58*, **P=0.003 dose males
104wk control group), 0.05, 6/60 (M) Second control group reported, but no
FDA (1999b) 0.3, or 2.0mg/kg bw in 1/60, 2/60, 3/60, 1/59, information on how it differed from the
1% methylcellulose 9/60** (F) first control group
60M and 60F/group
bw, body weight; F, female, M, male; wk, week
with BBN plus rosiglitazone (2 or 0.4mg/kg bw mean serum half-life of the pharmacologically
per day) group was 80% (24 out of 30; P<0.05) active metabolites M-III and M-IV ranged from
and 67% (19 out of 29), versus 48% (12 out of 25) 16 to 24 hours. Serum concentrations of total
in BBN-treated vehicle controls. When admin- pioglitazone (pioglitazone plus active metab-
istered alone (without prior exposure to BNN), olites) remained elevated 24 hours after dosing
rosiglitazone (10 mg/kg bw) did not induce carci- (Takeda Pharmaceuticals, 2013).
noma of the urinary bladder during the 8-month The apparent oral clearance (CL/F) of piogli-
observation period. [The Working Group noted tazone has been calculated as 57L per hour. The
that the predictive value of the BBN model for mean apparent volume of distribution (Vd/F) of
the identification of agents that can enhance or pioglitazone after administration of a single oral
promote cancer of the human bladder had not dose was 0.630.41 (meanstandard deviation)
been established.] L/kg bw. Pioglitazone binds extensively (>99%)
to protein in human serum, principally to serum
albumin. Pioglitazone also binds other serum
4. Mechanistic and Other proteins, but with lower affinity. Metabolites M-III
Relevant Data and M-IV also are extensively bound (>98%) to
serum albumin (Takeda Pharmaceuticals, 2013).
Steady-state serum concentrations of piogl-
4.1 Absorption, distribution, itazone and total pioglitazone were achieved
metabolism, and excretion of within 7days. At steady state, M-III and M-IV
pioglitazone reached serum concentrations equal to or greater
than that of pioglitazone. In healthy volunteers
4.1.1 Humans and in patients with type 2 diabetes, pioglitazone
(a) Absorption, distribution, and excretion comprised approximately 3050% of the peak
total pioglitazone serum concentrations and
In fasting individuals, pioglitazone was 2025% of the total area under the curve (AUC)
measurable in the serum within 30 minutes for serum concentrationtime. Maximum
after oral administration, with peak concentra- serum concentration (Cmax), AUC, and trough
tions observed within 2 hours. Administration serum concentration (Cmin) for pioglitazone
with food slightly delayed the time to peak serum and total pioglitazone increased proportionally
concentration (to 34hours), but did not alter the at doses of 15 mg and 30 mg per day (Takeda
extent of absorption. The mean serum half-life of Pharmaceuticals, 2013).
pioglitazone ranged from 3 to 7hours, while the
361
IARC MONOGRAPHS 108
After oral administration, 1530% of the 4.1.2 Experimental systems

administered dose of pioglitazone was recovered
in the urine. Most of the oral dose was excreted In pharmacokinetic studies with male rats,
into the bile either unchanged or as metabo- peak plasma concentrations of pioglitazone were
lites, and eliminated in the faeces. Renal elim- reported at 1hour, and the plasma terminal half-
ination of pioglitazone was negligible (Takeda life of pioglitazone was 7.5hours. The distribu-
Pharmaceuticals, 2013). tion of pioglitazone was not extensive; the tissue/
There was no significant difference in the phar- plasma ratio was low (< 0.5), except for the gastro-
macokinetic profile of pioglitazone in subjects intestinal tract (Krieter et al., 1994).
with normal or with moderately impaired renal The AUCs for pioglitazone metabolites M-III
function. In patients with moderate and severe and M-IV were higher in female rats than in
renal impairment, although mean serum concen- males, while levels of M-II were similar in both
trations of pioglitazone and its metabolites were sexes (Fujita et al., 2003).
increased, no dose adjustment is needed. After
repeated oral doses of pioglitazone, mean AUC 4.2 Absorption, distribution,
values were decreased in patients with severe metabolism, and excretion of
renal impairment compared with healthy
subjects with normal renal function for pioglita-
rosiglitazone
zone (Budde et al., 2003). 4.2.1 Humans
In a multi-dosing study, pioglitazone was
rapidly absorbed, with median time to maximal (a) Absorption, distribution, and excretion
serum concentration (Cmax) occurring within In a study in healthy volunteers, the absorp-
2 hours. Serum concentrations of pioglitazone tion of rosiglitazone was relatively rapid, with
and its active metabolites remained elevated 99% oral bioavailability after oral absorption
24hours after exposure (Christensen et al., 2005). (Cox et al., 2000).
Peak plasma concentrations were observed
(b) Metabolism about 1hour after single oral doses. Maximum
Pioglitazone is extensively metabolized by plasma concentration (Cmax) and the AUC of
hydroxylation and oxidation to its active metabo- rosiglitazone increased in a dose-proportional
lites, which are keto and hydroxy derivatives. The manner over the therapeutic dose range (National
active metabolites include M-II active (hydroxy), Library of Medicine, 2010).
M-III active (keto), and M-IV active (hydroxy) The mean oral volume of distribution of
(Fig.4.1; Takeda Pharmaceuticals, 2013). rosiglitazone was approximately 17.6 L, based
In-vitro data have demonstrated that multiple on a population pharmacokinetic analysis.
isoforms of cytochrome P450 (CYP) are involved Rosiglitazone is approximately 99.8% bound to
in the metabolism of pioglitazone, including plasma proteins, primarily albumin (National
CYP2C8 and, to a lesser degree, CYP3A4 Library of Medicine, 2010).
(Kirchheiner et al., 2005). CYP2C9 is not signif- The elimination half-life of rosiglitazone was
icantly involved in the elimination of pioglita- 34 hours and was independent of dose. The
zone (Jaakkola et al., 2006). Pioglitazone is not a time to Cmax and the elimination half-life for two
strong inducer of CYP3A4, and pioglitazone was metabolites in plasma were significantly longer
not shown to induce CYPs (Nowak et al., 2002). than for rosiglitazone itself (46 hours versus
0.51hours, and about 5days versus 37hours)
(Cox et al., 2000).
362
Fig.4.1 Pioglitazone and metabolites

O O
O N O N
S S
HN HN
COOH
COOH
O O
M-V M-VI
O O O OH
OH O N O N
S S S
HN HN HN
O O O
M-I Pioglitazone M-II
O O
O N O N
S S
HN HN
O O
OH O
M-IV M-III
After oral or intravenous administration of AUC for rosiglitazone (1020%), which were not
rosiglitazone maleate, approximately 64% and deemed to be clinically relevant (Chapelsky et al.,
23% of the administered dose was eliminated in 2003).
the urine and in the faeces, respectively (National In a placental transfer study, the risk of
Library of Medicine, 2010). No unchanged drug placental transfer of rosiglitazone was higher
was eliminated in the urine. after 10weeks of gestation (Chan et al., 2005).
In a pharmacokinetics study of adminis-
tration of rosiglitazone with food, absorption (b) Metabolism
measured via Tmax was delayed by 1.75hours. The Rosiglitazone is extensively metabolized
Cmax was reduced by approximately 20%, but the by CYP2C9 and CYP2C8, with no unchanged
geometric mean ratio of AUC for the fed/fasted drug excreted in the urine (Kirchheiner et al.,
state was 0.94. No dose adjustment is required for 2005). The major routes of metabolism were
administration of rosiglitazone with food (Freed N-demethylation and hydroxylation, leading to
et al., 1999). N-desmethyl-rosiglitazone and 3-hydroxy-ro-
The half-life values of rosiglitazone are similar siglitazone, followed by conjugation with sulfate
in fasted and fed subjects (Bulliman et al., 1995). and glucuronic acid. All the circulating metabo-
Ethnicity had no impact on the pharmacoki- lites were considerably less potent than the parent
netics of rosiglitazone among healthy subjects compound and, therefore, are not expected
(Chu et al., 2007). to contribute to the activity of rosiglitazone
In patients with mild, moderate, or severe (National Library of Medicine, 2010; see Fig. 4.2).
renal insufficiency there are slight increases in
363
IARC MONOGRAPHS 108
Fig.4.2 Rosiglitazone and metabolites

S S CH 3
O O
OH N
N O N O H
O O
H O H
Phenoxyacetic acid derivative N-Despyridinylrosiglitazone
S CH 3 S CH 3 S CH 3
O O O
N N N N N N
N O N O N O
O O O
H H H
OH HO
3-Hydroxyrosiglitazone Rosiglitazone 5-Hydroxyrosiglitazone
S H S H S H
O O O
N N N N N N
N O N O N O O
O O
H H H
OH HO
N-Desmethyl-3-hydroxyrosiglitazone N-Desmethylrosiglitazone N-Desmethyl-5-hydroxyrosiglitazone
4.2.2 Experimental systems aberrations, sister chromatid exchanges, and

increased levels of 8-oxodeoxyguanosine
Rosigilitazone was extensively metabolized (Table4.1; Alzoubi et al., 2012).
after oral administration in mice. Mean bioavail-
ability was found to be 100%, 60%, and 95% in
rats, dogs, and humans, respectively (EMEA,
2005). (a) DNA damage
The main metabolites observed in humans are Male Sprague-Dawley rats treated with
also observed in rats; however, the clearance in pioglitazone by gavage had a dose-dependent
rats was almost ten times higher than in humans, increase in the frequency of DNA damage in
probably due to the higher levels of CYP2C in rat peripheral blood lymphoctyes and liver cells, as
microsomes (EMEA, 2005; Calixto et al., 2011). measured by comet assays. The addition of an
enzyme mixture containing endonuclease III
4.3 Genetic and related effects and formamidopyrimidine glycosylase signifi-
cantly increased the frequency of DNA damage,
4.3.1 Humans suggesting that DNA damage was due to oxida-
DNA damage tion of DNA bases (Table4.1; Bedir et al., 2008).
Pioglitazone did not increase the frequency
Incubating pioglitazone (100 M) with of chromosomal aberrations in Chinese hamster
human peripheral blood lymphocytes signifi- lung cells. Pioglitazone did not induce unsched-
cantly increased the frequency of chromosomal uled DNA synthesis in primary rat hepatocytes,
364
Table 4.1 Genetic and related effects of pioglitazone
Test system Results Dose or concentration Reference

(LED or HID)
In vitro
Salmonella typhimurium, TA98, TA100, TA1535, 5000 g/plate FDA Drug Approval Package
TA1537, with 20-minute pre-incubation, reverse (1999a)
mutation
Salmonella typhimurium, TA98, TA100, TA1535, 2000 g/plate FDA Drug Approval Package
TA1537, TA1538, reverse mutation (1999a)
Escherichia coli WP2 uvrA, with 20-minute pre- 5000 g/plate FDA Drug Approval Package
incubation, mutation (1999a)
Gene mutation, Chinese hamster ovary cells, Hprt gene 200 g/mL S9; 500g/mL FDA Drug Approval Package
+ S9 (1999a)
Gene mutation, AS52 Chinese hamster cells, Xprt gene 200 g/mL S9; 200g/mL FDA Drug Approval Package
+ S9 (1999a)
Unscheduled DNA synthesis, male F344 rat primary NT 100 g/mL FDA Drug Approval Package
hepatocytes (1999a)
Chromosomal aberration, Chinese hamster lung cells 5 mM FDA Drug Approval Package
(1999a)
Chromosomal aberration, human peripheral blood + NT 100 M Alzoubi et al. (2012)
lymphocytes
Sister chromatid exchange, human peripheral blood + NT 100 M Alzoubi et al. (2012)
lymphocytes
8-Oxodeoxyguanosine, human peripheral blood + NT 100 M Alzoubi et al. (2012)
lymphocytes
In vivo
Micronucleus formation, CD-1 mice 5000 mg/kg bw per day, FDA Drug Approval Package
single intraperitoneal (1999a)
injection, up to 72 hours
Comet assay, male Sprague-Dawley rat peripheral + 10 mg/kg bw per day, by Bedir et al. (2008)
blood lymphocytes gavage, for 14 days
Comet assay, male Sprague-Dawley rat liver cells + 10 mg/kg bw per day, by Bedir et al. (2008)
gavage, for 14 days
+, positive; , negative; HID, highest ineffective dose; LED, lowest effective dose; NR, not reported; NT, not tested
365
IARC MONOGRAPHS 108
or micronucleus formation in CD-1 mice 4.4 Other mechanistic data

(Table4.1; FDA Drug Approval Package, 1999a).
Male Sprague-Dawley rats treated with Pioglitazone selectively stimulates PPAR,
rosiglitazone by gavage showed a dose-dependent and to a lesser extent PPAR (Smith, 2001).
increase in the frequency of DNA damage in Acidification of the urine, as a result of ammo-
peripheral blood lymphoctyes and liver cells, as nium chloride administration in male rats, did
measured by comet assays (Table4.2; Bedir et al., not alter PPAR, PPAR (PPAR), or PPAR
2006). mRNA or protein expression, PPAR- or PPAR-
In-vitro assays for chromosomal aberration regulated gene expression, total or phosphoryl-
and unscheduled DNA synthesis, and in-vivo ated epidermal growth factor receptor (Egfr)
assays for micronucleus formation gave negative protein, Egfr or Akt2 gene expression, or urothe-
results with rosiglitazone (Table4.2; FDA Drug lial-cell proliferation. These results suggested
Approval Package, 1999b). that the suppression of bladder tumorigenesis
Rosiglitazone gave negative results in by acidifying the urine of rats exposed to PPAR
the Growth Arrest and DNA Damage gene agonists, such as pioglitazone, was not due to
45-Green Fluorescent Protein (GADD45-GFP) alterations in PPAR, PPAR, or Egfr expression
GreenScreen Human Cell genotoxicity assay in or PPAR signalling in the bladder epithelium of
the presence or absence of metabolic activation rats (Achanzar et al., 2007; Sato et al., 2011).
(Table4.2; Luzy et al., 2012). Roglitazone significantly increased the inci-
dence of tumours of the bladder induced BBN in
(b) Gene mutations female F344 rats. The mechanism for the induc-
Pioglitazone and its metabolites M-I, M-IV, tion of these tumours was not known (Lubet
M-V, and M-VI were not mutagenic in Salmonella et al., 2008).
typhimurium strains TA98, TA100, TA1535, or Strain H Swiss mice exposed to mainstream
TA1537, or in Escherichia coli strain WP2 uvrA, cigarette smoke since birth for 4 months, and
in either the presence or absence of metabolic subsequently exposed to pioglitazone, had lower
activation. Pioglitazone was not mutagenic at levels of DNA damage in exfoliated bladder cells,
the Hprt gene of Chinese hamster ovary cells, or as measured by comet assays, than mice exposed
at the Xprt gene of AS52 Chinese hamster cells. to mainstream cigarette smoke and fed control
Pioglitazone metabolites M-I and M-VI were diet. However, the mice exposed to mainstream
mutagenic in mouse lymphoma L5178Y cells in cigarette smoke and then pioglitazone had an
the presence of metabolic activation; metabolites increased incidence of kidney tubular epithelium
M-IV and M-V gave negative results, and the hyperplasia, kidney adenoma, kidney lesions,
assay conducted with pioglitazone was consid- and/or urinary tract lesions, compared with mice
ered inadequate (Table4.1; Table4.3; FDA Drug exposed to mainstream cigarette smoke only, or
Approval Package, 1999a). sham-treated mice. These data suggested that
Rosiglitazone was not mutagenic in S. typhi- pioglitazone can act as a promoter of tumours
murium strains TA98, TA100, TA1535, TA1537, of the kidney in mice. Mice exposed to main-
or TA1538, or in E. coli strain WP2 uvrA, in stream cigarette smoke and pioglitazone had
either the presence or absence of an exogenous more acidic urine than sham-exposed mice (La
metabolic activation system. Rosiglitazone was Maestra et al., 2013). [There was not a group that
mutagenic in mouse lymphoma L5178Y cells in received pioglitazone only.]
the presence of metabolic activation (Table4.2; In male C57BL/6J-ApcMin/+ mice, a hete-
FDA Drug Approval Package, 1999b). rozygous mouse strain susceptible to intestinal
366
Table 4.2 Genetic and related effects of rosiglitazone
Test system Results Dose or concentration Reference

(LED or HID)
In vitro
Salmonella typhimurium, TA98, TA100, TA1535, 5000 g/plate FDA Drug Approval
TA1537, TA1538, reverse mutation Package (1999b)
Escherichia coli WP2 uvrA, mutation 5000 g/plate FDA Drug Approval
Package (1999b)
Gene mutation, mouse lymphoma L5178Y cells, ? + 100 g/mL S9; 200g/mL + S9 FDA Drug Approval
Tk locus Package (1999b)
Chromosomal aberration, human lymphocytes 240 g/mL FDA Drug Approval
Package (1999b)
GADD45-GFP GreenScreen Human Cell, NR Luzy et al. (2012)
genotoxicity assay
In vivo
Micronucleus formation, CD-1 mice 700 mg/kg bw, single intraperitoneal FDA Drug Approval
injection, up to 72 hours Package (1999b)
Unscheduled DNA synthesis, male Sprague- 2000 g/mL FDA Drug Approval
Dawley rat primary hepatocytes Package (1999b)
Comet assay, male Sprague-Dawley rat peripheral + 1.0 mg/kg bw per day, for 14 days Bedir et al. (2006)
blood lymphocytes
Comet assay, male Sprague-Dawley rat liver cells + 0.5 mg/kg bw per day, for 14 days Bedir et al. (2006)
+, positive; , negative; ?, inconclusive; GADD45, growth arrest and DNA damage gene; GFP, green fluorescent protein; HID, highest ineffective dose; LED, lowest effective dose; S9,
supernatant fraction of liver homogenate 9000 g
367
368
Table 4.3 Genetic and related effects of metabolites of pioglitazone
Test system Results Dose or Reference

concentration
Without exogenous With exogenous (LED or HID)
M-I
Salmonella typhimurium, TA98, TA100, TA1535, NR FDA Drug Approval
TA1537, reverse mutation Package (1999a)
Escherichia coli WP2 uvrA mutations NR FDA Drug Approval
Package (1999a)
IARC MONOGRAPHS 108
Gene mutation, mouse lymphoma L5178Y cells, Tk ? + NR FDA Drug Approval

locus Package (1999a)
M-IV
Salmonella typhimurium, TA98, TA100, TA1535, NR FDA Drug Approval
Escherichia coli WP2 uvrA, mutation NR FDA Drug Approval
Package (1999a)
Gene mutation, mouse lymphoma L5178Y cells, Tk NR FDA Drug Approval
M-V
Package (1999a)
Gene mutation, mouse lymphoma L5178Y cells, Tk NR FDA Drug Approval
M-VI
Package (1999a)
Gene mutation, mouse lymphoma L5178Y cells, Tk ? + NR FDA Drug Approval
+, positive; , negative;?, inconclusive; LED, lowest effective dose; HID, highest ineffective dose; NR, not reported; NT, not tested
neoplasms, PPAR agonists (troglitazone, PPAR agonist, increased the levels of Egr-1 in
150 mg/kg bw per day; rosiglitazone, 20 mg/kg bw the liver and heart (Egerod et al., 2010).
per day; or NID525, 150 mg/kg bw per day) signif-
icantly increased the multiplicity of neoplasms
(primarily adenomas) in the large intestine. The
4.5 Susceptibility
mechanism was not explored (Pino et al., 2004). No data were available to the Working Group.
The administration of rosiglitazone
(20 mg/kg bw per day for 8 weeks) to male
C57BL/6J-ApcMin/+ mice significantly increased 4.6 Mechanistic considerations
the multiplicity of tumours of the large intestine, Pioglitazone, a PPAR agonist, has been
and increased expression of PPAR in the large implicated in cancer of the urinary bladder in
intestine, and increased expression of -catenin. rats, and rosiglitazone, which is also a PPAR
A similar increase in tumour multiplicity, and agonist, promotes cancer of the urinary bladder
-catenin expression, was observed with trogl- in rats, and possibly tumours of the kidney in
itazone, another PPAR agonist (150 mg/kg bw mice. Four mechanisms of carcinogenicity have
per day) (Lefebvre et al., 1998). been considered in rats treated with PPAR
Male Sprague-Dawley rats were treated with agonists: (i) genotoxicity of metabolites formed
rosiglitazone at 8 mg/kg bw per day by gavage for from the agonists; (ii) cytotoxicity of the agonists
up to 16 days. Treatment with rosiglitazone had or their metabolites in the urothelium, causing
modest effects on levels of the transcription factor cancer due to a proliferation-driven chronic
Egr-1 in the bladder urothelium. Likewise, there wound-healing response; (iii) formation of
was minimal effect with fenofibrate, a PPAR urinary solids (urolithiasis), due to urinary
agonist. In contrast, ragaglitazar, a dual-action changes induced by the agonists or their metab-
agonist of PPAR and , and a rat bladder carcin- olites, which results in chronic irritation of the
ogen, caused a substantial increase in the level of urothelium; and (iv) a receptor-mediated effect
Egr-1. These data suggested that the co-activation of the agonists, with carcinogenesis caused by
of both PPAR and is required for the increased activation of PPAR transcription factors in
expression of Egr-1 (Egerod et al., 2005). the urothelium. These mechanisms may not be
The relationship between nuclear EGR-1 mutually exclusive.
protein levels and stage of human bladder tumour
was investigated using tissue microarrays. The 4.6.1 Genotoxicity
extent of nuclear EGR-1 immunostaining was
associated with a higher risk of progression to When assessed using a standard battery of
stage T2T4 cancer of the bladder (Egerod et al., assays for genotoxicity, pioglitazone and rosigl-
2009a). itazone have typically given negative results;
Male Sprague-Dawley rats given rosiglita- nonetheless, there are exceptions. Certain
zone plus fenofibrate expressed Egr-1 protein on metabolites of pioglitazone and rosiglitazone
both the dorsal and ventral regions of the urinary have given positive results in the assay for gene
bladder (Egerod et al., 2009b). mutation in mouse lymphoma cells and, more
In male Sprague-Dawley rats, rosiglitazone recently, pioglitazone has been reported to
(given by gavage at 8 or 20mg/kg bw per day for increase the frequency of chromosomal aberra-
7days) had minimal effects on the levels of the tion, sister chromatid exchange, and formation
transcription factor Egr-1 in the bladder urothe- of 8-oxodeoxyguanosine in human peripheral
lium, heart, or liver. In contrast, fenofibrate, a blood lymphocytes, and both pioglitazone and
369
IARC MONOGRAPHS 108
rosiglitazone gave positive results in comet assays and that acidifying the urine through the admin-
in liver cells and peripheral blood lymphocytes istration of ammonium chloride in the diet
from rats. Thus, while perhaps not the primary decreases the amount of urinary solids and the
mechanism, the contribution of genotoxicity extent of tumorigenesis in the urinary bladder.
to the carcinogenic activity of pioglitazone or Urinary acidification did not alter the expression
rosiglitazone in the urothelium of rats cannot of PPAR, PPAR, or epidermal growth factor
presently be excluded. receptor in the rat bladder urothelium, which
suggests that a receptor-mediated mechanism
4.6.2 Agonist cytotoxicity is not involved in the tumorigenic response. A
similar urolithiasis-based mechanism has been
Pioglitazone and rosiglitazone are lipophilic proposed for muraglitazar, a dual-action PPAR
drugs that are excreted to a limited extent in the and PPAR agonist (Achanzar et al., 2007).
urine of rats. Since urothelial carcinogenesis is
typically considered to be mediated by direct
4.6.4 Receptor-mediated effect
urinary exposure to the drugs or their metabo-
lites, rather than through systemic distribution Although a urolithiasis-based mechanism
in the blood, a mechanism involving cytotoxicity appears plausible for induction of tumours of
as a result of direct exposure to the agonists or the bladder in rats treated with pioglitazone,
their metabolites appeared unlikely. such a mechanism cannot explain the tumours
of the bladder observed in rats given rosiglita-
4.6.3 Urolithiasis zone after initiation with BBN (because urinary
solids were not observed; Lubet et al., 2008), the
The induction of tumours of the urinary promotion of intestinal neoplasms in susceptible
bladder in rodents as a consequence of the forma- mouse strains given pioglitazone, rosiglitazone,
tion of urinary solids has been documented for or other PPAR agonists such as troglitazone or
several compounds, including carbonic anhy- NID525 (Lefebvre et al., 1998; Pino et al., 2004),
drase inhibitors, HIV protease inhibitors, and the induction of tumours of the kidney in mice
sulfonamides. PPAR agonists are known to exposed to mainstream cigarette smoke and
cause fluid accumulation, oedema, cardiac then pioglitazone (La Maestra et al., 2013), or
enlargement, and heart failure, effects that can the induction of tumours of the urinary bladder
lead to significant changes in urine composi- in rats treated with naveglitazar, a -dominant
tion. The administration of pioglitazone to rats PPAR and PPAR agonist (Long et al., 2008).
results in the formation of urinary solids. This The promoting activity of rosiglitazone in the
occurs to a greater extent in rats than in mice, rat bladder has been attributed to an increased
and in male rats than in female rats; these trends expression of Egr-1, ribosomal S6 protein phos-
correspond to the greater susceptibility of rats phorylation, and c-Jun transcription factor
compared with mice, and of male rats compared phosphorylation, which can lead to hypertrophy,
with female rats, to the induction of urothelial hyperplasia, and subsequently urothelial-cancer
tumours upon the administration of pioglitazone. progression. While these responses appear to be
Further support for a urolithiasis-based mecha- greater with the dual-acting PPAR and PPAR
nism comes from the observation that tumours agonist ragaglitazar, a modest response does
arising from pioglitazone occur predominantly occur with rosiglitazone (Egerod et al., 2005,
on the ventral surface of the bladder, the region 2009b, 2010). Furthermore, pioglitazone also
where urinary solids would settle in rat bladders, shows modest PPAR agonist activity that may
370
contribute to the mechanism of induction of and bone fractures. While this agent is banned
tumours of the bladder (Sakamoto et al., 2000). in Europe and restricted in the USA, substan-
The induction of intestinal neoplasms in suscep- tial use continues in some countries, including
tible mouse strains treated with PPAR agonists China (global sales of US$41 million).
may be a consequence of increased expression
of -catenin protein, which activates transcrip-
tion factors associated with colon tumorigenesis
5.2 Human carcinogenicity data
(Lefebvre et al., 1998; Pino et al., 2004). 5.2.1 Cancer of the bladder
The risk of cancer of the bladder associated
5. Summary of Data Reported with the use of pioglitazone and rosiglitazone was
assessed in several studies, some with overlap-
ping populations, from Europe, North America
5.1 Exposure data and Asia. Some subjects may have received both
Thiazolidinediones are a unique class of drugs (in sequence) at some time during treat-
synthetic oral drug that exert direct effects on ment for diabetes.
the mechanisms of insulin resistance, and result Information for pioglitazone was evaluated
in improved insulin action and reduced hyperg- in one large randomized controlled trial, four
lycaemia. Two thiazolidinediones, rosiglitazone cohort studies, and three casecontrol studies,
and pioglitazone, initially showed great promise some with overlapping study populations.
as receptor-mediated oral therapy for type 2 Ever-use of pioglitazone was associated with
diabetes mellitus. an increased risk of cancer of the bladder in
Pioglitazone hydrochloride is approved in all studies except one casecontrol study from
some countries for the treatment of type 2 diabetes Taiwan, China, across all study designs and
mellitus. It is available both as a single agent and geographical regions, with risk ratios that ranged
in combination with other oral medications for from 1.2 in the observational studies to a nearly
diabetes. Until 2009, pioglitazone was among the threefold statistically significant increase in the
most widely used oral drugs for the treatment randomized controlled trial. In this trial, the
of type 2 diabetes mellitus. Use of pioglitazone Working Group noted the excess occurrence of
hydrochloride has declined following studies these cancers (14 in the treatment group versus
suggesting links to cancer of the bladder, heart 5 in the placebo group) within a short follow-up
failure, and bone fractures. While regulatory time (11 of the bladder cancers occurred within
agencies in France and Germany banned piogli- 1year of randomization), and the large number of
tazone in 2011, global sales remained substantial patients enrolled, and double-blind experimental
at US$3.3 billion in 2012. design resulting in the balance of confounding
Rosiglitazone maleate is approved in some factors at baseline.
countries for the treatment of type 2 diabetes Doseresponse relationships were assessed
mellitus. It is available both as a single agent and in five studies, three of which were high-quality
in combination with other oral medications for population-based studies (which adjusted for
diabetes. Until 2007, rosiglitazone was among smoking or chronic obstructive pulmonary
the most widely used oral drugs for treatment disease in the absence of data on smoking)
of type 2 diabetes. Use of rosiglitazone has conducted within the large health insurance
declined over the last few years following studies databases from the USA, United Kingdom, and
suggesting links to heart attack, heart failure, Taiwan, China. Greater risks were reported with
371
IARC MONOGRAPHS 108
higher dosage or longer use in the casecontrol pioglitazone and rosiglitazone has also been eval-
study in the United Kingdom, and in the cohort uated in studies using cohort and casecontrol
study in the USA. Observation of a doseresponse designs. No consistent pattern of increased risk
relationship helped to mitigate concerns about was reported for any other specific cancer site, or
potential confounding by most risk factors; for all cancers combined for either drug.
nevertheless, the magnitude of the excess risks
observed was modest and some estimates were
imprecise.
Among ever-users of rosiglitazone, with data 5.3.1 Pioglitazone
available from two casecontrol studies and
two cohort studies, risk ratios for cancer of the In a 2-year study in mice treated by gavage,
bladder were close to the null in all except one pioglitazone produced increases in the incidence
study from the United Kingdom. of benign pheochromocytoma of the adrenal
The Working Group was unable to consist- gland in males and increases in the incidence
ently rule out confounding, selection bias, detec- of leiomyosarcoma of the uterine cervix in
tion bias, and bias related to indication or severity females. Administration of pioglitazone in the
of disease in the populations studied as poten- feed caused a significant increase in the inci-
tial explanations for positive associations with dence of large intestine adenoma in one study
pioglitazone. Most of the studies were based on in genetically engineered male mice sensitive to
medical databases, which allowed for adjustment intestinal carcinogenesis. In a study in male and
for potential confounding by medical factors, female neonatal mice, pioglitazone in the feed
but did not permit direct control for cigarette promoted mainstream cigarette smoke-induced
smoking and other risk factors. However, for kidney adenoma in females.
pioglitazone, increased risks were consistently In a first 2-year study in rats treated by
seen in the studies that adjusted for smoking (one gavage, pioglitazone induced a significant posi-
cohort study from the USA, two studies from the tive trend in the incidence of transitional cell
United Kingdom, and one study from Taiwan, carcinoma of the urinary bladder and a signif-
China that adjusted for chronic obstructive icant positive trend in the incidence of subcu-
pulmonary disease), as well as those that did not. taneous fibrosarcoma of the subcutis in males.
The potential for confounding by smoking is also It also caused a significant positive trend in the
mitigated by the fact that, in the same studies, incidence of subcutaneous lipoma in females.
there was no consistent evidence of cancer of In a second 2-year study in male rats treated by
the lung and elevated risks were not found gavage, pioglitazone caused a significant increase
among rosigilitazone users in the same studies. in the incidence of transitional cell papilloma of
Furthermore, an excess of cancer of the bladder the urinary bladder.
among pioglitozone users, and not cancer of the
lung, was observed in the trial that randomized 5.3.2 Rosiglitazone
for potential confounders including smoking. Administration of diets containing rosigli-
tazone caused a significant increase in the inci-
5.2.2 Other cancer sites dence of large intestine adenoma in one study
The risk of cancers at several other sites, in genetically engineered male mice sensitive
including the liver, kidney, colorectum, lung, to intestinal carcinogenesis. In a 2-year study
prostate, and breast, among patients using in male and female mice treated by gavage, a
372
significant increase in the incidence of liver 6. Evaluation

haemangiosarcoma was observed in males, but
this was not treatment-related.
In a 2-year study in rats treated by gavage,
rosiglitazone induced significant increases in the There is limited evidence in humans for the
incidence of subcutaneous lipoma in males and carcinogenicity of pioglitazone. A positive asso-
females. ciation has been observed between pioglitazone
In two studies in female mice, coexposure and cancer of the bladder.
to N-butyl-N-(4-hydroxybutyl)nitrosamine There is inadequate evidence in humans for
plus rosiglitazone significantly increased the the carcinogenicity of rosiglitazone.
incidences of N-butyl-N-(4-hydroxybutyl)
nitrosamine-induced carcinoma of the urinary
bladder. 6.2 Cancer in experimental animals
There is sufficient evidence in experimental
5.4 Mechanistic and other relevant animals for the carcinogenicity of pioglitazone.
data There is limited evidence in experimental
animals for the carcinogenicity of rosiglitazone.
Pioglitazone and rosiglitazone undergo
extensive phase I metabolism. Although piogl-
itazone and rosiglitazone have typically given
negative results when assessed in standard Pioglitazone is probably carcinogenic to
batteries of genotoxicity assays, exceptions have humans (Group 2A).
been noted. Certain pioglitazone metabolites Rosiglitazone is not classifiable as to its
and rosiglitazone have given positive results in carcinogenicity to humans (Group 3).
assays in the mouse lymphoma cells; pioglitazone
increased the levels of chromosomal aberration,
sister chromatid exchange, and 8-oxodeoxy- References
guanosine in human peripheral blood lympho-
cytes; and both pioglitazone and rosiglitazone Achanzar WE, Moyer CF, Marthaler LT, Gullo R, Chen
gave positive results in comet assays in liver cells SJ, French MH et al. (2007). Urine acidification
and peripheral blood lymphocytes from rats. has no effect on peroxisome proliferator-activated
receptor (PPAR) signalling or epidermal growth factor
Four mechanisms have been considered for the (EGF) expression in rat urinary bladder urothelium.
induction of bladder tumours in rats admin- Toxicol Appl Pharmacol, 223(3):24656. doi:10.1016/j.
istered pioglitazone (genotoxicity of pioglita- taap.2007.06.015 PMID:17663016
AFFSAPS (2013). Use of medications containing piogli-
zone metabolites; cytotoxicity, urolithiasis, and tazone (Actos, Competact) suspended June 9th 2011.
PPAR and receptor-mediated effects). While Press release. Available from: http://ansm.sante.fr/var/
not mutually exclusive, data supporting urolith- ansm_site/storage/original/application/4e293bcd0814
iasis and receptor-mediated mechanisms appear c025b94d46d7502a0958.pdf. Accessed 10 July 2014.
Al-Ghamdi AF, Hefnawy MM (2012). Electrochemical
to be the strongest. Likewise, receptor-medi- determination of rosiglitazone by square-wave adsorp-
ated effects may play a role in the tumorigenic tive stripping voltammetry method. Arab J Chem,
response observed in other experimental models 5:383389. doi:10.1016/j.arabjc.2011.07.011
(e.g. intestinal neoplasia in mice). Alzoubi K, Khabour O, Hussain N, Al-Azzam S,
Mhaidat N (2012). Evaluation of vitamin B12 effects
on DNA damage induced by pioglitazone. Mutat Res,
373
IARC MONOGRAPHS 108
748(12):4851. doi:10.1016/j.mrgentox.2012.06.009 Cayman SDS (2013). Safety data sheet. Cayman Chemical
PMID:22790087 Company.
Avandia (2010). Rosiglitazone maleate. Tablet, film coated. Chan LY, Yeung JH, Lau TK (2005). Placental transfer
Anonymous. Total physicians care Inc. 127. Available of rosiglitazone in the first trimester of human
from: http://dailymed.nlm.nih.gov/dailymed/lookup. pregnancy. Fertil Steril, 83(4):9558. doi:10.1016/j.
cfm?setid=ec682aec-e98f-41a1-9d21-eb7580ea3a8a. fertnstert.2004.10.045 PMID:15820806
Accessed 1 October 2014. Chang CH, Lin JW, Wu LC, Lai MS, Chuang LM, Chan
Azoulay L, Yin H, Filion KB, Assayag J, Majdan A, Pollak KA (2012). Association of thiazolidinediones with
MN et al. (2012). The use of pioglitazone and the liver cancer and colorectal cancer in type 2 diabetes
risk of bladder cancer in people with type 2 diabetes: mellitus. Hepatology, 55(5):146272. doi:10.1002/
nested case-control study. BMJ, 344(may30 3):e3645 hep.25509 PMID:22135104
doi:10.1136/bmj.e3645 PMID:22653981 Chapelsky MC, Thompson-Culkin K, Miller AK, Sack
BDdrugs (2013). Online drug index of Bangladesh. M, Blum R, Freed MI (2003). Pharmacokinetics of
Available from: http://www.bddrugs.com/ rosiglitazone in patients with varying degrees of
Bedir A, Aliyazicioglu Y, Bilgici B, Yurdakul Z, Uysal M, renal insufficiency. J Clin Pharmacol, 43(3):2529.
Suvaci DE etal. (2008). Assessment of genotoxicity in doi:10.1177/0091270002250602 PMID:12638393
rats treated with the antidiabetic agent, pioglitazone. ChemicalBook (2013). Available from: http://www.
Environ Mol Mutagen, 49(3):18591. doi:10.1002/ chemicalbook.com/ProductMSDSDetailCB9257580_
em.20365 PMID:18213655 EN.htm. Accessed 08 July 2014.
Bedir A, Aliyazicioglu Y, Kahraman H, Yurdakul Z, Uysal ChemSpider (2013). ChemSpider: The free chemical
M, Suvaci DE etal. (2006). Genotoxicity in rats treated database. Royal Society of Chemistry. Available from:
with the antidiabetic agent, rosiglitazone. Environ http://www.chemspider.com. Accessed 10 July 2014.
Mol Mutagen, 47(9):71824. doi:10.1002/em.20261 Christensen ML, Meibohm B, Capparelli EV, Velasquez-
PMID:17078099 Mieyer P, Burghen GA, Tamborlane WV (2005). Single-
Bosetti C, Rosato V, Buniato D, Zambon A, La Vecchia and multiple-dose pharmacokinetics of pioglitazone
C, Corrao G (2013). Cancer risk for patients using in adolescents with type 2 diabetes. J Clin Pharmacol,
thiazolidinediones for type 2 diabetes: a meta-anal- 45(10):113744. doi:10.1177/0091270005279578
ysis. Oncologist, 18(2):14856. doi:10.1634/ PMID:16172178
theoncologist.2012-0302 PMID:23345544 Chu KM, Hu OY, Pao LH, Hsiong CH (2007).
BrFAM (2011). Important safety information on the Pharmacokinetics of oral rosiglitazone in Taiwanese
use of medicinal products containing pioglita- and post hoc comparisons with Caucasian, Japanese,
zone. Press release: Available from: http://www. Korean, and mainland Chinese subjects. J Pharm
bf a r m .d e /S h a re d D o c s / R i s i k oi n for m at ione n / Pharm Sci, 10(4):4119. PMID:18261363
Pharmakovigilanz/EN/RHB/2011/rhb-pioglitazon. Colmers IN, Bowker SL, Johnson JA (2012a).
pdf?__blob=publicationFile&v=3. Accessed 8 July Thiazolidinedione use and cancer incidence in
2014. type 2 diabetes: a systematic review and meta-anal-
Budde K, Neumayer HH, Fritsche L, Sulowicz W, Stompr ysis. Diabetes Metab, 38(6):47584. doi:10.1016/j.
T, Eckland D (2003). The pharmacokinetics of piogl- diabet.2012.06.003 PMID:23041441
itazone in patients with impaired renal function. Br Colmers IN, Bowker SL, Majumdar SR, Johnson JA
J Clin Pharmacol, 55(4):36874. doi:10.1046/j.1365- (2012b). Use of thiazolidinediones and the risk of
2125.2003.01785.x PMID:12680885 bladder cancer among people with type 2 diabetes: a
Bulliman S, Allen A, Harris AM etal. (1995). The influ- meta-analysis. CMAJ, 184(12):E67583. doi:10.1503/
ence of food on the pharmacokinetics of BRL 49653 in cmaj.112102 PMID:22761478
healthy male volunteers. Br J Clin Pharmacol, 40:517 Cox PJ, Ryan DA, Hollis FJ, Harris AM, Miller AK,
Calixto LA, de Oliveira AR, Jabor VA, Bonato PS (2011). In Vousden M et al. (2000). Absorption, disposition,
vitro characterization of rosiglitazone metabolites and and metabolism of rosiglitazone, a potent thiazoli-
determination of the kinetic parameters employing dinedione insulin sensitizer, in humans. Drug Metab
rat liver microsomal fraction. Eur J Drug Metab Dispos, 28(7):77280. PMID:10859151
Pharmacokinet, 36(3):15966. doi:10.1007/s13318-011- DHHS/FDA (2007). Electronic Orange Book-Approved
0039-8 PMID:21499911 Drug Products with Therapeutic Equivalence
Cantello BC, Cawthorne MA, Cottam GP et al. (1994). Evaluations. Available from: http://www.fda.gov/cder/
[[omega-(Heterocyclylamino)alkoxy]benzyl]-2,4-thi- ob/
azolidinediones as potent antihyperglycemic agents. Dormandy JA, Charbonnel B, Eckland DJ, Erdmann E,
J Med Chem, 37:39773985. doi:10.1021/jm00049a017 Massi-Benedetti M, Moules IK etal.; PROactive inves-
PMID:7966158 tigators (2005). Secondary prevention of macrovascular
events in patients with type 2 diabetes in the PROactive
374
Study (PROspective pioglitAzone Clinical Trial In FDA (1999b). Review and evaluation of pharmacology
macroVascular Events): a randomised controlled and toxicology data: Rosiglitazone. Pharmacology
trial. Lancet, 366(9493):127989. doi:10.1016/S0140- Reviews Application Number: 021071. Center for Drug
6736(05)67528-9 PMID:16214598 Evaluation and Research. Available from: http://www.
Drugbank (2013). Drug Bank: Open Data Drug & accessdata.fda.gov/drugsatfda_docs/nda/99/21071_
Drug Target Database. Available from: http://www. Avandia.cfm. Accessed 1 July 2014.
drugbank.ca/ FDA (2013a). Actos Product Information. Available from:
Egerod FL, Bartels A, Fristrup N, Borre M, rntoft TF, http://www.accessdata.fda.gov/drugsatfda_docs/
Oleksiewicz MB etal. (2009a). High frequency of tumor label/2011/021073s043s044lbl.pdf. Accessed 10 July
cells with nuclear Egr-1 protein expression in human 2014.
bladder cancer is associated with disease progression. FDA (2013b). Actos Risk Evaluation and Mitigation
BMC Cancer, 9(1):385 doi:10.1186/1471-2407-9-385 Strategy. Available from: http://www.accessdata.fda.
PMID:19878561 gov/drugsatfda_docs/appletter/2012/021073s045,021
Egerod FL, Brnner N, Svendsen JE, Oleksiewicz MB 842s016,022024s009,021925s012ltr.pdf. Accessed 10
(2010). PPARalpha and PPARgamma are co-expressed, July 2014.
functional and show positive interactions in the rat FDA (2013c). Joint Meeting of the Endocrinologic and
urinary bladder urothelium. J Appl Toxicol, 30(2):151 Metabolic Drugs Advisory Committee and the Drug
62. PMID:19757489 Safety and Risk Management Advisory Committee
Egerod FL, Nielsen HS, Iversen L, Thorup I, Storgaard T, Meeting Announcement Committee. Available
Oleksiewicz MB (2005). Biomarkers for early effects of from: http://www.fda.gov/AdvisoryCommittees/
carcinogenic dual-acting PPAR agonists in rat urinary CommitteesMeetingMaterials/Drugs/Endocrinologic
bladder urothelium in vivo. Biomarkers, 10(4):295309. andMetabolicDrugsAdvisoryCommittee/ucm331504.
doi:10.1080/13547500500218682 PMID:16240504 htm. Accessed 11 June 2013.
Egerod FL, Svendsen JE, Hinley J, Southgate J, Bartels FDA Drug Approval Package (1999a). Actos (Pioglitazone
A, Brnner N et al. (2009b). PPAR and PPAR Hydrochloride) Tablets, Application No. 021073.
coactivation rapidly induces Egr-1 in the nuclei of FDA Drug Approval Package (1999b). Avandia
the dorsal and ventral urinary bladder and kidney (Rosiglitazone Maleate) Tablets, Application No.
pelvis urothelium of rats. Toxicol Pathol, 37(7):94758. 21-071.
doi:10.1177/0192623309351723 PMID:20008548 Ferrara A, Lewis JD, Quesenberry CP Jr, Peng T, Strom
EMA; European Medicines Agency (2010). European BL, Van Den Eeden SK et al. (2011). Cohort study of
Medicines Agency recommends suspension of pioglitazone and cancer incidence in patients with
Avandia, Avandamet and Avaglim. Available from: diabetes. Diabetes Care, 34(4):9239. doi:10.2337/dc10-
http://www.ema.europa.eu/ema/index.jsp?curl=pages/ 1067 PMID:21447664
news_and_events/news/2010/09/news_detail_001119. Ferwana M, Firwana B, Hasan R, Al-Mallah MH, Kim
jsp&mid=WC0b01ac058004d5c1 S, Montori VM et al. (2013). Pioglitazone and risk of
EMA CHMP (2012). Assessment report Pioglitazone bladder cancer: a meta-analysis of controlled studies.
Accord. PM: EPA (2011). Estimation Program Interface Diabet Med, 30(9):102632. doi:10.1111/dme.12144
(EPI). United States, Environmental Protection Agency. PMID:23350856
EMEA; European Medecines Agency (2005). Freed MI, Allen A, Jorkasky DK, DiCicco RA (1999).
Rosiglitazone. Available from: http://www.ema. Systemic exposure to rosiglitazone is unaltered by
europa.eu/docs/en_GB/document_library/EPAR_-_ food. Eur J Clin Pharmacol, 55(1):536. doi:10.1007/
Scientific_Discussion/human/000522/WC500028918. s002280050592 PMID:10206085
pdf. Accessed 20 June 2014. Fujimoto K, Hamamoto Y, Honjo S, Kawasaki Y, Mori
EMEA; European Medicines Agency (2012). European K, Tatsuoka H et al. (2013). Possible link of pioglita-
Public Assessment Report. Pioglitazone. Available zone with bladder cancer in Japanese patients with
from: http://www.ema.europa.eu/docs/en_GB/ type 2 diabetes. Diabetes Res Clin Pract, 99(2):e213.
document_library/EPAR _-_Public_assessment_ doi:10.1016/j.diabres.2012.11.013 PMID:23228390
report/human/002277/WC500126045.pdf. Accessed Fujita Y, Yamada Y, Kusama M, Yamauchi T, Kamon J,
10 July 2014. Kadowaki T etal. (2003). Sex differences in the phar-
Enzo PDS (2012). Product data sheet: Enzo life sciences. macokinetics of pioglitazone in rats. Comp Biochem
FDA (1999a). Review and evaluation of pharmacology Physiol C Toxicol Pharmacol, 136(1):8594. doi:10.1016/
and toxicology data: Pioglitazone. Pharmacology S1532-0456(03)00194-7 PMID:14522601
Reviews Application Number: 021073. Center for Drug Giaginis C, Theocharis S, Tsantili-Kakoulidou A (2007).
Evaluation and Research. Available from: http://www. Investigation of the lipophilic behaviour of some
accessdata.fda.gov/drugsatfda_docs/nda/99/021073A_ thiazolidinediones. Relationships with PPAR-
Actos.cfm. Accessed 1 July 2014. activity. J Chromatogr B Analyt Technol Biomed Life
375
IARC MONOGRAPHS 108
Sci, 857:181187. doi:10.1016/j.jchromb.2007.07.013 pioglitazone in rats. Drug Metab Dispos, 22(4):62530.

PMID:17660053 PMID:7956739
Gowramma B, Babu B, Meyyanathan SN et al. (2012a). Krishna SR, Rao B et al. (2008). Isolation, synthesis
New chiral normal phase HPLC method for determi- and characterization of rosiglitazone maleate
nation of pioglitazone enantiomers on its marketed impurities. E-Journal of Chemistry, 5:562566.
formulation. J Pharm Res, 5:72 doi:10.1155/2008/308724
Gowramma B, Babu B, Meyyanathan SN et al. (2012b). Kumar YR, Reddy AR, Eswaraiah S, Mukkanti K, Reddy
Chiral HPLC Method for the Enantioselective Analysis MS, Suryanarayana MV (2004). Structural charac-
of Rosiglitazone Maleate in Formulations. J Pharm Res, terization of impurities in pioglitazone. Pharmazie,
5(1):7981. 59(11):8369. PMID:15587582
GSK (2012). AVANDIA product monograph: La Maestra S, Micale RT, De Flora S, DAgostini F, Ganchev
GlaxoSmithKline Inc. G, Iltcheva M etal. (2013). DNA damage in exfoliated
Hassan MM, Curley SA, Li D, Kaseb A, Davila M, Abdalla cells and histopathological alterations in the urinary
EK etal. (2010). Association of diabetes duration and tract of mice exposed to cigarette smoke and treated with
diabetes treatment with the risk of hepatocellular carci- chemopreventive agents. Carcinogenesis, 34(1):1839.
noma. Cancer, 116(8):193846. doi:10.1002/cncr.24982 doi:10.1093/carcin/bgs314 PMID:23042096
PMID:20166205 Langchem (2013). Langchem material safety data sheet.
Hillaire-Buys D, Faillie JL, Montastruc JL (2011). China Langchem Inc.
Pioglitazone and bladder cancer. Lancet, Lefebvre A-M, Chen I, Desreumaux P, Najib J, Fruchart
378(9802):15434, author reply 15445. doi:10.1016/ JC, Geboes K etal. (1998). Activation of the peroxisome
S0140-6736(11)61662-0 PMID:22035550 proliferator-activated receptor promotes the develop-
IMS Health (2012a). Multinational Integrated Data ment of colon tumors in C57BL/6J-APCMin/+ mice.
Analysis (MIDAS). Plymouth Meeting, Pennsylvania: Nat Med, 4(9):10537. doi:10.1038/2036 PMID:9734399
IMS Health. Lewis JD, Ferrara A, Peng T, Hedderson M, Bilker
IMS Health (2012b). National Disease and Therapeutic WB, Quesenberry CP Jr et al. (2011). Risk of bladder
Index (NDTI) 20042012. Plymouth Meeting, cancer among diabetic patients treated with pioglita-
Pennsylvania: IMS Health. zone: interim report of a longitudinal cohort study.
Jaakkola T, Laitila J, Neuvonen PJ, Backman JT Diabetes Care, 34(4):91622. doi:10.2337/dc10-1068
(2006). Pioglitazone is metabolised by CYP2C8 and PMID:21447663
CYP3A4 in vitro: potential for interactions with Lewis JD, Strom BL, Bilker W, Nessel L, Vaughn DJ,
CYP2C8 inhibitors. Basic Clin Pharmacol Toxicol, Ferrara A, et al. (2012). Pioglitazone HCl (ACTOS)
99(1):4451. doi:10.1111/j.1742-7843.2006.pto_437.x Clinical Study No. 0103-TL-OPI-524. Cohort Study
PMID:16867170 of Pioglitazone and Bladder Cancer in Patients with
Jain D, Jain S, Jain D, Amin M (2008). Simultaneous Diabetes Fourth Interim Analysis (8-Year). Report
estimation of metformin hydrochloride, pioglitazone with Data from 1 January 1997 to 31 December 2010.
hydrochloride, and glimepiride by RP-HPLC in tablet Available from: http://general.takedapharm.com/
formulation. J Chromatogr Sci, 46(6):5014. doi:10.1093/ Trial-Disclosure/01-03-TL-OPI-524-8-year-Interim-
chromsci/46.6.501 PMID:18647470 Report.pdf. Accessed 2 March 2012.
Jedlicka A, Klimes J, Grafnetterov T (2004). Reversed- Lin ZJ, Ji W, Desai-Krieger D, Shum L (2003). Simultaneous
phase HPLC methods for purity test and assay of determination of pioglitazone and its two active metab-
pioglitazone hydrochloride in tablets. Pharmazie, olites in human plasma by LC-MS/MS. J Pharm Biomed
59(3):17882. PMID:15074587 Anal, 33(1):1018. doi:10.1016/S0731-7085(03)00344-3
Julie NL, Julie IM, Kende AI, Wilson GL (2008). PMID:12946536
Mitochondrial dysfunction and delayed hepatotox- Loke YK, Singh S, Furberg CD (2009). Long-term use of
icity: another lesson from troglitazone. Diabetologia, thiazolidinediones and fractures in type 2 diabetes:
51(11):210816. doi:10.1007/s00125-008-1133-6 a meta-analysis. CMAJ, 180(1):329. doi:10.1503/
PMID:18726085 cmaj.080486 PMID:19073651
Kirchheiner J, Roots I, Goldammer M, Rosenkranz Long GG, Reynolds VL, Lopez-Martinez A, Ryan TE,
B, Brockmller J (2005). Effect of genetic polymor- White SL, Eldridge SR (2008). Urothelial carcino-
phisms in cytochrome p450 (CYP) 2C9 and CYP2C8 genesis in the urinary bladder of rats treated with
on the pharmacokinetics of oral antidiabetic drugs: naveglitazar, a -dominant PPAR / agonist: lack of
clinical relevance. Clin Pharmacokinet, 44(12):1209 evidence for urolithiasis as an inciting event. Toxicol
25. doi:10.2165/00003088-200544120-00002 Pathol, 36(2):21831. doi:10.1177/0192623307311757
PMID:16372821 PMID:18474944
Krieter PA, Colletti AE, Doss GA, Miller RR (1994).
Disposition and metabolism of the hypoglycemic agent
376
Lu YC, Li YC (2013). How doctors practice evidence-based activity. J Clin Pharmacol, 42(12):1299302.
medicine. J Eval Clin Pract, 19:4449. doi:10.1111/ doi:10.1177/0091270002042012009 PMID:12463723
j.1365-2753.2011.01765.x PMID:21883721 OMaille G, Pai SM, Tao X, Douglas GT Jr, Jenkins RG
Lubet RA, Fischer SM, Steele VE, Juliana MM, Desmond (2008). An improved LC-ESI-MS-MS method for simul-
R, Grubbs CJ (2008). Rosiglitazone, a PPAR gamma taneous quantitation of rosiglitazone and N-desmethyl
agonist: potent promoter of hydroxybutyl(butyl) rosiglitazone in human plasma. J Pharm Biomed
nitrosamine-induced urinary bladder cancers. Anal, 48(3):9349. doi:10.1016/j.jpba.2008.08.001
Int J Cancer, 123(10):22549. doi:10.1002/ijc.23765 PMID:18818043
PMID:18712722 ONeil (2006). Pioglithazone. The Merck Index - An
Luzy A-P, Orsini N, Linget J-M, Bouvier G (2012). Encyclopedia of Chemicals, Drugs, and Biologicals.
Evaluation of the GADD45-GFP GreenScreen HC Whitehouse Station (NJ): Merck & Co., Inc.
assay for rapid and reliable in vitro early genotoxicity Palem CR, Gannu R, Yamsani SK, Yamsani VV, Yamsani
screening. J Appl Toxicol, n/a doi:10.1002/jat.2793 MR (2011). Development of a high-performance liquid
PMID:22806210 chromatography method for simultaneous determina-
Madivada LR, Anumala RR, Gilla G, Alla S, Charagondla tion of pioglitazone and felodipine in pig serum: appli-
K, Kagga M etal. (2009). An improved process for piogl- cation to pharmacokinetic study. Biomed Chromatogr,
itazone and its pharmaceutically acceptable salt. Org 25(8):9528. doi:10.1002/bmc.1553 PMID:21058416
Process Res Dev, 13(6):11904. doi:10.1021/op900131m Pang W, Yang H, Wu Z etal. (2009). LC-MS-MS in MRM
Martn J, Buchberger W, Santos JL, Alonso E, Aparicio mode for detection and structural identification of
I (2012). High-performance liquid chromatography synthetic hypoglycemic drugs added illegally to natural
quadrupole time-of-flight mass spectrometry method anti-diabetic herbal products. Chromatographia,
for the analysis of antidiabetic drugs in aqueous envi- 70:13531359. doi:10.1365/s10337-009-1344-0
ronmental samples. J Chromatogr B Analyt Technol Perez AT (2013). Pioglitazone and risk of bladder
Biomed Life Sci, 895896:94101. doi:10.1016/j. cancer: clarification of the design of the French study.
jchromb.2012.03.023 PMID:22483984 Diabetologia, 56(1):227 doi:10.1007/s00125-012-2767-y
Milne GWA (2000). Drugs: Synonyms & Properties. PMID:23135221
Ashgate Publ Co. Physicians Desk Reference (2012) Pioglitazone. 66th ed.
Monami M, Lamanna C, Marchionni N, Mannucci E Montvale (NJ): PDR Network.
(2008). Rosiglitazone and risk of cancer: a meta-anal- Piccinni C, Motola D, Marchesini G, Poluzzi E (2011).
ysis of randomized clinical trials. Diabetes Care, Assessing the association of pioglitazone use and
31(7):145560. doi:10.2337/dc07-2308 PMID:18375416 bladder cancer through drug adverse event reporting.
Mostafa GA, Al-Majed A (2008). Characteristics of new Diabetes Care, 34(6):136971. doi:10.2337/dc10-2412
composite- and classical potentiometric sensors for PMID:21515844
the determination of pioglitazone in some pharmaceu- Pino MV, Kelley MF, Jayyosi Z (2004). Promotion of colon
tical formulations. J Pharm Biomed Anal, 48(1):5761. tumors in C57BL/6J-APC(min)/+ mice by thiazolidin-
doi:10.1016/j.jpba.2008.04.029 PMID:18555633 edione PPARgamma agonists and a structurally unre-
National Library of Medicine (2010). Avandia (rosigli- lated PPARgamma agonist. Toxicol Pathol, 32(1):5863.
tazone maleate) tablet, film coated. Bethesda (MD): doi:10.1080/01926230490261320 PMID:14713549
National Institutes of Health, Health & Human Pubchem (2013). Pubchem Databases: National Center
Services. Available from: http://dailymed.nlm.nih.gov/ for Biotechnology, Pioglitazone. In CID.4829 U.S.
dailymed/lookup.cfm?setid=ef14122b-7fff-45fa-b13b- National Library of Medicine. Available from: http://
0ea9e48bd57d. Accessed 20 June 2014. pubchem.ncbi.nlm.nih.gov/summar y/summar y.
Neumann A, Weill A, Ricordeau P, Fagot JP, Alla F, cgi?q=all&cid=4829#ec. Accessed 10 July 2014.
Allemand H (2012). Pioglitazone and risk of bladder Pubchem Substance (2013). Rosiglitazone (SID 46504556).
cancer among diabetic patients in France: a popula- Available from: http://pubchem.ncbi.nlm.nih.gov/
tion-based cohort study. Diabetologia, 55(7):195362. summary/summary.cgi?sid=46504556
doi:10.1007/s00125-012-2538-9 PMID:22460763 Radhakrishna T, Satyanarayana J, Satyanarayana A
Nissen SE, Wolski K (2007). Effect of rosiglitazone on (2002b). LC determination of rosiglitazone in bulk
the risk of myocardial infarction and death from and pharmaceutical formulation. J Pharm Biomed
cardiovascular causes. N Engl J Med, 356:24572471. Anal, 29:873880. doi:10.1016/S0731-7085(02)00209-1
doi:10.1056/NEJMoa072761 PMID:17517853 PMID:12093521
NLM (2013). Hazardous Substances Data Bank: National Radhakrishna T, Sreenivas Rao D, Om Reddy G (2002a).
library of medicine. Determination of pioglitazone hydrochloride in
Nowak SN, Edwards DJ, Clarke A, Anderson GD, bulk and pharmaceutical formulations by HPLC and
Jaber LA (2002). Pioglitazone: effect on CYP3A4 MEKC methods. J Pharm Biomed Anal, 29(4):593607.
doi:10.1016/S0731-7085(02)00036-5 PMID:12093488
377
IARC MONOGRAPHS 108
Ravikanth CH, Kumar AA, Kiran VU (2011). Sensitive Song SO, Kim KJ, Lee BW, Kang ES, Cha BS, Lee HC (2012).
and rapid HPLC method for the determination of The risk of bladder cancer in korean diabetic subjects
pioglitazone in rat serum. Int J Pharmaceutical Sciences treated with pioglitazone. Diabetes Metab J, 36(5):3718.
Drug Research, 3(1):3841. doi:10.4093/dmj.2012.36.5.371 PMID:23130322
Richter J, Jirman J, Havlcek J, Hrdina R (2007). Souri E, Jalalizadeh H, Saremi S (2008). Development
Pioglitazone impurities. Pharmazie, 62(8):5804. and validation of a simple and rapid HPLC method for
doi:10.1691/ph.2007.8.6720 PMID:17867551 determination of pioglitazone in human plasma and its
Ruiter R, Visser LE, van Herk-Sukel MP, Geelhoed- application to a pharmacokinetic study. J Chromatogr
Duijvestijn PH, de Bie S, Straus SM et al. (2012). Sci, 46(9):80912. doi:10.1093/chromsci/46.9.809
Prescribing of rosiglitazone and pioglitazone following PMID:19007483
safety signals: analysis of trends in dispensing patterns Sripalakit P, Neamhom P, Saraphanchotiwitthaya A
in the Netherlands from 1998 to 2008. Drug Saf, (2006). High-performance liquid chromatographic
35:471480. doi:10.2165/11596950-000000000-00000 method for the determination of pioglitazone in human
PMID:22540371 plasma using ultraviolet detection and its application
Rx List (2013). RxList - The internet drug index. Available to a pharmacokinetic study. J Chromatogr B Analyt
from: http://www.rxlist.com Technol Biomed Life Sci, 843(2):1649. doi:10.1016/j.
Sakamoto J, Kimura H, Moriyama S, Odaka H, Momose jchromb.2006.05.032 PMID:16815107
Y, Sugiyama Y et al. (2000). Activation of human Sultana N, Arayne MS, Shafi N, Siddiqui FA, Hussain
peroxisome proliferator-activated receptor (PPAR) A (2011). Development and validation of new assay
subtypes by pioglitazone. Biochem Biophys Res method for the simultaneous analysis of diltiazem,
Commun, 278(3):70411. doi:10.1006/bbrc.2000.3868 metformin, pioglitazone and rosiglitazone by RP-HPLC
PMID:11095972 and its applications in pharmaceuticals and human
Sato K, Awasaki Y, Kandori H, Tanakamaru ZY, Nagai serum. J Chromatogr Sci, 49(10):7749. doi:10.1093/
H, Baron D et al. (2011). Suppressive effects of acid- chrsci/49.10.774 PMID:22080805
forming diet against the tumorigenic potential of Sweetman SC (2011) Monographs on drugs and ancil-
pioglitazone hydrochloride in the urinary bladder lary substances. In: Martindale: The Complete Drug
of male rats. Toxicol Appl Pharmacol, 251(3):23444. Reference. 37th ed. Vol. A. London, UK: Pharmaceutical
doi:10.1016/j.taap.2011.01.006 PMID:21255596 Press, pp.494.
SciFinder (2013). SciFinder Databases. Registry, Chemcats. Takeda Pharmaceuticals (2013). Actos Product
American Chemical Society. Information. Deerfield (Il): Takeda Pharmaceuticals
Seedher N, Kanojia M (2008). Micellar solubilization of America, Inc. Available from: http://www.takeda.us/
some poorly soluble antidiabetic drugs: a technical products. Accessed 19 June 2014.
note. AAPS PharmSciTech, 9:431436. doi:10.1208/ Thevis M, Geyer H, Schnzer W (2005). Identification
s12249-008-9057-5 PMID:18431666 of oral antidiabetics and their metabolites in human
Seedher N, Kanojia M (2009). Co-solvent solubilization urine by liquid chromatography/tandem mass spec-
of some poorly-soluble antidiabetic drugs. Pharm Dev trometrya matter for doping control analysis. Rapid
Technol, 14:185192. doi:10.1080/10837450802498894 Commun Mass Spectrom, 19(7):92836. doi:10.1002/
PMID:19519190 rcm.1875 PMID:15747323
Singh S, Loke YK, Furberg CD (2007a). Thiazolidinediones Tseng CH (2011). Diabetes and risk of prostate cancer:
and heart failure: a teleo-analysis. Diabetes Care, a study using the National Health Insurance.
30(8):214853. doi:10.2337/dc07-0141 PMID:17536074 Diabetes Care, 34(3):61621. doi:10.2337/dc10-1640
Singh S, Loke YK, Furberg CD (2007b). Long-term PMID:21273499
risk of cardiovascular events with rosiglitazone: A Tseng CH (2012a). Pioglitazone and bladder cancer: a
meta-analysis. JAMA, 298:11891195. doi:10.1001/ population-based study of Taiwanese. Diabetes Care,
jama.298.10.1189 PMID:17848653 35(2):27880. doi:10.2337/dc11-1449 PMID:22210574
Sireesha MR Chandan S, Gurupadayya BM, Aswani Tseng CH (2012b). Diabetes, metformin use, and colon
Kumar CH (2011). Selective and validated spectro- cancer: a population-based cohort study in Taiwan.
photometric methods for determination of rosiglita- Eur J Endocrinol, 167(3):40916. doi:10.1530/EJE-12-
zone and pioglitazone with 2, 4-DNP. Department of 0369 PMID:22778198
Pharmaceutical Analysis, JSS College of Pharmacy, JSS Tseng CH (2012c). Thyroid cancer risk is not increased in
University, Mysore, India. Available from: http://www. diabetic patients. PLoS ONE, 7(12):e53096 doi:10.1371/
ijptonline.com/wp-content/uploads/2009/10/1554- journal.pone.0053096 PMID:23300866
1564.pdf. Accessed 1 October 2014. Tseng CH (2012d). Pioglitazone and bladder cancer in
Smith U (2001). Pioglitazone: mechanism of action. Int J human studies: is it diabetes itself, diabetes drugs,
Clin Pract Suppl, (121):138. PMID:11594239 flawed analyses or different ethnicities? J Formos Med
378
Assoc, 111(3):12331. doi:10.1016/j.jfma.2011.10.003 Appl, 677(1):1416. doi:10.1016/0378-4347(95)00440-8

PMID:22423665 PMID:8925086
Tseng CH (2013a). Rosiglitazone is not associated Yardmc C, Ozaltn N, Grlek A (2007). Simultaneous
with an increased risk of bladder cancer. Cancer determination of rosiglitazone and metformin in
Epidemiol, 37(4):3859. doi:10.1016/j.canep.2013.03.013 plasma by gradient liquid chromatography with
PMID:23619142 UV detection. Talanta, 72(4):141622. doi:10.1016/j.
Tseng CH (2013b). Benign prostatic hyperplasia is a signif- talanta.2007.01.042 PMID:19071778
icant risk factor for bladder cancer in diabetic patients: Zhong WZ, Lakings DB (1989). Determination of piogli-
a population-based cohort study using the National tazone in dog serum using solid-phase extraction and
Health Insurance in Taiwan. BMC Cancer, 13(1):7 high-performance liquid chromatography with ultra-
doi:10.1186/1471-2407-13-7 PMID:23286275 violet (229 nm) detection. J Chromatogr A, 490(2):377
Tseng CH (2013c). Oral cancer in Taiwan: is diabetes a risk 85. doi:10.1016/S0378-4347(00)82795-4 PMID:2768410
factor? Clin Oral Investig, 17(5):135764. doi:10.1007/ Zhu Z, Shen Z, Lu Y, Zhong S, Xu C (2012). Increased risk
s00784-012-0820-3 PMID:22895832 of bladder cancer with pioglitazone therapy in patients
Venkatesh P, Harisudhan T, Choudhury H etal. (2006). with diabetes: a meta-analysis. Diabetes Res Clin
Simultaneous estimation of six anti-diabetic drugs Pract, 98(1):15963. doi:10.1016/j.diabres.2012.05.006
glibenclamide, gliclazide, glipizide, pioglitazone, PMID:22705039
repaglinide and rosiglitazone: development of a novel
HPLC method for use in the analysis of pharmaceu-
tical formulations and its application to human plasma
assay. Biomed Chromatogr, 20:10431048. doi:10.1002/
bmc.635 PMID:16506282
Wang M, Miksa IR (2007). Multi-component plasma
quantitation of anti-hyperglycemic pharmaceutical
compounds using liquid chromatography-tandem mass
spectrometry. J Chromatogr B Analyt Technol Biomed
Life Sci, 856:318327. doi:10.1016/j.jchromb.2007.06.037
PMID:17689303
Wei L, MacDonald TM, Mackenzie IS (2013). Pioglitazone
and bladder cancer: a propensity score matched cohort
study. Br J Clin Pharmacol, 75(1):2549. doi:10.1111/
j.1365-2125.2012.04325.x PMID:22574756
WHO (2007). Cumulative List No. 14 of INN in Latin,
English, French, Spanish, Arabic, Chinese and Russian
(in alphabetical order of Latin name), with supllemen-
tary information. World Health Organisation.
Woodcock J, Sharfstein JM, Hamburg M (2010).
Regulatory action on rosiglitazone by the U.S. Food and
Drug Administration. N Engl J Med, 363(16):148991.
doi:10.1056/NEJMp1010788 PMID:20942663
Xavier CM, Basavaiah K (2012). Implementation of
Quality by Design for the Development and Validation
of Pioglitazone Hydrochloride by RP-UPLC with
Application to Formulated Forms Int School Res
Network, 111.
Xue YJ, Turner KC, Meeker JB, Pursley J, Arnold M,
Unger S (2003). Quantitative determination of pioglita-
zone in human serum by direct-injection high-perfor-
mance liquid chromatography mass spectrometry and
its application to a bioequivalence study. J Chromatogr
B Analyt Technol Biomed Life Sci, 795(2):21526.
doi:10.1016/S1570-0232(03)00575-0 PMID:14522026
Yamashita K, Murakami H, Okuda T, Motohashi M
(1996). High-performance liquid chromatographic
determination of pioglitazone and its metabolites
in human serum and urine. J Chromatogr B Biomed
379
DIGOXIN
1. Exposure Data The Working Group noted that most of what

has been used under the term digitalis in
Digoxin is a cardiac glycoside isolated from North America and Europe has been digoxin;
plants of the genus Digitalis. The use of prepa- however, there may be parts of the world where
rations of cardiac glycoside (synonyms: digi- crude extract of the digitalis plant is still in use.
talis, cardiac steroids) dates back to 1785, when No data on the use of digitalis extract were avail-
William Withering published his monograph An able to the Working Group.
account of the foxglove and some of its medical
uses (Withering, 1785; Albrecht & Geiss, 2000). 1.1 Chemical and physical data
Isolated digoxin has been used since the early
20th century (Cheng & Rybak, 2010). 1.1.1 Nomenclature
The Working Group noted that only four of
the many digitalis glycosides present in the plant Chem. Abstr. Serv. Reg. No.: 20830-75-5
remain important in the marketplace. These (SciFinder, 2013)
are digoxin, digitoxin, -acetyldigoxin and Chem. Abstr. Serv. Name: Card-20(22)-enolide,
methyldigoxin (Kleemann, 2012). Furthermore, 3-[(O-2,6-dideoxy--D-ribo-hexopyranosyl-
the term digitalis use found in many reports -(14)-O -2 ,6 -dideox y--D-r ibo-hexo
probably refers not to the use of plant mate- pyranosyl-(14)-2,6-dideoxy--D-ribo-
rial, which is not commercially available as a hexopy ranosyl)ox y]-12,14-dihydrox y-,
medicinal product, but to the use of the isolated (3,5,12)- (SciFinder, 2013)
compounds. Of the four medicinally available IUPAC Systematic Name: 3-[(3S,5R,8R,
compounds, digoxin is the most important and 9S,10S,12R,13S,14S,17R)-3-[(2R,4S,5S,6R)-
is exclusively available in some countries, such as 5 - [(2 S , 4 S , 5 S , 6 R ) - 5 - [(2 S , 4 S , 5 S , 6 R ) -
the USA (see Section 1.3). The Working Group 4 , 5 - d i h y d r o x y - 6 - m e t h y l o x a n -2 -y l ]
estimated that digoxin represents at least 90% of oxy-4-hydroxy-6-methyloxan-2-yl]oxy-4-
the world market for digitalis glycosides. hydroxy-6-methyloxan-2-yl]oxy-12,14-dihy-
While use of digitoxin worldwide is much droxy-10,13-dimethyl-1,2,3,4,5,6,7,8,9,11,12,1
less than that of digoxin, it may be significant 5,16,17-tetradecahydrocyclopenta[a]phenan-
in individual countries. Thus, studies reporting thren-17-yl]-2H-furan-5-one (PubChem,
use of digitalis should be carefully scrutinized 2013)
since the agent to which people were actually Synonyms: 12-hydroxydigitoxin
exposed could have been any one of the four
digitalis glycosides.
381
IARC MONOGRAPHS 108
Proprietary names for digoxin: Cardigox; Spectroscopy data: Specific optical rotation,
Cardiogoxin; Cardioxin; Cardixin; Cardoxin; ultraviolet, infrared, nuclear magnetic reso-
Chloroformic digitalin; Coragoxine; nance, and mass spectral data were reported
Cordioxil; Davoxin; Digacin; Digicor; Digitek; in the literature (Foss & Benezra, 1980; British
Digomal; Digon; Digosin; Digoxin Nativelle; Pharmacopoeia, 2009; HSDB, 2013)
Dilanacin; Dixina; Dokim; Dynamos; Solubility: In water, 64.8 mg/L at 25 C;
Eudigox; Fargoxin; Grexin; Homolles soluble in dilute alcohol, pyridine, or mixture
digitalin; Lanacordin; Lanacrist; Lanicor; of chloroform and alcohol; almost insoluble
Lanikor; Lanocardin; Lanorale; Lanoxicaps; in ether, acetone, ethyl acetate, chloroform;
Lanoxin; Lanoxin PG; Lenoxicaps; Lenoxin; slightly soluble in diluted alcohol, and very
Longdigox; Mapluxin; NSC 95100; Natigoxin; slightly soluble in 40% propylene glycol;
NeoDioxanin; Novodigal-Amp.; Purgoxin; (PubChem, 2013)
Rougoxin; Stillacor; Toloxin; Vanoxin (from Stability data: Digoxin is indefinitely stable
SciFinder, 2013). when kept in the dark in a tightly closed
container. No degradation is noted in tablets
1.1.2 Structural and molecular formulae and after 5 years when stored in tightly closed
relative molecular mass containers. A solution of digoxin hydrolyses
in the presence of acid, yielding digoxigenin
From USP (2007) bis-digitoxoside, digoxigenin mono-digitox-
O
O
oside and digoxigenin. A neutral solution in
OH
CH 3
H
ethanol and propylene glycol is stable for up
H3C H to 5 years. Digoxin solutions are relatively
H3C H3 C H3 C
O O O H OH stable to light, except when stored under
HO O O O
H intense light for long periods of time (Foss &
HO HO HO Benezra, 1980)
Storage: Digoxin preparations should be
C41H64O14 protected from light and stored at 1525 C
Relative molecular mass: 780.94 (HSDB, 2013)
Octanol/water partition coefficient (log P):
1.26 (HSDB, 2013)
Dissociation constant: pKa, basic = 3; pKa,
pure substance
acidic=7.15 (DrugBank, 2013)
Description: Odourless, colourless to white Vapour pressure: 3.31030 mm Hg at 25 C
crystals, or white crystalline powder, radi- (PubChem, 2013)
ally arranged four- and five-sided triclinic Flash point: 278.527.8 C (SciFinder, 2013)
plates from dilute alcohol or pyridine (British
Pharmacopoeia, 2009; PubChem, 2013)
1.1.4 Technical products and impurities
Melting point: Digoxin melts and decomposes
between 230 C and 265 C (Foss & Benezra, Since digoxin is isolated from plant materials,
1980; ChemicalBook, 2013) at least 21 other cardiac glycosides, including
Density: 1.360.1 g/cm3 (temperature, 20 C; digitoxin, may occur as impurities (British
pressure, 760 Torr) (SciFinder, 2013) Pharmacopoeia, 2009). The purity of digoxin is
382
Digoxin
typically at least 95% (see Section 1.5). According 1.3 Production and use
to the European Pharmacopoeia (2008), not more
than 0.5% digitoxin in relation to digoxin may be 1.3.1 Production
present as impurity. Digoxin is isolated from Digitalis lanata Ehrh.,
(a) Nomenclature for digitoxin the woolly foxglove, from the Scrophulariaceae
family. For the isolation of the therapeutically
Chem. Abstr. Serv. Reg. No.: 71-63-6 (SciFinder, important secondary glycosides, the finely
2013) ground material is moisturized and exposed to
Chem. Abstr. Serv. Name: 3-[(O-2,6-dide- glucosidase enzymes at 3037 C until glucose
oxy--D-ribo-hexopyranosyl-(14)-O-2,6- is completely removed. Extraction procedures,
dideoxy--D-ribo-hexopyranosyl-(14)- usually followed by precipitation of tannic
2,6 -dideox y--D-ribo-hexopy ra nosyl) acid and related phenolic products with lead
oxy]-14-hydroxy-4,14-card-20(22)-enolide. salts, afford a crude mixture of cardioactive
compounds, which is further purified by chro-
Proprietary names for digitoxin: Crystodigin,
matography and/or crystallization. Originally,
Digimed, Digimerck.
mixtures of glycosides or crude plant extracts
were used in therapy; these have been replaced by
(b) Structural and molecular formulae and chemically pure drugs today, which allow better
relative molecular mass of digitoxin control of therapy. Total syntheses of cardiac
O O
steroids and their corresponding glycosides have
CH 3 been accomplished but are not used commer-
H3C H
H
cially (Albrecht & Geiss, 2000).
H3C
O
H3 C
O
H3 C
O H OH
Digitoxin is isolated by extraction of the
HO O O O
H
leaves and seeds of Digitalis purpurea L. (purple
HO HO HO
foxglove) with 50% ethanol and subsequent
treatment with the enzyme digilanidase, which
effects cleavage of the -D-glucose moiety at the
C41H64O13 chain end of the main glycoside, purpureaglyco-
Relative molecular mass: 764.94 side A (Kleemann, 2012).
-Acetyldigoxin is prepared from digoxin by
acetylation with acetic acid. Methyldigoxin can
1.2 Analysis be prepared by methylation of digoxin, e.g. with
dimethyl sulfate (Kleemann, 2012).
Compendial methods to determine digoxin
and digitoxin in pharmaceutical preparations 1.3.2 Use
are typically based on liquid chromatography
with ultraviolet detection. For detection in (a) Indications
human plasma or urine, liquid chromatography Digoxin and digitoxin are therapeutically the
with mass spectrometric detection is required to most widely used digitalis glycosides. Table 1.2
achieve the necessary lower detection limits. The lists the most commonly reported clinical indica-
analytical methods are summarized in Table1.1. tions for digoxin in the USA. While digoxin was
once regarded as the drug of choice for conges-
tive heart failure with reduced left ventricular
383
384
Table 1.1 Analytical methods for digoxin
Sample matrix Sample preparation Assay method Detection Reference

limit
Compendial methods
Digoxin injection, digoxin LC-UV NR USP (2007)
Tablet and digoxin oral Column: Packing L1
solution Mobile phase: water and acetonitrile
Flow rate: 3 mL/min
Wavelength: 218 nm
IARC MONOGRAPHS 108
Digoxin injection, paediatric LC-UV NR BP (2009)

digoxin injection, paediatric Column: C18
digoxin oral solution, and Mobile phase: acetonitrile:water (10:90) and
digoxin tablets water:acetonitrile (10:90)
Wavelength : 220 nm
Human plasma, rat plasma Addition of DMA, addition of NaCl LC-MS-MS 0.1 ng/mL Hirabayashi et al.
and rat brain saturated 0.1 mol/L NaOH, collection of Column: C18 (LLOQ) (2011)
organic layer, centrifugation Mobile phases: ammonium carbonate, and
methanol
pH 9.0
SRM: 779.4 m/z, 649.4 m/z
Human plasma Deproteinization with perchloric acid in LC-ESI-MS 0.5 ng/mL Vlase et al.
water, mixing and centrifugation Column: C18 (LLOQ) (2009)
Mobile phase: mixture of methanol and formic
acid in sodium acetate
Flow rate: 1mL/min
SIM: 803.5 m/z
Human blood and tissues Mixing with sodium acetate buffer LC-ESI-MS 0.2 ng/g Frommherz et al.
pH7, homogenization, centrifugation, Column: C8 (LLOQ) (2008)
loaded on SPE column conditioned with Mobile phase: 0.1% formic acid in a mixture of 55%
methanol, water, and sodium acetate methanol and 45% water
buffer, washing with sodium acetate Flow rate: 0.2mL/min
buffer, dried under vacuum, second wash SIM: 803.4 m/z
with 20% isopropyl alcohol, drying,
addition of acetone, vacuum drying,
elution with acetone

limit
Human serum Addition of methyl tert-butyl ether, LC-ESI-MS 0.1 ng/mL Li et al. (2010)
centrifugation, evaporation, and Column: C18 (LLOQ)
reconstitution in methanol Mobile phase: 10 mM ammonium acetate/0.1%
formic acid in water and 0.1% formic acid in
acetonitrile
SRM transition: 798.6 m/z, 651.5 m/z
Human blood Mixing with ammonium carbonate LC-ESI-MS 0.08 ng/mL Oiestad et al.
buffer, extraction (ethyl acetate/ Column: C18 (LLOQ) (2009)
hepatene/dichloromethane=3:1:1), Mobile phase: 10 mM ammonium formate and 0.032 ng/mL
centrifugation, collection of organic acetonitrile (LOD)
layer, evaporation, reconstitution in pH 3.1
acetonitrile:water Flow rate: 0.3 mL/min
MRM transitions: 798.4 m/z, 651.3 m/z; 798.4 m/z,
633.3 m/z
Human plasma Mixing with 10% ammonium hydroxide, LC-ESI-MS 0.1 ng/mL Grabowski et al.
addition of chloroform, centrifugation, Column: UPLC AQUITY (LLOQ) (2009)
evaporation, reconstitution in 1mM Mobile phase: 30% 1mM ammonium
trifluoroacetic acid and acetonitrile (7:3). trifluoroacetate in acetonitrile and 100% water
SIM transition: 780.94 m/z, 893.5 m/z
Human plasma Addition of concentrated NaOH LC-ESI-MS 0.05 ng/mL Kirby et al.
and methyl t-butyl ether, shaking, Column: C8 (LLOQ) (2008)
centrifugation, evaporation, Mobile phase: 0.25 mM sodium acetate in water 0.025 ng/mL
reconstitution in mobile phase and 0.25 mM sodium acetate in methanol (LOD)
SIM: 803.4 m/z (positive mode)
Human plasma Addition of buffer solution pH 6.0, LC-ESI-MS 0.04 ng/mL Hashimoto et al.
loading into oasis HLB30 mg 96-well Column: C18 (LLOQ) (2008)
plate preconditioned with methanol:water Mobile phase: 10 mmol/L ammonium hydrogen
(40:60), elution of analyte with pure carbonate/methanol (1:9) and 10 mmol/L
methanol, evaporation and reconstitution ammonium hydrogen carbonate/methanol (9:1)
in methanol Flow rate: 0.6mL/min
SRM transition: 798.5 m/z, 651 m/z
385
Digoxin
386

limit
Human plasma and urine Addition of buffer solution pH6, loading LC-ESI-MS 0.2 ng/mL Salvador et al.
into oasis HLB30 mg 96-well plate Column: C18 (LLOQ) (2006)
preconditioned with methanol:water Mobile phase: 5 mM ammonium acetate and 1 ng/mL
(40:60), elution of analyte with pure acetonitrile (LLOQ)
methanol, evaporation, reconstitution in Flow rate: 250 L/min
methanol SRM transition: 798.5 m/z, 651.4 m/z (positive
mode)
IARC MONOGRAPHS 108
Drinking-water, ground SPE by using oasis HLB cartridge LC-MS-TOF 11000 ng/L Ferrer &
water, Column: C8 (LOD) Thurman (2012)
surface water, and waste Mobile phase: acetonitrile, water with 0.1% formic
water acid
Water, soil, sediment, and Extraction with solvents, and SPE LC-MS-MS 50 ng/L in EPA (2007)
biosolids water (LOD)
Plant extract Extracted from herbaceous plants of the LC-ESI-MS 38936 pg/g Josephs et al.
genus Digitalis Column: C18 in solution (2010)
Mobile phase: aqueous ammonium formate/ (LOD)
methanol (40/60% v/v), pure methanol
Rat plasma Addition of ammonium chloride buffer, LC-ESI-MS 0.1 ng/L Yao et al. (2003)
acetonitrile and methylene chloride, Column: C18 (LOQ)
vortexing, centrifugation, evaporation of Mobile phase: acetonitrile/ammonium formate
organic layer, reconstitution Flow rate: 0.2 mL/min
Human serum Incubation, centrifugation, supernatant IC Visual Omidfar et al.
loaded into a vial and frozen Colloidal gold mAb probe-colloidal gold conjugate detection (2010)
with IgG limit,
2 ng/mL
Detection
time, 25min
Human blood and urine Addition of water and ammonium acetate LC-ESI-MS 0.05 ng/mL Guan et al.
buffer (2M, pH9.5), centrifugation, Column: C18 (LLOQ) (1999)
collection of supernatant, clean-up by SPE Mobile phase: 20% acetonitrile in 80% 2 mM
ammonium formate and 80% acetonitrile in 20%
2mM ammonium formate

limit
Rat intestinal perfusion NR LC-UV 25 ng/mL Varma et al.
samples Column: C18 (LOQ) (2004)
Mobile phase: 10 mM ammonium acetate,
methanol, acetonitrile (50:25:25)
pH3.0
Wavelength: 220 nm
Human plasma NR LC-ESI-MS NR Tracqui et al.
Column: C18 (1997)
Mobile phase: acetonitrile and 2 mM ammonium
acetate
pH3.0
SRM transition:7 99 m/z
DMA, N,N-dimethylacetamide; IC, immunochromatography; IgG, immunoglobulin G; LC-ESI-MS, liquid chromatography electrospray ionization mass spectrometry; LC-MS-MS,
liquid chromatography tandem mass spectrometry; LC-TOF-MS, liquid chromatography time of flight mass spectrometry; LC-UV, liquid chromatography ultraviolet spectroscopy;
LLOQ, lower limit of quantification; LOD, limit of detection; LOQ, limit of quantification; mAb, monoclonal antibody; min, minute; MRM, multiple reaction monitoring; m/z, mass/
charge; NaCl, sodium chloride; NaOH, sodium hydroxide; NR, not reported; SIM, selected ion monitoring; SPE, solid-phase extraction; SRM, selected reaction monitoring
387
Digoxin
IARC MONOGRAPHS 108
Table 1.2 Most commonly reported clinical indications for digoxin in the USA, 20112012
Diagnosis ICD-9 codea Drug uses (in Percentage of

1000s) total
Atrial fibrillation 427.301 1595 42.3
Congestive heart failure 428.001 501 13.3
Other primary cardiomyopathy, NOS 425.402 113 3.0
Chronic ischaemic disease, unspecified 414.901 81 2.1
Surgery after heart disease treatment V67.038 53 1.4
Medical follow-up after atherosclerotic heart disease V67.533 50 1.3
Paroxysmal supraventricular tachycardia 427.001 50 1.3
a ICD-9 codes are a more detailed, proprietary version developed by IMS Health.
Prepared by the Working Group on the basis of data from IMS Health (2012b)
ICD-9, International Classification of Diseases Revision Nine; NOS, not otherwise specified
ejection fraction and for atrial fibrillation, it has in preference for other medications, particu-
been largely supplanted by other medications larly -blockers and non-dihydropyridine
(Sleeswijk et al., 2007). Digitoxin is useful for calcium-channel blockers. Digoxin is generally
maintenance therapy because its long half-life less effective than other drugs in producing
(59days) provides a sustained therapeutic effect consistent reduction of heart rate, particularly
even if a dose is missed. For the same reason toxic during exertion (McNamara et al., 2003). Joint
reactions are not easy to manage. Elimination is USA/European Union guidelines recommend
independent of renal function (Albrecht & Geiss, against use of digoxin as a first-line agent in most
2000). cases of atrial fibrillation (Fuster et al., 2006).
For congestive heart failure, use of digoxin
fails to improve survival (Digitalis Investigation (b) Dosage
Group, 1997) when compared with placebo, Administration is typically oral, although
unlike other leading therapies. It does, however, preparations for intravenous administration
provide symptomatic benefits in some cases exist. Typically, digoxin is used orally for months
and is associated with reduced risk of hospital- to years, while intravenous use requires careful
ization. USA guidelines suggest its use in situa- medical monitoring and is given only in the
tions where recommended therapies (diuretics, short-term. The absorption ratio was found to be
angiotensin-converting-enzyme inhibitors and 70%, the decay ratio is 20%, the effective dose
-blockers) fail to produce adequate symptom level is 2mg, and the maintenance dose is 0.5mg
relief (Hunt et al., 2009). European guidelines (Albrecht & Geiss, 2000).
continue to recommend digoxin as one of several For the treatment of heart failure, atrial fibril-
therapies used in combination for the manage- lation, the loading-dose regimen for intravenous
ment of congestive heart failure (Dickstein et al., administration is a single dose of 0.40.6 mg,
2008). with additional doses of 0.10.3 mg every
As for congestive heart failure, use of 68 hours to be given with caution until there
digoxin for atrial fibrillation has also declined is clinical evidence of adequate effect, and the
388
Digoxin
total dose should not exceed 0.0080.015mg/kg 2004). Trends in the European Union may have
bw. The oral dosage for this indication is a single lagged behind those in the USA, but use for both
dose of 0.50.75 mg, then additional doses of conditions has declined (Sturm et al., 2007).
0.1250.375 mg may be given cautiously every Use of digoxin may have been reduced between
68 hours until clinical evidence of adequate 1991 and 2004 in the USA, but not in the United
effect, up to a total dose of 0.751.25 mg (for a Kingdom (Haynes et al., 2008).
patient weighing 70kg). The maintenance dose is The Food and Drug Administration (FDA)
0.1250.5mg/day, intravenous or oral (Medscape reported that digitoxin and acetyldigitoxin are
(2013). no longer manufactured in the USA (FDA, 2013).
Most generic tablet preparations of digoxin Globally, there are 160 licensed products
average 7080% oral bioavailability, with containing digoxin, while there are only seven
90100% oral bioavailability for digoxin elixir licensed products containing digitoxin in
and the encapsulated gel preparation. Parenteral Germany, Austria, Hungary, and Norway (Index
digoxin is available for intravenous administra- Nominum, 2013).
tion, and is of value in patients who are unable Despite the introduction of new therapeutic
to take oral formulations. Caution to avoid over- strategies, cardiac glycosides are still widely used,
dosing is necessary in elderly patients or those and digoxin belongs to the 10 most frequently
with renal impairment (Li-Saw-Hee & Lip, 1998). prescribed drugs in the USA (Albrecht & Geiss,
In general, the therapeutic index for digoxin is 2000). In Estonia, the consumption of digoxin
narrow (Ehle et al., 2011). was very high in the times of the former Soviet
When digoxin is indicated, suggested thera- Union and decreased in the first years of inde-
peutic ranges of serum concentrations of digoxin pendence. When problems with drug availability
are lower now than in the past (Hunt et al., 2009), were overcome, the use of digoxin increased by
particularly given the report that mortality 35% in 199497 (Phkla et al., 1999).
among digoxin users was associated with higher While a rare event, the homicidal use of
serum concentrations of this drug (Rathore et al., digoxin has been described. Suicide by digoxin
2003). In a study of post-mortem cases, the range may have been more frequent in continental
of serum digoxin concentrations in cases of over- Europe, but has also occurred in the USA and
dose was 2.76.8 nmol/L (mean, 4.7 nmol/L) England (Burchell, 1983).
[2.15.3ng/L (mean, 3.7ng/L)] (Eriksson et al., Total worldwide sales of digoxin were
1984). US$ 142 million in 2012, with 33% occurring
Country-dependent differences in formula- in the USA (US$ 47 million). Other nations
tions may be correlated to the range of available reporting appreciable use of digoxin included
tablet strengths. For example, the dosage was Japan (US$14 million), Canada (US$11 million),
significantly higher in some hospitals in the USA and the United Kingdom (US$9 million) (IMS
and France than in the United Kingdom, and Health, 2012a).
significantly higher in France than in the USA In the USA in 2012, digoxin was reported by
(Saunders et al., 1997). office-based physicians in 1.85million drug uses,
and was being taken by approximately 700000
(c) Trends in use patients (IMS Health, 2012b). The trend in use of
Use of digoxin in the USA has declined digoxin in the USA is shown in Fig.1.1. According
substantially for treatment of congestive heart to the IMS Health National Prescription Audit
failure (Banerjee & Stafford, 2010) and of atrial Plus, there were a total of 9.6million prescriptions
fibrillation (Stafford et al., 1998; Fang et al.,
389
IARC MONOGRAPHS 108
for digoxin in 2012, down from 14.6 million 2. Cancer in Humans

prescriptions in 2008 (IMS Health, 2012c).
Beginning in the late 1970s, several small
1.4 Occurrence and exposure studies based on case series or chart reviews
reported a lower risk of cancer of the breast in
1.4.1 Natural occurrence women using digitalis (see introduction to
The principal natural occurrence of digoxin Section 1) (Stenkvist et al., 1979, 1982; Goldin &
is in the leaves of Digitalis lanata Ehrh., but it Safa, 1984). These reports, mostly in brief corre-
may also occur in some other Digitalis species spondence, have been cited as supporting the
(Hollman, 1985). After leaf-tissue damage or consideration of digitalis as a possible therapy
plant harvest, the primary glycoside lanatoside C for cancer of the breast (Stenkvist, 1999; Haux,
is converted to the secondary glycoside digoxin 1999); however, because so little information
by the endogenous enzyme, digilanidase, present was provided and larger studies with stronger
in the leaves, and by subsequent deacetylation. D. designs were available, these early studies were
lanata leaves were found to contain digoxin at judged to be uninformative and were not consid-
8.613.2g/100mg and its precursor, lanatoside ered further.
C, at 55.8153.2 g/100 mg, depending on the The studies reviewed by the Working Group
health of the plant material (Pellati et al., 2009). included a measure of relative risk, such as odds
Environmental factors that influence the digoxin ratios, hazard ratios, and incidence rate ratios.
content in D. lanata are carbon-dioxide enrich- Varied designs were used in these studies. Some
ment and water stress (Stuhlfauth et al., 1987). studies evaluated associations between risks of
cancers of all types and exposures to a wide range
of pharmaceuticals, or to a more restricted range
1.4.2 Occupational exposure
of cardiovascular drugs. Others examined risk
No data were available to the Working Group. factors for specific cancers, typically including
prescription drugs together with evaluation of
other demographic and health parameters. In
1.5 Regulations and guidelines recent years, national registries of prescription
Digoxin has been assigned classification as a drug use have yielded large data sets in which
water hazard in Germany and as an environ- follow-up can be linked to cancer outcomes in
mental hazard in several USA states (SciFinder cohort studies.
(2013). The United States Environmental Many reports described only digitalis
Protection Agency (EPA) assigned it to the list exposure, and therefore may refer to either
of extremely hazardous substances mandated digoxin (much more commonly used, especially
by Section 302 of the Emergency Planning in recent years) or digitoxin. Even when some
and Community Right-to-Know Act of 1986 epidemiological studies specified digoxin, the
(EPCRA), for which the reportable quantity subjects who were enrolled during years when
is 10 lbs [~4.5 kg] and the threshold planning digitoxin was more widely used might have also
quantity is 10/10000lbs [4.5/4536kg]. used digitoxin (e.g. because of renal failure). The
Digoxin is specified in several official phar- studies describing digitalis use are therefore
macopoeias (Table1.3). included, with the exposure type digoxin, digi-
toxin, or digitalis, indicated in the tables. Most
390
Digoxin
Fig.1.1 Trends in use of digoxin as a drug in the USA, 20042012
1500
1200
900
600
300
0
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
Prepared by the Working Group on the basis of data from IMS Health National Disease and Therapeutic Index, 200412 (IMS Health, 2012b).
of what has been used under the term digitalis digitalis than were women without cancer of the
in North America and Europe has been digoxin. breast (relative risk, RR, 2.67; 95% CI, 0.998.33;
in the subset restricted to 65 pairs with similar
follow-up time.). [Both cases and controls were
2.1 Cancer of the breast hypertensive and both were therefore at a high
2.1.1 Casecontrol studies risk of cardiovascular disease. This compara-
bility enhanced internal validity, but it may have
See Table2.1 reduced generalizability.]
Studies of the association of risk of cancer Lenfant-Pejovic et al. (1990) described risk
with use of digoxin and related drugs have factors for cancer of the breast in men in France
focused mainly on cancer of the breast. Aromaa and Switzerland, comparing 91 cases with 255
et al. (1976) reported a register-based case controls recruited from hospital cancer clinics in
control study in which use of digitalis (and France and a cancer registry in Switzerland, and
many other cardiovascular drugs) in the year matched for age and area of residence. Data on
before diagnosis was compared in 109 hyper- risk factors were limited to information available
tensive women with cancer of the breast and in physician interviews by mail or telephone, and
in 109 matched hypertensive women without clinical record reviews. Of all prescribed drugs,
cancer of the breast. Hypertensive women with only use of digitalis for at least 3months before
cancer of the breast were more likely to be using
391
392
Table 1.3 Regulations in pharmacopoeial monographs on digoxin
WHO International Pharmacopoeia, United States European Pharmacopoeia 7.0 Japanese Pharmacopoeia XVI
Regulation 4th edition Pharmacopeial
Convention30
Content C 41H64O14 95.0103.0% 95.0101.0% 96.0102.0% 96.0106.0%
(dried substance)
Identity tests Tests ABD or BCD may be applied: A. IR IR 1. Colour reaction with ferric
A. IR B. HPLC chloride hexahydrate/acetic acid/
B. TLC C. TLC sulfuric acid
C. Colour reaction with dinitrobenzene/ 2. IR
IARC MONOGRAPHS 108
ethanol
D. Colour reaction with ferric chloride/
glacial acetic acid/sulfuric acid
Specific optical +13.6 to +14.2 (0.10 g/mL in pyridine) +13.9 to 15.9 (0.50 g in 25 mL +10.0 to + 13.0
rotation methanol/methylene chloride (0.20 g in 10 mL pyridine)
50:50)
Sulfated ash Max. 1.0 mg/g Max. 0.1%
Loss on drying Max. 10 mg/g Max. 1.0% Max. 1.0% Max. 1.0%
Residue on ignition Max. 0.5% Max. 0.5%
Gitoxin Absorbance at 352 nm, max. 0.22 (about
40 mg/g)
Related substances/ TLC test, absence of spots that are TLC test, no spot that HPLC: specific limits for about HPLC: total area of peaks of
purity more intense than standard solution at is more intensive than 12 related substances are impurities is max. 3%
0.25 mg/mL gitoxin standard solution specified
(not more than 3% of
any related glycoside as
gitoxin)
Organic volatile General requirements,
impurities except limits for methylene
chloride and chloroform
are 2000 g/g
Bacterial Max. 200.0 IU of endotoxin per mg
endotoxins
HPLC, high-performance liquid chromatography; IR, infrared; IU, international units; TLC, thin-layer chromatography
Adapted from The United States Pharmacopoeial Convention (2006), European Pharmacopoeia (2008), The International Pharmacopoeia (2011), Pharmaceuticals and Medical Devices
Agency (2011)
Digoxin
diagnosis was associated with increased risk (11 drugs, confounding by indication, and frequency
users among cases; odds ratio, OR, 4.1; 95% CI, of mammography. [This large study was regarded
1.412.4). [The data from France and Switzerland as being of high quality. However, the Working
were collected in different ways and the Working Group noted that some important risk factors
Group questioned the quality of the data obtained of cancer of the breast, notably parity, obesity,
from medical records and physician interviews.] and alcohol drinking, were not controlled in the
In another study of risk factors for cancer of analysis.]
the breast in men, Ewertz et al. (2001) compared
156 incident cases in men in Norway, Sweden, 2.1.2 Cohort studies
and Denmark with 468 men matched for year
of birth, and country. Many variables were eval- See Table2.2
uated using self-administered questionnaires, Using data from persons enrolled in the
including use of prescribed drugs. Among all Kaiser Permanente Medical Care Programme,
drugs assessed, digoxin stood out most strongly, Friedman & Ury (1980) linked prescription-drug
with odds ratios for digoxin of 1.8 (95% CI, use for 95 drugs and drug classes between 1969
0.74.4) in men with < 5 years use and 2.0 and 1973 to subsequent cancer outcomes (56
(0.94.4) for 5years use. After adjustment for types) registered within this health-care system
body mass index determined from self-estimated until 1976. The drugs evaluated included digi-
weight and height 10 years before diagnosis, the talis as a group. A more detailed presentation
association between cancer of the breast and of digitalis-related associations used cancer-out-
digoxin use was still 1.8 (P=0.08). [Recalculated come data for 143594 subjects updated to 1980
by the Working Group from observed/expected (Friedman, 1984) (results provided in Table2.2).
data to be 1.9 (95% CI, 1.053.48).] The agesex standardized morbidity ratio for
Ahern et al. (2008) identified 5565 postmeno- cancer of the breast and ever-use of digitalis was
pausal women with incident cancer of the breast 1.2 [95% CI, 0.741.87]. [This study was large and
who used digoxin with a 10:1 birth year- and resi- was able to examine the association of cancer
dence area-matched population-control group with many different drugs; however, the preci-
in Denmark in 19912007. Use of digoxin was sion of specific drugcancer associations was
ascertained by county-level prescription registry limited and there was some concern about the
data, and by design, all subjects were required to large number of comparisons.]
have used digoxin for 2years before diagnosis Haux et al. (2001) used a database of plasma
(and use was likely to be current). Adjustments concentrations of digitoxin for 9271 women
included age, past use of hormone replacement and men in Trondheim, Norway, who were
therapy, nonsteroidal anti-inflammatory drugs undergoing their first treatment with digitoxin
(NSAIDs), and anticoagulants including aspirin. between 1986 and 1996. The risk of developing
Among the cases of cancer of the breast, 324 used cancer in people receiving their first treatment
digoxin compared with 2546 controls, yielding with digitoxin was compared with the incidence
an adjusted odds ratio of 1.30 (95% CI, 1.141.48). of cancers with at least 30 expected cases (all
Relative to non-users, the odds ratios increased sites, breast, prostrate, colorectum, lung, kidney/
with duration of use from 1.25 (95% CI, 1.031.52) urinary, melanoma, lymphoid/leukaemia) in
with 13years of use to 1.30 (95% CI, 1.051.61) the national population. Standardized incidence
with 46years of use to 1.39 (95% CI, 1.101.74) ratios (SIR) for most cancers, including cancer of
with >6years of use. The findings persisted after the breast, were higher (typically by about 25%)
adjustment for exposure to estrogen, use of other among digitoxin users. In an analysis of cancer
393
394
Table 2.1 Casecontrol studies on use of digoxin and cancer of the breast
Reference, Subjects Exposure Organ site Exposed Exposure Relative risk Adjustments Comments
study assessment cases category (95% CI) for potential
location and confounders
period
Aromaa Women with Prescription- Breast 28 Any digitalis 1.33(0.732.48) Age, Digitalis use was a
et al. (1976), breast cancers and acquired use vs no use geographical secondary outcome,
Finland, hypertension (n = cardiovascular area but the strongest
cases 109) compared with drugs 16 Any digitalis 2.67 (0.998.33) association seen
reported in matched women with use vs no use among prescription
IARC MONOGRAPHS 108
1973 hypertension only (casecontrol drug used; probably

(n=109) pairs with included some
comparable digitoxin users.
treatment
duration)
Lenfant- Men with breast Hospital chart Breast, 11 Any digitalis 4.1 (1.412.4) Controls Digitalis was the
Pejovic cancer (n = 91) abstracts and adeno- use vs no use matched only one of many
et al. (1990), identified in physician carcinoma by age and therapeutic drugs for
Switzerland, hospital or by interview; digitalis hospital which an association
197086, tumour registries specified was found. Probably
and France, compared with men included some
197588 with colorectal, digitoxin users.
haematolymphatic, or
skin cancers (n = 255)
Ewertz et al. Men with breast Self-reported Breast 20 Never digoxin 1.0 (ref.) Matched Multiple comparisons
(2001), cancer (n = 156) questionnaires Digoxin <5yr 1.8 (0.74.4) for sex, to diverse
Norway, compared with men including Digoxin 5yr 2.0 (0.94.4). age; overall demographic,
Sweden, in population registry prescription-drug analysis health, and drug-
Denmark, (n = 468) use and other adjusted for use variables, but
198791 demographic and BMI association for digoxin
health data appeared to be the
strongest among
drugs; probably
included some
digitoxin users.
P=0.08 for overall
association between
digoxin use and breast
cancer
Reference, Subjects Exposure Organ site Exposed Exposure Relative risk Adjustments Comments
study assessment cases category (95% CI) for potential
location and confounders
period
Ahern et al. Postmenopausal County-based Breast 5241 Never-user 1.0 (ref.) Age, location; Tumour ER status not
(2008), women with breast pharmacy 324 Ever used 1.30 (1.141.48) use of anti- examined. Association
North cancer (n = 5565) registries digoxin inflammatory not greatly changed
Jutland compared with (restricted to drugs, by adjustments;
and Aarhus matched women from casecontrol anticoagulants Suggestion of
Counties, population registry (n pairs with or HRT increased risk with
Denmark, = 55650) comparable longer duration of use.
19912007 treatment May have included
duration) some digitoxin
128 13yr 1.25 (1.031.52) users in early years,
although described as
103 46yr 1.30 (1.051.61)
digoxin users.
93 718 yr 1.39 (1.101.74) Adjusted for age,
county of residence,
and past receipt of
HRT, anticoagulants,
high- and low-dose
aspirin, and NSAIDs.
BMI, body mass index; ER, estrogen receptor; HRT, hormone replacement therapy; NSAIDs; non-steroidal anti-inflammatory drugs; ref., reference; vs, versus; yr, year
395
Digoxin
IARC MONOGRAPHS 108
incidence in people before their first use of digi- examine risk by estrogen-receptor status being
toxin, odds ratios for most cancers were similarly a particular strength. The study did not examine
increased. An analysis of the relationship between the effect of menopausal status; however, most
risk of cancer and serum concentration of digi- women included were postmenopausal (median
toxin did not show a coherent relationship for age, 79 years). Information on other covariates
cancer of the breast. [The Working Group noted was limited. While there are many risk factors
that the national population used as comparison for cancer of the breast, the inability to control
group was external to the study population and for alcohol drinking and obesity was likely to be
may differ in its underlying disease risk or in the of greatest concern.]
quality of cancer ascertainment. Elevated risk Biggar et al. (2013) examined features of
of cancer in the study population before begin- cancer of the breast in a casecase comparison of
ning treatment may be attributable to under- cancers developed in 369 women who were using
lying increases in the frequency of common risk digoxin at the time of diagnosis with 34 085
factors for cancer and for cardiovascular disease cancers in women not using digoxin. Tumours in
requiring digitoxin, rather than the use of digi- users were significantly more likely (P=0.002) to
toxin itself. In addition, estimates of digitoxin be estrogen receptor-positive (85%) than estrogen
dose were based on a single measurement at the receptor-negative (79%), and to have low versus
start of treatment and there was no information high histological grades, features suggesting
about ongoing exposure.] better prognosis. [The prognostic factors for
Biggar et al. (2011) reported a nationwide cancer of the breast in women receiving digoxin
cohort study in Denmark, evaluating incidence and in women receiving estrogen were similar
of cancer of the breast in women prescribed and more favourable, e.g. estrogen receptor-posi-
digoxin. Data were obtained by linking the tive tumours, than in women not receiving treat-
national Danish prescription-drug database ment (IARC, 2012).]
(available since 1995) and the nationwide Danish
cancer registry until 2008. Among 104 648
women using digoxin, 2144 developed cancer
2.2 Cancers of the uterus and ovary
of the breast. Risks associated with current and Cohort study
former use, and duration of current use among
new users only were analysed, with incidence See Table2.3
rate ratios for cancer of the breast adjusted for In a cohort study in Denmark, Biggar et al.
attained age at diagnosis and calendar year. (2012) evaluated the risk of cancer of the uterus.
The relative risk (RR) for current use was 1.39 The methods and data sources were identical
(95% CI, 1.321.46), with higher risk for devel- to those in the study of cancer of the breast
oping estrogen receptor-positive tumours described in Section 2.1.2 (Biggar et al., 2011).
(RR, 1.35; 95% CI, 1.261.45) than estrogen As with cancer of the breast, the incidence of
receptor-negative tumours (RR, 1.20; 95% CI, cancer of the uterus (n = 461 cases in digoxin
1.031.40) among digoxin users. Incidence was users) was increased among current users (RR,
not increased in women who had used digoxin in 1.48; 95% CI, 1.321.65). In addition, this study
the past (SIR, 0.91; 95% CI, 0.831.00). Increased also evaluated cancers of the ovary (n=277) and
incidence was not associated with duration of cervix (n=117) as control cancers, finding no
use, but declined to baseline within 1year after increase in the incidence of either cancer (RR for
use of digoxin had ceased. [This was regarded cancer of the ovary, 1.06; 95% CI, 0.921.22; RR
as a high-quality study, with the capacity to for cancer of the cervix, 1.00; 95% CI, 0.791.25)
396
Table 2.2 Cohort studies on use of digoxin and cancer of the breast
Reference, Subjects Exposure Organ site Exposed Exposure Relative risk Adjustments for potential confounders
location, and assessment cases category (95%CI) Comments
period
Friedman Members of a Pharmacy Lung 48 Digitalis 1.7[1.222.20] Age, sex
(1984), Kaiser private health- database from Colon 35 ever-use 1.5[1.022.04] Main summary for all drug-cancer
Permanente care insurance Health Plan Breast 20 (digoxin, 1.2[0.741.87] relationships reported by Friedman &
Medical Care programme digitoxin, Ury (1980). Updated to 1980: Friedman
Prostate 34 1.4[1.002.01]
Program (USA), (n = 143594) digitalis) (1984). Multiple comparisons.
196980 No association found for other cancers.
Haux People Digitoxin All sites 641 Digitoxin use 1.27(1.181.37) Age, year of birth, sex
et al. (2001), (n= 9271) in plasma Female breast 57 1.25(0.951.62) Incidence compared to population
Trondheim, undergoing measured Prostate 108 1.25(1.031.50) incidence when >30 cases were
Norway, their first in a central expected
Colorectum 127 1.29(1.061.51)
198696 digitoxin laboratory Use based on single assessment of
treatment Lung 63 1.35(1.041.74) digitoxin. A high risk of cancer
Kidney/urinary 59 1.14(0.871.47) diagnosed before digitoxin
Melanoma 61 1.23(0.941.58) measurement (not shown) suggested
high cancer risk preceded use. Expected
Leukaemia/ 53 1.41(1.061.85)
numbers of cancers obtained from
lymphoma (C81C
national registry rates.
85/C88/92)
Breast Digitoxin
concentration
(ng/mL):
<16 1.00(ref.) Doseresponse on the cohort on
1622 1.04(0.591.84) digitoxin users by different levels of
>22 0.90(0.481.67) digitoxin plasma concentration at first
measurement divided in tertiles
Biggar et al. Women Nationwide Breast 46872 Never 1.0 Attained age, calendar-year
(2011), aged 20 yr pharmacy 2144 Ever 1.24(1.181.30) Association found only with current use
Denmark, (n=2011381) registry for 454 Former 0.91(0.831.00) of digoxin and stronger when restricted
19952008 drug exposure to women with ER-positive tumours.
1690 Current 1.39(1.321.46)
Duration results apply to all breast
cancers, regardless of ER status.
397
Digoxin
398
Reference, Subjects Exposure Organ site Exposed Exposure Relative risk Adjustments for potential confounders
location, and assessment cases category (95%CI) Comments
period
Biggar et al. Duration of
(2011), use in new
Denmark, users only
19952008 (mo):
(cont.) 306 012 1.65 (1.471.86)
IARC MONOGRAPHS 108
147 1324 1.31 (1.121.55)

92 2536 1.13 (0.921.38)
265 37+ 1.31 (1.161.48)
ER, estrogen receptor; mo, month; ref., reference; vs, versus; yr, year
Digoxin
among current users. Patterns of risk with dura- high-quality study with robust findings adjusted
tion of digoxin use were not consistent by cancer for an extensive array of covariates. Although
type. For cancer of the uterus, stronger asso- exposure data were self-reported, reports by the
ciations were observed for digoxin use of 012 health professionals were assumed to be of rela-
months (RR, 1.60; 95% CI, 1.232.07) and >37 tively high quality. Cancer outcomes were also
months (RR, 1.91; 95% CI, 1.512.41) among self-reported, but validated by pathology-record
current users, while for cancer of the ovary the review in 95% of cases.]
strongest association was for digoxin use of 012 The association between cancer of the pros-
months among current users (RR, 1.37; 95% CI, tate and ever-use of drugs in the digitalis group
1.011.86) among current users. [The strengths was examined in the cohort study by Friedman
and limitations of this study were the same as & Ury (1980) and Friedman (1984), described in
for the study of cancer of the breast based on the Section 2.1.2. The standardized morbidity ratio
same cohort (Biggar et al., 2011).] was 1.4 [95% CI, 1.002.01; 34 cases].
An increased risk of cancer of the prostate was
also reported in the Norwegian cohort study by
2.3 Cancer of the prostate Haux et al., (2001). The relative risk was 1.25 (95%
Cohort studies CI, 1.031.50). [As noted in Section 2.1.2, relative
risks were elevated for most of the cancers exam-
See Table2.4 ined, leading to doubts about the appropriateness
Platz et al. (2011) examined the association of the comparison group.]
between incidence of cancer of the prostate
and use of digoxin in the USA-based Health
Professionals Follow-up Study, following 47884 2.4 Non-Hodgkin lymphoma
men from 1986 until 2006. Data on use of digoxin
Casecontrol study
were obtained by self-administered question-
naire at baseline and at 2-year intervals during See Table2.5
follow-up. Ever-users of digoxin had lower inci- To determine whether the development of
dence of cancer of the prostate compared with non-Hodgkin lymphoma is associated with
never-users, after adjustment for multiple risk medication use, Bernstein & Ross (1992) reviewed
factors, including race, body mass index, exer- prescription-medication use in 619 cases of
cise, and smoking (RR, 0.83; 95% CI, 0.720.94), non-Hodgkin lymphoma in Los Angeles, USA,
which was not changed by adjustment for between 1979 and 1982, that were matched to
other cardiovascular drugs (cholesterol-lowering 619 age, race, sex, and neighbourhood controls.
agents, aspirin). The inverse association was seen Among 49 medications evaluated (along with
regardless of indication for digoxin use (heart many other health conditions and immuniza-
failure or arrhythmia), present when digoxin tions), the odds ratios for use of digitalis were
was the only cardiac medication used (other 1.55 (95% CI, 0.992.43) for men and women
than aspirin), apparent at all stages of cancer of combined, 2.4 (95% CI, 1.314.38) for women
the prostate, and stronger in current than former and 0.75 (95% CI, 0.361.59) for men. A trend
users. The adjusted risk ratio for cancer of the with duration of use was found in women, but
prostate decreased with duration of use from not in men. [Multiple comparisons were made
0.87 (0.731.04) for those with <5years of use with many drug- and non-drug-related varia-
to 0.54 (0.370.79) for those with 10 years of bles, and the association with digitalis, seen only
use (P for trend <0.001). [This was regarded as a
399
400
Table 2.3 Cohort study on use of digoxin and cancer of the corpus uteri, cervix, and ovary
Reference, Subjects Exposure Organ sites Exposed Exposure Relative risk (95% Adjustments for potential confounders
location, assessment cases categories CI) Comments
and period
Biggar et al. See Table 2.2 Nationwide Corpus uteri 111 Former 1.20 (0.991.45) Attained age, calendar year
(2012), Denmark, and Biggar pharmacy 350 Current 1.48 (1.321.65) Association to digoxin found only for uterine
19952008 et al. (2011) registry Duration cancer and statistically significant only in
of use current users; marginal association for former
(mo): users.
For uterine cancer, increase greatest with
IARC MONOGRAPHS 108
59 012 1.60 (1.232.07)

prolonged use;
26 1324 1.19 (0.811.75) For all, a higher incidence was noted in the
11 2536 0.70 (0.391.27) first year after diagnosis, which could suggest
71 37+ 1.91 (1.512.41) confounding by indication
Ovary 70 Former 0.95 (0.751.21)
207 Current 1.06 (0.921.22)
Duration
of use
(mo):
42 012 1.37 (1.011.86)
20 1324 1.11 (0.711.72)
13 2536 1.01 (0.581.74)
30 37+ 1.02 (0.711.46)
Cervix uteri 36 Former 1.18 (0.851.65)
81 Current 1.00 (0.791.25)
Duration
of use
(mo):
18 012 1.44 (0.912.30)
8 1324 1.10 (0.552.20)
5 2536 0.96 (0.402.31)
8 37+ 0.66 (0.331.32)
mo, month
Table 2.4 Cohort study on use of digoxin and cancer of the prostate
Reference, Subjects Exposure Organ sites Exposed Exposure Relative risk Adjustments for potential
location, assessment cases categories (95% CI) confounders
and period Comments
Platz et al. (2011), Men aged Self-reported Prostate, 4923 Never 1.0 Age, race, calendar year,
Health Professionals 4075 years (n questionnaire data about invasive 243 Ever 0.83(0.720.94) BMI, height, smoking,
Follow-up Study, USA, = 47884) current use of digoxin cancer 175 Current 0.78 (0.670.90) diabetes, diet, exercise,
19852006 vitamin E supplement
Duration of
Cohort analysis undertaken
use (yr):
to assess effects observed in
Never 1.0 vitro (see Section 4).
125 <5 0.87 (0.731.04) Cancer self-report
90 59.9 0.87 (0.701.07) supplemented with death-
28 10 0.54 (0.370.79) certificate data; pathology-
record review: 94.5%
complete.
BMI, body mass index; yr, year
401
Digoxin
402
Table 2.5 Casecontrol study on use of digitalis and non-Hodgkin lymphoma
Reference, Subjects Exposure Organ sites Exposure categories Exposed Relative risk Adjustments for
location, assessment cases (95% CI) potential
and period confounders
Bernstein & Ross Cases, 619 Personal interview Non-Hodgkin No digitalis 35 1.00 Matched on age,
(1992), Controls, 619 and questionnaire lymphoma Digitalis (all) 52 1.55 (0.992.43) sex, race, and
Los Angeles (neighbourhood) including ever-use of Men 12 0.75 (0.361.59) neighbourhood
County (USA), digitalis
Women 40 2.40 (1.314.38)
197982
All (men and women)
IARC MONOGRAPHS 108
No digitalis 1.00
Digitalis 112 mo 23 1.35 (0.992.43)
Digitalis 13 mo 28 1.68 (0.923.08)
P for trend 0.063
Men
No digitalis 1.00
Digitalis 112 mo 7 1.00 (0.352.85)
Digitalis 13 mo 0.56 (0.191.66)
P for trend 0.34
Women
No digitalis 1.00
Digitalis 112 mo 16 1.72 (0.763.91)
Digitalis 13 mo 23 3.05 (1.356.87)
P for trend 0.042
mo, month
Digoxin
in women and not in men, could have been a 4. Mechanistic and Other
chance finding.] Relevant Data
2.5 Other cancer sites 4.1 Absorption, distribution,
See Table2.2 metabolism, and excretion
Elevated relative risks of cancers of the lung
and colorectum were observed in the cohort
4.1.1 Humans
study by Friedman & Ury (1980) and Friedman (a) Absorption and distribution
(1984), and in the cohort study by Haux et al. Digoxin exhibits first-order kinetics (Ehle
(2001) described in Section 2.1.2. The relative risk et al., 2011). In six healthy volunteers (average
of cancer of the lung was 1.7 [95% CI, 1.222.20] age, 20 2.5 years) given a single infusion of
in the former study, and 1.35 (95% CI, 1.041.74) digoxin of 750 g for 20 minutes (Finch et al.,
in the latter. For cancer of the colorectum, the 1984), digoxin had a half-life of 37.212 hours,
relative risks were 1.5 [95% CI, 1.022.04] and an area under the curve (AUC) of concentra-
1.29 (95% CI, 1.061.51) for the same studies, tiontime of 147.778.6ng/mL per hour, a large
respectively. Haux et al. (2001) also reported volume of distribution (311.494.0L) and clear-
an increased risk of leukaemia and lymphoma ance rate of 108.659.1 mL/minute. In a study in
combined (RR. 1.41; 95% CI, 1.061.85). [The four healthy men given 1mg of tritium-labelled
Working Group considered that the study by digoxin by intravenous injection (Marcus et al.,
Haux et al. (2001) may have used an inappro- 1964), the drug disappeared very rapidly from the
priate comparison group, as noted in Section circulation; 3minutes and 1hour after the injec-
2.1.2, and had limited confidence in the results. tion, only 15.9% and 2.8%, of the administered
The elevated relative risk of cancer of the lung dose, respectively, was detected in the blood. The
could be due to an association between smoking onset of pharmacological action, after intrave-
and cardiovascular disease for which digitalis nous administration, is detected within 1530
was prescribed.] minutes, and maximum effect within 14hours
(Ehle et al., 2011).
The distribution of digoxin follows a
3. Cancer in Experimental Animals two-compartment model (Reuning et al., 1973),
comprising plasma and rapidly equilibrating
No data were available to the Working Group. tissues (compartment one [small volume]), and
the more slowly equilibrating tissues (compart-
ment two [large volume]) (Currie et al., 2011).
Equilibrium between compartments is achieved
after a minimum of 6hours, distribution half-life
is 35 minutes, onset of action (oral) approximately
30120 minutes, and time to peak action (oral) is
68hours (Currie et al., 2011), or 26hours, as
reported by Ehle et al. (2011). Digoxin is 2025%
bound to plasma proteins (Ehle et al., 2011).
After oral administration of digoxin, half-
life and time to steady state vary significantly
403
IARC MONOGRAPHS 108
between individuals, and are also dependent compared with those with wildtype (C3435C)
on renal function (Ehle et al., 2011). In healthy alleles (Hoffmeyer et al., 2000).
subjects, the half-life is 1.52days (Currie et al., In eight volunteers, pre-treatment with
2011; Ehle et al., 2011), and steady state is reached rifampicin, an inducer of P-glycoprotein, altered
in 57days (Ehle et al., 2011). In anuric patients, absorption of digoxin. The rifampicin-induced
half-life is prolonged to 3.55days (Currie et al., mean concentration of digoxin in people carrying
2011; Ehle et al., 2011), and steady state is reached the T-allele single-nucleotide polymorphism was
in up to 1520 days (Ehle et al., 2011). The volume higher than that of the wildtype (CC) population
of distribution is 47 L/kg in healthy subjects (Hoffmeyer et al., 2000).
(Ehle et al., 2011), but is decreased in people with In healthy volunteers (with the TT and CC
renal disease and hypothyroidism, and increased genotypes [n=7 in each group]) given multiple
in people with hyperthyroidism (Currie et al., oral doses of digoxin (0.25 mg per day) to achieve
2011). A study of 32 men and 35 women receiving steady-state conditions, a statistically significant
long-term therapy with digoxin (in doses indi- difference (mean, 38%) was found in maximum
vidualized according to body weight), showed no serum concentration of digoxin (Cmax) between
sex-based differences in serum concentration of the two groups [read from Figure: CC, ~1.60 g/L;
digoxin (Lee & Chan 2006). TT, ~2.15 g/L]. This may reflect the importance
Oral bioavailability (F) of digoxin varies of genotype in determining absorption after oral
with formulation, and between individuals. administration of digoxin (Hoffmeyer et al.,
Bioavailability from digoxin capsules, elixirs, 2000).
or tablets are 90%, 80%, and 70%, respectively In 24 healthy Caucasian men who were
(Ehle et al., 2011), and almost 100% from gela- homozygous carriers of the wildtype exon 26
tine capsules (Currie et al., 2011). Bioavailability C3435T (CC), or heterozygous (CT), or homozy-
of digoxin is physiologically controlled by the gous mutant (TT) [n=8 in each group], AUC04h
transmembrane transporter, P-glycoprotein, (P=0.042) and Cmax (P=0.043) differed signif-
which has efflux pump function (Riganti icantly, with higher serum concentrations of
et al., 2011). P-glycoprotein controls bioavail- digoxin in men with the 3435TT genotype than
ability from its location on apical (or luminal) in those with wildtype C3435T (CC). No influ-
membranes of enterocytes of the small intestine, ence on digoxin parameters was detected for
by active extrusion of digoxin, back into the other single-nucleotide polymorphisms (Johne
lumen of gastrointestinal tract. A critical factor et al., 2002).
in intestinal absorption is the rate of apical efflux Genotypes deduced from single-nucleotide
(Riganti et al., 2011). polymorphism 2677G-T (exon 21) and 3435C-T,
substantiated by haplotype analysis, also showed
(i) Studies supporting an effect of MDR1
significant differences in AUC04h and Cmax.
polymorphism
These analyses indicated that haplotype 12
A study in 21 Caucasian individuals given a (2677G/3435T) was associated with high values
single oral dose of digoxin of 0.25 mg showed a of AUC04h and Cmax for orally administered
correlation between polymorphism of the MDR1 digoxin (Johne et al., 2002).
gene [the gene encoding P-glycoprotein, standard In homozygous carriers of TT, kinetic param-
nomenclature, ABCB1] at exon 26 (C3435T) and eters indicated a faster and more complete absorp-
significantly lower levels of duodenal expression tion of digoxin than in carriers of the wildtype.
and function of MDR1. Polymorphic individuals The digoxin plasma time course was evidenced by
had higher plasma concentrations of digoxin a 24% higher Cmax and by a 22% higher AUC04h,
404
Digoxin
considered to result from increased rate (indi- allele (CC), heterozygotes with a mutant T allele
cated by the steeper ascending phase of the curve (C3435T) (CT), and homozygotes for the mutant
in TT individuals) and extent of absorption (and allele (TT), values for AUC04h ( standard
not primarily of distribution) (Johne et al., 2002). deviation) were 4.11 0.57, 3.20 0.49, and
High doses of digoxin are thought to satu- 3.270.58 ng/hour per mL, respectively. There
rate P-glycoprotein transport, triggering addi- was a significant difference between CC and CT
tional mechanisms. Thus, it is likely that at low or TT.
doses, the pharmacokinetics of digoxin will be In a study in 39 Caucasian patients with
influenced by P-glycoprotein transport only, and congestive heart failure given digoxin at 0.25 mg
thus would be more greatly perturbed by genetic per day for at least 7days to reach steady state,
differences in P-glycoprotein activity (Johne Kurzawski et al. (2007) evaluated the effects of
et al., 2002). MDR1 gene polymorphism on serum concentra-
A study of elderly patients in the Netherlands tions of digoxin, and in 24 patients, the effects of
(n=195; mean age, 79.4years) who were taking coadministration of digoxin with P-glycoprotein
digoxin regularly also showed that the common inhibitors. Significantly higher (approximately
MDR1 variants, 1236C-T, 2677G-T, and 3435C-T 1.5-fold) (P<0.002) minimum serum concentra-
and the associated TTT haplotype were corre- tions of digoxin at steady state (Cmin ss) were shown
lated with higher serum concentrations of in patients given inhibitors of P-glycoprotein
digoxin (Aarnoudse et al., 2008). (0.8680.348ng/mL), compared with those not
To understand the relative contribution of given inhibitors (0.5240.281ng/mL); however,
environmental and genetic factors to the phar- in contrast to other studies, no association was
macokinetic variability of oral and intravenous found between 3435C > T and 2677G > A,T
digoxin, Birkenfeld et al. (2009) conducted a pilot MDR1 single-nucleotide polymorphisms and
study in 11 pairs of monozygotic twins (whose steady-state serum concentrations of digoxin
genes are almost identical), and 4 pairs of dizy- (Kurzawski et al., 2007).
gotic twins (control). Measures of peak plasma A higher (1mg) single oral dose of digoxin,
concentration and Tmax of digoxin, and calcu- without drug pre-treatment, in 50 healthy white
lated AUC, bioavailability, and renal clearance, men (aged 1840 years) showed no differences in
after oral or intravenous administration, demon- the AUC04h, Cmax, or tmax (as indices of digoxin
strated strong correlation between monozygotic absorption) among the genotype groups tested
twins, findings explained largely by inheritance (Gerloff et al., 2002). In contrast to previous reports
of P-glycoprotein function (Birkenfeld et al., (Hoffmeyer et al., 2000), no differences were seen
2009). between homozygous carriers of the C and T allele
in exon 26 3435 (AUC04h, 9.24 and 9.38mg/hour;
(ii) Studies not supporting an effect of MDR1
Cmax, 4.73 and 3.81g/L; tmax, 0.83 and 1.14 hours,
polymorphism
respectively). The MDR1 single-nucleotide poly-
Other studies have not shown an association morphisms studied, including that in exon 26,
between polymorphism in the MDR1 gene and did not affect the absorption of a single oral dose
increased plasma concentrations of digoxin. of 1 mg of digoxin, and it was suggested that the
A study in 114 healthy Japanese people given a higher dose (1mg) of digoxin may have caused
single oral dose of digoxin of 0.25 mg (Sakaeda saturation of the transport capacity of intestinal
et al., 2001) showed the serum concentration to P-glycoprotein. The pharmacokinetics of digoxin
be lower in those with a mutant allele (C3435T) showed substantial variation within each geno-
at exon 26 of the MDR1 gene. For the wildtype typic group, indicating that factors additional to
405
IARC MONOGRAPHS 108
P-glycoprotein may influence the absorption of 3-digoxigenin and its mono- and bis-digitoxo-
digoxin (Gerloff et al., 2002). sides, and 3-keto and 3(epi)-digoxigenin.
It is likely that passive diffusion (Gerloff This metabolic route comprised initial
et al., 2002) or other transporters (Johne et al., hydrolysis to 3-digoxigenin with release of
2002), in addition to P-glycoprotein, contribute sugars in the stomach or liver, followed rapidly by
to variations in the pharmacokinetics of digoxin. oxidation to 3-keto-digoxigenin, epimerization
Digoxin is a substrate for OATP8 (a member of to 3(epi)-digoxigenin and finally glucuronide
the organic anion-transporting polypeptide conjugation to polar species, 3-epi-glucuronide
group), for which genetic variants have been and 3-epi-sulfate. Results also indicated that
identified (Johne et al., 2002), the effects of conjugation of the mono-digitoxoside may occur,
which, have not yet been elucidated. In addi- with steroid-ring hydroxylation, producing two
tion, genetic variation in regulatory proteins, isomers. In individuals demonstrating extensive
for example, the pregnane X receptor, involved metabolism, the lactone ring may be opened
in regulation of P-glycoprotein, may also affect (possibly by a lactonase), forming a highly polar
digoxin disposition (Birkenfeld et al., 2009). The metabolite, or reduced, forming dihydro-metab-
absorption of digoxin may also be influenced by olites (Gault et al., 1984).
environmental factors (such as diet) by induction In studies using suspensions of freshly
or inhibition of P-glycoprotein activity (Johne isolated human hepatocytes in vitro, metabolism
et al., 2002; Gerloff et al., 2002), or by genetic of [3H]digoxin-12 has been shown to be very
variants governing its distribution and elimina- low (Lacarelle et al., 1991); after a 2-hour incuba-
tion (Gerloff et al., 2002). tion, extracellular radiolabel represented largely
unchanged digoxin (up to 93%), with a minor (5%
(b) Metabolism of the total extracellular radiolabel) unidentified
Gault et al. (1984) demonstrated a major polar metabolite. Similar results were obtained
metabolic sequence of digoxin hydrolysis, oxida- over a 24-hour exposure time in cultured human
tion, and conjugation, leading to polar end-me- hepatocytes, and also in human liver microsomal
tabolites. In this study, 10 patients with end-stage fractions, indicating that cleavage of digoxin
renal failure (who were dependent on dialysis), sugars is not dependent on the cytochrome P450
and 5 patients with comparatively normal renal (CYP) system that requires reduced nicotina-
function were given digoxin (as an oral dose of mide adenine dinucleotide phosphate (NADPH)
150 Ci of [3H]digoxin-12) and the metabolites (Lacarelle et al., 1991; also see Fig.4.1).
were analysed by high-performance liquid chro- Digoxigenin mono-digitoxoside was exten-
matography (HPLC). Of these patients, 13 were sively metabolized by human cultured hepato-
receiving maintenance therapy with digoxin and cytes to a single, more polar metabolite, which
were at steady state. The extent and time course of was subsequently completely hydrolysed by
metabolism of digoxin varied between subjects, -D-glucuronidase, and thus identified as the
but variation was not significant between the two glucuronide of digoxigenin mono-digitoxoside.
groups with different renal function. For all 15 The extent of glucuronidation analysed in human
patients, at 6 hours after drug administration, liver microsomal fractions prepared from 13
26% (range, 776%) of the radiolabel was in the different subjects was shown to vary among indi-
form of polar metabolites (quantitatively the viduals by a factor of 3 (Lacarelle et al., 1991).
most abundant metabolites), and 60% (range, Digoxigenin was also extensively biotrans-
1188%) was unchanged digoxin. Metabolites formed by cultured human hepatocytes. HPLC
usually found albeit in small amounts were peaks were shown for one or more glucuronides,
406
Digoxin
3-epi-digoxigenin, unchanged digoxigenin, gastrointestinal tract. These data also demon-

and possibly for unidentified metabolites. The strated a non-renal mechanism of elimination of
intracellular concentration of 3-epi-digoxigenin digoxin, entailing direct secretion into the small
decreased, due to conversion to polar compounds, intestine from the systemic circulation, which
which effluxed from the cells as formed. In had greater importance than elimination via bile
human liver microsomes, no metabolites were (Drescher et al., 2003).
observed in the absence of cofactor (NADPH or The organic anion transporter in human
uridine 5-diphospho-glucuronic acid, UDPGA); kidney (OATP4C1) may have an initial role in
however, with NADPH present, pre-digoxi- the transport of digoxin to the kidney. These
genin was detected. Formation of pre-digoxi- transporters have been isolated, and shown by
genin therefore appeared to be CYP-dependent, immunohistochemical analysis to be localized at
with a large variability observed among individ- the basolateral membrane of the proximal tubule
uals (Lacarelle et al., 1991; also see Fig.4.1). cell in the kidney. Both human OATP4C1 and rat
In contrast, formation of 3-epi-digoxigenin OATP4C1 transport digoxin in a sodium-inde-
did not depend on microsomal enzymes; it was pendent manner (Mikkaichi et al., 2004).
only observed after incubation of digoxigenin The role of OATPs in the disposition of digoxin
with hepatocytes, and not with microsomes. has not been clearly defined. Data from various
In the presence of both NADPH and UDPGA, in-vitro systems have indicated that digoxin is
only small quantities of polar compounds were not a substrate for human OATP1A2, OATP1B1,
observed. These findings confirmed that 3-epi-di- OATP1B3, or OATP2B1, although OATP4C1
goxigenin is formed before synthesis of polar may facilitate active uptake of digoxin into
compounds. Thus, the main metabolic route for human kidney and liver. Digoxin is a substrate
digoxigenin in vitro is the formation of 3-epi-di- for a sodium-dependent transporter, shown to
goxigenin, which is conjugated to a glucuronide be endogenously expressed in a human kidney
(Lacarelle et al., 1991; also see Fig.4.1). cell line (HEK29), and may, by its location in
proximal tubular cells, partially facilitate renal
(c) Elimination clearance of digoxin (Taub et al., 2011).
Recovery of digoxin in the urine was reported
as 7085% (Currie et al., 2011) and 5070% (Ehle (d) Interactions
et al., 2011). Drug recovery in the faeces was, on The bioavailability of digoxin is affected by
average, 14.8% of the administered dose, of which concurrent administration of many drugs which
14% comprised metabolic products (Marcus compete for binding to P-glycoprotein. Thus,
et al., 1964). digoxin auto-regulates its absorption. Many
In a study of the mechanisms of intestinal lipophilic P-glycoprotein-inducing drugs also
and biliary transport of digoxin, eight healthy promote CYP3A activity, and so a complex,
men (aged 2137 years), were given segmental and poorly understood, network of interactions
intestinal perfusion of a P-glycoprotein inhibitor between drugs or endogenous metabolites may
(quinidine) or inducer (rifampin), with intrave- affect transport and metabolic inactivation of
nous administration of digoxin (1 mg). Results digoxin (Riganti et al., 2011).
showed that intestinal P-glycoprotein mediates
the elimination of intravenously administered
digoxin from the systemic circulation into the
gut lumen, as well as the control of absorp-
tion of orally administered digoxin from the
407
IARC MONOGRAPHS 108
Fig.4.1 Structure of digoxin and proposed metabolic pathways

O O
OH
CH 3
H3C H
H3C H3C H3C
H OH
O O O
HO O O O
H
HO HO HO
Digitoxoses Digoxigenin
Reduction of
Digoxin (DG3) Dihydrodigoxin
the lactone
Stepwise cleavage of
the digitoxoses
Dihydrodigoxigenin
Digoxigenin
bis-digitoxoside
(DG2) (two sugars)
Digoxigenin
mono-digitoxoside
(DG1) (one sugar)
?
Digoxigenin (DGO)
(no sugar)
Epidigoxigenin Conjugates
(glucuronides, sulfates)
Unknown polar compounds
From Lacarelle et al. (1991), Copyright 1991, John Wiley and Sons
408
Digoxin
4.1.2 Experimental systems P-glycoproteins, are involved in the absorption

of digoxin (Yao & Chiou, 2006).
(a) Absorption An additional non-MDR1 component may
The pharmacokinetics of digoxin was studied contribute to active secretion of digoxin back
in male Sprague-Dawley rats given an intrave- into the lumen, to limit its intestinal absorption.
nous bolus dose at 1mg/kg bw. Plasma and urine In support of this, MDR1-transfected Madin-
samples were analysed by thin-layer chromatography Darby canine kidney (MDCKII) cell monolayers
to separate digoxin and its metabolites. Digoxin showed reduced secretion of digoxin by the
concentrations were described as a two-com- MDR1 inhibitor cyclosporin A, but not by the
partment model. Parent drug was rapidly elimi- MDR1 inhibitor MK-571 (Lowes et al., 2003).
nated from the plasma, with half-life of 2.5hours,
a volume of distribution of 3.6L/kg, and a total (b) Metabolism
body clearance of 5.77 mL/minute. Bile-duct A proposed metabolic pathway for digoxin is
ligation produced comparable pharmacokinetic shown in Fig.4.1 (Lacarelle et al., 1991).
parameters (with the exception of the total body In humans, more than 73% of an intravenous
clearance, 5.18 mL/minute). In rats with bilateral dose is excreted unchanged via the kidneys. In
ureter ligation, the plasma half-life of digoxin contrast, the rat metabolizes approximately 60%
was increased to 4 hours (Harrison & Gibaldi, of an intraperitoneal dose, and approximately
1976). 30% is excreted via biliary and urinary routes
The function of P-glycoprotein in vivo has (Harrison & Gibaldi, 1976).
been investigated pharmacokinetically, using Metabolism of digoxin follows a similar
mdr1a (/) mice [Abcb1a (/)] (Schinkel et al., metabolic pathway in humans and rats, i.e. step-
1995; Mayer et al., 1996; Kawahara et al., 1999). wise cleavage of the sugar residues to form the
These mice show no major pathology, but their digoxigenin bis- and mono-digitoxoside and the
intestinal epithelium and brain endothelial cells aglycone digoxigenin before conjugation and
have no detectable P-glycoprotein (Schinkel et al. elimination, but the rate is faster in rats (Harrison
1995). Schinkel et al. (1995) demonstrated that & Gibaldi, 1976).
concentrations of [3H]digoxin in plasma and The three sequential steps of oxidative metab-
most tissues were twofold, and in brain were olism of digoxin (to digoxigenin bis-digitoxoside,
35-fold, in mdr1a (/) mice given [3H]digoxin digoxigenin mono-digitoxoside, and digoxigenin)
intravenously compared with mdr1a (+/+) mice. were studied in rat liver microsomes (Salphati &
Similarly, Kawahara et al. (1999) reported that Benet, 1999). Inhibition of the CYP3A subfamily
digoxin accumulation in the brain was 68-fold with ketoconazole or triacetyloleandomycin,
higher. Mayer et al. (1996) further demonstrated or with antibodies specific to rat CYP3A2,
that the brain concentrations of [3H]digoxin affected oxidative metabolism; the formation of
continued to increase over 3 days after injec- digoxigenin bis-digitoxoside and digoxigenin
tion in mdr1a (/) mice, resulting in a 200-fold mono-digitoxoside decreased by up to 90%,
higher concentration than in mdr1a (+/+) mice. and the rate of oxidation of digoxin and digoxi-
However, Kawahara et al. (1999) reported that genin bis-digitoxoside was decreased by up to
disruption of the mdr1a gene did not to change 85%, respectively. These oxidation reactions
plasma-protein binding or the blood-to-plasma were unaffected by chemical or immunological
partition coefficient. inhibition of CYP2E1, CYP2C or CYP1A2. The
Inhibition studies in vitro have shown subsequent metabolic step, i.e. oxidation of digox-
that anionic transporters, in addition to igenin mono-digitoxoside, was not inhibited
409
IARC MONOGRAPHS 108
by triacetyloleandomycin or by antibodies to excretion in mdr1a (/) mice given [3H]digoxin

CYP3A2, CYP2C11, CYP2E1, CYP2B1/2B2 or (0.2 mg/kg bw) as a single intravenous or oral
CYP1A2, but was however reduced (by > 80%) bolus, i.e. lower faecal elimination of [3H]digoxin.
by inhibitors of human CYP3A. In summary, This was due to reduced drug excretion via intes-
these results indicated that CYP3A, most likely tinal epithelium, since biliary excretion was not
CYP3A2, is the primary enzyme responsible for decreased in mdr1a (/) mice, and suggested
metabolism of digoxin and digoxigenin bis-digi- that other transporters could be involved in the
toxoside in rat liver microsomes, but the enzyme biliary excretion of digoxin. Indeed, the capacity
that metabolizes digoxigenin mono-digitoxoside for renal excretion remained substantial, and
remains to be identified (Salphati & Benet, 1999). cumulative urinary excretion of digoxin in
mdr1a (/) mice was greater than in wildtype
(c) Elimination (+/+) mice. Thus, intestinal P-glycoprotein acts
Digoxin is eliminated primarily via the by directly excreting digoxin into the intestinal
kidney through glomerular filtration and tubular lumen, and also limiting the rate of its re-up-
secretion. P-glycoprotein has a role in the elimi- take from the intestine by biliary excretion, thus
nation of digoxin. Studies in vitro have demon- directing faecal excretion (Mayer et al., 1996).
strated that mouse mdr1a and human MDR1 [P-glycoprotein seems to have important roles in
P-glycoprotein actively transport digoxin across elimination of digoxin from the systemic circula-
a polarized kidney epithelial cell layer (Schinkel tion, and also in decreasing intestinal re-uptake
et al., 1995). Furthermore, experiments in vivo of digoxin after biliary excretion.]
showed that mdr1a (/) mice eliminated [3H]
digoxin-12 more slowly (Schinkel et al., 1995). 4.2 Genetic and related effects
The total body clearance was lower in mdr1a (/)
mice than in the wildtype (+/+) mice; however, No data were available to the Working Group.
disruption of the mdr1a gene did not change the
contributions of renal and bile clearances to total
clearance (Kawahara et al., 1999). 4.3 Other mechanistic data relevant
Digoxin is partly excreted via the biliary to carcinogenicity
system. In male Sprague-Dawley rats, total body
4.3.1 Effects on cell physiology
clearance values for digoxin were 10% lower in
rats with bile-duct ligation, and were reduced The physiological action of digoxin involves
by a further 30% by bilateral ureter ligation. binding to and inhibition of the -subunit of the
The approximately 60% of total body clearance Na+/K+ ATPase pump on the myocyte plasma
unaffected by ligations of bile duct or ureter were membrane. This causes an increase in intracel-
considered due to biotransformation of digoxin. lular concentrations of sodium and calcium ions.
A main excretory route for digoxigenin bis-dig- Digoxin shares some structural homology with
itoxoside was shown to be biliary as indicated steroid hormones, suggesting functional similar-
by high levels of this metabolite in plasma and ities (Schussheim & Schussheim, 1998; Newman
urine of rats with ligated bile ducts (Harrison & et al., 2008). There is evidence that digitoxin at
Gibaldi, 1976). concentrations of 0.52.0106 M competes with
Intestinal P-glycoprotein in mice has been estrogen for the estrogen cytosolic receptor in the
shown to contribute to excretion of [3H]digoxin rat uterus; however, no evidence for competition
via the gastrointestinal epithelium. Mayer by digoxin was obtained (Rifka et al., 1976; Rifka
et al. (1996) demonstrated a shift in balance of et al. 1978).
410
Digoxin
Other intriguing evidence for digoxin in part, to similar increases in P-glycoprotein

includes a case report of gynaecomastia (Aiman expression (Pinto et al., 2005).
et al., 2009), an increased relative risk of uraemic Evans et al., (1990) showed that age affects the
cancer in digoxin users (RR, 1.48; 45% CI: clearance of digoxin in rats. In male Fischer 344
1.321.65; n = 350) (Biggar, 2012), and lower rats (age, 4, 14, or 25 months) given [3H]digoxin
relative risks of cancer of the prostate (RR, 0.76; and unlabelled digoxin at a dose of 1mg/kg bw
95% CI, 0.610.95) among regular users versus as an intravenous bolus dose, total body clear-
non-users (Platz et al., 2011). ance was 14.2, 12.1, and 7.5mL/minute per kg,
respectively, indicating a significant decrease in
4.3.2 Effects on cell function clearance (P < 0.05). No difference was seen in
the terminal elimination half-life (2.0, 2.3, and
Digoxin reduces synthesis of the TP53 2.5hours respectively) or steady-state volume of
protein in human cancer cell lines; this appears distribution (1.51, 1.49, and 1.27 L/kg, respec-
to be triggered by activation of Src/mitogen-acti- tively) in rats aged 4, 14, and 25 months. Serum
vated protein kinase signalling as a consequence protein binding did not change with age; the
of inhibition of the Na+/K+ ATPase pump (Wang average percentage of unbound digoxin for all
et al., 2009). Digoxin also inhibits the action of rats was 61.35.3% (meanstandard deviation;
cellular DNA topoisomerases in MCF-7 cells n=15) (Evans et al., 1990).
(Bielawski et al., 2006), and inhibits synthesis of
hypoxia-inducible factor 1 (HIF-1) in human
4.4.2 Effects of renal failure on elimination
Hep3B-c1 hepatoblastoma cells (Zhang et al.,
2008). Digoxin may inhibit synthesis of steroids Tsujimoto et al. (2008) showed that, in contrast
(Kau et al., 2005). to normal serum, 10% uraemic serum inhibited
the hepatic uptake of digoxin by human isolated
hepatocytes (by 23%) and by rat hepatocytes (by
4.4 Susceptibility 50%). It was further shown that the uraemic toxins
4.4.1 Effects of age on elimination 3-carboxy-4-methyl-5-propyl-2-furanpropanoic
acid (CMPF), p-cresol, (both at 400mM, which
Since young children require higher doses is within the plasma concentration range for
of digoxin per kilogram of body weight than patients with renal failure) and hippuric acid (at
adults to achieve pharmacological effects, 3000 M) significantly inhibited the uptake of
there has been interest in whether expression digoxin. CMPF and p-cresol inhibited the uptake
of P-glycoprotein is age-dependent. Pinto et al. of digoxin into rat hepatocytes by 27% and 23%,
(2005) have studied mdr1a and mdr1b and the respectively, and into human hepatocytes by
clearance rates of digoxin (dose, 7g/kg bw) in 23% and 28%, respectively. These toxins were,
FVB mice of different ages (at birth, and age 7, 14, however, not wholly responsible for inhibition of
21, 28 or 45 days). At birth and day 7, gene expres- uptake. Indeed, 10% uraemic serum from patients
sion of mdr1a and mdr1b was very low, but mdr1b contained these toxins at concentrations (CMPF,
levels were significantly higher at day 21 than at 37.6 mM; hippuric acid, 26.8 mM; and p-cresol,
days 14 or 28. Digoxin clearance rates correlated 19.5 mM) that may not have been sufficient to
significantly with expression of P-glycoprotein, inhibit the uptake of digoxin. Additionally, the
showing highest clearance values at day 21. It was mechanism of inhibition of these toxins was
concluded that increases in digoxin clearance competitive, while the inhibition shown by 10%
rates after weaning may be attributed, at least uraemic serum was non-competitive. Thus, the
411
IARC MONOGRAPHS 108
inhibitory effects of 10% uraemic serum cannot Digitoxin, another glycoside isolated from
be fully explained by the three major uraemic D. purpurea, is used for the same indications as
toxins studied (Tsujimoto et al., 2008). digoxin in certain countries; it is also found as an
impurity in preparations of digoxin.
In most countries, use of digitalis would in
4.5 Mechanistic considerations practice almost always correspond to digoxin,
The increase in the incidence of cancers of unless digitoxin were specified.
the breast and uterus after long-term treatment Specifications for digitalis glycosides are
with digoxin (Biggar, 2012), and the observed provided in several international and national
estrogen-like side-effects of digoxin and digi- pharmacopoeia. In some countries, digoxin
toxin (Rifka et al., 1976, 1978; Schussheim & has been classified as a hazard to water, an
Schussheim, 1998), suggested that digoxin and environmental hazard, or as an extremely
digitoxin act via estrogen-signalling pathways hazardous substance.
to increase cell proliferation in the mammary
gland, potentially contributing to tumour devel- 5.2 Human carcinogenicity data
opment. However, mechanistic evidence was
limited to a demonstration that digitoxin inhib- Studies in humans have assessed the risk of
ited the binding of estradiol to specific, saturable cancer in patients who may have used digoxin,
binding sites in the rat uterine cytosol. Mammary digitoxin, or digitalis drugs as a group. The prin-
epithelial cells contain several estrogen-binding cipal cancer of interest is cancer of the breast.
proteins, including estrogen receptors (ER and Although risk of some other cancers has been
ER) and estrogen-related receptors (ERR and found to be increased, the literature on other
ERR), and the signalling pathways linking cancers was insufficient to establish patterns of
receptor activation to cellular proliferation are increased risk.
complex (Gibson & Saunders, 2012). The molec-
ular targets associated with the carcinogenic 5.2.1 Cancer of the breast
properties of digoxin and digitoxin have not yet
been defined. Information about the association of cancer
of the breast with use of digoxin and digitoxin
is available from four casecontrol studies
(including two studies in men) conducted in four
5. Summary of Data Reported
Nordic countries, France, and Switzerland, and a
nationwide cohort study of women in Denmark,
5.1 Exposure data and other cohort studies in the USA and Norway.
Statistically significant increases in the occur-
Digoxin is a glycoside isolated from Digitalis
rence of cancer of the breast in users of digoxin
lanata and is used in the treatment of chronic
were seen in three casecontrol studies; in one
heart failure and irregular heart rhythm. While
study in women, the odds ratio was 1.3, while
use may have declined over the past 30 years,
odds ratios were two- and fourfold in the two
digoxin is still frequently prescribed. Global sales
studies in men. The largest study, which included
of digoxin were US$ 142 million in 2012, with
all women using digitalis in Denmark, reported
33% occurring in the USA. Other countries with
an increased risk for current users (hazard ratio,
appreciable use included Japan, Canada, and the
1.39). The positive associations with exposure
United Kingdom.
to digoxin in this study were due to increased
412
Digoxin
risk in current users only: there was no associ- In a casecontrol study from southern
ation in former users and the number of new California, USA, a positive association was
tumours declined after discontinuing drug use. observed with non-Hodgkin lymphoma in
Doseresponse effects were difficult to examine women, but not in men.
because of the narrow dose range, and trends
in risk with duration of exposure were gener- 5.2.3 Synthesis
ally not observed. In a casecase comparison
among a subset of the same population, tumours Statistically significant associations of cancer
occurring in digitalis users were reported to have of the breast with use of digoxin were observed
more favourable prognostic features (estrogen consistently in women and men, across different
receptor-positive) than in non-users. Data on geographical regions, and with different study
the association of cancer of the breast with use designs. Cancer of the breast is rare in men and
of digitoxin were available from one cohort study strengthens the validity of association observed
in women in Denmark, which reported a positive for cancer of the breast in women. The record-
association (relative risk, 1.39). These studies had linkage studies that provided key evidence were
limited ability to account for other risk factors not able to adjust for many of the recognized risk
for cancer of the breast, with obesity and alcohol factors for cancer of the breast, notably obesity
drinking being of greatest concern. and alcohol drinking, although there was no
reason to believe these would be associated
with use of digoxin. Although clear effects with
5.2.2 Other cancer sites
duration and dose were not observed, a decline
Increases in the incidence of cancer of the in the detection of new tumours after cessation
uterus in current users of digoxin were found in of exposure was seen in the largest study from
one cohort study in Denmark. The same study Denmark, consistent with a possible promoting
found no increase in risk of cancers of the cervix effect of digoxin. The association was specific to
and ovary. The risk of cancer of the prostate, estrogen receptor-positive tumours of the breast
another cancer that is influenced by hormones, in the same study.
was reduced in one high-quality cohort study
from the USA, but increased in two others
(one study with methodological weaknesses
from Norway, and the other a very large data- No data were available to the Working Group.
base-screening programme from a health plan
in northern California, USA). The increased
risk of cancer of the uterus, and decreased risk 5.4 Mechanistic and other relevant
of cancer of the prostate, is also consistent with data
a hormone-related mechanism, adding to the
Oral bioavailability of digoxin is generally
plausibility of the epidemiological findings.
high, but varies due to interindividual genetic
Excess risks of cancers of the lung and
differences in expression of the efflux pump,
colorectum were also observed in the cohort
P-glycoprotein.
studies in Norway and northern California.
The metabolism of digoxin in rats and humans
The cohort study in Norway reported a posi-
involves stepwise hydrolytic cleavage of the digi-
tive association with leukaemia and lymphoma
toxoses to form digoxigenin bis- and mono-dig-
combined.
itoxosides and the aglycone digoxigenin before
conjugation and renal elimination.
413
IARC MONOGRAPHS 108
No data were available on genetic effects of 6.3 Overall evaluation

digoxin or its metabolites.
Digoxin has structural homology with Digoxin is possibly carcinogenic to humans
steroid hormones, suggesting functional simi- (Group 2B).
larities. The structurally related glycoside digi- The Working Group recognized a possible
toxin competes with estrogen for the rat uterine association between digoxin and an increased
estrogen cytosolic receptor; however, no evidence incidence of endocrine-related human cancers.
for competition by digoxin was found. However, the evidence that digoxin and digi-
Digoxin reduces synthesis of the TP53 protein toxin act through an estrogen-receptor mediated
in human cancer cells, inhibits cellular DNA mechanism was limited.
topoisomerases, inhibits the synthesis of hypoxia- Favouring a Group 2A classification, the
inducible factor 1, and may inhibit synthesis of epidemiological data associating increased risk
steroids. of cancer of the breast with use of digoxin were
The possible association between use of compelling. Consistent with an endocrine-medi-
digoxin and an increased incidence of endo- ated mechanism, the increase in risk was largely
crine-related human cancers (primarily breast) for estrogen receptor-positive tumours; further,
suggests a mechanism that is estrogen recep- risk of uterus cancer was increased and cancer
tor-mediated. However, evidence that digoxin of the prostate was decreased. The evidence in
and digitoxin act through estrogen-signalling humans favoured a promoter effect that is seen
pathways was limited to a demonstration that only in current users.
digitoxin inhibited the binding of estradiol to Favouring a Group 2B classification, not all
specific, saturable binding sites in rat uterine potential confounders were eliminated in the
cytosol. The molecular targets associated with epidemiological studies, in particular, obesity.
the carcinogenic properties of digoxin and digi- In addition, there were no available data from
toxin have not yet been identified. studies in experimental animals, and no known
molecular mechanism by which digoxin might
be a carcinogen. The weak evidence supporting
6. Evaluation an endocrine-mediated mechanism was noted as
a problem.
There is limited evidence in humans for the References
carcinogenicity of digoxin. A positive association
has been observed between use of digoxin and Aarnoudse AJ, Dieleman JP, Visser LE, Arp PP, van
der Heiden IP, van Schaik RH et al.(2008). Common
cancer of the breast. ATP-binding cassette B1 variants are associ-
ated with increased digoxin serum concentration.
Pharmacogenet Genomics, 18(4):299305. doi:10.1097/
6.2 Cancer in experimental animals FPC.0b013e3282f70458 PMID:18334914
Ahern TP, Lash TL, Srensen HT, Pedersen L (2008).
There is inadequate evidence in experimental Digoxin treatment is associated with an increased inci-
animals for the carcinogenicity of digoxin. dence of breast cancer: a population-based case-con-
trol study. Breast Cancer Res, 10(6):R102 doi:10.1186/
bcr2205 PMID:19055760
Aiman U, Haseeen MA, Rahman SZ (2009).
Gynecomastia: An ADR due to drug interaction. Indian
414
Digoxin
J Pharmacol, 41(6):2867. doi:10.4103/0253-7613.59929 ChemicalBook (2013). ChemicalBook-Chemical Search

PMID:20407562 Engine. Available from: http://www.chemicalbook.
Albrecht HP, Geiss KH (2000). Cardiac glycosides com/, accessed 10 February 2015.
and synthetic cardiotonic drugs. In: Ullmanns Cheng JW, Rybak I (2010). Use of digoxin for heart
Encyclopedia of Industrial Chemistry. Weinheim, failure and atrial fibrillation in elderly patients. Am
Germany: Wiley-VCH Verlag GmbH & Co. KGaA; J Geriatr Pharmacother, 8(5):41927. doi:10.1016/j.
pp. 118. http://dx.doi.org/10.1002/14356007.a05_271 amjopharm.2010.10.001 PMID:21335295
doi:10.1002/14356007.a05_271 Currie GM, Wheat JM, Kiat H (2011).
Aromaa A, Hakama M, Hakulinen T, Saxn E, Teppo Pharmacokinetic considerations for digoxin in
L, Id lan-Heikkil J (1976). Breast cancer and use older people. Open Cardiovasc Med J, 5(1):1305.
of rauwolfia and other antihypertensive agents in doi:10.2174/1874192401105010130 PMID:21769303
hypertensive patients: a nationwide case-control study Dickstein K, Cohen-Solal A, Filippatos G, McMurray JJ,
in Finland. Int J Cancer, 18(6):72738. doi:10.1002/ Ponikowski P, Poole-Wilson PA etal.; ESC Committee
ijc.2910180603 PMID:992904 for Practice Guidelines (CPG)(2008). ESC Guidelines
Banerjee D, Stafford RS (2010). Lack of improvement in for the diagnosis and treatment of acute and chronic
outpatient management of congestive heart failure in heart failure 2008: the Task Force for the Diagnosis and
the United States. Arch Intern Med, 170(15):1399400. Treatment of Acute and Chronic Heart Failure 2008
doi:10.1001/archinternmed.2010.270 PMID:20696970 of the European Society of Cardiology. Developed in
Bernstein L, Ross RK (1992). Prior medication use and collaboration with the Heart Failure Association of
health history as risk factors for non-Hodgkins the ESC (HFA) and endorsed by the European Society
lymphoma: preliminary results from a case-control of Intensive Care Medicine (ESICM). Eur Heart
study in Los Angeles County. Cancer Res, 52(19):Suppl: J, 29(19):2388442. doi:10.1093/eurheartj/ehn309
5510s5s. PMID:1394165 PMID:18799522
Bielawski K, Winnicka K, Bielawska A (2006). Inhibition Digitalis Investigation Group(1997). The effect of digoxin
of DNA topoisomerases I and II, and growth inhibi- on mortality and morbidity in patients with heart
tion of breast cancer MCF-7 cells by ouabain, digoxin failure. N Engl J Med, 336(8):52533. doi:10.1056/
and proscillaridin A. Biol Pharm Bull, 29(7):14937. NEJM199702203360801 PMID:9036306
doi:10.1248/bpb.29.1493 PMID:16819197 Drescher S, Glaeser H, Mrdter T, Hitzl M, Eichelbaum
Biggar RJ (2012). Molecular pathways: digoxin use and M, Fromm MF (2003). P-glycoprotein-mediated intes-
estrogen-sensitive cancersrisks and possible ther- tinal and biliary digoxin transport in humans. Clin
apeutic implications. Clin Cancer Res, 18(8):21337. Pharmacol Ther, 73(3):22331. doi:10.1067/mcp.2003.27
doi:10.1158/1078-0432.CCR-11-1389 PMID:22368159 PMID:12621387
Biggar RJ, Andersen EW, Kroman N, Wohlfahrt J, Melbye DrugBank (2013). DrugBank: Open Data Drug &
M (2013). Breast cancer in women using digoxin: Drug Target Database. Available from: http://www.
tumor characteristics and relapse risk. Breast Cancer drugbank.ca/, accessed 10 February 2015.
Res, 15(1):R13 doi:10.1186/bcr3386 PMID:23421975 Ehle M, Patel C, Giugliano RP (2011). Digoxin: clinical
Biggar RJ, Wohlfahrt J, Melbye M (2012). Digoxin use and highlights: a review of digoxin and its use in contem-
the risk of cancers of the corpus uteri, ovary and cervix. porary medicine. Crit Pathw Cardiol, 10(2):938.
Int J Cancer, 131(3):71621. doi:10.1002/ijc.26424 doi:10.1097/HPC.0b013e318221e7dd PMID:21988950
PMID:21913187 EPA (2007). Method 1694: Pharmaceuticals and Personal
Biggar RJ, Wohlfahrt J, Oudin A, Hjuler T, Melbye M Care Products in Water, Soil, Sediment, and Biosolids by
(2011). Digoxin use and the risk of breast cancer in HPLC/MS/MS. Washington (DC): U.S. Environmental
women. J Clin Oncol, 29(16):216570. doi:10.1200/ Protection Agency.
JCO.2010.32.8146 PMID:21422417 Eriksson M, Lindquist O, Edlund B (1984). Serum levels
Birkenfeld AL, Jordan J, Hofmann U, Busjahn A, Franke of digoxin in sudden cardiac deaths. Z Rechtsmed,
G, Krger N et al.(2009). Genetic influences on the 93(1):2932. doi:10.1007/BF00202981 PMID:6495885
pharmacokinetics of orally and intravenously admin- European Pharmacopoeia (2008). Digoxin. 7th ed.
istered digoxin as exhibited by monozygotic twins. Clin Strasbourg, France: European Directorate for the
Pharmacol Ther, 86(6):6058. doi:10.1038/clpt.2009.170 Quality of Medicines & HealthCare.
PMID:19776737 Evans RL, Owens SM, Ruch S, Kennedy RH, Seifen E
British Pharmacopoeia (2009) British Pharmacopoeia. (1990). The effect of age on digoxin pharmacokinetics in
London, UK: Medicines and healthcare products regu- Fischer-344 rats. Toxicol Appl Pharmacol, 102(1):617.
latory agency. doi:10.1016/0041-008X(90)90083-7 PMID:2296772
Burchell HB (1983). Digitalis poisoning: historical and Ewertz M, Holmberg L, Tretli S, Pedersen BV, Kristensen
forensic aspects. J Am Coll Cardiol, 1(2 Pt 1):50616. A (2001). Risk factors for male breast cancera
doi:10.1016/S0735-1097(83)80080-1 PMID:6338083
415
IARC MONOGRAPHS 108
case-control study from Scandinavia. Acta Oncol, Gerloff T, Schaefer M, Johne A, Oselin K, Meisel C, Cascorbi
40(4):46771. doi:10.1080/028418601750288181 I et al.(2002). MDR1 genotypes do not influence the
PMID:11504305 absorption of a single oral dose of 1 mg digoxin in
Fang MC, Stafford RS, Ruskin JN, Singer DE (2004). healthy white males. Br J Clin Pharmacol, 54(6):6106.
National trends in antiarrhythmic and antithrom- doi:10.1046/j.1365-2125.2002.01691.x PMID:12492608
botic medication use in atrial fibrillation. Arch Intern Gibson DA, Saunders PT (2012). Estrogen dependent
Med, 164(1):5560. doi:10.1001/archinte.164.1.55 signalling in reproductive tissues - a role for
PMID:14718322 estrogen receptors and estrogen related receptors.
FDA (2013). Drugs@FDA. FDA Approved Drug Products. Mol Cell Endocrinol, 348(2):36172. doi:10.1016/j.
Available from: http://www.accessdata.fda.gov/scripts/ mce.2011.09.026 PMID:21964318
cder/drugsatfda/, accessed 8 September 2014 Goldin AG, Safa AR (1984). Digitalis and cancer. Lancet,
Ferrer I, Thurman EM (2012). Analysis of 100 pharmaceu- 323(8386):1134 doi:10.1016/S0140-6736(84)92556-X
ticals and their degradates in water samples by liquid PMID:6144872
chromatography/quadrupole time-of-flight mass spec- Grabowski T, wierczewska A, Borucka B, Sawicka R,
trometry. J Chromatogr A, 1259:14857. doi:10.1016/j. Sasinowska-Motyl M, Gumuka SW et al.(2009). A
chroma.2012.03.059 PMID:22487190 rapid chromatographic/mass spectrometric method
Finch MB, Johnston GD, Kelly JG, McDevitt DG (1984). for digoxin quantification in human plasma. Pharm
Pharmacokinetics of digoxin alone and in the pres- Chem J, 43(12):7105. doi:10.1007/s11094-010-0384-y
ence of indomethacin therapy. Br J Clin Pharmacol, Guan F, Ishii A, Seno H, Watanabe-Suzuki K, Kumazawa
17(3):3535. doi:10.1111/j.1365-2125.1984.tb02353.x T, Suzuki O (1999). Identification and quantification of
PMID:6712868 cardiac glycosides in blood and urine samples by HPLC/
Foss PRB, Benezra SA (1980). Digoxin. In: Florey K, editor. MS/MS. Anal Chem, 71(18):403443. doi:10.1021/
Analytical Profiles of Drug Substances. Academic ac990268c PMID:10500489
press, vol 9, pp. 207243. Harrison LI, Gibaldi M (1976). Pharmacokinetics of
Friedman GD (1984). Digitalis and breast cancer. Lancet, digoxin in the rat. Drug Metab Dispos, 4(1):8893.
324(8407):875 doi:10.1016/S0140-6736(84)90915-2 PMID:3407
PMID:6148608 Hashimoto Y, Shibakawa K, Nakade S, Miyata Y (2008).
Friedman GD, Ury HK (1980). Initial screening for Validation and application of a 96-well format solid-
carcinogenicity of commonly used drugs. J Natl Cancer phase extraction and liquid chromatography-tandem
Inst, 65(4):72333. PMID:6932525 mass spectrometry method for the quantitation of
Frommherz L, Khler H, Brinkmann B, Lehr M, Beike J digoxin in human plasma. J Chromatogr B Analyt
(2008). LC-MS assay for quantitative determination of Technol Biomed Life Sci, 869(12):12632. doi:10.1016/j.
cardio glycoside in human blood samples. Int J Legal jchromb.2008.05.026 PMID:18515196
Med, 122(2):10914. doi:10.1007/s00414-007-0175-5 Haux J (1999). Digitoxin is a potential anticancer agent for
PMID:17569072 several types of cancer. Med Hypotheses, 53(6):5438.
Fuster V, Rydn LE, Cannom DS, Crijns HJ, Curtis AB, doi:10.1054/mehy.1999.0985 PMID:10687899
Ellenbogen KA etal.; American College of Cardiology/ Haux J, Klepp O, Spigset O, Tretli S (2001). Digitoxin
American Heart Association Task Force on Practice medication and cancer; case control and internal
Guidelines; ; European Society of Cardiology dose-response studies. BMC Cancer, 1(1):11
Committee for Practice Guidelines; ; European Heart doi:10.1186/1471-2407-1-11 PMID:11532201
Rhythm Association; ; Heart Rhythm Society(2006). Haynes K, Heitjan D, Kanetsky P, Hennessy S (2008).
ACC/AHA/ESC 2006 Guidelines for the Management Declining public health burden of digoxin toxicity
of Patients with Atrial Fibrillation: a report of the from 1991 to 2004. Clin Pharmacol Ther, 84(1):904.
American College of Cardiology/American Heart doi:10.1038/sj.clpt.6100458 PMID:18091761
Association Task Force on Practice Guidelines and Hirabayashi H, Sugimoto H, Matsumoto S, Amano N,
the European Society of Cardiology Committee for Moriwaki T (2011). Development of a quantification
Practice Guidelines (Writing Committee to Revise method for digoxin, a typical P-glycoprotein probe in
the 2001 Guidelines for the Management of Patients clinical and non-clinical studies, using high perfor-
With Atrial Fibrillation): developed in collaboration mance liquid chromatography-tandem mass spec-
with the European Heart Rhythm Association and the trometry: the usefulness of negative ionization mode to
Heart Rhythm Society. Circulation, 114(7):e257354. avoid competitive adduct-ion formation. J Chromatogr
PMID:16908781 B Analyt Technol Biomed Life Sci, 879(32):383744.
Gault MH, Longerich LL, Loo JC, Ko PT, Fine A, Vasdev doi:10.1016/j.jchromb.2011.10.031 PMID:22098716
SC et al.(1984). Digoxin biotransformation. Clin Hoffmeyer S, Burk O, von Richter O, Arnold HP,
Pharmacol Ther, 35(1):7482. doi:10.1038/clpt.1984.11 Brockmller J, Johne A et al.(2000). Functional poly-
PMID:6690172 morphisms of the human multidrug-resistance gene:
416
Digoxin
multiple sequence variations and correlation of one Kirby BJ, Kalhorn T, Hebert M, Easterling T, Unadkat
allele with P-glycoprotein expression and activity in JD (2008). Sensitive and specific LC-MS assay for
vivo. Proc Natl Acad Sci USA, 97(7):34738. doi:10.1073/ quantification of digoxin in human plasma and urine.
pnas.97.7.3473 PMID:10716719 Biomed Chromatogr, 22(7):7128. doi:10.1002/bmc.988
Hollman A (1985). Plants and cardiac glycosides. Br Heart J, PMID:18317988
54(3):25861. doi:10.1136/hrt.54.3.258 PMID:4041297 Kleemann A (2012). Cardiovascular Drugs. In: Ullmanns
HSDB (2013). Hazardous substance databank. National Encyclopedia of Industrial Chemistry. Weinheim,
library of medicine. Available from: http://toxnet. Germany: Wiley-VCH Verlag GmbH & Co. KGaA,
nlm.nih.gov/cgi-bin/sis/htmlgen?HSDB, accessed 10 pp. 118. http://dx.doi.org/10.1002/14356007.a05_289
February 2015. doi:10.1002/14356007.a05_289
Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis Kurzawski M, Bartnicka L, Florczak M, Grnik W,
GS, Ganiats TG et al.(2009). 2009 focused update Drodzik M (2007). Impact of ABCB1 (MDR1) gene
incorporated into the ACC/AHA 2005 Guidelines polymorphism and P-glycoprotein inhibitors on
for the Diagnosis and Management of Heart Failure digoxin serum concentration in congestive heart failure
in Adults: a report of the American College of patients. Pharmacol Rep, 59(1):10711. PMID:17377214
Cardiology Foundation/American Heart Association Lacarelle B, Rahmani R, de Sousa G, Durand A, Placidi
Task Force on Practice Guidelines: developed in M, Cano JP (1991). Metabolism of digoxin, digoxi-
collaboration with the International Society for Heart genin digitoxosides and digoxigenin in human hepat-
and Lung Transplantation. Circulation, 119(14):e391 ocytes and liver microsomes. Fundam Clin Pharmacol,
479. doi:10.1161/CIRCULATIONAHA.109.192065 5(7):56782. doi:10.1111/j.1472-8206.1991.tb00746.x
PMID:19324966 PMID:1778535
IARC (2012). Pharmaceuticals. Volume 100 A. A review of Lee LS, Chan LN (2006). Evaluation of a sex-based
human carcinogens. IARC Monogr Eval Carcinog Risks difference in the pharmacokinetics of digoxin.
Hum, 100:Pt A: 1401. PMID:23189749 Pharmacotherapy, 26(1):4450. doi:10.1592/
IMS Health (2012a). Multinational Integrated Data phco.2006.26.1.44 PMID:16509025
Analysis (MIDAS). Plymouth Meeting, Pennsylvania: Lenfant-Pejovic MH, Mlika-Cabanne N, Bouchardy C,
IMS Health; 2012. Auquier A (1990). Risk factors for male breast cancer:
IMS Health (2012b). National Disease and Therapeutic a Franco-Swiss case-control study. Int J Cancer,
Index (NDTI). Plymouth Meeting, Pennsylvania: IMS 45(4):6615. doi:10.1002/ijc.2910450415 PMID:2323842
Health; 200412. Li S, Liu G, Jia J, Miao Y, Gu S, Miao P et al.(2010).
IMS Health (2012c). National Prescription Audit Plus Therapeutic monitoring of serum digoxin for
(NPA). Plymouth Meeting, Pennsylvania: IMS Health; patients with heart failure using a rapid LC-MS/MS
200812. method. Clin Biochem, 43(3):30713. doi:10.1016/j.
Index Nominum (2013). International Drug Directory. clinbiochem.2009.09.025 PMID:19833118
Edited by the Swiss Pharmaceutical Society. Stuttgart, Li-Saw-Hee FL, Lip GY (1998). Digoxin revisited.
Germany: Medpharm Scientific Publishers. QJM, 91(4):25964. doi:10.1093/qjmed/91.4.259
Johne A, Kpke K, Gerloff T, Mai I, Rietbrock S, Meisel PMID:9666948
C et al.(2002). Modulation of steady-state kinetics of Lowes S, Cavet ME, Simmons NL (2003). Evidence for a
digoxin by haplotypes of the P-glycoprotein MDR1 non-MDR1 component in digoxin secretion by human
gene. Clin Pharmacol Ther, 72(5):58494. doi:10.1067/ intestinal Caco-2 epithelial layers. Eur J Pharmacol,
mcp.2002.129196 PMID:12426522 458(12):4956. doi:10.1016/S0014-2999(02)02764-4
Josephs RD, Daireaux A, Westwood S, Wielgosz RI PMID:12498906
(2010). Simultaneous determination of various cardiac Marcus FI, Kapadia GJ, Kapadia GG (1964). The metab-
glycosides by liquid chromatography-hybrid mass olism of digoxin in normal subjects. J Pharmacol Exp
spectrometry for the purity assessment of the ther- Ther, 145:2039. PMID:14214418
apeutic monitored drug digoxin. J Chromatogr A, Mayer U, Wagenaar E, Beijnen JH, Smit JW, Meijer DK,
1217(27):453543. doi:10.1016/j.chroma.2010.04.060 van Asperen J et al.(1996). Substantial excretion of
PMID:20537342 digoxin via the intestinal mucosa and prevention of
Kau MM, Kan SF, Wang JR, Wang PS (2005). Inhibitory long-term digoxin accumulation in the brain by the
effects of digoxin and ouabain on aldosterone synthesis mdr 1a P-glycoprotein. Br J Pharmacol, 119(5):103844.
in human adrenocortical NCI-H295 cells. J Cell Physiol, doi:10.1111/j.1476-5381.1996.tb15775.x PMID:8922756
205(3):393401. doi:10.1002/jcp.20415 PMID:15887230 McNamara RL, Tamariz LJ, Segal JB, Bass EB
Kawahara M, Sakata A, Miyashita T, Tamai I, Tsuji A (1999). (2003). Management of atrial fibrillation: review
Physiologically based pharmacokinetics of digoxin of the evidence for the role of pharmacologic
in mdr1a knockout mice. J Pharm Sci, 88(12):12817. therapy, electrical cardioversion, and echocar-
doi:10.1021/js9901763 PMID:10585223 diography. Ann Intern Med, 139(12):101833.
417
IARC MONOGRAPHS 108
doi:10.7326/0003-4819-139-12-200312160-00012 Reuning RH, Sams RA, Notari RE (1973). Role of

PMID:14678922 pharmacokinetics in drug dosage adjustment. I.
Medscape (2013). Medscape Drugs & Disease. Available Pharmacologic effect kinetics and apparent volume of
from: http://reference.medscape.com/drug/lanoxin- distribution of digoxin. J Clin Pharmacol New Drugs,
digoxin-342432, accessed 5 June 2013. 13(4):12741. doi:10.1002/j.1552-4604.1973.tb00074.x
Mikkaichi T, Suzuki T, Onogawa T, Tanemoto M, PMID:4487163
Mizutamari H, Okada M et al.(2004). Isolation and Rifka SM, Pita JC Jr, Loriaux DL (1976). Mechanism of
characterization of a digoxin transporter and its rat interaction of digitalis with estradiol binding sites in
homologue expressed in the kidney. Proc Natl Acad Sci rat uteri. Endocrinology, 99(4):10916. doi:10.1210/
USA, 101(10):356974. doi:10.1073/pnas.0304987101 endo-99-4-1091 PMID:976189
PMID:14993604 Rifka SM, Pita JC, Vigersky RA, Wilson YA, Loriaux DL
Newman RA, Yang P, Pawlus AD, Block KI (2008). Cardiac (1978). Interaction of digitalis and spironolactone with
glycosides as novel cancer therapeutic agents. Mol human sex steroid receptors. J Clin Endocrinol Metab,
Interv, 8(1):3649. doi:10.1124/mi.8.1.8 PMID:18332483 46(2):33844. doi:10.1210/jcem-46-2-338 PMID:86546
Oiestad EL, Johansen U, Stokke Opdal M, Bergan Riganti C, Campia I, Kopecka J, Gazzano E, Doublier
S, Christophersen AS (2009). Determination of S, Aldieri E et al.(2011). Pleiotropic effects of cardi-
digoxin and digitoxin in whole blood. J Anal Toxicol, oactive glycosides. Curr Med Chem, 18(6):87285.
33(7):3728. doi:10.1093/jat/33.7.372 PMID:19796507 doi:10.2174/092986711794927685 PMID:21182478
Omidfar K, Kia S, Kashanian S, Paknejad M, Besharatie Sakaeda T, Nakamura T, Horinouchi M, Kakumoto M,
A, Kashanian S etal.(2010). Colloidal nanogold-based Ohmoto N, Sakai T et al.(2001). MDR1 genotype-re-
immunochromatographic strip test for the detection of lated pharmacokinetics of digoxin after single oral
digoxin toxicity. Appl Biochem Biotechnol, 160(3):843 administration in healthy Japanese subjects. Pharm
55. doi:10.1007/s12010-009-8535-x PMID:19224402 Res, 18(10):14004. doi:10.1023/A:1012244520615
Phkla R, Irs A, Oselin K, Rootslane L (1999). Digoxin: PMID:11697464
use pattern in Estonia and bioavailability of the Salphati L, Benet LZ (1999). Metabolism of digoxin and
local market leader. J Clin Pharm Ther, 24(5):37580. digoxigenin digitoxosides in rat liver microsomes:
doi:10.1046/j.1365-2710.1999.00239.x PMID:10583701 involvement of cytochrome P4503A. Xenobiotica,
Pellati F, Bruni R, Bellardi MG, Bertaccini A, Benvenuti S 29(2):17185. doi:10.1080/004982599238722
(2009). Optimization and validation of a high-perfor- PMID:10199593
mance liquid chromatography method for the analysis Salvador A, Sagan C, Denouel J (2006). Rapid quanti-
of cardiac glycosides in Digitalis lanata. J Chromatogr tation of digoxin in human plasma and urine using
A, 1216(15):32609. doi:10.1016/j.chroma.2009.02.042 isotope dilution liquid chromatography-tandem mass
PMID:19268961 spectrometry. J Liquid Chromatogr Relat Technol,
Pharmaceuticals and Medical Devices Agency (2011). The 29(13):191732. doi:10.1080/10826070600757821
Japanese Pharmacopoeia Sixteenth Edition, English Saunders KB, Amerasinghe AK, Saunders KL (1997). Dose
Version. Tokyo, Japan: Ministry of Health, Labour and of digoxin prescribed in the UK compared with France
Welfare. Available from: http://jpdb.nihs.go.jp/jp16e/ and the USA. Lancet, 349(9055):8336. doi:10.1016/
jp16e.pdf, accessed 10 February 2015. S0140-6736(96)09057-5 PMID:9121258
Pinto N, Halachmi N, Verjee Z, Woodland C, Klein J, Schinkel AH, Wagenaar E, van Deemter L, Mol CA,
Koren G (2005). Ontogeny of renal P-glycoprotein Borst P (1995). Absence of the mdr1a P-Glycoprotein
expression in mice: correlation with digoxin renal in mice affects tissue distribution and pharmacoki-
clearance. Pediatr Res, 58(6):12849. doi:10.1203/01. netics of dexamethasone, digoxin, and cyclosporin A.
pdr.0000188697.99079.27 PMID:16306209 J Clin Invest, 96(4):1698705. doi:10.1172/JCI118214
Platz EA, Yegnasubramanian S, Liu JO, Chong CR, Shim PMID:7560060
JS, Kenfield SA etal.(2011). A novel two-stage, transdis- Schussheim DH, Schussheim AE (1998). Is digoxin a
ciplinary study identifies digoxin as a possible drug for designer oestrogen? Lancet, 351(9117):1734 doi:10.1016/
prostate cancer treatment. Cancer Discov, 1(1):6877. S0140-6736(05)77771-0 PMID:9734913
doi:10.1158/2159-8274.CD-10-0020 PMID:22140654 SciFinder (2013). Digoxin: Regulatory Information.
PubChem (2013). Pubchem database. NCBI. Available Columbus (OH): American Chemical Society. Available
from: https://pubchem.ncbi.nlm.nih.gov/, accessed 10 from: http://www.chemnet.com/cas/en/20830-75-5/
February 2015. digoxin.html, accessed 8 September 2014.
Rathore SS, Curtis JP, Wang Y, Bristow MR, Krumholz Sleeswijk ME, Van Noord T, Tulleken JE, Ligtenberg JJ,
HM (2003). Association of serum digoxin concen- Girbes AR, Zijlstra JG (2007). Clinical review: treat-
tration and outcomes in patients with heart failure. ment of new-onset atrial fibrillation in medical inten-
JAMA, 289(7):8718. doi:10.1001/jama.289.7.871 sive care patientsa clinical framework. Crit Care,
PMID:12588271 11(6):233 doi:10.1186/cc6136 PMID:18036267
418
Digoxin
Stafford RS, Robson DC, Misra B, Ruskin J, Singer DE Vlase L, Popa D-S, Muntean D, Mihu D, Leucuta S (2009).
(1998). Rate control and sinus rhythm maintenance A new, high-throughput high-performance liquid
in atrial fibrillation: national trends in medication chromatographic/mass spectrometric assay for thera-
use, 19801996. Arch Intern Med, 158(19):21448. peutic level monitoring of digoxin in human plasma. J
doi:10.1001/archinte.158.19.2144 PMID:9801182 AOAC Int, 92(5):13905. PMID:19916377
Stenkvist B (1999). Is digitalis a therapy for breast carci- Wang Z, Zheng M, Li Z, Li R, Jia L, Xiong X etal.(2009).
noma? Oncol Rep, 6(3):4936. PMID:10203580 Cardiac glycosides inhibit p53 synthesis by a mecha-
Stenkvist B, Bengtsson E, Dahlqvist B, Eriksson O, nism relieved by Src or MAPK inhibition. Cancer Res,
Jarkrans T, Nordin B (1982). Cardiac glycosides and 69(16):655664. doi:10.1158/0008-5472.CAN-09-0891
breast cancer, revisited. N Engl J Med, 306(8):484 PMID:19679550
doi:10.1056/NEJM198202253060813 PMID:7057849 Withering W (1785) An account of the foxglove and some
Stenkvist B, Bengtsson E, Eriksson O, Holmquist J, Nordin of its medical uses: with practical remarks on dropsy,
B, Westman-Naeser S (1979). Cardiac glycosides and and other diseases. London, UK: G.G.J. and J. Robinson.
breast cancer. Lancet, 1(8115):563 doi:10.1016/S0140- Yao HM, Chiou WL (2006). The complexity of intestinal
6736(79)90996-6 PMID:85158 absorption and exsorption of digoxin in rats. Int J Pharm,
Stuhlfauth T, Klug K, Fock HP (1987). The production 322(12):7986. doi:10.1016/j.ijpharm.2006.05.030
of secondary metabolites by Digitalis lanata during PMID:16781832
CO2 enrichment and water stress. Phytochemistry, Yao M, Zhang H, Chong S, Zhu M, Morrison RA (2003).
26(10):27359. doi:10.1016/S0031-9422(00)83581-5 A rapid and sensitive LC/MS/MS assay for quantita-
Sturm HB, van Gilst WH, Veeger N, Haaijer-Ruskamp FM tive determination of digoxin in rat plasma. J Pharm
(2007). Prescribing for chronic heart failure in Europe: Biomed Anal, 32(6):118997. doi:10.1016/S0731-
does the country make the difference? A European 7085(03)00050-5 PMID:12907263
survey. Pharmacoepidemiol Drug Saf, 16(1):96103. Zhang H, Qian DZ, Tan YS, Lee K, Gao P, Ren YR etal.
doi:10.1002/pds.1216 PMID:16528759 (2008). Digoxin and other cardiac glycosides inhibit
Taub ME, Mease K, Sane RS, Watson CA, Chen L, HIF-1 synthesis and block tumor growth. Proc
Ellens H et al.(2011). Digoxin is not a substrate for Natl Acad Sci USA, 105(50):1957986. doi:10.1073/
organic anion-transporting polypeptide transporters pnas.0809763105 PMID:19020076
OATP1A2, OATP1B1, OATP1B3, and OATP2B1 but is a
substrate for a sodium-dependent transporter expressed
in HEK293 cells. Drug Metab Dispos, 39(11):2093102.
doi:10.1124/dmd.111.040816 PMID:21849517
The International Pharmacopoeia (2011) 4th Edition.
Geneva, Switzerland: World Health Organization.
Tracqui A, Kintz P, Ludes B, Mangin P (1997). High-
performance liquid chromatography-ionspray mass
spectrometry for the specific determination of
digoxin and some related cardiac glycosides in human
plasma. J Chromatogr B Biomed Sci Appl, 692(1):1019.
doi:10.1016/S0378-4347(96)00462-8 PMID:9187389
Tsujimoto M, Kinoshita Y, Hirata S, Otagiri M, Ohtani H,
Sawada Y (2008). Effects of uremic serum and uremic
toxins on hepatic uptake of digoxin. Ther Drug Monit,
30(5):57682. doi:10.1097/FTD.0b013e3181838077
PMID:18708994
United States Pharmacopoeial Convention (2006) United
States Pharmacopeial Convention, 30. Rockville (MD):
The United States Pharmacopoeia Convention.
USP; US Pharmacopeia (2007) United States
Pharmacopeial Convention. Rockville (MD), USA:
United States Pharmacopeial Convention.
Varma MVS, Kapoor N, Sarkar M, Panchagnula R
(2004). Simultaneous determination of digoxin and
permeability markers in rat in situ intestinal perfu-
sion samples by RP-HPLC. J Chromatogr B Analyt
Technol Biomed Life Sci, 813(12):34752. doi:10.1016/j.
jchromb.2004.09.047 PMID:15556552
419
LIST OF ABBREVIATIONS
5-ASA 5-aminosalicylic acid

ACBS 2-amino-4-chloro-1,3-benzenedisulfonamide
AUC area under the curve
BBN N-butyl-N-(4-hydroxybutyl)nitrosamine
bw body weight
CI confidence interval
CMPF 3-carboxy-4-methyl-5-propyl-2-furanpropanoic acid
CYP cytochrome P450
DDD defined daily doses
EPA Environmental Protection Agency
ELS evaporative light scattering
ESI-MS electrospray ionization mass spectroscopy
FDA United States Food and Drug Administration
GC gas chromatography
HbA1c glycated haemoglobin
HCTZ hydrochlorothiazide
1
H-NMR proton nuclear magnetic resonance
HPLC high-performance liquid chromatography
HPLC/UV high-performance liquid chromatography/ultraviolet
HPTLC high-performance thin-layer chromatography
IBD inflammatory bowel disease
ICD International Classification of Diseases
KPNC Kaiser Permanente Northern California
LOQ limit of quantitation
MRT mean residence time
MS mass spectrometry
NADPH reduced nicotinamide adenine dinucleotide phosphate
NMR nuclear magnetic resonance
NQO NAD(P)H quinone oxidoreductase
NTP National Toxicology Program
OAT organic anion transporter
OTC over the counter
PEMA phenylethyl malonamide
421
IARC MONOGRAPHS 107
PPAR peroxisome proliferator-activated receptor

SMR standardized morbidity ratio
SSL solar simulated light
TLC thin-layer chromatography
TSH thyroid-stimulating hormone
UGT UDP glycosyltransferase
USP United States Pharmacopeia
UV ultraviolet
v/v volume for volume
w/w weight for weight
422
Design by Diizz.com
This volume of the IARC Monographs provides an assessment of the carcinogenicity
of 14 drugs and herbal products.
The IARC Monographs Working Group relied on epidemiological studies to evaluate

the carcinogenic hazard to humans exposed to the drugs digoxin (widely prescribed
for the treatment of chronic heart failure), pioglitazone (used for the treatment of type
2 diabetes mellitus), and hydrochlorothiazide (used to treat hypertension).
Other agents evaluated included the drugs primidone, sulfasalazine, pentosan

polysulfate sodium, and triamterene, and five herbal products (or their components):
Aloe vera whole leaf extract, goldenseal root powder, Ginkgo biloba leaf extract, kava
extract, and pulegone. In view of the limited agent-specific information available from
epidemiological studies, assessments of these agents relied mainly on carcinogenicity
bioassays to reach conclusions as to the carcinogenic hazard to exposed humans.
istockphoto.com

Mono 108

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mono 108

Uploaded by

Copyright:

Available Formats

SOME DRUGS AND

This publication represents the views and expert

Lyon, France - 2016

Co-funded by the European Union

ISBN 978 92 832 0174 8 (NLM Classification: W1)

NOTE TO THE READER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2. Cancer in Humans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

METHYLENE BLUE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

5. Summary of Data Reported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203

PENTOSAN POLYSULFATE SODIUM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

4. Mechanistic and Other Relevant Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

PIOGLITAZONE AND ROSIGLITAZONE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

LIST OF ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421

Esperanza J. Carcache de Blanco

Michael L. Cunningham [retired]

June K. Dunnick Dirk W. Lachenmeier

M. Matilde Marques (Subgroup Chair,

Margaret Karagas (Subgroup Chair, Cancer in David L. McCormick

Saranjit Singh [unable to attend] Shufeng Zhou

Hiroyuki Tsuda Observers

Paul Dolin 6 Yann Grosse (Responsible Officer, Rapporteur

Egon Koch 8 Administrative Assistance

an assessment of the biological plausibility of epi- 3. Studies of cancer in experimental

(d) Susceptibility data found on the Monographs programme web site

5. Summary (d) Mechanistic and other relevant data

1. Introduction specific therapeutic usage, a major aspect of the

2. Exposure to herbal products Monographs are specified with reference to indi-

4. Extrapolating from specific 5. Considerations beyond hazard

1. Exposure Data A glossary of commonly used terms for Aloe

Table 1.1 Definition of terms commonly used in the Aloe industry

(b) Description Aloe vera plants contain two major liquid

The Aloe vera inner leaf pulp is

Aloe vera Plant

From Boudreau et al. (2013a)

Table 1.2 Summary of chemical constituents of Aloe vera products

Aloe vera latex

A loin A (Barbaloin) A loin B (Isobarbaloin)

A cetylated mannan (A cemannan) HO O HO O

A loeresin A A loesin (A loeresin B )

Ac, acetyl group

contained either aloe-emodin or aloin A at 10 1.3.2 Dosage

Fig.1.4 Sales of dietary supplements containing Aloe vera in the USA

72 million in 2011 (Fig. 1.4; Nutrition Business 1.4.3 Consumption

max., maximum; min., minimum; TLC, thin-layer chromatography

Species, strain Dosing regimen Incidence of tumours Significance Comments

4. Mechanistic and Other (a) Components of Aloe vera gel: metabolism

Fig.4.1 Metabolites of aloin A and aloin B

A loin A (Barbaloin) A loin B (Isobarbaloin)

H ydrolysis of the -glycosidic

Aloin A and B are constituents of Aloe vera whole leaf latex.

Test system Results Dose Aloe vera preparation Reference

Test system Results Dose Aloe vera preparation Reference

b Toxic in TA98, without exogenous metabolic system

Constituent Test system Results Dose Reference

Constituent Test system Results Dose Reference

b Not positive in TA102 or TA1978

c Positive only in TA98, TA1537, and TA1538

d Positive only in TA1538

e Mutations associated with loss of heterozygosity.

undergoes sequential oxidation to aloe-emodin Although the carcinogenicity of Aloe vera

6.2 Cancer in experimental animals Phytother Res, 21(3):23944. doi:10.1002/ptr.2057

Chem Toxicol, 55:36370. doi:10.1016/j.fct.2013.01.012 Westendorf J, Marquardt H, Poginsky B, Dominiak

1. Exposure Data Common names: Hydrastis; Golden seal;

Fig.1.1 Goldenseal (Hydrastis canadensis L.), (b) Hydrastine