Booms Ma

The endometrial factor in human
embryo implantation
_________________________________________________
CAROLIEN M BOOMSMA
The endometrial factor in human embryo implantation
Thesis, Utrecht University, with a summary in Dutch.

Proefschrift, Universiteit Utrecht, met een samenvatting in het Nederlands.
ISBN 978-90-393-49496
Author Carolien M. Boomsma
Cover Painting by Gertjan Scholte-Albers, www.scholte-albers.nl
Lay-out & Print Gildeprint Drukkerijen B.V. Enschede, The Netherlands
© 2009 Carolien Boomsma
The author gratefully acknowledges financial support for printing this thesis by: Division
Women and Baby University Medical Center Utrecht, Jurriaanse Stichting,
GlaxoSmithKline, Merck Serono, Clindia Benelux, BMA BV (Mosos), Medical Dynamics,
Wyeth Pharmaceuticals, Ferring, Smiths Medical BV, Organon, Abbott BV.
The endometrial factor
in human embryo implantation
De endometrium factor in humane embryo

implantatie
(met een samenvatting in het Nederlands)
Proefschrift
ter verkrijging van de graad van doctor aan de Universiteit

Utrecht op gezag van de rector magnificus, prof. dr. J.C.
Stoof, ingevolge het besluit van het college voor promoties
in het openbaar te verdedigen op dinsdag 14 april 2009 des
middags te 2.30 uur.
door
Caroline Mariëtte Boomsma
geboren op 13 februari 1979 te Amersfoort
Promotoren:
Prof. dr. N.S. Macklon
Prof. dr. C.J. Heijnen
Co-promotor:
Dr. A. Kavelaars
Aim to achieve the mountain top,
but don’t forget to enjoy the journey!
Contents
____________________________________________
Chapter 1 Introduction 9
Chapter 2 Current clinical therapies to improve endometrial

receptivity and embryo implantation in IVF
2.1 What can the clinician do to improve implantation? 23
A literature review.
2.2 Peri-implantation glucocorticoid administration for 43
assisted reproductive technology cycles.
A literature review and meta-analysis.
Chapter 3 Studying the endometrial factor: cytokine profiling in

endometrial secretions
3.1 Cytokine profiling in endometrial secretions: 57
a non-invasive window on endometrial receptivity.
3.2 Analysis of the endometrial secretion cytokine profile 75
to predict pregnancy in in-vitro fertilization.
3.3 Ovarian stimulation for in-vitro fertilization alters the 95
intra-uterine cytokine, chemokine and growth factor
milieu encountered by the embryo.
3.4 Is bacterial vaginosis associated with a 111
pro-inflammatory cytokine profile in endometrial
secretions of women undergoing IVF?
Chapter 4 The endometrial factor in recurrent implantation failure

Determination of the peri-implantation endometrial 127
transcript profile in women with recurrent implantation
failure versus controls.
Chapter 5 General Discussion 151
References 161
Nederlandse Samenvatting 185
List of publications 193
List of authors and their affiliations 195
Dankwoord 197
Curriculum Vitae 201
List of abbreviations 203
Chapter 1
____________________________________________
General Introduction
10 | The endometrial factor in human embryo implantation
General introduction
In-vitro fertilization (IVF) is a process by which oocytes are fertilized by sperm
outside the women’s womb, in vitro. It still represents one of the most exciting
modern scientific developments and continues to have a tremendous impact on
people's lives. One in six couples have an unwanted delay in conception and half of
these couples remain subfertile or need complex treatment, such as IVF. In Europe
this technology accounts for 2-3 percent of births (The European IVF-monitoring
program, 2005) and due to the current tendency of women to postpone their
pregnancy until beyond their most fertile age, its use is increasing (Heffner et al.
2004; Wright et al. 2005).
Since the beginning of IVF, with the first baby born in 1978 (Steptoe and Edwards,
1978), much effort has been invested into developing techniques to improve
outcomes. However, despite the transfer of apparently healthy embryos, pregnancy
rates remain around 25-35 percent per embryo transfer (The European IVF-
monitoring program, 2005). Nowadays, implantation of the embryo after transfer
constitutes the major success rate limiting step in IVF. Transferring multiple embryos
is widely employed in order to achieve higher pregnancy rates per transfer (Pandian
et al. 2005; Gleicher and Barad, 2006). However, this occurs at the expense of a high
rate of multiple gestations. Developed societies have witnessed an iatrogenic
epidemic of twin, triplet, and higher order multiple births since the adoption of
assisted reproductive techniques (Fauser et al. 2005). Multiple gestations are
increasingly viewed as a complication of infertility treatment, as they are associated
with serious adverse pregnancy and fetal outcomes (Ombelet et al. 2005; Ombelet et
al. 2006, Verberg et al. 2007). The growing trend towards transferring fewer embryos
provides further incentives to improve implantation rates.
The low implantation rates in IVF should be considered in the context of the failure
of the majority of spontaneous human conceptions to complete implantation. The
maximum chance of a successful implantation is 30-40 percent per cycle. Increasing
evidence points to pre-clinical pregnancy loss rather than failure of conception as the
principal cause of the relatively low fecundity in humans (Macklon et al. 2002;
Wilcox et al. 1988). Data from published studies point to a 30 percent of loss due to
implantation failure of the embryo (Figure 1).
Chapter 1 General Introduction | 11
Figure 1: An overview of the outcome of human conceptions
Adapted from Macklon et al., 2002.
Reproductive immunology
The fetus has a genetic makeup derived from both paternal and maternal genes and is
therefore a semi-allograft to the maternal immune system. Intriguingly, the fetus
escapes immune rejection and is tolerated for the duration of pregnancy. This
phenomenon was first described in 1953 as ‘the paradox of pregnancy’ (Medawar,
1953). The mechanism behind this paradox of pregnancy remains a central question
in reproductive immunology.
The two principal components of the immune system are the innate or non-adaptive
immune response and the acquired or adaptive immune response. Essentially, the
innate response constitutes the first non-specific line of defence against infection,
whereas the acquired response allows for a stronger antigen specific immune
response and provides the basis of immunological memory. The proteins encoded by
the Major Histocompatibility Complex (MHC) display antigens involved in immune
recognition in the human immune system known as Human Leukocyte Antigens
(HLA). Trophoblast cells express a unique pattern of MHC class I molecules, namely
HLA-C, HLA-G and HLA-E molecules, which is different from any other human
tissue (Choudhury and Knapp, 2001). Soluble HLA-G (Hunt et al. 1988), expressed
by the embryo, has been associated with downregulation of immune responses (Long
et al. 1999; Lila et al. 2000). Leukocytes are an important part of the body’s defence
system. They increase in number in the endometrium during the menstrual cycle and
account for 30 percent of endometrial cells in the decidua during early pregnancy
(Bulmer et al. 1991). The decidua is primarily populated by three leukocyte
subpopulations, namely uterine Natural Killer (uNK) cells, macrophages and T cells.
B cells are virtually absent. uNK cells constitute the majority of the immune cells and
they populate the decidua more densely than NK cells do in the peripheral blood,
namely 70 versus 20 percent (Moffett-King et al. 2002). Although the specific role of
these CD56 brightCD16 - cells is still unknown, it is suggested that uNK cells induce
immune cell modulation, endometrial decidualization and regulate placental
development by mediating vascular remodelling and fetal extra embryonic
trophoblast invasion (Moffett-King et al. 2002). The number and population of uNK
cells in pre-implantation endometrium has been related to pregnancy failure (Quenby
et al. 1999; Ledee-Bataille et al. 2005), although this was not confirmed by other
studies (Tuckerman et al. 2007; Matteo et al. 2007). uNK cells produce a wide
variety of cytokines, such as granulocyte macrophage colony stimulating factor (GM-
CSF), interferon (IFN)-γ, tumor necrosis factor (TNF)-α, interleukin (IL)-2, IL-10
and Leukemia Inhibitory Factor (LIF) (Dosiou and Giudice, 2005). A cascade of
these locally acting cytokines, chemokines and growth factors intervene in the
dialogue at the maternal-embryonic interface (Chard, 1995; Dey et al. 2004; Giudice,
1994).
Human embryo implantation

Maternal-embryonic dialogue
The process of implantation involves complex synchronized interactions between
embryonic and uterine cells and can be subdivided in three stages (Figure 2). The
initial attachment of the blastocyst to the endometrial epithelial cells is called
apposition and is unstable. The blastocyst then adheres to the endometrial surface and
a firm adhesion is realized. Thereafter the blastocyst penetrates the epithelial layer
and invades the stroma and is completely embedded by the tenth day. Eventually
cytotrophoblasts invade the entire endometrium and the inner third of the
myometrium as well as the uterine vasculature which establishes the uteroplacental
circulation.
Figure 2. Different stages of the embryo implantation
Adapted from Chaouat et al 2003.
MUC1 is a candidate for selection of the implantation site. A MUC1 barrier at the
luminal epithelium impairs interaction between the embryo and the endometrium by
anti-adhesive properties. MUC1 either must be locally removed during the
attachment process or it has functions to promote the initial steps in embryo adhesion
(DeLoia et al. 1998). The embryo itself has been speculated to be involved in
signaling the endometrium to modify MUC1 expression (Aplin, 1999). When the
embryo reaches the implantation site, it is stopped by embryonic L-selectin
interaction with uterine oligosaccharides (Genbacev et al. 2003), which are
upregulated during the window of implantation (Aplin, 1999). Heparin-binding
Epidermal Growth Factor (Hb-EGF) expression is increased in the mid-secretory
phase and is also involved in the process of embryo attachment. Cells expressing the
transmembrane form of Hb-EGF adhere to blastocysts displaying cell surface ErbB4
(Chobotova et al. 2002). The expression of Hb-EGF slightly precedes the expression
of α vβ3 (Lessey et al. 1992), which is an important integrin (Lessey et al. 1992).
Interaction with these transmembrane glycoproteins leads to a firm embryo adhesion
(Hoozemans et al. 2004).
The invasion of cytotrophoblasts requires enzymatic digestion of the extracellular

matrix, by matrix metalloproteinases (MMPs), such as MMP-9. Transforming growth
factor (TGF)-β plays an inhibitory role on trophoblast invasion by inducing TIMP-1
secretion, a MMP tissue inhibitor (Graham et al. 1992). The simultaneous increase in
MMPs and the tissue inhibitors of these metalloproteinases appear to regulate and
restrict invasion of trophoblasts to a proper depth in the uterus (Schatz et al. 1999).
The current idea is that a compromised invasion of trophoblast in the maternal spiral
arteries and shallow interstitial invasion may lead to recurrent miscarriage or
preeclampsia (Meekins et al. 1994), whereas excessive invasion of the trophoblast
can lead to abnormally firm attachment to (placenta accreta), into (placenta increta),
or through (placenta percreta) the myometrium, which is associated with serious
haemorrhage complications (Pijnenborg et al. 2006; Khong and Robertson, 1987).
One hallmark of implantation is increased vascular permeability and
neoangiogenesis. Vasoactive agents, such as vascular endothelial growth factor
(VEGF) and angiopoietin are critical for early placentation. Chemokines are small
chemotactic cytokines, well known for their function in leukocyte recruitment and
activation. As described previously, immune cells accumulate in the endometrium
and are believed to be important modulators of implantation. There is evidence that
chemokines may also control embryo trafficking (Hannan et al. 2006; Jones et al.
2004). The embryo has receptors for a number of chemokines, such as Monocyte
Chemotactic Protein-1 (MCP-1), Eotaxin and Interferon-γ inducible 10 kD protein
(IP-10).
The embryo
Fertilization occurs in the fallopian tube within 24-48 hours after ovulation. The
embryo is encased in a non-adhesive protective coating, known as the zona pellucida.
The human blastocyst, marked by a fluid filled inner cavity within the mass of cells,
enters the uterine cavity on day 4 of development. The surface cells of the blastocyst
differentiate into trophoblast, giving rise to extra embryonic structures including the
placenta, and the inner cell mass giving rise to the embryo. Within 72 hours the
embryo hatches from the zona and implants six or seven days after conception
(Norwitz, 2001).
With the increasing trend to single embryo transfer in IVF, the selection of viable
embryos is becoming increasingly important. In IVF, embryos are generally selected
on their quality by morphological assessment. Most centers transfer embryos on day
2 or 3, prior to activation of the embryonic genome and before compaction and
blastulation, which are critical stages of embryonic development. While selection of
the embryo on morphological criteria (fragmentation, number of blastomeres,

multinucleation, uneven cleavage) is associated with higher implantation and
pregnancy rates (Ebner et al. 2003), embryo morphology on day three after
fertilization is not indicative of a normal chromosomal constitution (Magli et al.
2000; Munne, 2006). An increasing body of evidence suggests that the incidence of
chromosomal abnormalities in human embryos is high and may be the cause for the
inefficiency of human reproduction (Macklon et al. 2002). Transfer of embryos
shown to be euploid by Preimplantation Genetic Screening (PGS) has been proposed
as a way to increase live birth rates. The use of PGS is increasingly common,
although evidence supporting its use is limited (Twisk et al. 2005; Donoso et al.
2007). A recent randomized controlled trial (RCT) in women with advanced maternal
age reported a significant detrimental effect of PGS on ongoing pregnancy rates
(Mastenbroek et al. 2007). These detrimental effects may be brought about by the
invasive nature of the embryo biopsy.
Non-invasive methods to assess embryo quality in order to improve embryo selection

after IVF are therefore preferential. The uptake of specific nutrients by the embryo,
such as glucose, pyruvate and lactate, from the surrounding medium has been
investigated and correlated to embryo viability (Gott et al. 1990). Another approach
is metabolic profiling of culture medium. Significant correlations between embryo
viability and pregnancy rates have been shown by spectroscopy (Seli et al. 2007), or
amino acid turnover (Houghton et al. 2002; Brison et al. 2004). Cell viability is
suggested to be associated with a quieter metabolism, since viable embryos may
require less energy rectifying damage to the genome than compromised embryos
(Leese et al. 2007). The blastocyst actively participates in the process of implantation
by producing immunologically relevant molecules. Soluble HLA-G secreted by the
embryo in culture medium, which suppresses the effector functions of T cells and
uNK cells, has been shown to be significantly related to pregnancy rates (Rebmann et
al. 2007; Yie et al. 2005; Sher et al. 2005). Other mediators in the culture medium
have also been correlated to embryo viability and pregnancy rates, such as insulin-
like growth factor (IGF)-1 (Kowalik et al. 1999), human chorionic gonadotropin
(hCG) (Oum et al. 2006), and platelet activating factor (PAF) (Roudebush et al.
2002). Comparative protein-profile analysis of culture medium from implanted
versus non-implanted human blastocysts also showed some differences in protein
profiles (Dominguez et al. 2008). However, identification of mediators secreted by

the embryo has always been difficult. Several papers have reported proposed levels of
secretion of certain mediators which seem excessive compared to the total dry matter
of the embryo (Menezo et al. 2006).
The endometrium
On the maternal side, a so called ‘receptive’ endometrium is a prerequisite. The
endometrium essentially undergoes cyclical changes in order to support a potential
pregnancy. Pre-ovulatory, estradiol-17β stimulates the proliferation and
differentiation of uterine epithelial cells and angiogenesis in order to ensure nutrition
and also stimulates the expression of progesterone and estrogen receptors. After
ovulation during the secretory phase of the menstrual cycle, this tissue begins to
transform into the decidua under the influence of progesterone produced by the
corpus luteum. Decidualization denotes secretory transformation of the uterine
glands, influx of specialized uNK cells, and vascular remodeling and is accompanied
by the secretion of prolactin, insulin-like growth factor binding protein (IGFBP)-1
and tissue factor (Christian et al. 2001; Tseng and Mazella, 1999).
Uterine receptivity is defined as a restricted period when the uterus supports

blastocyst attachment and this maternally driven ‘window’ of receptivity takes place
from approximately day 19-24 in a normal menstrual cycle (Wilcox et al. 1999;
Navot et al. 1991). It was originally proposed by Wegmann et al. that successful
implantation requires a dominance of anti-inflammatory cytokines rather than pro-
inflammatory cytokines, which have been associated with pregnancy rejection
(Wegmann et al. 1993). In a CBA x DBA/2 murine abortion model, IL-10 has been
shown to prevent naturally occurring fetal loss, which is a major anti-inflammatory
cytokine. Moreover, injecting either anti- IFN-γ or pentoxifillin (anti-TNF-α), both
pro-inflammatory cytokines, partially reduces the resorption (Chaouat et al. 1995). In
addition, peripheral blood mononuclear cells in women with unexplained recurrent
abortion have been shown to produce pro-inflammatory cytokines in response to
trophoblast antigens (Hill et al. 1995). However, recent research has questioned this
paradigm and it is now suggested that the pro- and anti-inflammatory paradigm may
be an oversimplification of the role of cytokines in reproduction (Chaouat et al. 2004;
Chaouat et al. 2007). Indeed, it is noteworthy that all the mice knock-out studies
which described implantation defects concern pro-inflammatory cytokines.
Techniques to study endometrial receptivity

Understanding the determinants of successful implantation and endometrial
receptivity is of primary importance if we are to further improve implantation rates in
IVF therapy. For a number of reasons available tests for endometrial receptivity are
of limited value and the endometrial factor in human embryo implantation has been
difficult to study.
Morphological changes in the endometrium

As early as the 1950s, the histological morphological changes of the endometrium
throughout the menstrual cycle have been described by Noyes (Noyes et al. 1950).
The Noyes criteria, designed to date the endometrium by distinct histological
patterns, have since remained a gold standard approach. However, the usefulness of
these criteria for couples with infertility has been questioned because the incidence of
out-of-phase endometrial maturation was high among patients and controls and
therefore fails to discriminate between these two study groups (Coutifaris et al.
2004). Moreover, the value of applying Noyes criteria in a clinical context is limited
due to their subjective nature (Murray et al. 2004; Diedrich et al. 2007). Both intra-
and interobserver variability are high and intraobserver variability has been shown to
be highest among infertile women during the window of implantation. Electron
microscopy has revealed the cyclic appearance of endometrial surface structures in
the mid-luteal phase, termed pinopodes (Bentin-Ley et al. 1999) or uterodomes
(Murphy, 2000). These epithelial protrusions might influence the concentration of
endometrial fluids, thus facilitating apposition and adhesion of the embryo.
Therefore, their expression in the endometrium has been proposed as a marker of
endometrial receptivity. Pinopodes are however expressed for an extended period of
time and do not correlate with the putative ‘window of implantation’ (Quinn et al.
2007).
The use of animal models to study human embryo implantation

Much of our understanding of early human development is inferred from studies in
animals. A KO mouse is a genetically engineered mouse in which one or more genes
have been turned off through a gene knockout. KO mice are important animal models
to examine functions driven by single genes. This, however, is not the case for most
reproductive functions. It seems highly unlikely that a function critical to the survival
of species could be compromised by a deficient expression of a single cytokine
(Chaouat et al. 2007). Relatively few cytokine gene KO studies have shown effects
on reproduction. The best known mouse KO model resulting in total implantation
failure is that for LIF (Stewart et al. 1992). In humans LIF deficiency has also been
linked to reproductive failure (Laird et al. 1997; Delage et al. 1995), although
conflicting results were found. Further, KO mice for IL-11 (Robb et al. 1998) and
CSF-1 KO were observed to be infertile (Pollard et al. 1991). However, caution is
required when data from mice are extrapolated to humans. Although both species
have hemochorial placentas in which there is direct contact between maternal blood
and trophoblasts, there are also important differences. In mice, NK cells are not
present in the non-pregnant endometrium, decidualization occurs only after
implantation and in pregnancy there is minimal trophoblast invasion (Figure 3)
(Carson et al. 2000). Primate models are more representative, but research is
expensive, morally more difficult and availability is limited.
Addressing molecular redundancy

A number of possible markers of the receptive endometrium have been proposed.
Candidate molecules include members of the integrin family (Lessey et al. 1995;
Thomas et al. 2003), glycodelin (Chryssikopoulos et al. 1996), Hb-EGF (Yoo et al.
1997), CSF-1 (Kauma et al. 1991) and LIF (Ledee-Bataille et al. 2002). However, to
date no single marker has been identified that is specific and sensitive in predicting
pregnancy though (Hoozemans et al. 2004; Strowitzki et al. 2006). The search for
predictors of implantation has focused primarily on the analysis of individual
markers. There is now a movement towards technologies capable of quantifying
thousands of genes or gene products. DNA micro-array gene expression profiling
enables an approach from a global genomic perspective. Modern multiplex
immunoassays enable multiple markers to be quantified simultaneously in a small
sample volume (de Jager et al. 2003). These technologies may be particularly
relevant in the study of implantation, since numerous factors are involved with
multiple functions and with potentially a large amount of redundancy.
Figure 3: Implantation strategies among species
A Murid rodents. Penetration of the uterine epithelium by trophoblast cells, aided by apoptosis of uterine
epithelial cells.
B Guinea pig. The syncytial trophoblast penetrates through the zona pellucida, then between uterine epithelial
cells, after which the rest of the blastocyst is drawn into the uterine stroma.
C Rabbit. Syncytial trophoblast knobs fuse with one or two uterine epithelial cells creating pegs in the
epithelium prior to penetration into the stroma.
D Primates. Syncytial trophoblast formed near the inner cell mass intrudes between uterine epithelial cells
before penetrating the basal lamina.
Abbreviations: D, decidual cells; En, embryonic endoderm; Ep, uterine epithelium; ICM, inner cell
mass; S, stroma; T, trophoblast; ZP, zona pellucida.
From: Carson et al. 2000.
Non-invasive techniques
Most techniques to investigate endometrial receptivity require an endometrial biopsy.
However, an endometrial tissue biopsy cannot be obtained prior to embryo transfer
without disrupting the process of implantation itself (van der Gaast et al. 2003).
Investigations in a donor oocyte model offer a way around this problem (Damario et
al. 2001), but are limited by the lack of implantation as an endpoint, since the embryo
is transferred into the uterus of the recipient. A non-invasive approach to assess
endometrial receptivity, which can be applied during the window of implantation, is
by flushing of the uterine cavity (Berkkanoglu et al. 2006; Li et al. 2006; Olivennes
et al. 2003). However, this approach results in a higher and variable degree of
dilution of endometrial secretions, since 1-2 mL saline water is instilled and the
amount of fluid recovered differs. Therefore, this may account for the inconsistent
findings reported (Laird et al. 1997; Lédée et al. 2002; Olivennes et al. 2003). A
further non-invasive approach proposed is the analysis of endometrial markers of
receptivity in peripheral serum samples. However, since the expression of cytokines
differs in different compartments of the body due to the strong influence of the local
micro-environment, it is unlikely that this approach would generate data
representative of the intra-uterine milieu (Hoozemans et al. 2004).
Another possible approach is the analysis of endometrial secretions aspirated from

the uterine cavity during the window of implantation. We have previously shown that
endometrial secretion aspiration can be carried out immediately prior to embryo
transfer without affecting implantation rates (van der Gaast et al. 2003). One-
dimensional gel electrophoresis (1D PAGE) demonstrated protein profiles in the
endometrial secretions which change with the phase of the cycle (Beier-Hellwig et al.
1994). Endometrial secretion analysis may open a new ‘window’ on endometrial
receptivity, implantation and the factors which modulate this complex and elusive
process.
Aims and outline of this thesis

Since implantation failure in IVF is a considerable cause of frustration to both
patients and clinicians, it is unsurprising that clinicians are prone to intervene in order
to improve implantation. Moreover, the competitive commercial context in which
IVF is usually practised has often led to the early adoption of promising new
interventions by clinicians before clear evidence for their efficacy and safety has been
established.
 In chapter 2.1 the various clinical strategies which may be employed by
clinicians to increase endometrial receptivity and the chance of embryo
implantation are reviewed.
 It has been proposed that glucocorticoids may improve the intra-uterine
environment by acting as immuno-modulators. However, the RCTs on
glucocorticoid administration in the peri-implantation period to improve
embryo implantation are insufficiently powered and reveal inconsistent
results. In chapter 2.2 the results of a meta-analysis of these RCTs are
described and the available evidence is reviewed in order to clarify the

clinical value of adjuvant glucocorticoid administration in IVF treatment.
In order to improve implantation in IVF, the ability to investigate endometrial

receptivity and the implantation process is of vital importance. Endometrial secretion
analysis is a novel means of assessing the intra-uterine milieu.
 Chapter 3.1 describes the establishment of the quantification of a profile of
cytokines, chemokines and growth factors involved in the implantation
process and endometrial maturation in endometrial secretions aspirated prior
to embryo transfer. Further aims of this study were to confirm the safety of
the technique and to assess possible contamination of the aspirate by cervical
mucus and the influence of blood contamination on the results.
 In chapter 3.2 the hypothesis that the intra-uterine cytokine expression profile,
identified at the time of embryo transfer following IVF/ intra cytoplasmic
sperm injection (ICSI) treatment can predict successful implantation, will be
tested. This study also describes the rate of preclinical pregnancy loss in
women undergoing IVF.
 The intra-uterine milieu may be influenced by ovarian stimulation and ovarian
response after IVF or the existence of a bacterial vaginosis. These two
questions will be addressed in chapter 3.3 and 3.4 respectively.
Microarray analysis enables an approach from a global genomic perspective.

 In chapter 4 we have investigated the global gene expression during the
window of implantation in fertile controls compared to normo-ovulatory
women with idiopathic recurrent implantation failure after IVF, which
represents a major clinical problem in IVF centres.
Finally, in chapter 5 the findings from the conducted studies, implications for
clinical practice and future research are discussed.
Chapter 2.1
____________________________________________
What can the clinician do to improve implantation?
Carolien M Boomsma1
Nick S Macklon1
1
Department of Reproductive Medicine and Gynaecology, University Medical Center
Utrecht
(Reproductive BioMedicine Online 2006; 13: 845-55)

Abstract
Implantation is a complicated process that requires the orchestration of a series of
events involving both the embryo and the endometrium. Even with the transfer of
high quality embryos, implantation rates remain relatively low. The growing
tendency towards transferring fewer embryos provides further incentives to improve
implantation rates. In this article, the various clinical strategies employed to increase
the chance of implantation are reviewed. Embryo transfer technique is a critical step
in assisted reproductive technology cycles. Recent studies have shown significant
improvements in clinical pregnancy rates resulting from careful embryo transfer
technique, appropriate catheter type and placing for embryo transfer. Increasingly,
adjuvant pharmaceutical therapies are also being applied with the aim of improving
embryo implantation. However, the evidence for their efficacy and safety is limited.
Recent evidence suggests that adoption of milder ovarian stimulation regimens may
provide a more effective clinical approach to improving implantation, since beneficial
affects have been shown for both endometrial receptivity and embryo quality.
Chapter 2.1 | 25
Introduction
Implantation is a complicated process that requires the orchestration of a series of
events involving both the embryo and the endometrium. Even when embryos are to
be of high quality using morphological and chromosomal criteria, implantation rates
remain around 25%-35% (The European IVF-monitoring program 2005). High rates
of implantation failure and early pregnancy loss in in-vitro fertilisation (IVF) are a
considerable cause of frustration to both patients and clinicians and serve to
encourage the transfer of multiple embryos. However, in contemporary practice, the
transfer of multiple embryos is seen less as being the ‘solution’ to low implantation
rates, but rather more as the cause of the most important problems arising from IVF
treatment: that of multiple pregnancy (Fauser et al. 2005). The widespread adoption
of single embryo transfer is recognized to be necessary if the morbidity associated
with IVF multiple pregnancies is to be reduced. Already faced with frustratingly low
success rates following embryo transfer, IVF providers are therefore searching for
strategies and interventions designed to increase the chance of successful
implantation.
In recent years, clinicians have watched their laboratory colleagues make progress in
improving embryo quality and selection for transfer. Similarly, clinicians have sought
to develop clinical interventions aimed at enhancing implantation rates from IVF. The
competitive commercial context in which IVF is usually practised has often led to the
early adoption of promising new interventions before clear evidence for their efficacy
and safety has been established.
In this article, the various clinical strategies which may be employed by clinicians to
increase the chance of implantation are reviewed.
Embryo transfer technique

In recent years, the technique of embryo transfer (ET) has been shown to be an
important factor in determining the outcome of assisted reproductive technology
cycles. A number of studies have shown that significant improvements in clinical
pregnancy rates can be achieved by giving due attention to ET technique.
When difficulty in passing the catheter through the cervix is experienced, stiff
catheters make catheter placement easier, but may be associated with more bleeding
and trauma. Furthermore, cervical manipulation may result in an increase of
contractions of the uterus, which has been observed to hinder IVF outcome, possibly
by expelling embryos out of the uterine cavity (Fanchin et al. 1998). In a recent meta-
analysis of seven randomized controlled studies (RCTs) comparing stiff and soft ET
catheters, significantly increased pregnancy rates were observed with the latter, Odds
ratio (OR) 1.34, 95% confidence intervals (CI) 1.18–1.54 (Buckett et al. 2006). A
further systematic review on the effect of both the embryo transfer catheter type and
volume and constitution of the transfer medium on IVF outcome is in progress (Al-
Inany et al. 2006).
Traditionally, embryo transfer after IVF has been performed ‘blindly’, with the aim
of placing the embryos 1 cm below the fundus of the uterus (Schoolcraft et al. 2001).
However, it has been suggested that transferring embryos lower in the uterine cavity
may improve implantation rates. In a prospective investigation of effect of both the
distance from the fundus and the relative position in the uterus of the catheter tip,
significantly better results were obtained when the catheter tip was positioned close to
the middle of the endometrial cavity (Oliveira et al. 2004). In this study, the absolute
distance from the fundus appeared less important. However, another randomized
study revealed significantly higher implantation rates when embryos were deposited
1.5 or 2 cm from the fundus compared to 1 cm (Coroleu et al. 2002). In addition, a
retrospective cohort study showed increased pregnancy rates when the distance of the
fundus was increased, OR 1.11 (95% CI: 1.07–1.14); the authors suggesting that for
every additional millimetre embryos are deposited away from the fundus, the odds of
clinical pregnancy increased by 11% (Pope et al. 2004). Moreover, increasing the
distance to the fundus resulted in significantly lower ectopic pregnancy rates. As a
result of these and other similar studies, many centres have adjusted their ET
procedures. It has been recently postulated that embryo transfer in the lower uterine
segment may result in an increased risk of placenta praevia, since a six-fold higher
risk of placenta praevia in singleton pregnancies conceived by assisted fertilization
compared with naturally conceived pregnancies has been reported (Romundstad et al.
2006).
Chapter 2.1 | 27
The blind nature of traditional ‘clinical touch’ embryo transfer had led to the
suggestion of a role for ultrasound in improving IVF outcomes. Following initial
encouraging reports (Strickler et al. 1985), there have been numerous studies
evaluating the use of ultrasound-guided embryo transfer. A meta-analysis of four
RCTs comparing ultrasound guided embryo transfer versus clinical touch showed a
significant higher pregnancy rate and implantation rate after ultrasound guided
transfer (1.38, 95% CI 1.20-1.60) (Buckett et al. 2003).
During embryo transfer it is likely that bacteria from the cervix may be introduced
into the uterine cavity. Bacterial vaginosis (BV) is characterized by an overgrowth of
anaerobic organisms, the prevalence among women undergoing IVF is approximately
25% (Liversedge et al. 1999). There is growing evidence that the pathogenic effects
of bacterial vaginosis may not be confined to the lower genital tract. Almost half of
the patients with symptomatic BV showed hystopathological evidence of plasma cell
endometritis (Korn et al. 1995). There is no consensus in the literature whether BV is
associated with the success of embryo implantation. Salim et al. reported a
significantly higher pregnancy rate among women without cervical colonization in
the cervix versus women with bacterial colonization (30.7% versus 16.3%). However,
other studies have not shown this correlation (Liversedge et al. 1999; Gaudoin et al.
1999). At present, routine screening for bacterial vaginosis in the hope of improving
the success of IVF treatment is not justified.
Adjuvant pharmaceutical therapies

Drug treatments adjuvant to those required for ovarian hyperstimulation are
frequently applied in an empirical manner with the aim of improving embryo
implantation. One of the most attractive in terms of rationale, cost and presumed
safety is low dose aspirin.
Aspirin
The rationale for the use of aspirin as an adjuvant drug in IVF is based on its
vasodilatation and anticoagulant properties. Its main method of action is the
inhibition of cyclo-oxygenase (the rate-limiting enzyme in the prostaglandin synthesis
pathway) and subsequent reduction of platelet aggregation. The aim of therapy in the
context of IVF is to improve blood perfusion to the ovaries and the endometrium.
Kuo et al. reported a significant improvement in the uterine blood perfusion
(reduction of the pulsatility index of the uterine artery) in the peri-implantation period
after aspirin supplementation (Kuo et al. 1997).
Aspirin has been shown to be effective, either alone or in combination with heparin,
in the treatment of recurrent miscarriage in women with antiphospholipid antibody
syndrome (APS) (Empson et al. 2005; Rai et al. 2002). However, efficacy in
recurrent miscarriage in women without APS has not been proven (Di Nisio et al.
2005). In the context of IVF treatment, a RCT was performed comparing aspirin plus
heparin treatment from the time of embryo transfer with placebo in 143
antiphospholipid or antinuclear antibody-seropositive women with a previous history
of IVF implantation failure (Stern et al. 2003). No significant differences in
implantation or pregnancy rates were observed.
RCTs investigating the use of aspirin as an empirical therapy in non-selected IVF

populations have shown conflicting results (Table 1). In a study of 374 women,
randomised to receive either placebo or aspirin administration from ovarian
stimulation onwards, no differences in ovarian response and implantation rates were
observed (Pakkila et al. 2005). These results are in line with Urman et al. (Urman et
al. 2000). In contrast, significant improvements in ovarian response and implantation
rates were found by Rubinstein et al., who administered aspirin or a placebo from day
21 of the preceding menstrual cycle in 298 women (Rubinstein et al. 1999). A recent
large RCT of 1380 women reported a live birth rate of 27.2% following
administration of luteal phase aspirin versus 23.2 % in the placebo group. However,
when adjusted for the number of embryos transferred, the difference was not
statistically different (OR1.2, 95% CI 1.0-1.6) (Waldenstrom et al. 2004). Recently, a
further study investigating the efficacy of aspirin administered during the luteal
phase, also did not show significant differences, reporting a pregnancy rate of 29 %
versus 40 % in the placebo group (Duvan et al. 2006). A meta-analysis of 10 RCTs
showed no statistically significant improvement in clinical pregnancy rates with
aspirin versus no or placebo treatment (OR 1.18, 95% CI 0.86-1.61) (Daya et al.
Chapter 2.1 | 29
2006). At present there is insufficient evidence to support the use of aspirin outwith
the context of randomised controlled trials.
Table 1: Overview of RCTs investigating the use of aspirin as an empirical therapy in non-
selected IVF populations
Study n Dose Timing Pregnancy rate p-

aspirin (mg) Aspirin: Placebo: value
Rubinstein 298 100 From cycle day 21 (preceding 45 % 28 % <0.05
1999 menstrual cycle) onwards
Urman 279 80 From ovarian stimulation onwards 40 % 43 % NS
2000
Waldenström 1380 75 From ET onwards 35 % 30 % NS
2004
Pakilla 374 100 From ovarian stimulation onwards 25 % 27 % NS
2005
Duvan 100 100 From ET onwards 29 % 40 % NS
2006
NS: Non Significant.
Nitric oxide donors

Nitric oxide (NO) has been shown to be an important modulator of folliculogenesis
(Hefler et al. 2002), fertilisation (Kuo et al. 2000), decidualization and implantation
(Chwalisz et al. 2000). NO acts as a relaxant of arterial and smooth muscle and
inhibits platelet aggregation. High resistance to uterine blood flow on the day of
either human Chorionic Gonadotropin (hCG) administration or embryo transfer is
reported to be correlated with a poor clinical outcome for patients undergoing IVF
(Coulam et al. 1995). These observations suggested that uterine vasodilatation
induced by a NO donor during IVF might improve endometrial receptivity. However,
while initial studies suggested beneficial effects on ovarian response and implantation
(Battaglia et al. 1999), more recent studies suggested a detrimental effect of NO on
implantation (Ohl et al. 2002; Battaglia et al. 2002). Moreover, a recent prospective
study of women undergoing IVF for tubal or male factor infertility reported an
association between high follicular NO levels and advanced embryo fragmentation
and implantation failure (Lee et al. 2004). Although subgroups of women may be
identified who benefit from NO donor therapy, at present the available data demand
caution in its use, which at present should be restricted to well-designed studies.
Aromatase Inhibitors
Aromatase inhibitors block the conversion of androstenedione and testosterone to
estriol and estradiol, respectively (Cole et al. 1990). Rather than antagonizing
estrogen feedback activity at the hypothalamic-pituitary axis as with clomiphene
citrate, this approach reduces the amount of estrogens synthesised, thereby increasing
gonadotropin secretion and stimulating follicular growth (Mitwally et al. 2001). By
preventing excessive estradiol synthesis, it has been hypothesised that adjuvant
treatment with aromatase inhibitors during gonadotropin ovarian stimulation may
result in less disruption of endometrial receptivity. One RCT has been performed on
women with a poor ovarian response (defined by Goswami et al. as less than two
dominant follicles). A significantly lower total dose of follicle stimulating hormone
(FSH) was required for women using aromatase inhibitors, however, the pregnancy
rates were comparable (Goswami et al. 2004). These results are in line with three non
randomised trials investigating aromatase inhibitors as an adjunct treatment in normal
responders (Healey et al. 2003; Mitwally et al. 2003; Mitwally et al. 2004). Only a
subgroup of women with polycystic ovary syndrome (PCOS) showed significantly
higher pregnancy rates after addition of an aromatase inhibitor (Mitwally et al. 2004).
While of considerable potential value, further studies are required to confirm the
value and safety of aromatase inhibitors in IVF.
Ascorbic Acid
Ascorbic acid (AA) appears to be involved in normal folliculogenesis (Luck et al.
1995), ovulation (Igarashi et al. 1977) and luteal formation and regression (Luck et
al. 1993). An imbalance of oxidative stress and antioxidant defence has been
implicated in the pathogenesis of several diseases, including recurrent abortion,
unexplained infertility and defective embryo development. Transient high plasma
levels can be achieved by high dose intake of AA, which can exert anti-inflammatory
and immunostimulant effects. These effects might benefit embryo implantation.
However, a RCT investigating the effect of 1, 5 or 10 mg of AA versus a placebo
during the luteal phase in 620 women undergoing IVF showed no difference in
implantation rates (Griesinger et al. 2002).
Chapter 2.1 | 31
Prolonged progesterone
An important regulator of endometrium receptivity is the corpus luteum, the primary
function of which is the production of progesterone. Luteinizing Hormone (LH)-like
gonadotropins are the primary stimulatory factors of corpus luteum function: firstly
by means of the midcycle LH-surge, secondly by the pulsatile secretion of pituitary
LH during the luteal phase (Macklon et al. 2006). Following ovarian stimulation and
IVF, the luteal phase is abnormal compared to the natural cycle, with characteristic
features being elevated progesterone levels during the early luteal phase, followed by
a dramatic and premature fall in the unsupported mid-luteal phase (Jones et al. 1996)
(Figure 1). Following cessation of GnRH agonist treatment for the prevention of
premature luteinisation, pituitary recovery in the luteal phase takes around 14 days
(Macklon et al. 2006). The luteal phase can be restored by stimulating the corpora
lutea with hCG (luteal phase support) or supplementation with progesterone.
Figure 1: Serum progesterone levels during the luteal phase following ovarian stimulation
and IVF and normal cycles.
Schematic representation of changes in luteal phase length and endocrine profile induced by ovarian
hyperstimulation for IVF. [HW Jones Jr Hum Reprod 11(Suppl 1):7–24, 1996. © The European Society of
Human Reproduction and Embryology. Reproduced by permission of Oxford University Press/Human
Reproduction.]
The optimal duration of progesterone administration remains to be clarified. Many

centres continue with progesterone supplementation throughout the first trimester of
pregnancy. However, the rationale for this approach is unclear. Proponents point to
the uterine relaxing properties of progesterone, elegantly demonstrated by a reported
negative correlation between uterine contractility frequency and progesterone
concentrations (Fanchin et al. 1998). Moreover, Fanchin et al. reported a significantly
decreased uterine contraction frequency on the day of ET after vaginal progesterone
administration starting on the day of oocyte retrieval versus the day of embryo
transfer (Fanchin et al. 2001). Secondly, progesterone has been shown to have
potentially beneficial immunomodulatory properties. Studies in mice demonstrated
that progesterone administration abrogated the abortogenic effects of stress exposure
by decreasing the frequency of pro-inflammatory cytokines (Blois et al. 2004).
Previous studies suggested that successful pregnancy is more likely when anti-
inflammatory rather than pro-inflammatory cytokines are predominant (Wegmann et
al. 1993).
Although providing an appealing rationale for extended progesterone

supplementation, these studies need to be viewed within the context of the
endocrinology of early pregnancy after IVF. Studies of luteal function in the early
pregnancy among women pregnant after IVF reported markedly raised progesterone
levels for up to nine weeks of gestation compared with spontaneous pregnancies
(Costea et al. 2000) (Figure 2). It is perhaps therefore not surprising that a RCT
comparing five weeks versus two weeks of progesterone therapy (600mg tds per
vaginum) from oocyte retrieval onwards, showed no difference in outcomes (Nyboe
et al. 2002). Since the multiple corpora lutea resulting after IVF produce high levels
of progesterone in early pregnancy, extending therapy beyond the luteal phase is
probably superfluous.
Although supplementation of progesterone is widely used to improve implantation

rates, the application of luteal E2 supplementation remains controversial. A meta-
analysis of three RCTs (Farhi et al. 2000; Smitz et al. 1993; Lewin et al. 1994), using
a long GnRH agonist protocol, reported no difference in pregnancy rates when
estrogen was added to progesterone in the luteal phase (Pritts et al. 2002). Only Farhi
et al. investigated the effect of E2 supplementation on embryo implantation rates;
significantly higher implantation rates were observed in women using a long GnRH
agonist protocol (RR .49, 95% CI 1.02–2.19) (Farhi et al. 2000). On the contrary, a
Chapter 2.1 | 33
recent RCT of 166 women undergoing ICSI reported significantly higher pregnancy
and implantation rates after E2 supplementation (Lukaszuk et al. 2005). These data
require further substantiation, and the role of estradiol in supplementing the luteal
phase after IVF remains uncertain (Younis et al. 1994; Ghosh et al. 1994).
Figure 2: Serum progesterone levels in normal and IVF pregnancies during the first
trimester
Levels expressed as mean ± standard error.

• Normal pregnancy; ○ IVF pregnancy; *P<0.05. [DM Costea et al, Int J Gynaecol Obst 68:123-9, 2000.
Reproduced by permission of the International Journal of Gynaecology and Obstetrics]
Glucocorticoids
Uterine receptivity is controlled by locally-acting growth factors, cytokines and
uterine Natural Killer (uNK) cells (Dey et al. 2004). It has been shown that uNK cells
may have an important role in early implantation, since they accumulate around
arteries supplying the implantation site (Croy et al. 2002). Murine knock-out models
have been used to study the effect of NK cells on fertility and pregnancy outcome.
These mice achieved pregnancies, however, an increased fetal loss has been reported
(Guimond et al. 1998). This finding has not been substantiated by other studies
(Miyazaki et al. 2002). A defect in the integrity of the number of uNK cells has also
been implicated in implantation failure. Ledee et al. reported higher numbers of NK
cells in endometrial biopsies from women with implantation failure versus fertile
controls (Ledee-Bataille et al. 2005). In women with recurrent miscarriage
prednisolone has been shown to reduce the expression of uNK cells in the
endometrium (Quenby et al. 2005).
There is therefore evidence to support a possible role for glucocorticoids in

improving the intra-uterine environment by acting as immunomodulators. Studies
addressing the effect of glucocorticoid administration to women undergoing IVF have
been mostly small, and reveal contradictory results. A RCT of 206 patients,
investigating the use of glucocorticoids from oocyte retrieval onwards, reported no
differences in embryo implantation or pregnancy rates (Moffitt et al. 1995). These
results are in line with another RCT addressing the effect of adjuvant glucocorticoids
(Mottla et al. 1996). The effect of glucocorticoids on embryo implantation is
currently the subject of a Cochrane Systematic review (Boomsma et al. 2006). On the
basis of the data published thus far, and considering the potential deleterious effects
of prolonged glucocorticoid therapy on pregnancy with premature delivery (Empson
et al. 2002), there is insufficient evidence to support the empirical use of
glucocorticoids to improve implantation in IVF. In certain patient groups, such as
women with antiphospholipid syndrome, their may be a specific role for this therapy.
Two RCTs investigating implantation rates among women with positive anti-nuclear,
anti-double-stranded DNA, anti-cardiolipin antibodies and lupus anticoagulant,
reported significantly higher pregnancy rates after glucocorticoid administration
(Geva et al. 2000; Ando et al. 1996). Further research is necessary to clarify the role
of glucocorticoid therapy as an aid to implantation.
Insulin sensitizing drugs

The contention that insulin resistance may play a key role in the pathogenesis of
ovarian dysfunction in PCOS patients led to the application of insulin sensitizing
drugs to improve outcomes from ovarian stimulation in IVF. The most studied insulin
sensitizing drug in the treatment of anovulation is metformin, which has been shown
to be effective in achieving ovulation in women with PCOS (Lord et al. 2003).
However, no effect on implantation rates has been reported. A meta-analysis of eight
RCTs investigating metformin in women with PCOS demonstrated no significant
differences in pregnancy rates, although the risk of Ovarian Hyperstimulation
Syndrome (OHSS) was significantly reduced by metformin (Costello et al. 2006).
One recent RCT of 101 women with PCOS undergoing IVF, included in this meta-
Chapter 2.1 | 35
analysis, demonstrated lower rates of miscarriage and OHSS in the group receiving
metformin (Tang et al. 2006). Caution should be applied before insulin sensitizing
drugs are prescribed in the context of adjuvant treatment for IVF, since there is no
evidence supporting their use in a non-selected IVF population.
GnRH agonist
It has been postulated that the LH-releasing property of a GnRH agonist could be
exploited as luteal support in non-down regulated cycles. Pirard et al. randomised
patients to either hCG followed by progesterone versus different doses of an
intranasal GnRH agonist. Two study groups using less frequent administration were
prematurely discontinued, due to a short luteal phase. However, the trial has shown
that a regimen of three intranasal administrations per day could be at least as effective
as hCG followed by progesterone vaginally. Moreover, this regimen is compatible
with normal implantation and pregnancy (Pirard et al. 2006). On the contrary,
preliminary data from a RCT administrating progesterone and a GnRH agonist on the
day of ET and three days after ET versus no GnRH agonist did not show any
beneficial effect. A trend for higher hormonal levels at early stages of implantation is
observed among GnRH agonist treated women, which may interfere with embryonic
development rather than corpus luteum function (Hugues et al. 2006).
Ovarian stimulation regimens

The contemporary approach to ovarian stimulation in IVF treatment is based on the
perceived need to maximise the number of oocytes available for fertilization, in order
to generate multiple embryos for selection and transfer. In order to obtain multiple
oocytes during IVF treatment, ovaries are stimulated with exogenous FSH. Urinary
FSH has been widely replaced by recombinant FSH, which offers improved purity,
consistency and assured availability. However, its benefits in terms of improving
pregnancy rates appear limited. For a recent review see Macklon et al. (Macklon et
al. 2006).
Ovarian hyperstimulation and the resultant supra-physiological estradiol levels have

been shown to impact negatively on endometrial receptivity (Macklon et al. 2000;
Simon et al. 1995). This may be due to advanced postovulatory endometrial
maturation and defective induction of progesterone receptors (Devroey et al. 2004).

Elevated estrogen concentrations may increase sensitivity to progesterone action and
thus lead to secretory advancement. In one study of endometrial histology,
advancement on the day of oocyte retrieval exceeding three days was associated with
no subsequent pregnancies (Devroey et al. 2004; Ubaldi et al. 1997). Studies of the
impact of ovarian stimulation on endometrial maturation several days after ovulation
have shown either no effect or endometrial delay (Basir et al. 2001). However, the
magnitude of estrogen dose to which the endometrium is exposed, has been shown to
affect the duration of the receptive phase (Kolibianakis et al. 2002; Ma et al. 2003).
Increasing awareness of the possible detrimental effects of standard ovarian

stimulation regimens and their burdens on the patient, have stimulated a reassessment
of the optimal approach to ovarian stimulation for IVF (Macklon et al. 2000; Fauser
et al. 1999; Edwards et al. 1996). Increasing knowledge regarding the physiology of
ovarian follicle development together with the clinical availability of GnRH
antagonists, which allows ovarian stimulation to be commenced in the undisturbed
menstrual cycle, have presented the opportunity to develop novel, milder approaches
for ovarian stimulation for IVF (Macklon et al. 2006; Fauser et al. 2004).
In a RCT, it has been shown that the initiation of exogenous FSH (fixed dose, 150
IU/day, GnRH antagonist co-treatment) as late as cycle day 5 results in a comparable
clinical IVF outcome, despite a reduced duration of stimulation (and total dose of
FSH given) and increased cancellation rates (Hohmann et al. 2003). In this study, the
quality of embryos obtained after mild stimulation was significantly greater than
following the conventional long protocol. Moreover, almost all the pregnancies after
minimal stimulation were observed in patients with a relatively low oocyte yield,
whereas no pregnancies were observed when a similar yield was obtained after
conventional IVF.
Adjusting the dose of ovarian stimulation to the individual patient with the aim of
inducing a mild ovarian stimulation can be challenging, because the ovarian response
varies substantially between patients. Pharmaco-genetics has emerged as an attractive
tool to individualize FSH dosing. More clinical studies are warranted to investigate
the usefulness of genotyping as a routine diagnostic test before ovarian stimulation
Chapter 2.1 | 37
(Greb et al. 2005). Models have been developed which enable the response dose to
FSH to be determined on the basis of initial screening parameters. A model
developed by Popovic et al. included the number of antral follicles, ovarian volume,
age and smoking habits (Popovic-Todorovic et al. 2003). A RCT of an individualised
dose of FSH versus a standard dose of 150 IU/ day using this prediction model,
showed a higher ongoing pregnancy rate in the individual dose group (Popovic-
Todorovic et al. 2003).
The optimal number of growing follicles or oocytes retrieved in order to maximize

IVF outcome has been shown in a recent study to be around 13 (Gaast van der et al.
2006). However, these data were derived from conventional long protocol cycles, and
the number is likely to be lower following mild ovarian stimulation. The stimulation
of large numbers of follicles may result in the generation of poorer quality oocytes,
and hence lower quality embryos. In a recent randomized study preimplantation
genetic screening (PGS) was employed to investigate the chromosomal constitution
of embryos obtained after conventional ovarian stimulation compared to embryos
obtained after mild ovarian stimulation (Baart et al. 2006). Although more oocytes
were obtained in the conventional ovarian stimulation group, the mild group
demonstrated an increased percentage of euploid embryos per number of oocytes
retrieved. These observations support the concept of a ‘natural’ selection of oocytes
during follicular development which may be overridden by conventional ‘maximal’
stimulation protocols. Employing mild stimulation regimens may aid embryo
selection by increasing the chance that the transferred embryo is euploid.
Selecting the optimal embryo, endometrium and patient

Improved selection of both the embryo and the optimal uterine environment for
transfer remains the key to improving implantation. Our embryologist colleagues
have seen a number of developments in recent years which may be shown to improve
outcomes. Preimplantation Genetic Screening (PGS) for embryo aneuploidy is being
increasingly employed to aid the selection of embryos for transfer (Sermon et al.
2005; Verlinsky et al. 2004). Numerous other interventions have been developed in
current laboratory practice to improve implantation rates, for example assisted
hatching (Seif et al. 2005), the use of sequential media (Macklon et al. 2002), in vitro
maturation (Mikkelsen et al. 2005) and the use of embryo-endometrial co-culture

(Mercader et al. 2003) for selected patients.
Parallel to these techniques, preconceptional assessment of endometrial receptivity

offers the potential of correcting and optimizing receptivity prior to embryo transfer.
A reliable clinical test for endometrial receptivity remains elusive however. An ideal
molecular marker of receptivity would be obtained non-invasively and would be
highly specific and sensitive. While a number have been proposed such as integrins,
glycodelin and leukemia inhibitory factor (Thomas et al. 2003; Chryssikopoulos et al.
1996; Ledee-Bataille et al. 2002), no single marker has been identified that meets
these criteria. The complexity and molecular redundancy observed in human
implantation indicate the need to identify molecular profiles conductive to
implantation. In present clinical practice, assessment of endometrial appearance and
perfusion by ultrasound techniques are employed despite poor predictive values for
pregnancy (Ng et al. 2006).
Until high pregnancy rates can be widely achieved following single embryo transfer
(SET), clinicians will limit its application to ‘good prognosis’ patients. In recent years
a number of models have been published which allow the clinician to predict the
chance of conception and multiple pregnancy following the transfer of one or more
embryos. Based on a number of patient derived factors, such as age and response to
stimulation, and on parameters of embryo quality, it is possible to identify those
treatment cycles at particular risk of leading to multiple pregnancy. Identifying those
women for whom single ET would not reduce the chance of achieving a singleton
pregnancy, would prevent a reduction in implantation rates due to single ET (Hunault
et al. 2005). An increasing number of IVF centres are now transferring a single
embryo in women under 37 years, with at least one top quality embryo.
Conclusions
Improving embryo implantation continues to pose a major challenge to clinicians.
The growing trend towards transferring fewer embryos further increases the need to
improve implantation rates. Possible methods for a clinician to improve implantation
rates have been addressed in this review and summarised in Table 2. The technique of
Chapter 2.1 | 39
embryo transfer itself has been shown to be critical in assisted reproductive

technology cycles. Significant improvements in clinical pregnancy rates can be
achieved by giving due attention to the type of catheter used and embryo transfer
position. On the contrary, although increasingly adjuvant pharmaceutical
interventions are being applied with the aim of improving embryo implantation, they
have usually been of unproven benefit. The evidence for their efficacy and safety is
limited, therefore their use should be restricted to well-designed studies.
Increasing awareness of the possible detrimental effects of ovarian hyperstimulation

and their burdens on the patient, have stimulated development of milder ovarian
stimulation regimens. Standard ovarian hyperstimulation and the resultant
supraphysiological estradiol levels have been shown to impact negatively on
endometrial receptivity and embryo quality. Studies on mild ovarian stimulation
regimens have shown encouraging results. Although, fewer embryos are obtained, an
increased percentage of euploid embryos per number of oocytes retrieved has been
reported. In addition to causing less disruption of endometrial receptivity, mild
ovarian stimulation may also improve embryo quality.
Despite numerous new interventions to improve the success rate of ART, the role of
preconceptional care has been largely ignored. Increasing evidence suggests that
preconceptional interventions designed to optimize factors related to lifestyle and
nutrition may improve outcomes from fertility treatment. In a recent large
retrospective analysis, body mass index and smoking were both found to impact
significantly on ovarian response to IVF stimulation. Smoking was observed to have
a similar effect as an additional ten years of aging; as a result, smokers require
approximately twice as many IVF cycles to conceive as non-smokers (Lintsen et al.
2005). Moreover, recent studies have highlighted the importance of nutritional
factors. Folic acid supplementation was shown to alter the vitamin microenvironment
of the oocyte (Boxmeer et al. 2005), while seminal plasma cobalamin levels were
demonstrated to effect sperm concentration (Boxmeer et al. 2006). Furthermore, a
high intake of caffeine has been linked to an increase risk of spontaneous abortions
and lower live birth rates after IVF treatment (Tolstrup et al. 2003; Klonoff-Cohen et
al. 2002). It is likely that increased attention to preconceptional counselling, lifestyle
and nutritional factors in order to optimise reproductive health could contribute to

improving implantation rates and pregnancy outcomes after IVF.
Embryo implantation failure can be caused by multiple factors; as a result, no single

additional treatment is likely to be the key solution. Until our understanding of the
factors which determine the ability of an embryo to successfully implant are better
understood, it is unlikely that additional medical interventions, such as those
addressed in this article, will be shown to have anything but a marginal effect on IVF
outcomes.
Chapter 2.1 | 41
Table 2: Summary of possible methods for a clinician to improve implantation rates

addressed in this review.
What can the clinician do (Presumed) method of Empirical use in
to improve implantation? action non-selected IVF population
Careful embryo transfer Minimise trauma and cervical Soft ET catheter: significantly higher
Technique manipulation pregnancy rates.
Adjuvant pharmaceutical therapies
Aspirin Vasodilatation and anticoagulant No significant beneficial effect on
properties pregnancy rates in a non-selected IVF
population
Nitric oxide donors Uterine vasodilatation No significant beneficial effect on
pregnancy rates in a non-selected IVF
population. Possible detrimental
effects on embryo development.
Aromatase inhibitors Reduction estrogen synthesis: No significant beneficial effect on
improving endometrial receptivity pregnancy rates in a non-selected IVF
population
Ascorbic acid Exerts anti-inflammatory and No significant beneficial effect on
immunostimulant effects pregnancy rates in a non-selected IVF
population
Prolonged progesterone Luteal phase supplementation in a No significant beneficial effect on
down-regulated cycle pregnancy rates in a non-selected IVF
population
Luteal E2 supplementation Luteal phase supplementation in a Application of luteal E2
down-regulated cycle supplementation remains
controversial
Glucocorticoids Immunomodulatory effect, No significant beneficial effect on
reduction NK cell expression pregnancy rates in a non-selected IVF
population
Insulin sensitizing drugs Minimising insulin resistance No significant beneficial effect on
pregnancy rates, but it may reduce the
miscarriage and OHSS rate in women
with PCOS; empirical use in a non-
selected IVF population has not been
investigated.
GnRH agonist LH-releasing properties Can possibly be exploited as luteal
support: no beneficial effect on
implantation rates described
Ovarian stimulation
Mild stimulation regimens Improvement of endometrial Comparable IVF outcomes, despite
receptivity and increased fewer oocytes
percentage of euploid embryos,
despite less embryos
Prediction of optimal starting dose Optimal stimulation level Prediction models may have a role. In
FSH in the individual patient the future, pharmacogenetics is likely
to become important
Selecting optimal embryo, endometrium and patient
Preimplantation Genetic Selection of euploid embryos No RCTs in a non-selected IVF
Screening population have been published yet
Marker for endometrial Correcting and optimizing A reliable test remains elusive
receptivity receptivity prior to embryo
transfer
Selecting patients for SET Identification of patients in which No prospective analysis of prediction
SET does not reduce the chance of models has yet been performed.
achieving a singleton pregnancy
Chapter 2.2
____________________________________________
Peri-implantation glucocorticoid administration for

assisted reproductive technology cycles
Carolien M Boomsma1
Nick S Macklon1
1
Utrecht
(Curr Opin Obstet Gynecol. 2008; 20: 249-56)

Abstract
Purpose of the review
Failure of the embryo to implant remains the major limiting step in assisted
reproductive techniques success rates. There is evidence to support a possible role for
glucocorticoids in improving the intra-uterine environment and therefore embryo
implantation. The present review aims to summarise the available evidence and
recommendations on the use of glucocorticoids will be made.
Recent findings
A recent meta-analysis on glucocorticoids to improve embryo implantation showed
no beneficial effect in the routine IVF/ ICSI population, although a subgroup analysis
of women undergoing IVF did show a borderline significant improvement in
pregnancy rates. Studies on women with autoantibodies or treatment cycles in which
assisted hatching is performed have also indicated a benefit from glucocorticoid co-
treatment. However, no significant improvement in pregnancy rates have been
demonstrated in women with recurrent implantation failure in whom assisted
hatching is performed combined with glucocorticoids. Conflicting results have been
reported on the use of glucocorticoids to improve the ovarian response.
Summary
There is insufficient evidence to support the empirical use of glucocorticoids to
improve implantation, although it may be beneficial in specific patient groups. More
studies are required to confirm the efficacy and safety of adjuvant glucocorticoid
therapy, and they should only be empirically used in the context of randomized
controlled trials.
Chapter 2.2 | 45
Introduction
Failure of the embryo to implant is the major limiting step determining Assisted
Reproductive Techniques (ART) success rates. A number of strategies and
interventions aimed at increasing the chance of successful implantation, such as Pre-
implantation Genetic Screening, assisted hatching, co-culture with human
endometrial epithelial cells and adjuvant medical treatments have been proposed
(Boomsma et al. 2006). A role for glucocorticoids in improving the intra-uterine
environment and therefore embryo implantation rates has been suggested as this drug
has potent anti-inflammatory and immunosuppressive properties. The present review
aims to summarise the available evidence and to provide recommendations regarding
the use of this therapy.
Glucocorticoids in the general IVF population

Pregnancy represents a unique situation in which foreign fetal tissue is allowed to
implant rather than be rejected by the mother. Uterine receptivity to the implanting
embryo is controlled by numerous locally-acting growth factors, chemokines and
cytokines. It has been suggested that uterine Natural Killer (uNK) cells have an
important role in regulating the endometrial immune response to the embryo (Dey et
al. 2004). Their role in implantation is supported by the observation that they
accumulate around arteries supplying the implantation site (Croy et al. 2002). A
deviant number of uNK cells has been implicated in implantation failure. Higher
numbers of NK cells in endometrial biopsies were reported in women with
implantation failure versus fertile controls (Ledee-Bataille et al. 2005). In women
with recurrent miscarriage, glucocorticoids have been shown to reduce the number of
uNK cells in the endometrium (Quenby et al. 2005), suggesting a possible
mechanism of action of glucocorticoids.
Recently, a meta-analysis has been performed of randomized controlled trials (RCTs)

in which the efficacy of supplementary systemic administration of glucocorticoids in
the peri-implantation period was compared with a placebo or no glucocorticoids in
women undergoing IVF or intracytoplasmic sperm injection (ICSI) (Boomsma et al.
2007). The analysis was restricted to women with a standard IVF or ICSI indication.
Studies in men or women with auto-antibodies were excluded. Thirteen studies (Ando
et al. 1996; Bider et al. 1996a; Bider et al. 1996b; Botti et al. 1998; Catt et al. 1994;
Duvan et al. 2006; Ezzeldin et al. 2003; Kim et al. 1997; Moffitt et al. 1995; Mottla
et al. 1996; Tan et al. 1992; Ubaldi et al. 2002) were eligible for inclusion in the
meta-analysis, involving a total number of trial participants of 1759. Study
characteristics are summarized in Table 1.
Three studies reported the live birth rate per couple (Ando et al. 1996; Bider et al.
1996a; Moffitt et al. 1995). No significant difference was observed between the
intervention and control group (OR 1.21, 95% CI 0.67 to 2.19). The pregnancy rate
per couple was reported in all the included studies. No significant differences were
observed between the intervention and control group (OR 1.16, 95% CI 0.94 to 1.44)
(Figure 1). However, an analysis of studies which only included couples undergoing
IVF rather than ICSI (Ando et al. 1996; Bider et al. 1996a; Kemeter et al. 1986; Kim
et al. 1997; Moffitt et al. 1995; Mottla et al. 1996; Tan et al. 1992) revealed a
marginally significant higher pregnancy rates among those women receiving
glucocorticoid treatment (OR 1.50, 95% CI 1.05 to 2.13). Women undergoing ICSI or
subzonal sperm injection (Catt et al. 1994; Duvan et al. 2006; Ezzeldin et al. 2003;
Tan et al. 1992; Ubaldi et al. 2002) did not show benefit from glucocorticoid
administration (OR 1.00, 95% CI 0.74 to 1.36).
Further subgroup analyses specific for cause of infertility and dosage or timing of
glucocorticoids, did not reveal significant differences in pregnancy rates. Seven
studies reported the miscarriage rate per couple (Ando et al. 1996; Bider et al. 1996a;
Kemeter et al. 1986; Kim et al. 1997; Moffitt et al. 1995; Mottla et al. 1996; Ubaldi
et al. 2002). Again, no statistically significant effect on miscarriage was evident, OR
1.48, 95% CI 0.86 to 2.54. One study was undertaken to address the effectiveness of
glucocorticoids to prevent Ovarian Hyper Stimulation Syndrome (OHSS) (Tan et al.
1992). However, no significant difference in the incidence of this serious
complication was observed (OR 0.93, 95% CI 0.22 to 3.91).
When interpreting these results, a number of points should be considered. The study
of Ezzeldin et al. was most heavily weighted in the meta-analysis, due to its large
sample size of 526 participants (Ezzeldin et al. 2003). However, this was only
published in an abstract form and did not report the method of randomization and
Chapter 2.2 | 47
concealment of allocation. Only three studies were double blinded and placebo
controlled. Moreover, a funnel plot of precision versus odds ratio showed a
suggestion of publication bias for pregnancy rate per couple, suggesting that the
studies available for the meta-analysis are not fully representative. Publication bias
arises from the greater tendency for researchers and editors to publish experimental
results that are positive rather than negative or inconclusive.
Figure 1: Odds ratio for pregnancy rates in IVF and ICSI with adjuvant glucocorticoid
administration versus no glucocorticoids.
Study Glucocorticoids Control OR (fixed) Weight OR (fixed)

or sub-category n/N n/N 95% CI % 95% CI
01 Pregnancy rate after IVF

Kemeter 1986 16/73 6/73 9.19 3.13 [1.15, 8.54]
Moffitt 1995 42/103 37/103 42.98 1.23 [0.70, 2.16]
Ando 1996 12/23 16/35 11.90 1.30 [0.45, 3.72]
Bider 1996-1 9/54 4/24 9.05 1.00 [0.28, 3.63]
Mottla 1996 17/39 12/36 13.81 1.55 [0.60, 3.95]
Kim CH 1997 33/43 29/44 13.07 1.71 [0.66, 4.38]
Subtotal (95% CI) 335 315 100.00 1.50 [1.05, 2.13]
Total events: 129 (Glucocorticoids), 104 (Control)
Test for heterogeneity: Chi² = 3.09, df = 5 (P = 0.69), I² = 0%
Test for overall effect: Z = 2.24 (P = 0.02)
02 Pregnancy rate after ICSI

Tan 1992 7/17 5/14 3.95 1.26 [0.29, 5.42]
Catt 1994 8/56 6/55 6.35 1.36 [0.44, 4.22]
Ubaldi 2002 21/50 24/50 17.05 0.78 [0.36, 1.73]
Ezzeldin 2003 66/267 65/259 60.84 0.98 [0.66, 1.46]
Duvan 2006 19/50 14/40 11.81 1.14 [0.48, 2.70]
Subtotal (95% CI) 440 418 100.00 1.00 [0.74, 1.36]
Total events: 121 (Glucocorticoids), 114 (Control)
Test for heterogeneity: Chi² = 0.84, df = 4 (P = 0.93), I² = 0%
Test for overall effect: Z = 0.00 (P = 1.00)
0.1 0.2 0.5 1 2 5 10

Favours control Favours steroids
Adapted from Boomsma et al., Cochrane Database of Systematic Reviews 2007.
Table 1: Characteristics of included studies regarding peri-implantation glucocorticoid administration in ART.
Article Methods Participants ART Interventions Outcomes Notes
Ando Single centre 58 patients: 23 patients in group IVF. Glucocorticoid: LBR/ couple A: 24.1%, B: 24.6 %. This study also included a
et al. Trial design: A, 35 patients in group B. COH: FSH, GNRH analogue. HCG 5000 Prednisolone 5 mg or PR/ couple A: 52.2% B: 45.7 %. group of antibody positive
parallel RCT Cause infertility: variety. IU. Luteal support: progesterone 25 mg dexamethasone 0.5 mg AR/ couple A: 17,2 % B: 3,5 %. women, these results were
1996
IM for 7 days, 50 mg IM for 8 days. from COH onwards for IR/ ET A: 15.5% B: 14.6%. not included in the analysis.
4 weeks.
Bider Single centre 78 patients: 54 patients in group A IVF. Glucocorticoid: PR/ couple A1: 16.7%, A2: 18.5%,
et al. Trial design: (A1:27 receiving 0.5 mg, A2: 27 COH: 225 IU hMG, GnRH analogue. Dexamethasone 0.5 mg/ cumulative PR (A total) 16.7%,
parallel RCT receiving 1 mg), 24 patients group HCG 10 000 IU. Luteal support 1 mg during 5 days B: 16.7%.
1996a
B. Cause infertility: tubal factor. progesterone 50 mg vaginally twice a from COH onwards,
day. until ET+1.
Bider Single centre 99 patients. 52 patients group A, cryo-thaw cycle after IVF (with hMG Glucocorticoid: LBR/ couple A: 5.8% B: 8.5%. PR/ couple Data from Live Birth rate
et al. Trial design: 47 patients group B. and GnRH agonist). ET 48 h. after Dexamethasone 0.5 mg A: 13.5%, B: 12.8%. Ectopic PR/ couple A: are preliminary.
parallel RCT. Cause infertility: tubal factor. ovulation. during 5 days from 0%, B: 0%. AR/ couple A: 1.9%, B: 2.1%.
1996b
ovulation onwards. Multiple PR/ couple A: 0%, B: 0%.
Side effects steroids A: 0% B: 0%.
Botti Single centre 91 patients: 39 patients in group IVF and ICSI Glucocorticoid: PR/ couple A: 33.3% B: 32.7% Abstract.
et al. Trial design: A, 52 patients in group B. COH, luteal support: not stated. 16 b- IR/ ET A: 9.5% B: 15.2%
parallel RCT. Cause infertility: not stated. methylprednisolone 16
1998
mg/day during 4 days
from OPU-1 until ET+1.
Catt Single centre 111 patients: 56 patients group A, Subzonal sperm insertion (SUZI) Glucocorticoid: PR/ couple A: 14.3%, B: 10.9%
et al. Trial design: 55 patients group B. Cause COH, luteal support: not reported. ET 48 methylprednisolone 16 Multiple PR/ couple A: 3.6%, B: 0%
one cycle. infertility: male factor. h after OPU. mg/ day during 4 days IR/ ET A: 13.2%, B: 9.1%
1994
RCT. from OPU+1.
Duvan Single centre 90 patients: 50 patients group A, ICSI Glucocorticoid: PR/ couple A: 38%, B: 35% Not included were 2 study
et al. Trial design: 40 patients group B. COH: 225-450 IU/day FSH, GnRH Prednisolone 10 mg IR/ ET A: 13.3% B: 10.9% arms using aspirin, and
parallel RCT. Cause infertility: variety. analogue. HCG 10 000 IU. Luteal until clinical pregnancy Side effects steroids A: 0% B: 0%. prednisolone+aspirin.
2006
support vag progesterone 3x 200 mg/dy confirmed from ET
onwards.
Ezzeldin Single centre. 526 patients: 267 patients in ICSI. Glucocorticoid: PR/ couple A: 24.7%, B: 25.1%. Abstract.
et al. Study design: group A, 259 patients in group B. COH: HMG. No further details. Prednisolone 15 mg IR/ ET A: 9.46%, B: 10.2%.
parallel RCT Cause infertility: male factor. following ET during
2003
luteal phase.
Kemeter Single centre 146 patients: 73 patients in group IVF. Glucocorticoid: PR/ couple A: 21.9%, B: 8.2%.
et al. Trial design: A, 73 patients in group B. COH: Clomid+ hMG. HCG 5000 U. No Prednisolone 7.5 mg 4 AR/ couple A: 5.5%, B: 4.1%.
parallel RCT. Cause infertility: male/ tubal. GNRH anta/agonist. weeks from start COH. Ectopic PR/ couple A: 2.7%, B: 0%.
1986
Kim Single centre 87 patients: 43 patients in group IVF. Glucocorticoid: PR/ couple A: 76.7%, B: 65.9%. 2 patient populations
et al. Trial design: A, 44 patients in group B. COH: HMG+hFSH+GNRH agonist. prednisolone 10 mg Multiple PR/ couple A:14%, B:6.8% included in this study:
parallel RCT. Cause infertility: tubal. HCG 10 000 IU. ET 3 days after OPU. during COH, from OPU AR/ couple A: 9.3%, B: 6.8%. 1: 38% autoantibodies.
1997
Luteal support progesterone 50 mg. onwards 60 mg for 4 Ectopic PR/ couple A: 0%, B: 0%. 2:tubal factor. Only group 2
days. is included.
Moffit Multi centre IVF 206 patients: 103 patients in IVF or cryo-thawed embryo transfer Glucocorticoid: IVF: PR/ couple A: 42.7%, B: 40.8%. A cumulative pregnancy rate
et al. Trial design: group A, 103 patients in group B. cycle 6-alfa-methyl AR/ couple A: 10.7% B: 8.7%. IR/ ET A: for both groups has been
one cycle. Cryo-thawed embryo transfer COH: FSH + hMG, leuprolide acetate prednisolone 4 mg for 4 12.6% B: 13.0%. LBR/ couple A: 17.5% B: calculated.
1995
RCT cycle 61 patients: 28 patients in long/ short protocol. days from OPU/ day 11.7%. Cryo-thawed embryo transfer cycle:
group A, 33 patients in group B. 10 000 IU hCG. Luteal support: before thawing PR/ couple A: 25.0%, B: 30.3%.
Cause infertility: variety. progesterone 50 mg IM/day. onwards. AR/ couple A: 10.7% B: 6.1%.
ET: 2 days after OPU. IR/ ET A: 7.4% B: 9.9%.
LBR/ couple A: 0% B: 0%.
Mottla Single centre 75 patients. 39 patients group A, IVF. ET: 2 days after OPU. Glucocorticoids: PR/ couple A: 43.5% B: 32.3%. AR/ couple
et al. Trial design: 36 patients group B. COH: hMG + GnRH agonist. 1,2-dehydrocortisone 60 A: 12.8% B: 5.6%. Multiple PR/ couple
parallel RCT. Cause infertility: not stated. 10 000 IU hCG. Luteal support: mg/ day for 4 days from A:15.4%, B: 11.1%. Ectopic PR/couple
1996
Inclusion criteria: younger than 40 progesterone vaginal 600 mg/ day + 3 OPU onwards. A:2.6%, B:2.8%. IR/ ET A: 16.0%, B:
years. times hCG 3.300 IU IM. 11.0%. Side effects/ couple A: 0%, B: 0%.
Tan Single centre 31 patients. 17 patients in group ICSI. Glucocorticoid: PR/ couple A: 41.2%, B: 35.7%. Power calculation showed
et al. Trial design: A, 14 patients in group B. COH: hMG. GNRH agonist. HCG 10 Hydrocortisone 100 mg OHSS/ couple A: 41.2%, B: 42.9%. 15 women were needed in
parallel RCT. Inclusion criteria: high risk for 000 IU. The embryos were transferred 2 iv after OPU. IR/ ET A: 20.5%, B: 21.7%. each study group.
1992
OHSS (>20 follicles/ E2 > 10 days after oocyte retrieval. Luteal Prednisolone 30 mg for
000). Cause infertility: Tubal/ support: 100 mg gestone IM. 5 days, 20 mg 3 days,
unexplained/ male factor. 10 mg 2 days.
Ubaldi Single centre 100 patients: 50 patients group A, ICSI. Glucocorticoid: PR/ couple A: 42%, B: 48%. Ccross-over design.
et al. Trial design: 50 patients group B. Cause All patients: aspirin 100 mg/day from Prednisolone 10 mg for AR/ couple A: 4.7% B: 8.3%. Preliminary data from
cross-over infertility: male factor. COH onwards for 4 weeks. 4 weeks from COH IR/ ET A: 21.1% B: 18.5%. abstract (ESHRE 2000)
2002
RCT. Inclusion criteria: < 39 years, COH: 200-225 IU (r)FSH, GNRH onwards. included 1cycle/ patient.
normal hormonal profile and agonist. HCG 10000 IU. Luteal support: These data included in
uterine cavity, < 4 ICSI attempts. progesterone 50 mg/day IM. analysis.
AR: Abortion rate, COH: controlled ovarian hyperstimulation, ET: embryo transferred, FSH: follicle stimulating hormone, GnRH: gonadotrophin releasing hormone, hCG:
human chorionic gonadotrophin, hMG: human menopausal gonadotrophin, IM: intra muscular, IR: implantation rate, OHSS: ovarian hyperstimulation syndrome, OPU: ovum
pick up, PR: pregnancy rate. Group A: glucocorticoids. Group B: no treatment / placebo. Adapted from Boomsma et al., Cochrane Database of Systematic Reviews 2007.
The use of glucocorticoids in combination with aspirin has also been investigated. In
a RCT, not included in the meta-analysis, which investigated the use of
glucocorticoids and aspirin from oocyte retrieval onwards in women undergoing
ICSI, no significant differences in pregnancy rates were found (Mollo et al. 2002).
These results are in line with another RCT, also in women with an ICSI indication
(Duvan et al. 2006). However, significantly more miscarriages were reported in non-
treated women (6% versus 28%).
Overall there is no clear evidence to support empirical administration of

glucocorticoids during the peri-implantation period in subfertile women undergoing
IVF/ ICSI. Nevertheless, the use of glucocorticoids in women undergoing IVF (rather
than ICSI) was associated with an improvement in pregnancy rates of borderline
statistical significance.
Glucocorticoids for couples with autoantibodies

The presence of autoantibodies is associated with reproductive failure, including
implantation failure and repeated pregnancy loss (Stern et al. 1998). Endometriosis
and unexplained infertility are also associated with the presence of autoantibodies
(Reimand et al. 2001; Taylor et al. 1989). A retrospective analysis in women with
autoantibodies showed more than 50 percent lower pregnancy rates following IVF
treatment compared to women without autoantibodies (Dwowski et al. 1995;
Tanuguchi et al. 2005). It has been suggested that uteroplacental thrombosis and
vasoconstriction, resulting from binding of the antibodies to platelet and endothelial
membrane phospholipids, may underlie adverse reproductive outcomes in these
women (Lockwood et al. 1986). Corticosteroids are known to suppress
autoantibodies, and may therefore support successful embryo implantation in these
women.
A RCT investigated the effect of glucocorticoids in women with anti-nuclear, anti-

double-stranded DNA, anti-cardiolipin antibodies or lupus anticoagulant and in
women without autoantibodies (Ando et al. 1996). Corticosteroid treatment was
provided during the entire IVF cycle in a low dose (predisolone 5 mg daily).
Although no significant decrease in autoantibody titers was observed, pregnancy rates
Chapter 2.2 | 51
in women with autoantibodies were three fold increased and significantly higher after
the use of glucocorticoids. A non-significant tendency towards higher pregnancy rates
was observed in women without autoantibodies. Another RCT investigated the use of
glucocorticoids in women with endometriosis, of whom 38% were also seropositive
for non-organ specific autoantibodies (Kim et al. 1997). Significantly higher
pregnancy rates were reported in treated women. A beneficial effect from
glucocorticoids was also indicated in two retrospective analyses in autoantibody
seropositive women with implantation failure (Geva et al. 2000), and in an analysis of
women with antinuclear antibodies (Taniguchi et al. 2005). Another retrospective
analysis investigated the use of glucocorticoids in combination with aspirin in women
with autoantibodies, and reported almost three times higher pregnancy rates in treated
versus non-treated women (41% versus 15%) (Hasegawa et al. 1998).
Women with antithyroid antibodies do not appear to benefit from adjuvant

glucocorticoid treatment. In an abstract reporting results from a RCT in hypothyroid
women with antithyroid antibodies, no significant differences in success rates were
found (Shohayeb et al. 2005). Antiovarian antibodies (AOA) have also been
associated with reproductive failure (Gobert et al. 1990). AOA may interfere with
oocyte maturation and causes a decrease in fertilization rates (Narayanan et al. 1995).
A non-controlled study investigated the use of prednisolone (0.5 mg/kg per day) from
ovarian stimulation onwards, in women with implantation failure and AOA (Forges et
al. 2006). Post oocyte retrieval antiovarian IgG levels significantly decreased
following glucocorticoid treatment and high pregnancy rates were achieved (38.8%).
Another study investigated the incidence of AOA in 80 women with implantation
failure and compared subsequent pregnancy rates in seropositive women treated with
prednisone (10mg/day, starting one month prior to ovulation induction) and non-
treated seronegative women (Geva et al. 1999). AOA were found in 12.5% of women
with implantation failure versus no cases among normal fertile women. A significant
higher pregnancy rate was found in 9 treated seropositive women versus 70 non-
treated seronegative women, 33.3% and 8.6% respectively. A positive effect of
glucocorticoid treatment is plausible, although a randomized controlled study design
would have been preferential. This may not be solely an effect of improved embryo
implantation, as an improvement of embryo grading was noted.
Antisperm antibodies (ASA), either maternal or paternal, are associated with reduced
pregnancy rates. The prevalence of antisperm immunity in infertile couples has been
reported to be as high as 9-36% (Naz et al. 2004). When ASA are present in a female
partner, it may be treated by ART. However, sperm bound antibodies present in a
male partner are difficult to remove by sperm washing procedures.
Immunosuppressive therapies have yielded contradictory results. A retrospective
analysis of 3 months treatment with prednisone (15 mg/day) did not improve overall
pregnancy rates although those women who responded with a reduction in cytotoxic
antibody titers had improved pregnancy rates (Smarr et al. 1988). Two placebo
controlled RCTs have been performed. One study provided high dose corticosteroids
from cycle day 1-12 in the female partner and significantly higher pregnancy rates
were observed in treated couples (31% versus 9%) (Hendry et al. 1990). Another
study randomized men with ASA-bound on sperm to methylprednisolone or placebo
for three cycles. Despite a significant reduction in antibody titer, there was no
significant difference in pregnancy rates (Haas et al. 1987).
Glucocorticoids in zona pellucida dissected embryos

Five days after fertilization, the human blastocyst hatches from the zona pellucida.
Zona thinning during the first cleavage divisions appears to be an important factor
correlated with hatching and implantation (Cohen et al. 1989). Therefore, assisted
hatching has been proposed to improve implantation in ART. However,
micromanipulation of the zona may compromise its protective properties. The
incisions in the zona are small and it is unlikely that immune cells will invade the
embryo. However, these gaps may expand after embryo transfer due to considerable
growth of the embryo. Glucocorticoids can reduce the concentration of immune cells
in the vicinity of the embryo, and may therefore support implantation.
A four times higher and significantly increased implantation rate of zona dissected
embryos in women receiving glucocorticoids versus women who did not receive
immunosuppression has been reported in a non-randomized controlled trial (Cohen et
al. 1990). Another study randomized women with increased maternal age or FSH >
10 IU/l, implantation failure, or cryopreserved embryo transfers in cycles where
assisted hatching was performed to glucocorticoids and antibiotic treatment or a
Chapter 2.2 | 53
placebo (Primi et al. 2004). The results indicated that assisted hatching appears to
have a negative effect on pregnancy rates in women with advanced maternal age or
women undergoing cryocycles, unless supported by a combination of glucocorticoids
and antibiotic treatment. On the contrary, pregnancy rates in women with
implantation failure showed a non-significant increase in pregnancy rates after
assisted hatching and no benefit from glucocorticoids and antibiotics. In both studies
methylprednisolone (16 mg/day) and tetracycline in the peri-implantation period was
prescribed for 4 days (Cohen et al. 1990) or 7 days (Primi et al. 2004).
ICSI treatment also requires micromanipulation of the zona to aid the passage of
sperm to the oocyte. As described previously, pooling the results of studies
investigating the use of glucocorticoids in women undergoing ICSI or subzonal sperm
insertion rather than IVF did not reveal a significant difference in pregnancy rates
(OR 1.00, 95% CI 0.74 to 1.36) (Boomsma et al. 2007).
Glucocorticoids to improve ovarian response

Glucocorticoids have also been proposed as a means of increasing ovarian response
and oocyte quality in IVF. The rationale for their use in this context is based on a
number of possible mechanisms (Tan et al. 2004). The physiological follicular rise in
cortisol prior to ovulation may suggest that steroids exert a role in final oocyte
maturation (Harlow et al. 1997). Indeed, higher follicular cortisol:cortisone ratios
have been found in successful treatment cycles (Keay et al. 2002). However, not all
studies support the concept of cortisol playing a role in oocyte maturation. Follicles
from which a pregnancy was derived were found to contain similar amounts of both
free and total cortisol compared with follicles from which the oocyte failed to implant
or cleave in vitro (Andersen et al. 1994). Glucocorticoids may also reduce androgen
levels. Increased androgen concentrations, frequently observed in women with
polycystic ovary syndrome, are thought to be detrimental for oocyte and embryo
quality. Circulating adrenal androgens are converted to testosterone in the ovarian
follicles. Although glucocorticoids have been employed to suppress androgen levels,
serum and follicular levels of biologically active androgens were not significantly
reduced (Fridstrom et al. 1999). Therefore, the rationale for a positive effect of
glucocorticoids on ovarian stimulation may be questioned.
The administration of glucocorticoids to improve ovarian response has been tested in

ovulation induction studies either alone, in combination with clomiphene citrate or
recombinant Follicle Stimulating Hormone (rFSH), or in women undergoing ovarian
stimulation for IVF/ ICSI treatment. Different effects of adjuvant glucocorticoids on
the ovarian response have been observed. A RCT of women using glucocorticoids
during ovulation induction in combination with human Menopausal Gonadotropin
(hMG) (Bider et al. 1997), and two RCTs of women with higher levels of
dehydroepiandrosterone (DHEAS) or PCOS undergoing IVF (Fridstrom et al. 1999;
Rein et al. 1996), did not show a beneficial effect of adjuvant glucocorticoids. Other
studies have however suggested beneficial effects. Women undergoing intrauterine
insemination (IUI) with ovarian stimulation (Kim et al. 1996), did show improved
pregnancy rates when adjuvant glucocorticoids were provided. A Cochrane
Systematic Review on glucocorticoids and ovarian response is in progress (Tan et al.
2004).
Adverse effects of glucocorticoids

Glucocorticoid therapy, even in a low dose, is not without side effects. Complications
range from minor skin bruising to infections, peptic ulceration and impaired glycemic
control. The risk is clearly greater as the dose or duration increases. Steroid dose is
commonly characterised into low dose (< 10 mg/day prednisone or equivalent dose
other drug), medium dose (10-20 mg/day) or high dose (>20 mg/day). Most infertility
studies have prescribed low dosages (Ando et al. 1996; Bider et al. 1996a; Bider et
al. 1996b; Duvan et al. 2006; Kemeter et al. 1986; Moffitt et al. 1995; Ubaldi et al.
2002), or medium dosages (Botti et al. 1998; Catt et al. 1994; Ezzeldin et al. 2003;
Mottla et al. 1996). However, a number of studies employed high dosages of 30-60
mg prednisone per day (Kim et al. 1997; Tan et al. 1998). The low dose of
glucocorticoids and short duration of treatment minimises the risk of serious side
effects. No side effects were reported in any of the included studies in our meta-
analysis (Boomsma et al. 2007). At present, it is advisable for women with a history
of peptic ulceration or diabetes mellitus to avoid glucocorticoid treatment.
Chapter 2.2 | 55
Currently there is limited information regarding the toxicities and teratogenetic

potential of glucocorticoids during pregnancy. The most commonly used
glucocorticoids are prednisone, prednisolone, and methylprednisolone, and the longer
acting dexamethasone and betamethasone. Prednisone and prednisolone can cross the
placenta and appears in cord blood, although dexamethasone and betamethasone
reach higher concentrations in the fetus because they are less efficiently metabolized
by the placenta (Beitins et al. 1972). A systematic review of studies of women who
used glucocorticoids during pregnancy concluded that there was an increased risk of
cleft palate (Odds Ratio 3.4, 95% CI 1.97-5.69) (Park-Wyllie et al. 2000), which is in
line with animal studies (Pinsky et al. 1965). A RCT of prednisolone (0.5-0.8 mg/kg
per day) combined with aspirin for the duration of pregnancy in women with
recurrent pregnancy loss and autoantibodies, showed no effect in promoting live
birth. However, significantly more infants were born prematurely in the treatment
group than in the placebo group (62 % versus 12 %) and significant more mothers
had hypertension (13% versus 5%) and diabetes mellitus (15% versus 5%). A meta-
analysis on prednisolone and aspirin for women with recurrent pregnancy loss and
autoantibodies also showed a significant increase in prematurity (relative risk 4.83,
95% CI 2.85, 8.21) but no significant reduction in pregnancy loss (Empson et al.
2002).
Glucocorticoids during pregnancy have also been linked to a higher incidence of

growth-restricted fetuses and an intrauterine programming of cardiovascular,
metabolic and neuroendocrine disorders in adult life (Seck et al. 2004). A current
model for the pathogenesis of fetal growth restriction implicates too shallow
trophoblast invasion into the maternal decidua in the first trimester of pregnancy as
the underlying cause. An in-vivo study has indeed shown that synthetic
glucocorticoids may have the potential to adversely affect trophoblast development
and differentiation. Therefore, we recommend that care must be exercised when
women are given glucocorticoids in the peri-implantation period.
In summary, the use of glucocorticoids should not be continued during pregnancy,

since their use has been associated with pregnancy complications. When
administrated in the peri-implantation period, a low glucocorticoid dose, namely 5-15
mg/day of prednisone, is recommended.
Conclusions
On the basis of the data published thus far, there is insufficient evidence to support
the empirical use of glucocorticoids to improve implantation in IVF. The causes for
implantation failure are multifactorial and therefore striving for immunosuppression
in all IVF cycles is over-simplistic and may be detrimental. Until our understanding
of the process of implantation improves, it is unlikely that additional medical
interventions, such as glucocorticoids, will be shown to have anything but a marginal
effect on IVF outcomes. The present review indicated that glucocorticoids in the peri-
implantation period may be beneficial in women with auto-antibodies, and for some
indications when assisted hatching is performed. Outwith these indications, data to
support their use are not convincing. Given that the safety is also uncertain, we would
recommend that adjuvant glucocorticoids should only be used in these patient groups
or in the context of well designed randomized controlled studies (Table 2).
Table 2: Summary of recommendations for adjuvant glucocorticoid treatment
Evidence supports their use in:

- Women with non-organ specific autoantibodies
- In combination with assisted hatching (+ antibiotics)
 Women with advanced maternal age/ FSH > 10
 Cryopreserved embryo transfer cycles
Evidence does not support their use in:
- Routine use in standard IVF/ ICSI population
- Women undergoing IVF/ ICSI for:
 tubal factor infertility
 male factor infertility
- Women in order to improve ovarian response
- Women with antithyroid antibodies
Unknown. No evidence from RCTs or conflicting results reported in:
- Women with antiovarian antibodies
- Couples with antisperm antibodies
- Women with endometriosis
Chapter 3.1
____________________________________________
Cytokine profiling in endometrial secretions:

a non-invasive window on endometrial receptivity
Carolien M Boomsma1
Annemieke Kavelaars2
Marinus JC Eijkemans1,3
Karima Amarouchi2
Gijs Teklenburg1
Dagmar Gutknecht1
Bart CJM Fauser1
Cobi J Heijnen2
Nick S Macklon1
1
Utrecht
2
Laboratory of Psychoneuroimmunology, University Medical Center Utrecht
3
Department of Public Health, Erasmus Medical Center Rotterdam
(Reproductive BioMedicine Online 2009; 18: 85-94)

Abstract
Investigation of human embryo implantation requires non-disruptive means of
studying the endometrium during the window of implantation. In this study we
describe a novel approach of cytokine profiling in endometrial secretions.
Endometrial secretions aspirated prior to embryo transfer from 210 women
undergoing IVF or ICSI were analyzed by a multiplex immunoassay. Ten mediators
(Interleukin-1ß, IL-6, IL-12, IL-18, Tumor Necrosis Factor-α, Macrophage Migration
Inhibitory Factor, Eotaxin, Monocyte Chemotactic Protein-1, Interferon-γ Inducible
Protein-10, Vascular Endothelial Growth Factor) were detectable in 90-100% of the
samples. Heparin-binding Epidermal Growth Factor, IL-5, IL-17, IL-10, Dickkopf
homolog-1 and IL-15 were detected in 23-76%, whereas Interferon-γ was not
detectable in any of the samples. To assess possible contamination of samples,
cervical mucus was also aspirated for comparative analysis in 22 women. The
endometrial cytokine profile differed significantly from cervical mucus. Pregnancy
rates of the study participants who underwent endometrial secretion aspiration were
compared with 210 controls matched for important prognostic variables; no
significant differences were found. In conclusion, cytokine profiling in endometrial
secretion offers an objective, non-disruptive means of analysing the in-vivo milieu
encountered by the embryo and offers a new and potentially valuable approach to
studying the endometrial factor in human embryo implantation.
Chapter 3.1 | 59
Introduction
Despite improvements in clinical and laboratory in-vitro fertilization (IVF)
techniques, pregnancy rates remain around 30% per cycle (The European IVF
monitoring program, 2005). Failure of the embryo to implant now constitutes the
major limiting step in IVF treatment (Macklon et al. 2006). The growing trend to
transfer fewer embryos provides further incentives to improve implantation rates
(Fauser et al. 2005). Although the quality of the embryo is considered to be the
principle determinant of successful implantation, appropriate endometrial maturation
and receptivity are required if the molecular cross-talk between uterus and embryo is
to result in normal implantation.
Many approaches to assessing endometrial maturation and receptivity have been

described (Diedrich et al. 2007). Still the most widely used method remains that
introduced by Noyes et al., who defined maturation in terms of histological criteria
(Noyes et al. 1950). As a result of the subjective assessment of these features there
are difficulties in the interpretation and the significant inter-observer variation limits
its use in a clinical setting (Murray et al. 2004). In recent years, global gene
expression studies have characterised the endometrial transcriptome throughout the
menstrual cycle (Talbi et al. 2006; Kao et al. 2002; Borthwick et al. 2003; Riesewijk
et al. 2003; Horcajadas et al. 2004) and after ovarian stimulation (Horcajadas et al.
2005; Mirkin et al. 2004; Macklon et al. 2008). This approach has identified a
number of mediators likely to play a role in regulating endometrial receptivity, and in
the subsequent embryo-endometrium molecular cross-talk (for review Horcajadas et
al. 2007). However, there is not always a direct relation between gene expression and
actual protein levels due to post-transcriptional factors including rate of protein
synthesis and turnover. Pregnancy has been associated with a predominant anti-
inflammatory cytokine profile (Wegmann et al. 1993). However, the complex pattern
of cytokines at the feto-maternal interface observed is difficult to reconcile with the
pure pro- and anti-inflammatory paradigm and this impression is further strengthened
by the observation that certain pro-inflammatory cytokines are mandatory for
successful implantation (Chaouat et al. 2001). Immunocytochemical techniques have
demonstrated the sequential temporal expression of a number of mediators of
implantation in endometrial tissue, such as integrin B3 and glycodelin, which appear
during the putative window of implantation and have been proposed as markers of
endometrial receptivity (Lessey et al. 1998; Damario et al. 2001). However, obtaining
representative material for the study of human implantation and in-vivo embryo-
endometrium interaction remains a challenge (Diedrich et al. 2007), since an
endometrial tissue biopsy cannot be obtained during the implantation window without
disrupting the process of implantation (van der Gaast et al. 2003). Investigations in a
donor oocyte model offer a way around this problem (Damario et al. 2001), but are
limited by the lack of implantation as an endpoint, since the embryo is transferred
into the uterus of a different woman. Our understanding of human embryo
implantation may increase if a non-disruptive objective clinical test of endometrial
receptivity becomes available, which can be applied during the window of
implantation in conception cycles.
One possible approach is the analysis of endometrial secretions aspirated from the
uterine cavity during the window of implantation. Endometrial secretions represent
the intra-uterine environment with which the embryo interacts, and we have
previously shown that endometrial secretion aspiration can be carried out
immediately prior to embryo transfer without affecting implantation rates (van der
Gaast et al. 2003). Endometrial secretions analyzed by one-dimensional gel
electrophoresis (1D PAGE) demonstrate protein profiles which change with the phase
of the cycle (Beier-Hellwig et al. 1994). However, the resolution of this means of
analysis is limited, and this technique does not provide quantitative data on specific
markers of endometrial receptivity. The complexity and inherent molecular
redundancy which characterizes human implantation suggest that a single molecular
marker of the receptive endometrium is unlikely to be identified (Tranguch et al.
2005; Hoozemans et al. 2004; Laird et al. 2006). Rather, it is more likely that a pro-
implantation profile of markers will be characterized which could ultimately form the
basis of a test of endometrial receptivity.
The aim of the present study was to establish whether a profile of mediators involved
in implantation and endometrial maturation can be quantified in endometrial
secretions aspirated prior to embryo transfer. Secondary aims of the study were to
compare the cytokine profile in cervical mucus and endometrial secretions, in order to
assess potential contamination of samples by cervical mucus during aspiration. In
addition, we wished to investigate how blood contamination of the endometrial
Chapter 3.1 | 61
aspirate might alter the concentration of mediator measured by the multiplex

immunoassay. Although our group has previously reported endometrial secretion
aspiration not to be disruptive to implantation (van der Gaast et al. 2003), a further
aim of this study was to confirm that pregnancy rates are not reduced in women
undergoing this procedure prior to embryo transfer in a larger cohort of patients.
Materials and Methods

Ethical approval was received from the local institutional Ethics Committee and
informed consent was obtained from all participants between December 2005 and
October 2006. Two hundred and eleven patients with maximally one prior embryo
transfer, attending the University Medical Center Utrecht (The Netherlands) for IVF/
intra cytoplasmic sperm injection (ICSI) treatment consented to participate in this
prospective cohort study. No additional exclusion criteria were applied.
Assisted Reproductive Technique

Ovarian stimulation was applied with recombinant FSH (rFSH) (100-300 IU/day),
starting on cycle day 2 or 3 and co-treatment with Gonadotrophin Releasing Hormone
(GnRH) antagonist (0.25 mg/day) in a fixed protocol from stimulation day 5 in IVF
cycles. All ICSI cycles were performed using a long GnRH agonist suppression
protocol, following down-regulation rFSH was administrated concurrently. Final
oocyte maturation was triggered with Human Chorionic Gonadotrophin (hCG: 5000-
10.000 IU). A maximum of 2 embryos were transferred 4 days after oocyte retrieval.
Women under the age of 36 years were only allowed to have double embryo transfer
when embryo quality was suboptimal. Luteal phase support was provided by
intravaginal Progestan (600 mg daily) or three doses of hCG (5000 IU).
Endometrial secretion aspiration procedure

With the patient lying in lithotomy position, the cervix was cleansed after insertion of
the speculum. A 2 ml syringe was connected to an embryo transfer catheter (Wallace:
SIMS Portex Ltd., Hythe, Kent, UK). The empty catheter was gently introduced 6 cm
transcervically into the uterine cavity. Suction was gradually applied. In order to
prevent contamination by cervical mucus during catheter removal, the outer sheath of
the embryo transfer catheter was advanced to a depth of 4 cm from the external
cervical os, following the application of suction. The inner catheter was then
withdrawn through the outer sheath, which prevented contact with the cervix. The
outside of the catheter was cleaned to remove any potential cervical mucus. The tip of
the catheter was cut off and snap frozen in liquid nitrogen in an Eppendorf tube and
stored at -80°C. The routine embryo transfer procedure was carried out thereafter. In
25 patients, cervical mucus was aspirated prior to endometrial secretion aspiration for
within patient comparison, in order to verify whether the aspirates represented
cervical mucus rather than endometrial secretions or were significantly contaminated
by cervical mucus.
Sample analysis
The endometrial secretion and cervical mucus samples were analysed using a
multiplex immunoassay. Key soluble regulators of implantation were identified as
candidate mediators for inclusion in the assay. A number of mediators were excluded
from the panel, either because appropriate antibodies were not available (Glycodelin),
or because of problems arising from cross-interference (IL-11, Leukemia Inhibitory
Factor [LIF] and Macrophage Colony Stimulating factor [M-CSF]). The final panel
therefore included IL-1ß, IL-5, IL-6, IL-10, IL-12, IL-15, IL-17, IL-18, TNF-α, IFN-
γ, MIF, Eotaxin, IP-10, MCP-1, Dkk-1, Hb-EGF and VEGF (Table 1). The antibodies
were purchased from commercial sources, as outlined by de Jager et al. (2003) or
otherwise from R&D Systems (R&D systems, Abingdon, United Kingdom).
Phosphate-buffered saline (PBS) was added to recover the samples from the frozen tip
catheter. Aspirate volumes could not be reliably measured, since the aspirates are
highly viscous, but were between 1 and 4 microliter. Therefore, total protein content
of the aspirate was measured by NanoOrange Protein Quantitation Kit (Molecular
probes Europe B.V., Leiden, The Netherlands) and was used for normalization
purposes. The samples were then centrifuged at maximum speed (13000g) for 1
minute. 50 µL of each sample was used for the multiplex immunoassay. Antibodies
were covalently coupled to the microspheres as described previously (de Jager et al.
2003). Assays were performed in a 96 well filter plate at room temperature and
protected from light. A mixture containing 500 microspheres per mediator (total
volume 10 µL/well) was incubated together with 50 µL of a standard calibration
sample, a study secretion sample or a ‘blank’ sample for 20 minutes under continuous
shaking. Beads were washed twice with PBS supplemented with 1% bovine serum
Chapter 3.1 | 63
albumin (BSA) and 0.5% Tween 20. After 10 minutes of incubation with streptavidin
R-phycoerythrin (50 ng/well; BD Biosciences, San Diego, California, USA) and
washing with PBS 1%, BSA 0.5% and Tween 20, the fluorescence intensity of the
beads was measured using the Bio-Plex array reader and Bio-Plex Manager software.
Five parametric curve fitting was used for data analysis. To control for interfering
factors, we spiked three endometrial secretion aspirate samples with known
concentrations of the soluble mediators tested. Recovery for all mediators was
between 90 and 110%. It has been shown previously that the intra-assay variance of
the multiplex immunoassay was less than 10% whereas inter-assay variance ranged
between 6-16% (de Jager et al. 2003).
Since concentrations of MIF in most samples exceeded the maximal detectable

concentration of the multiplex assay, MIF was determined in diluted samples using a
commercially available enzyme linked immunosorbent assay (ELISA) (R&D
systems, Abingdon, United Kingdom).
Statistical analysis
Since the data were not normally distributed, they were log transformed for statistical
analysis. For analysis of concentration values below the detection limit of the assay,
we used the lower detection limit divided by the square root of 2, as recommended
for Log transformed data by Schisterman et al. (Schisterman et al. 2006). A
probability (p) less than 0.05 was considered statistically significant. In case of
multiple testing a Bonferroni correction was applied. Analyses were performed using
SPSS version 12.0 (SPSS Inc., Chicago, IL, USA, 1999).
The extent of blood contamination in the endometrial secretion samples was visually
graded as none (0), minimal (1), moderate (2) or severe (3). Linear regression
analysis on Log transformed data of the soluble mediators was performed to
investigate the potential confounding effect of blood contamination on results. Within
patient comparisons of cervical mucus and endometrial secretion aspirations were
performed using the paired samples t-test for continuous data.
In order to further assess whether endometrial secretion aspiration prior to embryo

transfer reduces IVF/ ICSI pregnancy rates, the pregnancy rate of the 210 study
participants who underwent endometrial secretion aspiration was compared with the
pregnancy rate in a cohort of 210 controls, matched for type of treatment (IVF/ ICSI),
primary or secondary infertility, age, number of oocytes collected, and number of
previous attempts. The control patients underwent embryo transfer during the same
period as study patients. Clinical and treatment parameters were compared between
the study and control group by Student’s paired t- test and Mc Nemar test. A power
calculation indicated that with 210 subjects in the study we could detect a reduction
in pregnancy rate from 30 to 18% with 80% power at a significance level of p<0.05.
Table 1: Characteristics of soluble mediators included in the multiplex immunoassay in relation to implantation.
Brief description of role in Data from mouse knock-out Correlations to human fertility References
implantation studies status
IL-1ß Induction of integrin expression at Conflicting results. Impaired implantation. Significantly higher levels in women with (Simon, 1994; Simon, 1997;
endometrial epithelial surface. Others found reduction in growth or litter recurrent implantation failure. Stewart, 1997; Horai, 1998;
Stimulation of MMPs. size. Inagaki, 2003)
IL-5 Differentiation, recruitment and No overt reproductive defects. Reduction No data available. (Robertson, 2000a; Piccinni,
activation of eosinophils, which have number uterine eosinophils and 2000)
a role in endometrial tissue fetal:placental weight ratio, moderate
remodelling. increase placental size and birth weight.
IL-6 Immunomodulatory properties. Conflicting results. Significantly lower levels in women with (Nishino, 1990; Poli, 1994;
Activation MMPs. Contribution to Decreased litter size and reduced fertility. recurrent implantation failure. No Stewart and Cullinan, 1997;
trophoblast growth and placental Others found no overt reproductive defects. difference in women with unexplained Robertson, 2000b; Wolff von,
development. infertility versus fertile controls. 2002; Sherwin, 2002)
IL-10 Inhibits proliferation and secretion No overt reproductive defects. Significantly lower expression by (Chaouat, 1995; Szereday, 1997;
of Th1 leucocytes. peripheral lymphocytes in women with White, 2004)
recurrent miscarriage. Lower peripheral
blood levels in infertile women.
IL-12 Vascular remodelling. No data available Required in low levels for successful (Croy, 1997; Barber and Pollard,
uNK cells activation. implantation. Peripheral blood levels 2003; Ledee-Bataille, 2004a;
increased in women with recurrent Ledee-Bataille, 2005)
miscarriage.
IL-15 Growth factor and activator of uNK Lack of uNK cells. Vascular abnormalities No data available. (Croy, 1997; Barber and Pollard,
cells. and reduced fetal weight. Implantation not 2003; Ledee-Bataille, 2005)
Vascular remodelling. affected.
IL-17 Angiogenesis and immunoregulation. No data available Significantly higher in peritoneal fluid in (Zhang X, 2005; Beklen, 2007;
women with endometriosis and infertility. Pongcharoen, 2007)
IL-18 Vascular remodelling. Preterm birth and pregnancy loss in Required in low levels for successful (Croy, 1997; Barber and Pollard,
Activation uNK cells. response to intrauterine inflammation implantation. However, significantly lower 2003; Ledee-Bataille, 2004a+b;
Induction of production MMPs. enhanced. levels of IL-18 observed in women who Ledee-Bataille, 2005; Wang,
Abortifacient in high doses. No overt reproductive defects described. achieved pregnancy. Peripheral blood 2006)
levels decreased in women with recurrent
miscarriage.
TNFα Transformation uNK-cells into LAK No overt reproductive defects. Significantly higher levels in women with (Stewart and Cullinan, 1997;
cells. Inhibitor of embryo division recurrent implantation failure. Increased Wuu, 1999; Dey, 2004)
and attachment. production by peripheral blood
mononuclear cells in women with
recurrent miscarriage.
IFNγ Activator of NK cells. Vascular Vascular anomalies. No effect on Increased production by peripheral blood (Hill, 1995; Chaouat, 2007)
remodeling. Inhibition trophoblast implantation. mononuclear cells in women with
cell proliferation. Abortifacient in recurrent miscarriage.
high doses.
MIF Immunomodulatory and angiogenic No data available. Spermiogenesis was Significantly higher in women with (Kats, 2005; Akoum, 2006;
properties. Upregulation MMPs. disrupted in the male MIF knock out mice. endometriosis. Anahara, 2006)
Role in cell proliferation and
differentiation.
Eotaxin Chemoattractant to eosinophils, No data available. Significantly higher in pelvic fluid in (Hornung, 2000; Dimitriadis,
(CCL11) which have a role in endometrial women with endometriosis. 2005)
tissue remodelling.
MCP-1 Attractant and activator of immune No data available. No data available. (Caballero-Campo, 2002;
(CCL1) cells. Dimitriadis, 2005)
Dkk-1 Wnt signalling protein inhibitor. Wnt-7-/- : infertility and abnormal uterine No data available. Upregulation in gene (Parr and McMahon, 1998; Miller
development. expression studies during window of and Sassoon, 1998; Riesewijk,
Wnt 5 -/-: no cervix, no uterine glands. implantation. 2003; Mericskay, 2004)
IP-10 Attractant and activator of immune Neutralizing antibodies to IP-10 reduced No data available. (Ledee-Bataille, 2005)
(CCL10) cells. Stimulator of integrin ability to induce trophoblast migration.
expression by trophoblast.
Hb- Mediation embryo attachment. Mutant mice die early, precludes No data available. (Imai, 1995; Threadgill, 1995;
EGF Growth factor for IVF derived examination of implantation. Early Martin, 1998; Wang, 2000;
embryos. implantation seems unaffected in Lessey, 2002; Iwamoto, 2003;
blastocysts deficient in ErbB1 or ErbB4. Reis, 2005)
VEGF Influences vascularization, Embryonic death in utero during No data available. (Gargett, 1999; Smith, 2001)
angiogenesis and vascular midgestation with aberrant blood vessel
permeability. Endothelial cell formation.
mitogen.
IL, interleukin; TNF α, tumor necrosis factor α; IFN γ, interferon γ; MIF, macrophage migration inhibitory factor; MCP-1, monocyte chemotactic protein 1; IP10, interferon-
gamma-inducible 10 kD protein; DKK-1, dickkopf homolog 1; Hb-EGF, heparin-binding epidermal growth factor; VEGF, vascular endothelial growth factor; MMPs, matrix
metalloproteinases; RIF, recurrent implantation failure; uNK cells, uterine Natural Killer cells; LAK cells, lymphokine-activated killer cells; KO, knock out.
Chapter 3.1 | 67
Results
Endometrial secretion aspiration
Two hundred and ten patients were included in the analysis, since data from one
patient were excluded because insufficient material was available to measure the total
protein content required for normalization purposes. All other samples contained
sufficient material for analyses. No discomfort or side effects were reported by any of
the patients. The treatment outcome of our cohort of 210 women was compared with a
cohort of 210 matched controls. Endometrial secretion aspiration prior to embryo
transfer did not have a detectable effect on pregnancy and ongoing pregnancy rates
(Table 2).
Table 2: Patient characteristics and treatment outcome in women who underwent endometrial
secretion aspiration versus a matched control group
Treatment results Study Matched control p-

participants group value
N = 210 n = 210
Age (yrs)a 34.9 ± 4.1 35.0 ± 3.9 0.5 b
Number of oocytes collected a 8.1 ± 4.7 8.0 ± 4.4 0.6 b
Number of embryos available a 3.7 ± 2.4 3.6 ± 2.4 0.8 b
Number of embryos transferred (%) 0.3 c
Single ET 110 (52.4) 110 (52.4)
Double ET 100 (47.6) 100 (47.6)
IVF/ ICSI (%) 1.0 c
IVF 124 (59) 124 (59)
ICSI 86 (41) 86 (41)
Primary or secondary infertility (%) 0.6 c
Primary 124 (59) 122 (58)
Secondary 86 (41) 88 (42)
Pregnancy rate / embryo transfer (%) 68 (32.4) 62 (29.5) 0.6 c

Ongoing pregnancy rate (%) 51 (24.3) 50 (23.8) 1.0 c
a
Mean ± SD/ (range) b Paired samples t-test. c McNemar test.
Data on concentrations of the soluble mediators are expressed as pg per mg of total

protein in endometrial secretion samples and are presented in Table 3. Ten mediators
(IL-1ß, IL-6, IL-12, IL-18, TNF-α, MIF, Eotaxin, MCP-1, IP-10, and VEGF) of the 17
analyzed were detectable in 90-100% of the samples. Hb-EGF was detectable in 29%,
IL-5 in 37%, IL-17 in 56%, IL-10 in 60%, Dkk-1 in 69% and IL-15 in 78% of the
samples. IFN-γ was not detectable in any of the samples. In endometrial aspirate
samples spiked with a known amount of IFN-γ, the observed concentrations matched
those expected. It was therefore unlikely that IFN-γ was not detectable due to an
artefact.
Table 3: Soluble mediator concentrations in endometrial secretion aspirations prior to embryo

transfer in 210 women undergoing IVF/ ICSI.
Mediator Endometrial Below

secretion detection
(n=210) limit (%)
pg /mg protein
IL-1ß 6.25 (3.34; 14.4) 1.4
0.29-1679
IL-5 6.22 (3.00; 17.6) 63
0.62-1610
IL-6 30.8 (12.6; 68.9) 3.3
0.46 – 1220
IL-10 4.13 (2.03; 10.4) 40
0.16-84.8
IL-12 125 (56.3;285) 0
9.92-3880
IL-15 4.17 (2.22; 7.73) 22
0.20-150
IL-17 4.63 (1.96; 10.7) 44
0.20-548
IL-18 62.4 (35.7; 119) 0.5
0.88-2479
TNF-α 20.8 (10.0 ; 47.0) 5.2
1.62-458
MIF 71303 (44867; 113617) 1.0
18.9-2262000
Eotaxin 22.0 (9.18; 53.6) 7.6
(CCL11) 0.93-697
MCP-1 57.1 (27.0;116) 0.5
(CCL2) 1.15-12788
Dkk-1 728.7 (368;1394) 31
38.1-5735
IP10 101 (30.8;369) 0.5
(CCL10) 0.19-3934
Hb-EGF 2.22 (0.89;5.47) 71
0.15-73.8
VEGF 30.4 (16.0;53.8) 0.5
1.09-1541
Values represent median, 25-75 percentiles between brackets, and range.

Data include values below detection limit (detection limit/ √2).
Chapter 3.1 | 69
Nature of the endometrial secretion samples

All samples contained a volume of mucous fluid which was too small to be reliably
quantified. The relative protein content in the diluted samples varied from 40-3150 µg
per mL (median 535, 25-75 percentiles 280-943 µg per mL). Seventy four percent of
samples showed no visual signs of blood contamination. The other samples were
considered to be severely (3%), moderately (4%) or minimally (19%) contaminated.
Linear regression analysis revealed that blood contamination was a significant
confounder for IL-1 ß (p= 0.000), IL-12 (p= 0.004), IL-17 (p= 0.007), IL-18 (p=
0.008), TNF-α (p= 0.000) and Eotaxin (p= 0.001). After Bonferroni correction for
multiple testing, a probability p<0.0031 was considered statistically significant. The
results of IL-1ß, TNF-α and Eotaxin remained significantly affected by blood
contamination of the sample after this correction. Blood contamination was associated
with a lower observed concentration of these mediators. Protein measurements were
not influenced by blood contamination. In order to investigate the reason for the
altered measurements in blood contaminated samples, samples with a given amount of
mediators were spiked with increasing amounts of blood, representing minimally,
moderately or severely contaminated samples. The data showed that blood
contamination interfered with the measurement of IL-1β, IL-12, IL-17, IL-18, MCP-1
and Eotaxin. In blank control samples with severe blood contamination, all
concentrations of mediators were below the detection limit. A χ² test was performed to
asses whether blood contamination of the sample was related to the chance of
achieving pregnancy. No effect of blood contamination was observed (p=0.5).
Within patient comparison of endometrial aspirates and cervical mucus

The aspiration technique employed was designed to avoid sample contamination by
cervical mucus during passage of the catheter. To assess whether contamination was
avoided, cervical mucus and subsequently endometrial secretions were aspirated in 25
patients for within patient comparisons. Three patients were excluded from analysis
due to blood contamination of the endometrial secretions. The cervical mucus samples
showed no visual signs of blood contamination. The concentrations of the soluble
mediators in cervical mucus and endometrial secretion are presented in Table 4. Figure
1 illustrates the relative differences in mediator concentrations of significantly
differentially expressed mediators. The concentrations of IL-1ß, IL-6, IL-10, IL-18,
VEGF and MCP-1 were all significantly higher in cervical mucus, whereas the
concentrations of IL-12, IL-15, MIF and DKK-1 were significantly lower in cervical
mucus than in endometrial aspirates (p<0.05). In order to verify these results we
excluded all values which were below the detection limit of the assay. After exclusion
of these data, the results remained similar with the exception of IL-10, which no
longer demonstrated a significant difference in concentration between cervical mucus
and endometrial secretion. After Bonferroni correction for multiple testing a
probability p<0.0031 was considered statistically significant. The concentrations of
IL-6, IL-15, IL-18, MIF, MCP-1, Dkk-1, and VEGF remained significantly different
in endometrial secretion compared to cervical mucus after this correction.
Figure 1: Ratios of the concentrations of soluble mediators significantly differently expressed

(p<0.05) in endometrial secretion compared to cervical mucus.
32
16
8 4.67
5.20
4 3.00
2.43
Fold difference in concentration
2
1
0.5
0.25 0.24 0.29
0.32
0.125
0.11
0.064
0.031
0.016 0.03
0.008
0.004
0.002
IL1 IL6 IL12 IL15 IL18 MIF MCP1 DKK1 VEGF
Box: median, 25-75% interquartile range (IQR). Whiskers 10-90% IQR. Number: mean.
Chapter 3.1 | 71
Table 4: Within patient comparison of soluble mediator concentrations in cervical mucus and
endometrial secretion aspirations of 22 women undergoing IVF/ ICSI treatment.
Endometrial secretion Cervical mucus p-

(n=22) (n=22) value*
pg /mg protein pg /mg protein
IL-1ß 6.63 22.6 0.006
(2.46; 15.5) (7.53;108)
0.76-63.8 2.08-359
IL-5 3.61 1.85 0.078
(1.83; 18.8) (1.01; 7.01)
0.66-185 0.50-24.3
IL-6 5.03 409 0.000
(2.31; 18.4) (179; 1222)
0.22-489 0.18-2418
IL-10 2.46 7.87 0.012†
(0.84; 6.39) (4.42; 20.4)
0.16-41.7 0.69-115
IL-12 131 55.5 0.024
(54.8;511) (26.9;96.1)
8.57-2512 12.8-1497
IL-15 4.67 1.28 0.002
(3.49; 9.19) (0.88;2.51)
0.20-25.1 0.32-7.86
IL-17 10.1 8.17 0.5
(4.69; 31.8) (5.10-15.3)
0.89-96.0 1.09-126
IL-18 73.8 234 0.001
(34.1; 133) (132; 419)
18.8-434 25.0-3968
TNF-α 15.8 12.1 0.3
(9.38; 71.7) (7.25; 32.7)
3.53-415 1.75-2030
MIF 66609 9729 0.000
(42464; 86030) (7458;30740)
9439-2262000 323-437010
Eotaxin 19.1 7.81 0.17
(CCL11) (13.9;110) (1.69;38.4)
3.11-400 0.15-2571
MCP-1 52.6 129 0.003
(CCL2) (16.1;109) (82.4;289)
3.22-1356 32.1-964
Dkk1 1057 176 0.000
(523;1907) (81.0;522)
50.7-4011 16.3-789
IP10 77.1 150 0.2
(CCL10) (18.1;252) (35.3;493)
0.19-1538 3.80-1359
Hb-EGF 1.61 1.79 0.5
(0.84;4.51) (0.84;5.05)
0.19-8.46 0.28-68.3
VEGF 49.5 545 0.001
(22.4 ; 94.9) (277; 1473)
1.29-404 0.12-3698
Values represent median, 25-75 percentiles between brackets, and range. *Two-tailed paired t-test on Log
transformed data. † These results could not be confirmed after excluding all cases with values under the
detection limit.
Discussion
As far as we are aware, this is the first time that the profile of 17 soluble cytokines,
chemokines and growth factors in endometrial secretions at the time of embryo
transfer have been characterized. The present study confirms earlier observations by
our group (van der Gaast et al. 2003), which indicate that this technique can be safely
carried out prior to embryo transfer during IVF/ ICSI treatment, and may therefore
offer a clinically useful approach to study the endometrial factor in embryo
implantation. Endometrial secretion aspiration immediately prior to embryo transfer is
well tolerated, easy to perform, and the aspiration provides sufficient material for
analysis in 99.5% of cases.
The present study revealed significant differences in the cytokine and chemokine
profile in cervical mucus compared with endometrial secretions, which implies that the
aspirate is indeed endometrial secretion and does not represent cervical mucus or a
significant contamination during the procedure. The observed differences also make it
unlikely that the differences represent an artefact due to differing total protein
concentrations, which were observed to be generally higher in cervical mucus samples
compared to endometrial aspirates.
The mean and range of concentrations of individual mediators in endometrial

secretions showed considerable differences. The cellular origin of some of the
measured mediators is not known. A number of sources may contribute, such as
epithelial cells, immune cells infiltrating the endometrium and stromal cells. Mediators
may passively diffuse from the stromal compartment towards to glandular lumen, or
may be actively transported. However, shedding from the membrane may also occur.
The function of Hb-EGF in implantation has been primarily described for its trans-
membrane form (Chobotova et al. 2002). Whether the soluble Hb-EGF we measured
in our samples constitutes the shedded membrane-bound form or represents a secreted
soluble form is unknown. IFN-γ was not detectable in any of the samples in spite of a
comparable detection limit of the assay. This lower concentration may reflect the
origin of IFN-γ, which is produced by immune cells only (Schroder et al. 2004). In
contrast, the other mediators are also secreted by endometrial stromal or epithelial
cells. The observation that concentrations of certain mediators were below the
Chapter 3.1 | 73
detection limit of the assay may reflect the small sample volumes obtained, although
the dispersion in data is also broad.
An alternative means of obtaining endometrial secretions for analysis during the

window of implantation is by flushing the endometrial cavity (Oliviennes et al. 2003;
Li et al. 1993; Berkkanoglu et al. 2006). However, this approach results in a higher
and variable degree of dilution of endometrial secretions, since 1-2 mL saline water is
instilled and the amount of fluid recovered differs. Therefore, this may account for the
inconsistent findings reported (Laird et al. 1997; Lédée et al. 2002; Olivennes et al.
2003). A further non-invasive approach proposed is the analysis of endometrial
markers of receptivity in peripheral serum samples. However, since the expression of
cytokines differs in different compartments of the body due to the strong influence of
the local micro-environment, it is unlikely that this approach would generate data
representative of the intra-uterine milieu (Hoozemans et al. 2004).
Although the analysis of aspirated endometrial secretions constitutes a potentially

valuable means of assessing endometrial receptivity, certain limitations need to be
addressed. In the present study, one quarter of aspirates showed visual signs of blood
contamination, which was shown to affect the results of a number of mediators. The
reason for the altered measurements was investigated and blood contamination can
indeed affect the results of measurements of IL-1β, IL-12, IL-17, IL-18, MCP-1 and
Eotaxin. Since no mediators were detectable in blank samples with severe blood
contamination, the altered cytokine concentrations measured in contaminated samples
were most likely due to interference of laser readings of the microspheres in the
cytokine assay by haemoglobin, rather than cytokines present in blood. Therefore,
blood contamination should be included as a confounder when analysing the results of
cytokine measurements. A second problem was that concentrations of certain
mediators were frequently below the reliable detection limit, which may complicate
statistical analyses.
Despite these caveats, the approach we describe offers an objective, non-invasive

technique which allows the characterization of the in-vivo milieu encountered by the
embryo. Multiplex immunoassays enable the simultaneous determination of a number
of mediators in small samples, addressing the molecular redundancy present in
embryo-endometrial interactions. Further studies are required to assess how the

intrauterine cytokine milieu varies through the cycle and in particular the luteal phase,
to measure intercycle variability of cytokine expression and, most importantly, to
ascertain whether it is possible to identify a cytokine profile predictive of
implantation. In conclusion, endometrial secretion analysis may open a new ‘window’
on endometrial receptivity, implantation and the factors which modulate this complex
and elusive process.
Chapter 3.2
____________________________________________
Analysis of the endometrial secretion cytokine profile to

predict pregnancy in in-vitro fertilization
Carolien M Boomsma1
Eef G Lentjes4
Bart CJM Fauser1
Cobi J Heijnen2
Nick S Macklon1
1
Department of Reproductive Medicine and Gynaecology, University Medical Center Utrecht
2
3
4
Laboratory of Endocrinology, University Medical Center Utrecht
(Human Reproduction 2009; 1: 1-9)

Abstract
Background
The study of human endometrial-embryonic interactions is complicated by the
disruptive impact of endometrial sample collection on the process of implantation
itself. Endometrial secretion analysis is a novel technique, non-disruptive to
implantation. The primary aim of this prospective cohort study was to explore whether
a cytokine profile predictive of implantation and clinical pregnancy can be identified
in endometrial secretions aspirated immediately prior to embryo transfer following in-
vitro fertilization (IVF).
Methods
Endometrial secretions, aspirated prior to embryo transfer from 210 women
undergoing IVF, were analyzed using a multiplex immunoassay for 17 soluble
regulators of implantation, namely interleukin (IL)-1ß, IL-5, IL-6, IL-10, IL-12, IL-15,
IL-17, IL-18, Tumor Necrosis Factor (TNF)-α, Interferon-γ, Macrophage migration
Inhibitory Factor (MIF), Eotaxin, Interferon-γ inducible 10 kD protein (IP-10),
monocyte chemo attractant protein-1 (MCP-1), Dickkopf homolog 1 (Dkk-1),
Heparin-binding Epidermal Growth Factor (Hb-EGF) and Vascular Endothelial
Growth Factor (VEGF). In order to detect implantation, daily urine samples were
collected after embryo transfer, and hCG concentrations were analyzed by an
immunoassay.
Results
Multivariable logistic regression analysis revealed significant associations between
MCP-1 (p = 0.005), IP-10 (p = 0.037) levels and implantation and IL-1β (p = 0.047),
TNF-α (p = 0.023) levels and clinical pregnancy. The predictive value for pregnancy
of IL-1β and TNF-α was observed to be equivalent and additive to that of embryo
quality.
Conclusions
Endometrial secretion cytokine profiling offers a novel, non-disruptive approach to
study the role of the endometrium in human embryo implantation, and identifies a
profile conducive to clinical pregnancy.
Chapter 3.2 | 77
Introduction
Embryo implantation failure remains the major rate limiting step in in-vitro
fertilization (IVF) success. Successful implantation and pregnancy require a vital
embryo and an effective molecular dialogue with a ‘receptive’ endometrium (Wang
and Dey, 2006). However, what precisely constitutes a receptive human endometrium
remains poorly defined. The human endometrium appears to be receptive to
implantation for a limited time period during the mid-luteal phase of the normo-
ovulatory cycle (Ma et al. 2003). The duration of this putative ‘window of
implantation’ is primarily determined by sex steroids (Dey et al. 2004), which regulate
the expression of locally acting growth factors, transcription factors, cytokines and
chemokines (Wang and Dey, 2006; Dey et al. 2004). In recent years, our
understanding of the molecular pathways regulating implantation has been increased
by the identification of genes whose expression is cyclically regulated in the
endometrium (Talbi et al. 2006). Studies in mice have identified a number of key
regulators, which appear to be crucial for normal implantation (Stewart et al. 1992).
However, up to now, no single specific factor has been identified to be crucial for
implantation in humans (Strowitzki et al. 2006). The continuing high rate of
implantation failure observed in assisted reproduction technologies has driven the
search for clinical markers of endometrial receptivity. Many candidate markers have
been proposed, including integrins, glycodelin and Leukaemia Inhibitory Factor (LIF)
(Lessey et al. 1995; Chryssikopoulos et al. 1996; Ledee-Bataille et al. 2002).
However, none have as yet been shown to be valuable clinically.
While global gene array and histochemical studies are increasing our knowledge of the
molecular regulation of the receptive endometrium, unravelling the molecular
pathways of human implantation is hampered by the requirement of appropriately
timed endometrial biopsies, which themselves disrupt the necessary endpoint of
implantation (van der Gaast et al. 2003). Recently, our group has developed a novel
means of assessing the intra-uterine milieu by endometrial secretion analysis. We have
shown this approach to be non-disruptive to implantation and representative of the in-
vivo milieu encountered by the embryo (van der Gaast et al. 2003; Boomsma et al.
2009). Uterine glands produce a glycoprotein rich secretion, which is believed to
support the embryo during the pre-implantation period (Burton et al. 2002). In
addition to providing nutrition for the conceptus, it contains a complex array of growth
factors and cytokines. Modern multiplex immunoassays enable multiple markers to be
quantified simultaneously in a small volume (de Jager et al. 2003). Recently we have
reported the application of a multiplex immunoassay to determine multiple cytokine,
chemokine, growth and signalling factors in a single endometrial secretion sample
(Boomsma et al. 2009). A brief description of the roles in implantation of these
mediators is provided in Table 1 in chapter 3.1.
Increasing evidence points to pre-clinical pregnancy loss rather than failure of

implantation as a principal cause for the relatively low fecundity observed in humans.
In natural cycles up to 60 percent of conceptions are estimated to be lost due to pre-
clinical pregnancy loss (Macklon et al. 2002). It is uncertain to what extent failure of
IVF treatment can be attributed to early rejection of the implanted embryo. The
primary aim of this study was to explore whether a profile of these markers, identified
in endometrial secretions aspirated immediately prior to embryo transfer following
IVF/ intra cytoplasmic sperm injection (ICSI) treatment, is predictive of pregnancy.
However, since not all embryo implantations proceed into a clinical pregnancy, the
early implantation of embryos was also assessed. The secondary aim of the study was
to discern the contribution of early rejection of the implanting embryo to the failure to
achieve pregnancy after IVF.

Ethical approval was obtained from the local institutional Committee and informed
consent was obtained from all participants. Between December 2005 and October
2006, we recruited 211 women attending the University Medical Center Utrecht (The
Netherlands) for IVF/ ICSI treatment to participate in this prospective cohort study.
None had undergone more than one prior unsuccessful embryo transfer. No additional
exclusion criteria were applied.
As described previously in detail (Boomsma et al. 2009), ovarian stimulation to obtain

multiple oocytes was carried out by daily injections of recombinant follicle stimulating
hormone (FSH) (100-300 IU/day) from cycle day two or three of a spontaneous cycle
Chapter 3.2 | 79
onwards. To prevent premature luteinization, co-treatment with a Gonadotropin

Releasing Hormone (GnRH) antagonist (0.25 mg/day) was provided from stimulation
day five in IVF cycles. No oral contraceptives for cycle regulation were provided in
GnRH antagonist cycles. The majority of ICSI cycles were performed using a long
GnRH agonist suppression protocol from cycle day 21 of the previous cycle for
logistic considerations. Final oocyte maturation was triggered with a single
subcutaneous injection with human Chorionic Gonadotropin (hCG) (Pregnyl: 10.000
IU). A maximum of two embryos were transferred four days after oocyte retrieval.
Women under the age of 36 years were restricted to single embryo transfer if a high
quality embryo was available, which was defined as a compacted morula stage embryo
four days after oocyte retrieval with 0-10% fragmentation. Following embryo transfer,
the luteal phase was supported with three injections of hCG in five days (Pregnyl;
5000 IU), or by daily vaginal progesterone treatment (Progestan: 600 mg/day).
Endometrial secretion aspiration procedure

We aspirated endometrial secretions immediately prior to embryo transfer, as
described previously (Boomsma et al. 2009). Briefly, an embryo transfer catheter
(Wallace: SIMS Portex Ltd., Hythe, Kent, UK) was introduced transcervically.
Suction was gradually applied with a 2 mL syringe. Contamination by cervical mucus
was prevented as described previously (Boomsma et al. 2009). The aspirate was snap-
frozen in liquid nitrogen and stored at -80°C.
Sample analysis
Determination of the cytokine profile
The procedure employed to analyse the endometrial secretions has been previously
described in detail (Boomsma et al. 2009). We analysed endometrial secretion samples
using a multiplex immunoassay detecting 17 soluble mediators. Mediators were
eligible for inclusion in the panel when appropriate antibodies were available and no
cross-interference occurred. It has been shown previously that the intra-assay variance
of the multiplex immunoassay was less than 10% whereas inter-assay variance ranged
between 6-16% (de Jager et al. 2003). Since concentrations of MIF in most samples
exceeded the maximal detectable concentration of the multiplex assay, MIF was
determined in diluted samples using a commercially available ELISA (R&D systems,
Abingdon, United Kingdom). The volumes of the aspirates could not be exactly
measured, since their viscous quality and small volumes (around 3 µL) did not allow
an accurate measurement. Therefore, the total protein concentration was measured by
NanoOrange Protein Quantitation Kit (Molecular probes Europe B.V., Leiden, The
Netherlands) for normalization purposes. To recover samples from the frozen catheter
tip, 70 µL PBS was added, 50 µL of each sample was used for the multiplex
immunoassay. The lower detection limits of the multiplex immunoassay were as
follows, IL-β (0.6 pg/mL), IL-5 (2.4 pg/mL), IL-6 (1.2 pg/mL), IL-10 (1.2 pg/mL), IL-
12 (4.9 pg/mL), IL-15 (1.2 pg/mL), IL-17 (2.4 pg/mL), IL-18 (1.2 pg/mL), TNF-α (2.4
pg/mL), IFN-γ (2.4 pg/mL), Hb-EGF (1.2 pg/mL), VEGF (1.2 pg/mL), IP-10 (1.2
pg/mL), MCP-1 (1.2 pg/mL), DKK-1 (78 pg/mL), Eotaxin (2.4 pg/mL) and MIF was
performed by a separate ELISA (32 pg/mL).
Detection of implantation
In order to detect an embryo implantation, all participants were asked to freeze a urine
sample from day 9 until day 18 after oocyte retrieval. Following collection and
thawing, we dephosphorylated urine samples with calcium chloride and incubated
0.5mL of this sample overnight in a test unit. After fluid aspiration from the units, the
test units were analysed in the Immulite 1000® immunoanalyser (Siemens). The
functional sensitivity of the assay was 0.06 mU/mL with 20% coefficient of variation.
The day-to-day variation at three hCG concentration levels was: 0.14 mU/mL ±6.0%,
1.2 mU/mL ±3.9% and 4.9 mU/mL ±3.6%. The assay was calibrated against the
NIBSC hCG standard 99/688. In order to correct for variations in urinary
concentration, hCG levels were corrected for creatinine concentrations.
Since subjects received a standard dose of exogenous hCG to trigger final oocyte
maturation 36 hours prior to oocyte retrieval, it was necessary to differentiate between
endogenous hCG produced by an implanting conceptus and residual hCG derived
from exogenous administration. Therefore we generated hCG wash-out curves by
analysis of urine samples collected from nine-15 days after oocyte retrieval by 10
oocyte donors who had received a similar dose of hCG (10.000 IU) (Figure 1). We
calculated a cut off value from the wash-out curves by adding two standard deviations
by the mean concentration hCG on each day. Implantation was considered to have
occurred when urinary hCG levels were more than two standard deviations above the
mean wash-out curve value on two consecutive days. Women who received hCG
Chapter 3.2 | 81
rather than progesterone for luteal support were excluded from this analysis. A urinary
pregnancy test was performed 18 days after oocyte retrieval. When a rise in urinary
hCG levels indicative of implantation was followed by a negative pregnancy test, a
pre-clinical pregnancy loss was considered to have occurred. A clinically recognized
pregnancy was defined by a positive urinary pregnancy test. In case this resulted in a
menstruation shortly after, this was described as a biochemical pregnancy loss. A
miscarriage was defined as a loss of a histologically or ultrasonographic recognized
pregnancy in the first trimester. An ongoing pregnancy was diagnosed when fetal heart
activity was present on ultrasound at nine weeks of gestation.
Figure 1: Wash-out curve of exogenous hCG in oocyte donors following administration of

10.000 IU hCG (Pregnyl) 36 hours prior to oocyte retrieval.
Data analysis
We log transformed non-normally distributed data prior to analysis. Primary analyses
were performed using SPSS version 12.0 (SPSS Inc., Chicago, IL, USA). Comparisons
of characteristics and IVF outcomes between pregnant and non-pregnant women were
performed using the t-test for continuous data and the χ2-test for binary variables
unless stated otherwise. A probability (p) less than 0.05 was considered statistically
significant. As previously reported, 26% of samples showed minimal (19%), moderate
(4%) or severe (3%) signs of blood contamination, which has been shown to affect the
measurement of a number of cytokines (Boomsma et al. 2009). The extent of blood
contamination in the endometrial secretion samples was visually graded on a four
point scale (no, minimal, moderate, severe) and included as a covariate in the analyses.
Blood contamination was not associated with the chance of conceiving (Boomsma et
al. 2009).
Multivariate analyses were applied since mutual correlations and molecular

redundancy between the mediators is present. Concentrations of mediators were below
the detection limit of the assay in a number of cases. Given that multivariate analyses
only includes cases with data from all included variables available, it was necessary to
impute these missing data on the basis of the available measurements of the other
cytokines in the panel. Data were imputed by WinBUGS (MRC Biostatistics Unit,
Cambridge, UK), which enables to set a condition to the imputed data, since imputed
data needed to be lower than the detection limit of each assay. Mediators with more
than 50% of values missing were excluded from analyses, due to a certain level of
uncertainty of imputed data (Harrell, 2001).
Multivariable logistic regression analysis was performed with a backward elimination

procedure in S-Plus (Insightful Corporation) to identify factors predictive of embryo
implantation or pregnancy. A p-value > 0.1 was used as a criterion for exclusion
according to literature on multivariate prognostic modeling (Steyerberg et al. 2000).
Since this study exploratory in nature, we have applied no correction for multiple
testing as would be required for a confirmatory hypothesis based study. Thereafter, the
following covariates were included in a forward stepwise model: maternal age,
embryo quality, and blood contamination of the sample. Only covariates which
significantly contributed to the model were included in the final analysis. The ability
Chapter 3.2 | 83
of this model of included factors and covariates to predict pregnancy was assessed by
determining the area under the receiver operating characteristics (ROC) curve. In order
to achieve a maximal and more reliable predictive ability (Steyerberg et al. 2000), a
ROC curve was also constructed for a panel of mediators that were selected by
multivariable regression analysis with backward elimination with a relaxed criterion
for exclusion, namely a p-value > 0.3 rather than 0.1.
Results
Of the two hundred and forty eight women who consented to participate, 210 women
underwent endometrial secretion aspiration prior to embryo transfer and were subject
to analysis (Figure 2). Sixty eight women (32.4%) achieved a clinically recognized
pregnancy.
Figure 2. Flow-chart outlining the reasons for exclusion of women from analysis.
Univariate analysis comparing clinical parameters between pregnant and non pregnant
women showed pregnancy to be associated with a significantly higher number and
quality of embryos available for transfer (Table 1). 183 women collected urine
samples after embryo transfer for hCG measurements to detect an implantation event.
In four women hCG analysis revealed atypical patterns which did not allow
discrimination as to whether a pre-clinical pregnancy loss had occurred or not and

these data were excluded from further analysis. Figure 3 illustrates the outcome of the
IVF/ ICSI treatment and the incidence of pre-clinical pregnancy loss in 179 women.
51.4% of embryo transfers resulted in embryo implantation as detected by a rise in the
urinary hCG level. Of these implantations, 33.7% resulted in pre-clinical pregnancy
loss, whereas the other implantations resulted in a clinically recognized pregnancy.
Table 1: Clinical parameters of pregnant and non-pregnant women after embryo transfer.
Parameter Pregnant Non- p-

pregnant value
(n= 68) (n=142 )
Age (yrs) 34.2 ± 4.0 35.3 ± 4.2 0.077 a
(27-43) (26-44)
Primary type of infertility (%) 0.6 b
Andrological 51 49
Unknown 18 22
Tubapathology 20 20
Endometriosis 3 1
Maternal age (>40 yrs) 1 3
Ovulation induction 7 3
Cervical 0 2
Smoking (%) 14.1 15.5 0.79 b
IVF (%) 60 61 0.13 b
ICSI (%) 40 39
GnRH antagonist (%) 52 52 0.9 a
GnRH agonist (%) 48 48
Oocytes retrieved (n) 8.5 ± 5.2 7.9 ± 4.5 0.3 a
Embryos transferred (%) 0.8 b
Single ET 52 50
Double ET 48 50
Number of embryos available for 4.2 ± 2.4 3.4 ± 2.3 0.008 c
transfer (n)
High quality embryo transferred D (%) 50 28 0.002 b
Luteal phase support (%) 0.3 b
hCG 9 13
Progestan 91 87
All values are mean ± SD. Range between brackets.

ET: Embryo transfer. COH: controlled ovarian hyperstimulation.
a
Independent t-test. b χ² test. c Mann-Whitney U test. d Defined as a compacted morula stage embryo with 0-10%
fragmentation, four days after oocyte retrieval.
Chapter 3.2 | 85
Figure 3: Treatment outcome after embryo transfer (left bar) and subsequent implantation
(right bar).
This figure shows the outcome of embryo transfer (left bar) and pregnancy (right bar) in women who collected
urine samples 9-19 days after oocyte retrieval (n=179).
Intra-uterine cytokine profile and outcome of IVF/ ICSI

No discomfort or side effects of the aspiration were reported by any of the participants.
In 99.5% of samples, sufficient material was obtained for analysis. As reported before,
ten mediators (IL-1ß, IL-6, IL-12, IL-18, TNF-α, MIF, Eotaxin, MCP-1, IP-10,
VEGF) were detectable in >90% of the samples. IL-15 was detectable in 76%, Dkk-1
in 68%, IL-10 in 56%, IL-17 in 54%, IL-5 in 35%, and Hb-EGF was detectable in
23% of the samples. IFN-γ was not detectable in any of the samples and we have
previously shown that this was not due to technical problems (Boomsma et al. 2009).
Due to the high number of missing values, Hb-EGF, IL-5 and IFN-γ were excluded
from further analyses. No significant differences in total protein content comparing
samples from pregnant and non-pregnant women were found (p = 0.5).
In order to determine the association between implantation or a clinical pregnancy and

the cytokine profile in the secretions, multivariable logistic regression with backward
stepwise elimination of the 14 included mediators was performed. Blood
contamination, embryo quality and maternal age were entered as covariates into the
analysis in a forward stepwise model. Since the contribution of blood contamination to
implantation was not statistically significant, only embryo quality and maternal age
were included in the final model. Multivariable logistic regression revealed a
significant positive association between the concentration of IP-10 and embryo
implantation (p = 0.037), as well as a significant negative association between MCP-1
and implantation (p = 0.005).
Implantation embraces both pre-clinical pregnancy losses and women achieving a

clinical pregnancy. Therefore, the relation between the intra-uterine cytokine profile
and a clinical pregnancy was also assessed. The median concentration, and the 25th
and 75 th percentile of concentrations of each mediator, identified in endometrial
secretions from women who achieved a clinical pregnancy versus women who did not,
are presented in Table 2. Since embryo quality was the only covariate with a
statistically significant contribution to pregnancy, only embryo quality was included in
the final model. Multivariable logistic regression revealed a significant positive
association between the concentration of TNF-α and clinical pregnancy (p = 0.023), as
well as a significant negative association between IL-1β and clinical pregnancy (p =
0.047). Univariate analysis of individual mediators showed no significant associations
with pregnancy. Interaction analysis showed no modifying effect of age on the relation
of IL-1β and TNF-α with pregnancy. Secondly, and consistent with data from clinical
studies (Daya et al. 2004), the type of luteal phase support did not predict pregnancy
chances when added to a model containing IL-1β and TNF-α (p = 0.47), nor did it
change the significance of the two cytokines in the model. In order to identify a profile
of mediators predictive of clinical pregnancy, the criterion for exclusion in the
multivariable regression analysis was then relaxed (p-value > 0.3 rather than 0.1). Four
mediators were thus identified, namely IL-1ß (p = 0.058), TNF-α (p = 0.030), MIF (p
= 0.27), and MCP-1 (p = 0.22). Higher concentrations of IL-1ß and MCP-1 were
negatively associated with pregnancy, whereas those of TNF-α and MIF were
positively associated with a clinical pregnancy.
Chapter 3.2 | 87
The predictive value of endometrial secretion mediator concentrations for a clinical

pregnancy was investigated by constructing a ROC curve. The area under the curve
(AUC) for IL-1ß and TNF-α levels as predictors of pregnancy was 0.61 (95%
confidence interval [CI] 0.52-0.69), which is similar to the AUC for predicting
pregnancy by embryo quality (0.61; 95% CI 0.54-0.68). Logistic regression analysis
demonstrated that the combination of IL-1ß, TNF-α and embryo quality was a
significantly better predictor of pregnancy compared to embryo quality alone (p =
0.049, likelihood ratio test), resulting in an AUC of 0.67 (95% CI 0.60-0.75) (Figure
4). Adding MCP-1 and MIF to IL-1ß, TNF-α and embryo quality (AUC 0.69 95% CI
0.61-0.76), did not significantly improve the model (p = 0.22, likelihood ratio test).
Figure 4: Receiver Operating Characteristics Curve showing the area under the curves to
predict pregnancy by the concentration of IL-1ß, TNF-α and embryo quality (high quality
embryo yes/ no).
Table 2: Soluble mediator concentrations in endometrial secretion aspirations in pregnant and

non-pregnant women, expressed as pg mediator /mg protein.
Parameter Pregnant Non- p-

women pregnant valuea
women
(n= 68) (n=142 )
IL-1ß 6.87 6.14 0.6
(3.12; 13.5) (3.52; 14.6)
IL-5 15.7 16.8 0.5
(7.44; 34.7) (5.82; 37.9)
IL-6 23.1 31.5 0.8
(9.35; 68.9) (14.8; 77.7)
IL-10 2.72 2.35 0.9
(0.89; 4.54) (0.86; 5.27)
IL-12 146 120 0.6
(63.6; 302) (53.0; 287)
IL-15 4.16 4.04 0.8
(2.37; 6.15) (2.21; 7.88)
IL-17 4.90 2.90 0.071
(1.97; 10.1) (1.21; 8.27)
IL-18 58.3 63.5 0.9
(32.4; 138) (39.0; 104)
TNF-α 23.2 19.3 0.4
(9.95; 49.6) (10.0; 44.7)
MIF 73182 71087 0.3
(46667; (45228;
119841) 113478)
Eotaxin 26.5 21.1 0.9

(CCL11) (9.15; 54.7) (9.10; 53.8)
MCP-1 45.5 60.7 0.3
(CCL2) (17.2; 113) (27.4; 120)
Dkk-1 564 613 0.9
(231;1160) (154;1367)
IP10 98.6 106 0.9
(CCL10) (34.1; 390) (28.9; 359)
Hb-EGF 3.74 4.20 1.0

(1.61; 6.25) (1.60; 10.3)
VEGF 30.0 31.0 0.4
(15.4; 52.4) (18.0; 54.5)
Values are median, (25;75th percentiles) from mediator concentrations after imputation of data for values
below detection limit. Hb-EGF and IL-5 have not been imputed.
a
Logistic regression of Log transformed data with blood contamination included as a covariate.
Chapter 3.2 | 89
Discussion
The present study suggests that a molecular fingerprint of the receptive human
endometrium can be identified in endometrial secretions aspirated immediately prior
to embryo transfer. For the first time a significant association between the
concentrations of TNF-α and IL-1β in endometrial secretion during the window of
implantation and the likelihood of a clinical pregnancy occurring in the same cycle has
been shown.
Association between TNF-α and IL-1β and clinical pregnancy

Clinical pregnancy was associated with a significantly higher concentration of TNF-α
in endometrial secretions. Previous data largely derived from animal studies have
indicated that TNF-α is a pro-inflammatory cytokine involved in triggering
immunological pregnancy loss (Wegmann et al. 1993; Arck et al. 1997). Our findings
appear to be at odds with the prevailing paradigm that the establishment of pregnancy
requires an anti-inflammatory cytokine milieu. However, this concept is now
considered to represent an oversimplification of the immunological regulation of the
tolerance of the conceptus, which requires more than a simple predominance of anti-
inflammatory mediators (Chaouat et al. 2004). Knock out studies in mice have shown
that TNF-α may act as a protector of embryos exposed to teratogenic stress (Toder et
al. 2003). TNF-α has also been shown to be an important regulator of trophoblastic
matrix metalloproteinase (MMP)-2 (Meisser et al., 1999) and MMP-9 (Meisser et al.,
1999; Cohen et al., 2006), which have been shown to mediate the invasive properties
of trophoblastic cells into matrigel (Staun-Ram et al. 2004). In the present study,
clinical pregnancy was also associated with lower secretion levels of IL-1β. This
finding is consistent with a previous report comparing women with implantation
failure after IVF and controls, which demonstrated significantly higher IL-1β levels in
endometrial flushings from women with implantation failure (Inagaki et al. 2003). In
addition, significantly lower levels of IL-1β mRNA in the endometrium of individuals
with habitual abortion compared to fertile controls have been described (von Wolff et
al. 2000). It has been postulated that women with habitual abortion may have an ‘over-
receptive’ endometrium resulting in recurrent miscarriages of abnormal embryos
which were allowed to implant, rather than implantation failure of these embryos
(Quenby et al. 2002). However, others have suggested that IL-1β may play a
facilitating role in implantation by up-regulation of integrins at the luminal epithelial

surface (Simon et al. 1997) and MMPs (Karmakar and Das, 2002), which mediate
embryo adhesion and invasion. The observation that the levels of IL-1β and TNF-α are
significantly related to achieving a clinical pregnancy and not embryo implantation,
suggests that these mediators may have a more important role in later stages of
implantation rather than initial apposition and adhesion of the embryo. However, it
should be emphasized that in the analysis of mediators correlated to implantation, IL-
1β and TNF-α were only eliminated in the last steps of backward elimination in the
logistic regression analysis. Vice versa, MCP-1, IP-10 and also MIF were eliminated
in the last steps of elimination in the analysis on pregnancy rather than implantation.
Therefore, the results suggest a role for these mediators in both early and later steps of
implantation, rather than a difference in mediators predictive of implantation and
clinical pregnancy.
Association between MCP-1 and IP-10 and embryo implantation

Initial embryo implantation was observed to be significantly associated with
endometrial secretion concentrations of MCP-1 (negatively) and IP-10 (positively).
Both of these chemokines are known to play key roles in leukocyte migration (Jones et
al. 2004). Large populations of immune cells accumulate in the endometrium during
the luteal phase and are believed to be important modulators of implantation. There is
evidence that a clear parallel can be established between the different steps in human
embryo implantation and steps in leukocyte migration (Jones et al. 2004). The embryo
has been shown to express receptors for IP-10 and MCP-1 (Dominguez et al. 2003;
Drake et al. 2004; Hanna et al. 2006). MCP-1 is up-regulated by the administration of
progesterone and increased levels are found in secretory rather than proliferative
endometrium (Caballero-Campo et al. 2002). Moreover, it has been shown to be a
potent attractant and activator of uterine Natural Killer (uNK) cells (Loetscher et al.
1994). High numbers of uNK cells have been related to miscarriage and infertility
(Moffett et al. 2004), which may be consistent with lower secretion levels of MCP-1
being associated with higher implantation rates, as observed in the present study.
Secretion concentrations of IP-10 were positively associated with embryo

implantation. IP-10 is primarily expressed in the human endometrium during the
secretory phase and uNK cells present during the luteal phase are potent secretors of
Chapter 3.2 | 91
IP-10 (Kitaya et al. 2004). Evidence derived from animal studies also suggests a
possible function for IP-10 in the process of implantation. The ability of uNK cells to
induce trophoblast migration was shown to be significantly reduced by neutralizing
antibodies to IP-10 in an in vivo model of human trophoblast migration in mice (Drake
et al. 2004). Moreover, IP-10 was demonstrated to stimulate the expression of integrin
subunits in trophoblast cells and recombinant IP-10 promoted adhesion of caprine
trophoblast cells to endometrial epithelial cells (Nagaoka et al. 2003).
Pre-clinical pregnancy loss

Studies of spontaneous conception have shown that a significant proportion of
embryos which begin to implant fail to proceed to a clinically recognized pregnancy
(Wilcox et al. 1999). In the present study, the rate of pre-clinical pregnancy loss
following IVF was determined by analysing the frequency with which treatment
cycles in which implantation was detected, failed to result in a clinical pregnancy.
More than 50% of women undergoing embryo transfer demonstrated hCG rises
indicating an implanting embryo. However, approximately one third of these
implantations resulted in pre-clinical pregnancy loss. One previous study has reported
the incidence of pre-clinical pregnancy loss in an IVF population. 48% of the
implanted embryos were shown to be lost before they were clinically recognized as a
pregnancy (Simon et al. 1999), compared to 34% in the present study. This difference
in incidence may be due to differences in methodology and study population. The
incidence of pre-clinical pregnancy loss in IVF cycles appears to be higher than
reported following spontaneous conceptions (Wilcox et al. 1988). While embryo
quality is an important determinant of initial and ongoing implantation, the
endometrium may also have a role. Specific endometrial factors may be involved in
apposition and initial implantation, and others more important in facilitating later
invasion. We therefore explored whether those markers present in the endometrial
secretion predictive of implantation differed from those predictive of a clinical
pregnancy. This was observed to be the case.
Limitations of the study

Endometrial secretion analysis at the time of embryo transfer offers certain essential
advantages over other techniques. Most techniques require endometrial tissue, which
is not possible to obtain during the window of implantation without disrupting the
process of implantation (van der Gaast et al. 2003). It is uncertain to what extent
biopsies from the peri-ovulatory phase or preceding menstrual cycles reflect the mid-
luteal endometrium in the study cycle. Studies of uterine cavity flushings have shown
inconsistent data, possibly due to the high degree of dilution and differences in the
timing of aspiration between the studies (Strowitzki et al. 2006; Ledee-Bataille et al.
2002). While a promising technique, endometrial secretion analysis has a number of
limitations amenable to improvement. As yet, we have not been able to include certain
proposed markers of the window of implantation in the multiplex immunoassay
(Lessey et al. 1995; Chryssikopoulos et al. 1996; Ledee-Bataille et al. 2002), either
because appropriate antibodies were not available (Glycodelin, Osteopontin), or
because of problems arising from cross-interference (IL-11, Leukemia Inhibitory
Factor [LIF] and Macrophage Colony Stimulating factor [M-CSF]). In order to refine
the array, novel potential markers for endometrial receptivity should be included in
future studies. Endometrial secretions also contain cellular material, mostly epithelial
and immune cells. In order to minimise interference in the assay by contaminating
cells, the aspirate was centrifuged after dilution and the pellet was discarded.
However, it should be recognized that the released contents of lysed cells also
contribute to the intra-uterine environment encountered by the embryo in-vivo. At
present, endometrial secretion analysis is not yet clinically useful to predict pregnancy.
Significance of findings and conclusions

Despite these limitations, we have for the first time identified profiles of mediators in
aspirated endometrial secretions significantly correlated to implantation and to a
clinically recognized pregnancy. Our results suggest that the ratio of TNF-α and IL-1β
may serve as an indicator of endometrial receptivity, rather than individual absolute
values of these mediators, since both coefficients are fairly similar in magnitude but
opposite in direction. Interestingly, IL-1β and TNF-α are both pro-inflammatory
cytokines. Pregnancy has been postulated to be an event requiring a predominance of
anti-inflammatory mediators. However, implantation is a process of aggression in
which maternal and embryonic signals induce a state of controlled inflammation. Our
findings, showing an important role for two pro-inflammatory mediators, endorse the
view that the Th1/Th2 paradigm represents an over simplification of the role of
cytokine as determinants of successful implantation (Chaouat et al. 2004). ROC
analysis showed that the ability to predict clinical pregnancy by IL-1β and TNF-α
Chapter 3.2 | 93
levels in endometrial secretions was similar to the predictive value of embryo quality.
Although the predictive value of the markers selected in the present study is too low to
be clinically useful, it is noteworthy that the predictive values of the secretion cytokine
profile and embryo quality were additive, indicating that the endometrium is not
simply facilitatory in the implantation process, but is also itself an independant
determinant of outcome. The present study has also emphasized the importance of
chemokines at the fetal-maternal interface, presumably as a result of their function in
leucocyte and embryo migration. Since our study is exploratory in nature, we have
applied no correction for multiple testing as would be required for a confirmatory
hypothesis based study. Our results require confirmation, and the array should be
further refined. Moreover, issues such as inter-cycle and intra-cycle variation should
be explored. The intra-uterine milieu in natural cycles of normal fertile women is also
interesting to explore. However, this may be considered ethically difficult given the
risk of possibly aspirating an embryo already present in the uterus. The non-invasive
nature of endometrial secretion analysis enables endometrial receptivity to be
investigated prior to embryo transfer in IVF cycles using pregnancy as an endpoint,
rather than surrogate markers for endometrial receptivity, such as endometrial dating
by histology. Employing this novel approach will increase our understanding of the
complex molecular cross-talk by which the embryo engages the intrauterine
environment.
Chapter 3.3
____________________________________________
Ovarian stimulation for in-vitro fertilization alters the

intra-uterine cytokine, chemokine and growth factor
milieu encountered by the embryo
Carolien M Boomsma1
Bart CJM Fauser1
Cobi J Heijnen2
Nick S Macklon1
1
2
3
(Submitted)
Abstract
Objective
Supraphysiological sex steroid levels resulting from ovarian stimulation are
considered to be detrimental to endometrial receptivity in in-vitro fertilization. The
objective of this study was to elucidate the impact of ovarian stimulation on the intra-
uterine milieu represented by the cytokine, chemokine and growth factor profile in
endometrial secretions aspirated prior to embryo transfer.
Design en setting
Prospective cohort study. Fertility center in tertiary referral university hospital.
Patients
204 patients undergoing ovarian stimulation with gonadotropin releasing hormone
(GnRH) analogues co-treatment were recruited. A subgroup of 42 women participated
in both a natural cycle and an ovarian stimulated cycle for within patient comparisons.
Methods
Endometrial secretion aspiration was performed immediately prior to embryo transfer.
The concentrations of 17 mediators known to be involved in human embryo
implantation were assessed by multiplex immunoassay.
Results
After correction for multiple testing, significantly higher concentrations of interleukin
(IL)-1ß, IL-5, IL-10, IL-12, IL-17, Tumor Necrosis Factor (TNF)-α, Heparin-binding
Epidermal Growth Factor (Hb-EGF), Eotaxin, and Dickkopf homolog (Dkk)-1 were
present in endometrial secretions obtained in stimulated compared to natural cycles.
Multivariable analysis in 203 patients revealed a significant relationship between intra-
uterine cytokine concentrations and the number of oocytes retrieved and between the
intra-uterine cytokine profile and type of GnRH analogue used.
Conclusions
The in-vivo intra-uterine milieu encountered by the embryo after transfer is
significantly altered by ovarian stimulation. This effect increases along with the
number of oocytes retrieved. Minor differences in protein profiles in endometrial
secretions were found in GnRH antagonist compared to GnRH agonist co-treatment
cycles.
Chapter 3.3 | 97
Introduction
Ovarian stimulation is employed in in-vitro fertilization (IVF) programs in order to
collect multiple oocytes to produce multiple embryos per cycle of treatment (Macklon
et al. 2006). While this procedure enables selection of high quality embryos for
transfer, ovarian stimulation also results in supraphysiological levels of progesterone
and estradiol. These high sex steroid concentrations may disrupt the luteal phase and
impair endometrial receptivity (Bourgain et al. 2003). Evidence for a direct effect of
sex steroids on the endometrium comes from histological, histochemical and gene
expression studies. Estrogen and progesterone act through endometrial nuclear
receptors, which are usually down-regulated in response to a physiological increase in
estrogen and progesterone levels in the early luteal phase (Lessey et al. 1988). Ovarian
stimulation has been shown to advance this down-regulation of steroid receptors
(Bourgain et al. 2002). Supraphysiological concentrations of progesterone in
stimulated cycles also cause an accelerated transformation to secretory endometrium at
the time of embryo transfer (Lass et al. 1998), which has been shown to be detrimental
to implantation rates (Ubaldi et al. 1996). Immunohistochemical studies have
indicated a disruptive effect of ovarian stimulation on molecular processes involved in
implantation. A number of proteins considered to be key regulators in the implantation
process, such as integrins (Creus et al. 2003), glycodelin-A (Brown et al. 2000), and
the L-selectin ligand (Lai et al. 2006) were shown to be differentially expressed in
stimulated compared to natural cycles. In recent years, the emergence of micro array
technology has enabled the study of the entire gene expression profile of the
endometrium and several studies have reported dysregulated gene expression
following ovarian stimulation (Horcajadas et al. 2005; Macklon et al. 2008).
In addition to the impact of supraphysiological sex steroid levels as a result of

exogenous gonadotropins, the concurrent administration of gonadotropin-releasing
hormone (GnRH) analogues to prevent a premature rise in luteinzing hormone
concentrations may also have direct and indirect effects on the endometrium. A meta-
analysis showing lower implantation rates after the use of GnRH antagonists versus
agonists, suggested a possible detrimental impact of GnRH antagonist directly on the
endometrium (Al-Inany et al. 2006).
Understanding and ameliorating the impact of ovarian stimulation on the intra-uterine

environment with which the embryo must interact may result in improved
implantation rates in the future. Our group has developed a novel means of assessing
the intra-uterine milieu by endometrial secretion analysis (van der Gaast et al. 2003;
Boomsma et al. 2009), which can be safely aspirated prior to embryo transfer.
Endometrial secretions contain a variety of molecules secreted by the endometrial
epithelia, which are presumed to support the embryo during the pre-implantation
period (Gray et al. 2001; Burton et al. 2002). In addition to providing nutrition for the
conceptus, endometrial secretions contain growth factors and cytokines involved in the
implantation process. In the present study we utilized endometrial secretion analysis to
elucidate the impact of ovarian stimulation with exogenous gonadotropins on the
levels of key regulatory cytokines, chemokines and growth factors in endometrial
secretion at the time of embryo transfer. Secondary objectives were to assess the
relation between the ovarian response to stimulation (defined as the number of oocytes
retrieved) and the intra-uterine milieu and to study whether the co-treatment with a
GnRH antagonist versus an agonist demonstrated an effect on the endometrial
secretion protein profile.

Ethical approval was received from the local institutional committee and written
informed consent was obtained from all patients. Between December 2005 and
October 2006, two hundred and four patients attending the University Medical Center
Utrecht (The Netherlands) for IVF or intracytoplasmatic sperm injection (ICSI)
treatment were included in the study. None had undergone more than one prior
embryo transfer and none had hydrosalpinges or anatomic uterine abnormalities. All
consented to undergo endometrial secretion aspiration prior to embryo transfer (6 days
after human Chorionic Gonadotropin [hCG] injection). In order to study the impact of
ovarian stimulation, forty two of the participants with regular menstrual cycles of 25-
35 days underwent an additional endometrial secretion aspiration in their spontaneous
cycle 6 days after the luteinizing hormone (LH) surge. Urinary LH was monitored by a
home LH ovulation predictor kit (Ovulady®, Clindia Benelux, The Netherlands).
Chapter 3.3 | 99
Assisted Reproductive Technique

Ovarian stimulation was applied with recombinant FSH (rFSH) (Puregon®: NV
Organon, Oss or Gonal-F®: Serono Benelux B.V. Amsterdam, The Netherlands; 100-
300 IU per day, s.c.), starting on cycle day 2 or 3. In order to prevent a premature rise
in LH, GnRH antagonist (Orgalutran®: N.V. Organon; 0.25 mg/day s.c.) co-treatment
was provided in a fixed protocol from stimulation day 5 in IVF cycles. The majority of
ICSI cycles were performed using a long GnRH agonist (Lucrin®: Abbot B.V.,
Amstelveen, The Netherlands; 0.2 mg/day s.c.) suppression protocol, starting on
menstrual cycle day 21. After 14 days down-regulation was established and rFSH was
started. Final oocyte maturation was triggered with a single subcutaneous injection
with hCG (Pregnyl®: N.V. Organon; 10.000 IU, s.c.) when the largest follicle had
reached a diameter of at least 18 mm and at least one additional follicle larger than 14
mm was observed. A maximum of 2 embryos were transferred 4 days after oocyte
retrieval. Following embryo transfer luteal phase support was as described previously
(Boomsma et al. 2009).
Endometrial secretion aspiration

The aspiration technique employed has been described previously (Boomsma et al.
2009). Briefly, an embryo transfer catheter (Wallace®: SIMS Portex Ltd., Hythe, Kent,
UK) was introduced transcervically. Suction was gradually applied with a 2 mL
syringe. In order to prevent contamination by cervical mucus during catheter removal,
the outer sheath of the embryo transfer catheter was advanced to a depth of 4 cm from
the external cervical os following the application of suction. The inner catheter was
then withdrawn through the outer sheath, which prevented contact with the cervix. In
the treatment cycle, the embryo transfer was carried out immediately thereafter. The
aspirate was snap frozen in liquid nitrogen and stored at -80°C.
Endometrial secretion analysis

The procedures employed to analyse the endometrial secretion aspirates have been
previously described in detail (Boomsma et al. 2009). Briefly, endometrial secretion
samples were analysed using a multiplex immunoassay to measure the concentrations
of 17 soluble mediators of human implantation, namely IL-1ß, IL-5, IL-6, IL-10, IL-
12, IL-15, IL-17, IL-18, TNF-α, IFN-γ, MIF, Eotaxin, IP-10, MCP-1, Dkk-1, Hb-EGF
and VEGF. MIF was determined in diluted samples using a commercially available
enzyme linked immunosorbent assay (ELISA) (R&D systems, Abingdon, United

Kingdom). The protein content was measured by NanoOrange Protein Quantitation Kit
(Molecular probes Europe B.V., Leiden, The Netherlands) and was used for
normalization purposes.
The lack of previous indicative data made it difficult to carry out an appropriate power
calculation for this study. However, previous studies of the effect of ovarian
stimulation on endometrial markers of receptivity showed significant findings with
sample sizes of 12-38 (Brown et al. 2000; Lai et al. 2006; Thomas et al. 2002; Lai et
al. 2006). Moreover, studies on the effect of ovarian stimulation on endometrial gene
expression have demonstrated multi-fold dysregulation of genes coding for a number
of mediators in the panel. Given these expected large differences and the paired design
of the study, a sample size of forty was considered to be sufficient to demonstrate
significant differences at a p-value of 0.05. Secondary aims were to assess the
association between the ovarian response as well as co-treatment with GnRH
antagonists versus agonists on the intra-uterine cytokine profile. Data from the
complete cohort of 204 women were analyzed to study these secondary outcomes. A
probability (p) less than 0.05 was considered statistically significant. A Bonferroni
correction was applied in case of multiple testing. Analyses were performed using
SPSS version 12.0 (SPSS Inc., Chicago, IL, USA, 1999).
Where necessary, data were log transformed to obtain a normal distribution prior to
analysis. For statistical analysis of concentration values below the reliable detection
limit of the assay, we calculated the lower detection limit divided by √2, as
recommended for Log transformed data by Schisterman et al. (Schisterman et al.
2006). Results from multivariable analyses were confirmed after exclusion of
mediators for which more than 50% of values were below the detection limit (IFN-γ,
Hb-EGF and IL-5). As we have previously shown, blood contamination of the
endometrial secretion can affect the measurement of a number of cytokines (Boomsma
et al. 2009). The extent of blood contamination in the endometrial secretion samples
was therefore visually graded and included as a covariate in the analyses.
Chapter 3.3 | 101
Within patient comparisons of endometrial secretion aspirations in the natural and

stimulated cycle were performed using a multivariate analysis of variance of the
difference in both values, with blood contamination included as a covariate. The
differences in cytokine profiles after GnRH antagonists versus agonist co-treatment or
along with the ovarian response were analysed by multivariable analysis, which is the
preferred method of analysis considering the complex mutual correlations between the
mediators. In order to investigate the relationship between ovarian response and the
cytokine profile and GnRH antagonist or agonist co-treatment and the cytokine profile,
a linear regression analysis and a multivariable logistic regression analysis were
performed respectively, with a backward elimination procedure of the soluble
mediators (p > 0.1 as an exclusion criterion). Covariates (e.g. blood contamination)
were entered in a forward stepwise model. Ovarian stimulation characteristics, namely
duration of stimulation and total amount of rFSH, were included as covariates in the
analysis on GnRH analogues. This enabled a possible direct effect of GnRH
antagonists on the endometrium to be assessed, rather than an indirect effect secondary
to modulation of ovarian response by the GnRH analogues.
Results
Of the 204 women who underwent endometrial secretion aspiration in the IVF or ICSI
cycle, 203 were included in the analysis. One patient was excluded, since insufficient
material was available to measure the total protein content in the aspirate required for
normalization purposes. A subgroup of forty two patients were included in the within
patient comparison of the cytokine profile in a natural and an ovarian stimulated cycle.
Again, one patient was excluded since the total protein content of the aspirate could
not be determined. Patient and treatment characteristics are provided in Table 1 for
both the total cohort of participants and the subgroup of patients participating in both a
natural and stimulated cycle. The concentrations of the soluble mediators in
endometrial secretion samples have been previously reported (Boomsma et al. 2009).
Nine mediators (IL-1ß, IL-6, IL-12, IL-18, TNF-α, MIF, MCP-1, IP-10, VEGF) were
detectable in 90-100% of the samples. Eotaxin was detectable in 89% of the samples,
IL-15 was detectable in 76% of the samples, Dkk-1 in 68%, IL-10 in 56%, IL-17 in
54%, IL-5 in 35% and Hb-EGF was detectable in 24% of the samples. IFN-γ was not
detected in any of the samples (Boomsma et al. 2009).
Table 1: Clinical parameters and treatment characteristics of all participants and the subgroup
for within patient comparisons.
Parameter Total cohort of Subgroup p-

participants value
(n= 203) (n= 41)
Age (yrs) 34.9 ± 4.2 34.5 ± 4.3 0.4 a
Duration of infertility (yrs) 3.2 ± 1.8 3.2 ± 1.9 0.9 a
Primary infertility (%) 57 52 0.7 b
Primary cause of infertility (%) 0.6 b
Andrological 50 60
Unknown 22 12
Tubapathology 18 24
Endometriosis 2 2
Maternal age (> 40 yrs) 3 2
Cervical 4 0
BMI 24.4 ± 4.5 24.5 ± 4.9 0.9 a
Smoking (%) 15 2 0.026 b
IVF (%) 58 52 0.6 b
ICSI (%) 42 48
Duration ovarian stimulation 9.9 ± 2.3 9.3 ± 2.1 0.13 a
(days)
Total amount of rFSH (IU) 1,696 ± 687 1,528 ± 594 0.19 c
GnRH antagonist (%) 52 52 0.8 b
GnRH agonist (%) 48 48
Oocytes retrieved (n) 8.1 ± 4.7 (1-23) 8.9 ± 5.1 (1-23) 0.3 c
hCG 10 14
Progestan 90 86
Values are means ± standard deviations (SDs). Range: between brackets.

a
Independent t-test. b χ² test. c Mann-Whitney U test.
Effect of ovarian stimulation on the cytokine profile

Of the 41 samples obtained in a stimulated cycle from women who also underwent an
endometrial secretion aspiration in the natural cycle, only four samples showed visual
signs of minimal blood contamination. In the natural cycle, 29 of the 41 samples
showed no contamination. Of the 12 contaminated samples, only 1 was severely
contaminated with blood. The concentrations of the soluble mediators (expressed in pg
Chapter 3.3 | 103
per mg protein) in endometrial secretion samples from a natural and a stimulated cycle
are presented in Table 2. Significantly higher concentrations of IL-1ß, IL-5, IL-10, IL-
12, IL-17, TNF-α, Eotaxin, Dkk-1 and Hb-EGF (p-values < 0.003) in endometrial
secretions in the stimulated versus the natural cycle were observed (Table 2 and Figure
1). VEGF was the only mediator which demonstrated a reduced concentration
following ovarian stimulation, although not significant after correction for multiple
testing.
Table 2: Soluble mediator concentrations in endometrial secretion aspirations on LH+6

(natural cycle) and hCG + 6 (stimulated cycle), expressed as pg mediator /mg protein.
Natural Stimulated Fold change in p-
cycle† cycle† concentrations value**
(n=41) (n=41) after stimulation*
IL-1ß 2.32 5.06 2.67(1.57-4.55) <0.001
(1.43;4.63) (3.38;17.6) ↑
IL-5 2.39 5.86 2.90 (1.74-4.85) <0.001
(1.23;6.00) (2.64;24.1) ↑
IL-6 6.16 12.2 3.07 (1.43-6.58) 0.012
(0.95;20.3) (3.97;37.2) ↑
IL-10 1.03 2.65 3.50 (1.99-6.15) <0.001
(0.37;2.30) (1.42;6.48) ↑
IL-12 53.4 151 3.45 (2.08-5.73) <0.001
(28.4;112) (75.7;479) ↑
IL-15 3.69 4.35 1.04 (0.68-1.58) 0.9
(2.67;5.94) (2.26;7.41)
IL-17 3.29 9.91 2.60 (1.55-4.38) 0.001
(1.40;8.87) (4.02;25.6) ↑
IL-18 64.5 91.0 1.42 (0.91-2.23) 0.19
(41.3;116) (58.9;171)
TNF-α 11.8 19.7 2.69 (1.66-4.37) <0.001
(3.27;21.6) (10.8;75.3) ↑
MIF 50913 58667 2.04 (1.00-4.13) 0.044
(38782;65027) (37175;124432) ↑
Eotaxin 8.79 22.3 3.53 (2.23-5.58) <0.001
(CCL11) (4.09;19.1) (10.5;89.2) ↑
MCP-1 29.7 53.1 1.41 (0.79-2.52) 0.2
(CCL2) (15.0;71.4) (22.6;145)
Dkk-1 263 753.26 3.01 (1.87-4.84) <0.001

(127;511) (450;1379) ↑
IP10 125 112 1.10 (0.54-2.24) 0.7
(CCL10) (29.3;283) (41.8;351)
Hb-EGF 0.64 1.60 2.23 (1.33-3.74) 0.003
(0.36;1.49) (0.67;4.47) ↑
VEGF 51.3 32.3 0.56 (0.35-1.74) 0.029
(33.0;127) (19.9;61.6) ↓
†
Values are median and (25;75 percentiles).
* Mean and (95% Confidence Interval). Blood contamination included as covariate.
** Multivariate analysis of variance with blood contamination as a covariate.
Figure 1: Concentrations of mediators which are significantly differentially expressed (p <

0.0031) in natural versus stimulated cycles, expressed as pg mediator /mg protein.
Pro-inflammatory cytokines
40 800 IL-12 120 TNF -α

IL-1 ß
100
30 600
80
20 400 60
40
10 200
20
0 0 0
NAT STIM NAT STIM NAT STIM
Anti-inflammatory cytokines Pro- and anti-

inflammatory properties
50 12 IL-10 50 IL-17
IL-5
40 10 40
30 8
30
6
20 20
4
10 10
2
0 0 0
Wnt signalling Growth Factor Chemokine

inhibitor
2.000 Dkk-1 8 HbEGF Eotaxin

150
1.500 6
100
1.000 4
50
500 2
0 0 0
NAT: natural cycle. STIM: ovarian stimulated cycle.
Association between ovarian response and intra-uterine cytokine profile

The relation between the ovarian response to ovarian stimulation, expressed as the
number of oocytes retrieved, and the endometrial secretion cytokine profile, was
Chapter 3.3 | 105
investigated in the cohort of 203 women who underwent endometrial secretion

aspiration prior to embryo transfer. Linear regression with backward stepwise
elimination of mediators, and blood contamination included as a covariate, was
performed and revealed significant correlations between the number of oocytes
retrieved and the concentration of IL-12 (p = 0.002), Dkk-1 (p = 0.002) (positive
correlation) and IL-15 (p = 0.001) (negative correlation).
Differences in cytokine profile after co-treatment with GnRH antagonist versus

agonist
The influence of the use of a GnRH agonist versus antagonist on the intra-uterine
cytokine expression profile was also investigated in the cohort of 203 women who
underwent endometrial secretion aspiration prior to embryo transfer. As shown in
Table 1, 52% of participants underwent treatment using GnRH antagonist co-
treatment, whereas the other 48% were treated with a GnRH agonist. Characteristics of
patients and their treatment cycles are provided in Table 3. The GnRH agonist
protocol was primarily applied in ICSI cycles. Therefore couples in whom women
were treated with a GnRH agonist showed for the most part severe semen
abnormalities, and a lower incidence of tubal pathology or unknown cause of
infertility. The cause of infertility is therefore a confounder which cannot be corrected
for by inclusion as a covariate. Maternal age, primary or secondary infertility and
stimulation characteristics (duration of stimulation and total amount of rFSH) were
included as covariates in a forward stepwise model, since significant differences in
these parameters were observed between both study groups (Table 3). Multivariable
analysis revealed a significantly lower concentration of Eotaxin (p = 0.024), and a
higher IL-1β (p = 0.012) concentration after GnRH antagonist versus agonist
treatment. Univariate analyses revealed no significant differences in cytokine
concentrations according to the GnRH stimulation protocol employed.
Table 3: Clinical parameters and treatment characteristics of women treated with a GnRH
agonist versus antagonist co-treatment.
Parameter GnRH GnRH p-

agonist antagonist value
(n= 98) (n= 105)
Age (yrs) 34.0 ± 3.9 35.8 ± 4.3 0.003 a
Duration of infertility (yrs) 3.2 ± 1.7 3.2 ± 1.8 0.9 a
Primary infertility (%) 66 49 0.011 b
Primary cause of infertility (%) <0.001 b
Andrological 77 26
Unknown 9 33
Tubapathology 8 28
Endometriosis 2 2
Maternal age (> 40 yrs) 0 5
Cervical 0 3
BMI 24.5 ± 4.6 24.3 ± 4.4 0.8 a
Smoking (%) 15 15 1.0 b
IVF (%) 20 95 <0.001b
ICSI (%) 80 5
Duration ovarian stimulation 11.0 ± 2.2 8.8 ± 1.9 <0.001 a
(days)
Total amount of rFSH (IU) 1,886 ± 704 1,512 ± 620 <0.001 c
Oocytes retrieved (n) 8.3 ± 4.2 (1-19) 7.9 ± 5.1 (1-23) 0.19 c
hCG 14 8
Progestan 86 92
Values are means ± standard deviations (SDs). Range: between brackets.

a
Independent t-test. b χ² test. c Mann-Whitney U test.
Discussion
The present study demonstrates for the first time that ovarian stimulation causes
significant disruption to the intra-uterine secretome, which constitutes the milieu into
which the embryo is transferred and with which it must first interact. Significantly
higher concentrations of IL-1ß, IL-5, IL-10, IL-12, IL-17, TNF-α, Hb-EGF, Dkk-1 and
Eotaxin in the endometrial secretion in the stimulated versus the natural cycle were
observed. Whether these changes in the intra-uterine milieu affect endometrial
receptivity is unknown. However, previously we have shown that intra-uterine
cytokine levels in ovarian stimulated cycles are associated with the likelihood of
conceiving (Boomsma et al. 2009). The marked differences in cytokine profiles in
Chapter 3.3 | 107
ovarian stimulated cycles compared to natural cycles observed in the present study are
therefore likely to influence endometrial receptivity.
Multivariable analysis showed that IL-12 and Dkk-1 concentrations significantly

increased in relation to the number of oocytes retrieved. In contrast, the concentration
of IL-15, an important regulator of uterine Natural Killer (uNK) cell activity, was
significantly lower in women with a higher number of oocytes retrieved. These
observations suggest that that the effect of ovarian stimulation on endometrial
receptivity may be related to the individual ovarian response. This is consistent with
clinical studies which showed higher implantation rates when milder stimulation was
applied in women who previously had a high response to ovarian stimulation (Simon
et al. 1995).
Gene expression studies have also reported disruption of the mid-luteal endometrial
transcriptome following ovarian stimulation (Horcajadas et al. 2005; Macklon et al.
2008). Interestingly, for many of the genes studied, the expression levels in hCG+7
samples were more comparable with those of LH+2 than with LH+7 patterns. This
observation suggests a delay in the regulation of gene expression due to ovarian
stimulation. Of the 17 mediators examined in the present study, only the gene Dkk-1
has previously been shown to be significantly up-regulated by ovarian stimulation in
global gene expression studies (Horcajadas et al. 2005). Dkk-1, an inhibitor of Wnt
signaling, also showed a marked up-regulation during the window of implantation in
four studies on gene expression throughout the menstrual cycle (Carson et al. 2002;
Kao et al. 2002; Borthwick et al. 2003; Riesewijk et al. 2003). The upregulation of
Dkk-1 during the luteal phase and during ovarian stimulated cycles suggests its
progesterone dependent expression, which has also been shown by an in vitro study of
Dkk-1 mRNA and protein expression by endometrial stromal cells after progesterone
treatment (Tulac et al. 2006).
In the present study, cytokine concentrations were expressed relative to total protein
content in the aspirations. The observation that the concentrations of some mediators
are increased after stimulation, whereas others were not altered, excludes the
possibility that differences in mediator concentrations simply arose due to differences
in total protein content of aspirates obtained during natural versus stimulated cycles.
An explanation for the increase in the concentrations of most mediators after ovarian
stimulation may be an increased cytokine expression by endometrial stromal and
epithelial cells. However, it may also be caused by an increased number of uNK cells
present in the endometrium. The number of uNK cells increase under the influence of
supraphysiological levels of estrogen (DeLoia et al. 2002) and after ovarian
stimulation (Lukassen et al. 2004). uNk cells are the most abundant leucocytes in the
luteal phase endometrium and they can secrete an array of cytokines (Quenby et al.
2006), such as VEGF, IFN-γ, TNF-α, MIF, MCP-1, Eotaxin and IP-10 (Moffett-King
et al. 2006; Hanna, 2006). VEGF, a major modulator of vascular growth and
remodelling required for the cyclical blood vessel proliferation in the endometrium
(Smith, 2001), was the only studied mediator which demonstrated a reduced
concentration following ovarian stimulation though. Our finding that this regulator of
angiogenesis appears to be downregulated at the protein level following ovarian
stimulation, is consistent with clinical studies which have demonstrated endometrial
and sub-endometrial blood flow to be lower in ovarian stimulated cycles than in
natural cycles (Ng et al. 2004). Moreover, ovarian stimulation has been shown to be
associated with lower levels of endometrial VEGF in the mouse uterus during the peri-
implantation period (Sibug et al. 2002). A study in humans showed significantly
higher mRNA levels of VEGF in stimulated cycles, although no differences were
observed in VEGF protein expression in endometrial biopsies (Lee et al. 2008). A
previous study investigated secretion levels of leukaemia inhibitory factor (LIF) and
glycodelin A (GdA), both key regulators of implantation (van Mourik et al. 2008). No
significant differences were observed when comparing natural with stimulated cycles.
However, this may be attributed to the low number of patients included (n = 8) (van
der Gaast et al. 2008).
In recent years, GnRH antagonists have been introduced into IVF treatment. A meta-
analysis of early studies suggested that implantation rates were reduced after the use of
GnRH antagonists compared to the use of GnRH agonists and raised concerns on a
possible detrimental effect on the endometrium by GnRH antagonists (Al-Inany et al.
2006). In the present study, GnRH antagonist co-treatment was associated with a
significant lower concentration of Eotaxin, and higher IL-1β concentration in
endometrial secretions compared to women using GnRH agonist co-treatment.
However, differences in cytokine profiles after GnRH antagonist or agonist treatment
Chapter 3.3 | 109
were minor compared to those observed comparing natural and stimulated cycles.
Previous studies reported no differences in endometrial maturation or endometrial
gene expression between GnRH antagonist and agonist co-treatment (Mirkin et al.
2008; Saadat et al. 2004). A study on the cytokine profile in serum and follicular fluid
in women using a GnRH antagonist versus agonist co-treatment showed no significant
differences in IL-1β, IL-6, TNF-α and VEGF levels on the day of oocyte retrieval
(Asimakopoulos et al. 2006). When interpreting the findings of our analysis, the
possible effect of confounding factors needs to be addressed. The majority of patients
treated by a GnRH agonist underwent ICSI for severe male factor infertility.
Therefore, a confounding effect by cause of infertility on the observed impact of the
type of GnRH analogue on the endometrial secretion cytokine profile cannot be ruled
out.
In conclusion, it is demonstrated that the in-vivo milieu encountered by the embryo

after transfer is significantly altered by ovarian stimulation as compared to the natural
cycle. Significantly higher concentrations of IL-1ß, IL-5, IL-10, IL-12, IL-17, TNF-α,
Hb-EGF, Eotaxin, and Dkk-1 were present in endometrial secretions obtained in the
stimulated cycle. In women with a higher number of oocytes retrieved some of these
differences were more marked, suggesting a further disruption of the intra-uterine
milieu in high responders. Minor differences in the secretion protein profiles were
observed in GnRH antagonist or agonist co-treatment cycles. Understanding how the
endometrium is disrupted by ovarian stimulation, is of primary importance if we are to
understand the determinants of successful implantation and further improve
implantation rates in IVF therapy. Employing this novel technique to investigate the
effect of ovarian stimulation on the intra-uterine environment during the window of
implantation may help to develop less disruptive ovarian stimulation protocols for IVF
in the future.
Chapter 3.4
____________________________________________
Is bacterial vaginosis associated with a

pro- inflammatory cytokine profile in endometrial
secretions of women undergoing IVF?
Carolien M Boomsma1
Nuray Bozkurt3
Bart CJM Fauser1
Cobi J Heijnen2
Nick S Macklon1
1
2
4
Departments of Obstetrics and Gynecology, Gazi University Ankara, Turkey
3
(Submitted)
Abstract
Objective
To elucidate whether bacterial vaginosis (BV) is associated with a pro-inflammatory
cytokine profile and to assess whether there is a relation between BV and the levels of
a number of key regulatory cytokines, chemokines and growth factors in endometrial
secretions aspirated prior to embryo transfer.
Design and setting
Prospective cohort study. Fertility centre in a tertiary referral university hospital.
Patients
198 women undergoing IVF or ICSI treatment.
Interventions
Prior to embryo transfer participants underwent screening for BV according to Nugent
criteria by a Gram stained cervical smear. Cytokines were determined in endometrial
secretions aspirated prior to embryo transfer.
Main outcome measures
The concentrations of 17 soluble mediators of human implantation measured by
multiplex immunoassay in endometrial secretions.
Results
8.6% of women had BV as indicated by a Nugent score > 6. Multivariable linear
regression analysis showed a significant positive association between IL-1β and the
Nugent score and a significant negative association between IL-10 and Eotaxin and the
Nugent score. No significant differences were found in the ratios of distinct pro- and
anti-inflammatory cytokines in endometrial secretions from women with or without
BV.
Conclusion
A higher Nugent score is associated with a more pro-inflammatory cytokine profile in
endometrial secretions compared to women without BV. However, an effect on
endometrial receptivity is unlikely, since differences are small.
Chapter 3.4 | 113
Introduction
Bacterial Vaginosis (BV) represents a common imbalance of the vaginal bacterial
flora. An overgrowth of anaerobic organisms, such as Gardnerella, Bacteroides and
Mobinculus, leads to replacement of lactobacilli and an increase in vaginal pH from
less than 4.5 to as high as 7.0 (Morris et al. 2006). While BV may be associated with a
malodorous vaginal discharge, up to 40-50% of women are asymptomatic (Klebanoff
et al. 2004). BV is known to be a risk factor for preterm birth (Leitich et al. 2007) and
antibiotic treatment before 20 weeks of gestation has been shown to reduce this risk
(McDonald et al. 2007).
There is evidence that the pathogenic effects of BV may not be confined to

complications during pregnancy, but may also interfere with embryo implantation. BV
has been reported to be associated with a significant increased risk of miscarriage in
the first trimester in women undergoing IVF, independent of other risk factors (Ralph
et al. 1999). Another study concluded that women with BV compared to women with
a normal flora may have decreased conception rates, namely 30% and 52%
respectively (Eckert et al. 2003). However, these results were of borderline
significance (p = 0.06) in a cohort of 91 patients and could not be confirmed by others
(Liversedge et al. 1999; Gaudoin et al. 1999).
The mechanism by which BV may diminish IVF outcome is unclear. The term
‘bacterial vaginosis’ rather than ‘vaginitis’ is used, because of the absence of
associated inflammation. However, there is evidence for an inflammatory response to
BV. It has been shown that almost half of patients with symptomatic BV showed
histopathological evidence of plasma cell endometritis (Korn et al. 1995). In addition,
a number of studies have shown correlations between the concentrations of cytokines
in both vaginal fluid and cervical mucus and the presence of BV (Platz-Christensen et
al. 1993; Mattsby-Baltzer et al. 1998; Sturm-Ramirez et al. 2000; Spandorfer et al.
2001; Cauci et al. 2003; Hedges et al. 2006). The effects of BV on the endometrium
may be brought about by endotoxins, such as Lipopolysaccharides (LPS), a
component of the outer membrane of Gram-negative bacteria. Significantly
diminished pregnancy rates after IVF have been reported when endotoxin levels > 200
pg/mL were found in menstrual effluent (Kamiyama et al. 2004). Microbial flora of
the cervix could be identified on embryo transfer catheters in up to 50% of cases, and
has been reported to adversely affect IVF outcome (Egbase et al. 1996; Fanchin et al.
1998; Moore et al. 2000).
It is unclear whether cytokine patterns in the uterine cavity are altered in the presence
of BV. Successful implantation appears to be associated with a predominance of anti-
inflammatory cytokines (Wegmann et al. 1993). Any disturbance of the delicate
immune balance in pro-and anti-inflammatory cytokines at the maternal fetal interface
may result in failure of implantation or early pregnancy loss (Chaouat et al. 1991).
Next to a shift in pro- and anti-inflammatory cytokines, a dysregulation in the
expression of other mediators involved in embryo implantation and endometrial
maturation (Arcuri et al. 1999; Caballero-Campo et al. 2002; Dey et al. 2004; Ledee-
Bataille et al. 2005; Lessey et al. 2002) may also reduce endometrial receptivity in
women with BV. The objective of this study was to elucidate whether BV colonizing
the cervix is associated with a more pro-inflammatory cytokine profile and to assess
the correlation between the Nugent score and the levels of a number of key regulatory
cytokines, chemokines and growth factors in endometrial secretions aspirated prior to
embryo transfer.

Study design
Ethical approval was received from the local institutional Ethics Committee and
informed consent was obtained from all participants. Between December 2005 and
October 2006, one hundred and ninety eight patients attending the University Medical
Center Utrecht (The Netherlands) for IVF or intracytoplasmatic sperm injection (ICSI)
treatment, who had undergone no more than one prior embryo transfer and showed no
signs of hydrosalpinges on the ultrasound, consented to participate in this prospective
cohort study.
In-vitro fertilization
The IVF procedure has been described previously in detail (Boomsma et al. 2009). A
maximum of 2 embryos were transferred 4 days after oocyte retrieval. Women with
Chapter 3.4 | 115
endometriosis or tubal pathology received an oral dose of ampicillin with clavulanic

acid (875/125 mg) and doxycycline (200 mg), two hours prior to oocyte retrieval. A
urinary pregnancy test was performed 14 days after embryo transfer and if positive, an
ultrasound was performed at nine weeks of gestation.
Study procedure
Immediately prior to embryo transfer, participants underwent an endometrial secretion
aspiration and screening for BV with a Gram stained smear. BV was assessed from an
(endo) cervical smear, since colonization of the cervix rather than the vagina may
increase the risk of intra-uterine disturbances.With the patient lying in lithotomy
position and after insertion of the speculum without lubrication, a sterile cotton swab
was used to obtain a sample from the cervix and endocervix for diagnosis of BV. The
swab was rolled onto a glass slide, fixed with acetone, and allowed to dry in air.
Afterwards, the cervix was cleansed and endometrial secretion was aspirated using the
technique described previously in detail (Boomsma et al. 2009; van der Gaast et al.
2003). Briefly, an embryo transfer catheter (Wallace: SIMS Portex Ltd., Hythe, Kent,
UK) was introduced transcervically. Suction was gradually applied with a 2 mL
syringe. Cervical mucus contamination was prevented by covering the cervix with the
outer layer of the embryo transfer catheter, while removing the inner part of the
catheter after aspiration. The embryo transfer procedure was carried out thereafter. The
aspirate was snapfrozen in liquid nitrogen and stored at -80°C. All patients were
questioned on the day of embryo transfer regarding symptoms associated with BV,
namely an abnormal aspect or malodorous vaginal discharge, or itching and burning
symptoms.

The procedure employed to analyse the endometrial secretions has been previously
described in detail (Boomsma et al. 2009). Briefly, the endometrial secretion aspirates
were analysed using a multiplex immunoassay detecting 17 soluble cytokines,
chemokines and growth factors. The factors included in the assay are known to be
involved in the process of implantation and a description of these mediators is
provided in Boomsma et al. Distinct pro-inflammatory cytokines are interleukin (IL)-
1ß, IL-12, IL-18, interferon (IFN)-γ, Tumour Necrosis Factor (TNF)-α and
Macrophage migration Inhibitory Factor (MIF), whereas IL-5 and IL-10 have more
anti-inflammatory properties. IL-6, IL-15 and IL-17 have both pro- and anti-
inflammatory characteristics. Three chemokines involved in implantation were
included in the assay, namely Eotaxin, Monocyte Chemotactic Protein (MCP)-1 and
Interferon-γ inducible 10 kD protein (IP-10). Vascular Endothelial Growth Factor
(VEGF) and Heparin binding Epidermal Growth Factor (Hb-EGF) are growth factors.
Lastly, Dickkopf homolog (Dkk)-1 inhibits Wnt signaling proteins, which comprise a
family of molecules with pivotal roles in cell proliferation and differentiation. MIF
was determined in 400x diluted samples using a commercially available enzyme
linked immunosorbent assay (ELISA) (R&D systems, Abingdon, United Kingdom).
To investigate whether BV was associated with a shift towards a pro- or anti-
inflammatory profile, the ratios of TNF-α or IL-1β divided by IL-10 or IL-5 were
compared in women with or without BV. The total protein concentration was
measured by NanoOrange Protein Quantitation Kit (Molecular probes Europe B.V.,
Leiden, The Netherlands) and was used for normalization purposes.
Endocervical slides for diagnosis of BV

The slides were Gram-stained and examined microscopically under oil immersion at x
1.000 magnification. Specimens were scored according to a standardized scoring
system providing a 0 to 10 point scale for the evaluation of vaginal flora (Nugent et al.
1991). The scale is based on a weighted sum of different bacteria with certain
morphology, namely Lactobacilli (large gram-positive rods), Gardnerella Vaginalis
and Bacteroides (small gram-negative to -variable rods) and Mobinculus spp (curved
gram-negative rods). Each morphotype was quantitated from 1 to 4+ with regard to the
number of morphotypes per oil immersion field (Table 1). A Nugent score from 0-3 is
considered normal, 4-6 intermediate and 7-10 is the criterion for BV. The slides were
independently scored by two trained observers (NB and CB), who were blinded for
clinical parameters or pregnancy outcome.
The Gram stain technique employed in this study is considered to be the gold standard
for the diagnosis of BV (Nugent et al. 1991). Its use is usually restricted to research
settings since it is more cumbersome to use than the Amsel criteria (Amsel et al.
1983), which are more commonly applied in clinical practice. These criteria use
surrogate markers, such as a homogeneous vaginal discharge, an amine odour when
potassium hydroxide solution is added, the presence of clue cells on microscopy and a
Chapter 3.4 | 117
vaginal pH greater than 4.5. Compared to the Amsel criteria, the Nugent score allows
for assessment of alteration in vaginal flora as a continuum rather than a dichotomy.
Culture of vaginal discharge has no role in diagnosis of BV, since there are no bacteria
that are specific for BV. DNA probes for Gardnerella Vaginalis are expensive and
cannot be used to diagnose bacterial vaginosis because it can be cultured from the
vagina of more than 50% of normal women (Amsel et al. 1983). Nugent score has
been favored for diagnosing BV due to superior reproducibility and sensitivity (Krohn
et al. 1989). The Papanicolaou smear has been shown to be poorly reliable for the
diagnosis of BV (Greene et al. 2000).
Table 1: Scoring system of the Nugent criteria for diagnosis of bacterial vaginosis
Score Lactobacillus Gardnerella and Curved gram-

Bacteroides spp. variable rods
0 4+ 0+ 0
1 3+ 1+ 1+ or 2+
2 2+ 2+ 3+ or 4+
3 1+ 3+
4 0+ 4+
Average number of bacteria with a certain morphology (morphotypes) /field: 0+ equals 0 morphotypes, 1+
equals < 1 morphotypes, 2+ equals 1-4 morphotypes, 3+ equals 5-30 morphotypes, 4+ equals > 30 morphotypes.
Adapted from Nugent et al. (1991).
On the basis of data previously derived and reported and the incidence of BV
previously reported in literature, a power calculation was performed to assess the
feasibility of the study. The value of the ratio of TNF-α and IL-10 concentrations in
endometrial secretions ranges from less than 0.12 up to 32, with an observed standard
deviation (SD) of 7.5. Highest values may be expected in women with BV. With an
expected incidence of BV of 25% (Liversedge et al. 1999; Ralph et al. 1999; Wilson
et al. 2002), 197 patients would allow us to identify a significant difference in the ratio
of TNF-α over IL-10 of 0.46 SD (Cohen’s d) with 80% power. Therefore, the present
study was considered to be sufficiently powered for this outcome. The ratio of TNF-α
and IL-10 was utilized because of the well defined pro-inflammatory character of
TNF-α, and since the other anti-inflammatory cytokine in our panel (IL-5) showed a
higher number of values below the detection limit.
Analyses were performed using SPSS version 12.0 (SPSS Inc., Chicago, IL, USA). A
p-value less than 0.05 was considered to be statistically significant. Comparisons of
clinical parameters in women with and without BV were performed using the
independent t-test or Mann-Whitney U test for continuous data and the χ2-test for
binary variables. Before analysis, all cytokine values were corrected for the total
protein content in the sample. Thereafter, data were log transformed to obtain a normal
distribution. Nine mediators (IL-1ß, IL-6, IL-12, IL-18, TNF-α, MIF, MCP-1, IP-10,
VEGF) were detectable in 90-100% of the samples. Eotaxin was detectable in 88% of
samples, IL-15 in 76% of samples, Dkk-1 68%, IL-10 56%, IL-17 55%, IL-5 34% and
Hb-EGF was detectable in 23% of samples. IFN-γ was not detectable in any of the
samples, and this was not due to technical problems (Boomsma et al. 2009). For
statistical analysis of values below the detection limit of the assay, we imputed data by
using the lower detection limit divided by √2, as recommended for Log transformed
data by Schisterman et al. (Schisterman et al. 2006). Blood contamination of the
endometrial secretion aspirate can affect the quantitation of a number of cytokines by
the multiplex immunoassay method, as reported previously (Boomsma et al. 2009).
The extent of blood contamination in the endometrial secretion samples was therefore
visually graded and included as a covariate in the analyses.
The ratios of TNF-α or IL-1β divided by IL-10 or IL-5 were calculated and
subsequently log transformed to obtain a normal distribution. In order to investigate
whether the ratios of distinct pro- and anti-inflammatory cytokines were different in
women with or without BV one-way analysis of covariance (ANCOVA) was
performed with blood contamination (polynomial scale) and BV (Nugent score > 6
versus ≤ 6) as fixed factors. In order to include all participants, women with normal
and intermediate flora were combined. However, since the mediators included in the
assay are known to be highly mutually correlated, and molecular redundancy is
present, multivariable analysis with backward elimination of mediators is the preferred
method of analysis. The correlation between the concentration of mediators and the
Nugent score on a 0-10 point scale was assessed by multivariable linear regression
analysis with a backward elimination procedure of the soluble mediators. A p-value >
Chapter 3.4 | 119
0.1 was used as a criterion for exclusion. Blood contamination was included as a
covariate.
Results
From 198 women who underwent endometrial secretion aspiration and a cervical
smear for diagnosis of BV, 197 were included in the analysis. One patient was
excluded, since insufficient material was available to measure the total protein content
required for normalization purposes. Samples from seventeen women (8.6%) met the
criteria for diagnosing BV (Nugent score > 6). The flora of 43 women (21.8%) was
considered intermediate according to the Nugent score (> 4) and the flora of 137
women (69.5%) was considered to be normal. None of the 197 women displayed
symptoms of abnormal discharge at inspection during embryo transfer. On
questioning, 7.6% of women complained of abnormal vaginal discharge or itching or
burning symptoms. BV was diagnosed on Gram staining in 13.3% of these cases,
which did not significantly differ from the 8.2% prevalence of BV in women without
symptoms on questioning (p = 0.5). Univariate analyses of clinical parameters of
women with BV versus women with normal or intermediate flora are presented in
Table 2. No significant differences in demographic parameters between both groups
were observed. With regard to treatment outcome, significantly less oocytes were
retrieved (p = 0.015) and less embryos developed (p = 0.062) in women with BV.
However, with the number of oocytes retrieved included as a covariate, no significant
difference in the number of embryos developed was found by ANCOVA analysis (p =
0.6). No significant group differences were observed in the quality of the embryo
transferred, pregnancy rates and early pregnancy loss rates in women with or without
BV, but it should be noted that the study was not sufficiently powered to discern such
a difference. The number of oocytes retrieved was significantly lower among women
with BV. It was therefore included as a covariate in further analyses, since the ovarian
response has been shown to significantly alter the intra-uterine cytokine profile
(unpublished data). The number of women who smoked did not significantly differ
between women with or without BV. However, since the observed difference was
sizeable (0% versus 17% in women with BV), smoking was also included as a
covariate.
Table 2: Clinical parameters of women with BV (Nugent score ≥7) versus women with
normal or intermediate flora (Nugent score <7)
Parameter Bacterial Normal flora and p-

Vaginosis Intermediate flora value
(n= 17) (n=180)
Age (years) 36.7 ± 3.9 (28-44) 34.8 ± 4.1 (26-44) 0.080a
Duration of infertility (years) 2.8 ± 1.8 3.2 ± 1.7 0.3a
Primary type of infertility (%) 0.4b
Andrological 65 48
Unknown 12 23
Tuba pathology 5.9 19
Endometriosis 5.9 1.7
Other 12 7.8
BMI 25.8 ± 5.3 24.3 ± 4.4 0.20a
Smoking (%) None 16 0.073b
Number of women treated with 18 17 1.0b
antibiotics prior to oocyte retrieval
(%)
IVF (%) 60 58 1.0b
ICSI (%) 40 42
Number of oocytes retrieved (n) 5.7 ± 3.7 8.4 ± 4.8 0.015c
Number of embryos available (n) 2.6 ± 1.3 3.8 ± 2.4 0.062c
Number of embryos transferred (n) 1.7 ± 0.5 1.5 ± 0.5 0.18b
High quality embryo (%) 24 37 0.3b
Pregnancy rate (%) 47 33 0.2b
Miscarriage / chemical pregnancy 25 31 0.6b
rate per pregnancy (%)
Number of women with subjective
complaints on questionnaire (%) 0.5b
Complaints absent 88 93
Complaints present 12 7
All values are mean ± SDs. Range between brackets. a Independent t-test. b χ² test. c Mann-Whitney U test.
Intra-uterine cytokine profile and presence of BV

To elucidate whether BV is associated with a pro-inflammatory cytokine profile, ratios
of distinct pro- and anti-inflammatory cytokines were compared for women with and
without BV. ANCOVA analysis, with blood contamination, smoking and BV (Nugent
score > 6 versus ≤ 6) as fixed factors and the number of oocytes retrieved as a
covariate, did not show any significant differences. Figure 1: TNF-α / IL-10 (p = 0.5),
IL-1β/ IL-10 (p = 0.2), TNF-α / IL-5 (p = 0.7) and IL-1β/ IL-5 (p = 0.99). These
results did not alter when women with BV (Nugent score > 6) were compared with
Chapter 3.4 | 121
women without BV (Nugent score 0-3), neither when all values below the detection
limit, which have been imputed, were excluded.
For women with or without BV, the concentrations of pro- and anti-inflammatory
cytokines are presented in Figure 2, cytokines with both pro- and anti-inflammatory
characteristics in Figure 3 and other mediators in Figure 4. An independent t-test
showed a significantly lower concentration of Eotaxin in women with BV versus
women without BV (Nugent score > 6 versus ≤ 6). Multivariable linear regression
analysis with a backward elimination procedure of the soluble mediators and smoking,
the number of oocytes retrieved and blood contamination included as covariates,
showed a significantly positive association between IL-1β (p = 0.001) and the Nugent
score (1-10 scale). A significantly negative association between Eotaxin (p = 0.005) as
well as IL-10 (p = 0.033) and the Nugent score was found. All results from
multivariable analyses were confirmed when IL-5 and Hb-EGF with more than 50%
imputed data were excluded from analyses.
Figure 1: Ratios of distinct pro- and anti-inflammatory cytokines in women with or without
BV.
BV: Bacterial Vaginosis (Nugent score ≥ 7), n=17. No BV: normal or intermediate flora (Nugent store < 7),
n=180.
Figure 2: Concentrations of pro- and anti inflammatory cytokines in women with or without
BV.
n=180. Concentration mediator expressed in pg/mg protein. p-values provided for independent t-test of Log
transformed concentrations of mediators in women with or without BV.
Figure 3: Concentrations of cytokines with both pro- and anti-inflammatory characteristics in

women with or without BV.
Chapter 3.4 | 123
Figure 4: Concentrations of other mediators in women with or without BV.
Discussion
The hypothesis tested in this study was that BV, assessed from the cervix, may be
associated with a pro-inflammatory cytokine profile in the uterus. No significant
differences in the ratios of distinct pro- and anti-inflammatory cytokines were
observed comparing endometrial secretions obtained from women with or without BV.
However, taking into account the strong mutual correlations between the mediators,
multivariable analysis is the preferred method of analysis to reveal deviatons in
cytokine profiles. This analysis revealed a significantly positive correlation between
the endometrial secretion concentration of IL-1β and the Nugent score and a
significantly negative correlation between IL-10 as well as Eotaxin and the Nugent
score. IL-1β is a cytokine with pro-inflammatory characteristics, whereas IL-10 has
anti-inflammatory characteristics. Therefore, a more pro-inflammatory cytokine
profile in endometrial secretions in women with a higher Nugent score, assessed from
the cervix, is present. Previously, it has been shown that almost half of patients with
symptomatic BV showed histopathological evidence of plasma cell endometritis (Korn
et al. 1995). A marked difference in absolute numbers of cytokine concentrations in
lochia of symptomatic women with endometritis has been reported previously
(Sukhikh et al. 2005). The concentration of IL-1β and TNF-α increased 1.4 and 2.7
fold respectively in women with endometritis versus controls. In the present study of
women with mainly asymptomatic BV, we observed cytokine concentrations which
were within the normal range compared to controls with a tendency to a more pro-
inflammatory profile in women with a higher Nugent score.
Cervical smears from just seventeen women (8.6%) met the criteria for BV, namely a
Nugent score between 7 and 10. Complaints of abnormal vaginal discharge or itching
or burning symptoms by the participants of this study, was not associated with a
higher or lower prevalence of BV. This observation may be explained by the fact that
BV was screened from the cervix rather than the vagina. This percentage of women
with BV was lower than the prevalence of 25% in subfertile women previously
reported (Liversedge et al. 1999; Ralph et al. 1999; Wilson et al. 2002), although other
studies also reported lower incidences of BV, namely 16% (Eckert et al. 2003) or
4.2% (Spandorfer et al. 2003). However, in the latter study all patients received
tetracycline prophylaxis at the time of oocyte retrieval (Spandorfer et al. 2001). In the
present study 17% of women with either endometriosis or tubal pathology diagnosed
on laparoscopic inspection, received a prophylactic dose of ampicillin with clavulanic
acid and doxycycline prior to oocyte retrieval. The use of prophylactic antibiotics in
this group may explain the lower prevalence of BV observed. However, the same
percentage of women received antibiotics in women with or without BV and the
percentage of women with BV was similar among women who did receive antibiotics
versus women who did not, 8.8% and 8.6% respectively. Moreover, doxycycline
treatment at oocyte retrieval has been reported to have no substantial impact on the
presence of BV (Moore et al. 2000). Although ampicillin is significantly less effective
than the recommended therapy for BV by metronidazole or clinadamycine (Piot et al.
1984; Joesoef et al. 1999), ampicillin with clavulanic acid has been shown to be
effective against bacterial colonization of the cervix in women undergoing IVF (Salim
et al. 2002). The difference in the prevalence of BV between the present study and the
prevalence reported in literature may also be related to the method of diagnosis. The
Chapter 3.4 | 125
above mentioned studies used a Gram stained vaginal smear, whereas we assessed a
Gram stained cervical smear since colonization of the cervix rather than the vagina
was considered more likely to increase the risk of intra-uterine disturbances. The
power calculation used in this study assumed a higher incidence of BV than observed.
With an incidence of 8.6% BV, 197 patients would still allow us to identify a
significant difference in the ratio of TNF-α and IL-10 of 0.7 SD (Cohen’s d).
Studies have shown associations between BV and cytokine levels in vaginal fluid or
cervical mucus. As yet, the influence on the intra-uterine milieu has not been assessed.
In women with BV, higher levels of endotoxins in both vaginal fluid and cervical
mucus have been observed (Larsson et al. 1991), which can induce the production of
pro-inflammatory cytokines, such as TNF-α and IL-1β (Romero et al. 1989; Romero
et al. 1991), and also anti-inflammatory cytokines, such as IL-5 and IL-10 (Masuda et
al. 2002). Cervical mucus from women with BV containing endooxins was shown to
induce IL-6 expression in monocytic cells in vitro (Sturm-Ramirez et al. 2000). In
vivo, the concentrations of IL-1β (Cauci et al. 2003; Hedges et al. 2006; Cauci et al.
2002) and IL-1α (Platz-Christensen et al. 1993) are reported to be significantly higher
in vaginal fluid from women with BV, whereas no differences in IL-8 (Cauci et al.
2003; Hedges et al. 2006), TNF-α and IL-6 (Hedges et al. 2006) were observed. In
cervical mucus rather than vaginal fluid, BV has also been shown to be correlated to a
higher concentration of IL-1β (Mattsby-Baltzer et al. 1998; Sturm-Ramirez et al.
2000; Spandorfer et al. 2001), IL-1α (Platz-Christensen et al. 1993), but also to IL-8
(Spandorfer et al. 2001) and TNF-α (Sturm-Ramirez et al. 2000).
The previously reported positive association between BV and the concentration of IL-
1β in cervical mucus as well as vaginal fluid (Mattsby-Baltzer et al. 1998; Sturm-
Ramirez et al. 2000; Spandorfer et al. 2001; Cauci et al. 2003; Hedges et al. 2006;
Romero et al. 1989; Cauci et al. 2002) is consistent with our findings of an increased
concentration of IL-1β in endometrial secretions from women with BV. We also
observed a significantly negative association between both IL-10 and Eotaxin and the
Nugent score. A decreased concentration of IL-10 is consistent with a pro-
inflammatory cytokine profile in endometrial secretions in women with BV and may
be detrimental to implantation. IL-10 inhibits type 1 immunity at the fetal-maternal
interface and has been shown to be associated with successful pregnancy in mice
(Chaouat et al. 1995). An increased placental production of IL-10 was observed in
normal pregnancies, whereas intraperitoneal injection of anti-IL-10 increased the
resorption rate. Moreover, IL-10-/- mice fail to control intrinsic inflammatory
responses (Kuhn et al. 1993), which may lead to fetal resorption or intra-uterine
growth restriction in response to very low doses of LPS (Murphy et al. 2005). BV has
been reported to be associated with a significant increased risk of first trimester
miscarriage in women undergoing IVF (Ralph et al. 1999). Eotaxin was initially
identified as a specific chemoattractant protein for eosinophils. Eosinophils are
considered the main effector cells in allergic responses and asthma. These eosinophils
have been shown to accumulate in the endometrium just prior to and during
menstruation (Jeziorska et al. 1995) and may therefore also have a role in endometrial
tissue remodeling.
Whether or not BV coincides with an intra-uterine pro-inflammatory cytokine profile

is of interest since this may be detrimental to endometrial receptivity. However, other
mechanisms by which BV may impair pregnancy rates, such as introduction of
bacteria into the uterus at embryo transfer or diminished oocyte fertilization rates due
to contaminated follicular fluid (Nagata et al. 1993; Deb et al. 2004), cannot be
excluded. The presented data indicate that BV does affect the intra-uterine cytokine
profile, but only marginal differences towards a pro-inflammatory milieu were
observed in our study. Therefore, it can be concluded that the impact of BV on
endometrial receptivity is likely to be minor. Screening for BV in women with
implantation failure is not recommended on the basis of these results.
Chapter 4
____________________________________________
Determination of the peri-implantation endometrial

transcript profile in women with recurrent implantation
failure versus controls
Carolien M Boomsma1
Marian JA Groot-Koerkamp 2
Nathalie ACH Brabers2
Dik van Leenen2
Sander R van Hooff2
Bart CJM Fauser1
Cobi J Heijnen4
Frank CP Holstege2
Nick S Macklon1
1
2
Department for Physiological Chemistry, University Medical Center Utrecht
3
4
(Submitted)
Abstract
Objective
Implantation failure constitutes the major limiting step in in-vitro fertilization (IVF)
success rates. The aim of this study was to identify possible differences between the
endometrial transcriptome in women with recurrent implantation failure and controls.
Design en setting
Prospective cohort study. Fertility center in tertiary referral university hospital.
Patients
24 women with idiopathic recurrent implantation failure and 24 controls who achieved
pregnancy after ICSI for severe male factor infertility only.
Methods
Endometrial biopsies were carried out 7 days after the LH surge in a natural cycle.
mRNA from each sample was analysed by DNA microarrays.
Results
75 genes were significantly differentially expressed, 63 genes were upregulated,
whereas 12 genes were downregulated in women with implantation failure. qRT-PCR
for 4 genes showed a good correlation between the microarray and PCR results. Genes
observed to be significantly differently expressed included FOXO-1, MUC16 and
ADAMTS8, involved in decidualization, embryo adhesion and angiogenesis.
Supervised hierarchical clustering was performed to classify samples and was
independently validated on a test set of samples. However, the predictor was unable to
correctly determine whether the endometrial sample is taken from a normal or less
receptive endometrium in on average 45.3% of cases (95% CI: 43.4- 47.2%),
consistent with a heterogeneous aetiology of recurrent implantation failure.
Conclusions
The present study reports the largest set of microarray data on the endometrial
transcriptome in women with implantation failure to date. Dysregulated pathways in
the endometrium in women with recurrent implantation failure have been identified
meriting further investigation.
Chapter 4 | 129
Introduction
Since the birth of the first in-vitro fertilization (IVF) baby in 1978 (Steptoe and
Edwards, 1978), IVF has become a central part of fertility treatment, with more than 3
million IVF children being born to date. However, many couples seeking IVF remain
childless even after multiple attempts. Failure of morphologically sound embryos to
implant constitutes the major limiting step in IVF success rates. Recurrent
implantation failure has been defined as three or more unsuccessful IVF cycles or the
failure of conception after the replacement of 10 or more good quality embryos
(ESHRE PGD Consortium Steering Committee, 2002). The chance of pregnancy per
cycle tends to remain stable in the first three attempts, but declines thereafter
(Templeton and Morris, 1998;Croucher et al., 1998), suggesting an underlying cause
for implantation failure in these patients rather than simply a chance effect. Multiple
aetiologies for implantation failure have been proposed and include embryo
chromosomal abnormalities, elevated numbers of uterine Natural Killer (NK) cells,
autoimmune conditions, thrombophilic gene mutations, structural uterine anomalies,
inadequate embryo transfer technique or a disturbed endometrial receptivity (Dicker et
al., 1992;Gianaroli et al., 1997;Hornstein et al., 2000;Matsubayashi et al.,
2001;Ledee-Bataille et al., 2004;Coulam et al., 2006). If good quality embryos have
been replaced on each occasion and no concurrent abnormalities have been found on
haematological and ultrasound investigation, functional abnormalities of the
endometrium may underlie the failure of implantation.
The endometrium undergoes dramatic cyclical changes in response to changing sex

steroids levels. In the mid-luteal phase, the endometrium becomes receptive to
implantation from approximately cycle day 19 to 24 in a normal menstrual cycle,
which is referred to as the putative ‘window of receptivity’ (Wilcox et al., 1999;Navot
et al., 1991). Although a multitude of complex molecular pathways modulate this
process, no specific individual factor in the endometrium has been shown to be crucial
for implantation in the human (Hoozemans et al., 2004;Dey et al., 2004). The
development of DNA microarray gene expression profiling has made it possible to
investigate the endometrium from a global genomic perspective. Altered expression of
endometrial regulatory genes has been postulated as an underlying cause of infertility
and implantation failure. A number of studies have demonstrated that endometrial
gene expression profiles change throughout the menstrual cycle (Riesewijk et al.,
2003; Carson et al., 2002; Ponnampalam et al., 2004; Borthwick et al., 2003;
Horcajadas et al., 2004; Kao et al., 2002; Mirkin et al., 2005; Talbi et al., 2006). In
addition, endometrial gene expression has been investigated under non-physiological
and pathological conditions, such as ovarian stimulated cycles (Mirkin et al., 2004;
Horcajadas et al., 2005; Macklon et al., 2008), endometriosis (Kao et al., 2003) or
during the use of an intrauterine device (Horcajadas et al., 2006). These studies have
provided indirect evidence of genes involved in endometrial receptivity. However,
many of the gene transcripts changing throughout the menstrual cycle may not be
involved in endometrial receptivity and many dysregulated genes observed in cycles
stimulated with exogenous gonadotropins, or in women with endometriosis may
represent different dysregulated pathways compared to those occurring in women with
implantation failure after IVF treatment.
In the present study we have investigated the global gene expression during the
window of implantation in 24 normo-ovulatory women with idiopathic recurrent
implantation failure after IVF compared with control subjects with proven normal
implantation. The primary aim of the study was to identify genes whose dysregulation
was associated with implantation failure. Secondly, we wished to identify biological
processes affected in women with implantation failure.

Study design and tissue collection
Ethical approval was received from the local institutional Committee and written
informed consent was obtained from all patients. Between June 2006 and November
2007, 25 patients treated by IVF or intracytoplasmatic sperm injection (ICSI) who had
undergone at least 3 consecutive fresh embryo transfers without achieving a positive
pregnancy test and with regular menstrual cycles of 25-35 days, participated in this
prospective cohort study. Patients in which an embryo factor was expected to be the
cause of implantation failure, or with poor ovarian response (<4 oocytes retrieved) on
adequate ovarian stimulation, were excluded. All patients underwent extensive
screening for thrombophilias and were screened for abnormalities in prothrombin time
(PT-INR), antithrombin activity, Protein C or S activity, activated protein C (APC)
resistance, and factor VIII activity, presence of a Factor V Leiden mutation,
Chapter 4 | 131
Prothrombin (factor II) mutation, lupus anticoagulant, and anticardiolipin antibodies.

Patients were also screened for abnormalities in glycosylated haemoglobin (Hb A1c),
and thyroid-stimulating hormone (TSH) levels. Only one patient showed a
heterozygous Prothrombin mutation. This patient was not excluded from the study,
since no association with implantation failure has been found in previous studies
(Qublan et al., 2006; Azem et al., 2004; Martinelli et al., 2003), except for one study
with a smaller sample size including only 18 women with implantation failure
(Grandone et al., 2001). 25 women who were formerly treated by ICSI for severe male
factor infertility only and who achieved a healthy live birth after the first or second
treatment cycle were invited to participate as controls. Women with any possible
fertility undermining factors were excluded. These supposed normal fertile women had
normal menstrual cycles and did not use oral contraceptives or an intra-uterine device.
All patients and controls were aged ≤38 years. An endometrial biopsy was scheduled 6
or 7 days after the putative luteinizing hormone (LH) surge in a natural cycle in
patients and controls, which is considered to be the representative day of the window
of implantation. Urinary LH was monitored by a home LH ovulation predictor kit
(Ovulady, Clindia Benelux, The Netherlands). Biopsies were collected using an
Endobiops endometrium sampling device (Gynotec, The Netherlands) under sterile
conditions. The sample was snap frozen in liquid nitrogen and stored at -80 °C until
use. Among patients, the biopsy was not scheduled in the first natural cycle occurring
after an unsuccessful IVF or ICSI cycle.
Gene expression profiling

RNA isolation
Total RNA was isolated from individual tissue samples using Trizol reagent
(Invitrogen) following the manufacturer’s protocol, followed by a purification using
the RNeasy Mini Kit (Qiagen) and a DNAse treatment using the Qiagen DNA-free kit.
The yield and quality of total RNA was checked by spectrophotometry and by the
Agilent Bioanalyser (Agilent). Two biopsies, derived from one patient and one
control, were excluded on the basis of the RNA yield and cRNA yield respectively.
The remaining 48 samples showed good RNA quality, with clear 18S and 28S
ribosomal bands and RNA integrity numbers (RIN) from 7.1 to 10 (mean 8.9, standard
deviaton [SD] 0.6).
cRNA synthesis and fluorescent labelling

All amplification and labelling procedures were performed in 96 wells plates (4titude,
Bioke) on a customized Sciclone ALH 3000 Workstation (Caliper LifeSciences),
supplemented with a PCR PTC-200 (Bio-Rad Laboratories), SpectraMax 190
spectrophotometer (Molecular Devices), and a magnetic bead-locator (Beckman).
cRNA products were purified and concentrated with RNAClean (Agencourt,
Beckman) according to manufacturers protocol. mRNA was amplified by in vitro
transcription using T7 RNA polymerase on 1 µg of total RNA. First a double stranded
cDNA template was generated including the T7 promoter. Next, this template was
used for in vitro transcription with the T7 megascript kit (Ambion) to generate cRNA.
During the in vitro transcription, 5-(3-aminoallyl)-UTP (Ambion) was incorporated
into the single stranded cRNA. Samples with a yield less than 5000 ng or with small
cRNA fragments (median less than 500 bp) were not used. Cy3 or cy5 fluorophores
(GE Healthcare) were coupled to 2500 ng of cRNA. The yield and label incorporation
(1.5-3%) of the cy-labelled cRNA was checked using spectrophotometry. Before
hybridization, 2500 ng of cy-labelled cRNA from one biopsy was mixed with an equal
amount of reverse colour cy-labelled material from the reference sample.
Microarray hybridization
For each sample, two expression profiles in dye-swap experiments were performed.
The samples of patients and controls were compared against a commercial reference
(Universal Human Reference RNA catalog #740000, Stratagene). The Human Array-
Ready Oligo set (version 2.0) was purchased from Qiagen and spotted on Codelink
slides (GE Healthcare) according to the manufacturer’s protocol. The microarrays
contained 70-mer oligo-nucleotides representing 21.329 human genes and expressed
sequence tags (ESTs), as well as 3871 additional spots for control purposes. Gene
annotations were updated by BLAST analysis of all feature sequences to build 49 of
ENSEMBL, including 1000nt UTR-space. Arrays were hybridized on a Tecan
HS4800PRO hybridization station, using the protocol described previously (Peppel
van de, Kemmeren et al., 2003). Hybridized slides were scanned on an Agilent scanner
(G2565AA) at 100% laser power and 30% PMT.
Chapter 4 | 133
Data accessibility
In accordance with proposed MIAME (Minimum information about a microarray
experiment) standards (Brazma et al., 2001), primary data are deposited in Array
Express (http://www.ebi.ac.uk/microarray-as/aer) under accession number E-TABM-
567.
Data analysis
After automatic data extraction using Imagene 8.0.1 (BioDiscovery), printtip Loess
normalization was performed on mean spot intensities (Yang et al., 2002). The
normalized data were then subjected to statistical analysis and fold change filtration.
Patient and controls were compared statistically by using ANOVA (R version
2.4.0/MAANOVA version 1.4.1, http://www.r-project.org/) (Wu H et al., 2003). In a
fixed effect analysis, sample, array and dye effects were modelled. P-values were
determined by a permutation F2-test in which residuals were shuffled 5000 times
globally. Genes with a p-value <0.05 after Benjami-Hochberg multiple testing
correction were considered significantly differentially expressed. After the primary
analyses of 24 patients and 24 controls, a subgroup analysis of patients and controls
meeting a stricter criterion of recurrent implantation failure and healthy control
respectively, was performed. Seven patients with more than 10 embryos transferred in
total without an implantation and who did not conceive after study participation, were
compared to 16 controls who did not have any embryo transferred without achieving
pregnancy.
Differentially expressed genes between controls and women with implantation failure
(either the primary analysis or the two subgroup analysis) were queried for their gene
ontology classifications. A web-based tool (http://integromics.holstegelab.nl/) was
used to perform the statistical analysis. A given biological process is considered
enriched in the input gene list compared to the reference gene list (21.329 genes),
when the observed number of genes in that category is significantly greater than the
expected number, with a p-value <0.05 using a Benjami-Hochberg correction for
multiple testing.
An unsupervised hierarchical clustering was performed on samples to group them

according to similarities in their gene expression patterns (GeneSpring 7.2 Silicon
Genetics). In addition, we attempted to construct a gene set, a classifier, able to

distinguish between women with implantation failure and controls using k-nearest
neighbour (knn). The data was divided in a training set and a test set, which was not
used in gene selection to avoid a selection bias (van 't Veer et al., 2002; Clark et al.,
2000). The first step in gene selection was based on a signal-to-noise-ratio (SNR)
ranking (Golub et al., 1999) combined with forward selection to find the optimum
gene set. This procedure was done using a 5-fold cross-validation repeated 10 times
within the training set. In a second gene selection step the genes were ranked based on
the frequency of occurrence in the 50 gene sets resulting from the first step. The final
optimum gene set was determined again using forward selection in a 5-fold cross-
validation repeated 10 times within the training set. Finally the independent test set
was used to asses the performance of the final classifier. In order to investigate
whether the classification is better than chance, the procedure described above was
repeated using data with randomized class labels.
Real-time PCR verification of gene expression determined by microarray

analyses
Microarray findings were validated using quantitative RT-PCR. Four genes were
selected because they exhibited a significant change in expression in women with
implantation failure versus controls or because they were of particular biological
interest. The housekeeping gene RPL19 was used as a normaliser gene because the
microarray data displayed a low variation in expression between subjects with a level
of expression comparable with the genes of interest for PCR confirmation. Primers
were designed using Beacon Designer Software. The oligonucleotides sequences used
for FOXO1, MUC16, Cyp26A1, SULT1E1 and RPL19, were 5’-
AAGAGCGTGCCCTACTTCAA-3’, 5’- GCCTCTACCTTAACGGTTACAATGAA-
3’, 5’- AATGACCCGCAATCTCTTCTC-3’, 5’- TTGCCACCTGAACTTCTTC-3’,
5’- CCATGAGTATGCTCAGGCTTCA -3’, for sense primers, and, 5’-
CTGTTGTTGTCCATGGATGC-3’, 5’- GGTACCCCATGGCTGTTGTG-3’, 5’-
CCTCTCCCACGAGTGCTC-3’, 5’-GGAACCATAAGGAACCTGTC-3’, and 5’-
CTGACGGGAGTTGGCATTG-3’ for anti-sense primers respectively. Real-time one-
step RT-PCR analysis was performed as described elsewhere (Eijkelkamp et al. 2007).
Differences in expression between patients and controls were analyzed by an
Chapter 4 | 135
independent t-test (SPSS version 12.0, Chicago, IL, USA, 1999). Bivariate correlations
between microarray and PCR results were analysed by Pearson correlation.
Results
Endometrial biopsies suitable for analysis were obtained from 24 women with
recurrent implantation failure and 24 controls with severe male factor infertility only.
The male partners of control patients had a history of sterilization, varicocele,
orchiopexy, chemotherapy, congenital bilateral absence of the vas deferens (CBAVD)
or epididymitis in 58% of cases. The clinical characteristics of the women studied are
given in Table 1. Among controls, 53 embryo transfer procedures of a total of 91
embryos were performed. These resulted in 41 pregnancies from 49 implanted
embryos.
Table 1: Characteristics of patients and controls included in this study and treatment history
Parameter Patients Controls p-

(n=24) (n=24 ) value
Age (years) 33.9 ± 3.0 (27-38) 35.1 ± 2.0 (32-38) 0.12 A
Body mass index 23.2 ± 4.7 (19-37) 24.7 ± 5.2 (19-46) 0.3 A
Smoking (n) 1 3 0.3 B
Primary infertility (%) 67 Not applicable
Primary cause of infertility (%)
Andrological 50 100
Unknown 17
Tubal pathology 33
Treatment history before biopsy Not applicable
Previous embryo transfers (n) 5.1 ± 1.5 (3-10)
Transferred embryos (n) 8.5 ± 3.5 (4-20)
High quality C embryos transferred (n) 2.9 ± 2.3 (0-9)
All values are mean ± SDs. Range between brackets.

A
Independent t-test. B χ² test. C Defined as a compacted morula stage embryo with 0-10% fragmentation,
four days after oocyte retrieval.
The primary analysis revealed that 75 genes were significantly differentially expressed
in 24 women with recurrent implantation failure. The expression of 63 genes was
significantly upregulated in women with implantation failure (Table 2), and the
expression of 12 genes was significantly downregulated (Table 3). A subgroup
analysis of 7 patients and 16 controls meeting stricter criteria for both recurrent
implantation failure (>10 embryos transferred) or healthy control (pregnant after first
embryo transfer) respectively, was performed. In these clinically more strictly defined
subgroups, 76 genes were significantly upregulated and 13 genes significantly
downregulated among patients compared to controls. Table 4 shows 16 genes with an
increased expression among women with implantation failure with an average fold
change >1.5 and a p-value < 0.05. 17 genes among these 89 genes were also
significantly differentially expressed in the primary analysis of all patients and
controls and are shown in bold in Table 2-4.
In the primary analysis, average fold changes in gene expression of women with
implantation failure and controls were relatively small, ranging from a fold change of
1.47 up to 1.22 down. However, in the subgroup analysis in which a stricter criterion
of recurrent implantation failure and healthy control was applied, average fold changes
in gene expression were higher, namely up to 2.67 fold change. In order to test
whether the variance in gene expression was larger among patients compared to
controls, possibly due to a large variability in patients, an F-test (ratio of variances
test) was performed (p-value <0.05 after Benjami-Hochberg multiple testing). Since
one control showed an aberrant gene expression profile and was identified as an
outlier, this sample was excluded from this analysis. It was, however, not excluded
from the primary analysis since no clinical or procedural abnormalities could be
identified. A higher number of genes demonstrated significantly greater variance in
expression among patients versus controls than vice versa, namely 1143 genes versus
134 genes respectively. Figure 1 illustrates this, demonstrating individual fold changes
in expression of 4 genes in patients and controls compared to the commercial reference
RNA.
Chapter 4 | 137
Table 2: Genes significantly upregulated in women with implantation failure versus controls,
identified in the primary analysis of 24 patients and 24 controls.
Ensembl Gene Description (n=63) Function/ grouping p- Fold
Symbol value change↑
ENSG00000 PRG4 Megakaryocyte-stimulating factor G-protein coupled receptor protein signaling <0.0001 1.21
116690 pathway
Cell proliferation
ENSG00000 SNAI2 Zinc finger protein SLUG Transcription <0.0001 1.30
019549
ENSG00000 COL3A1 Collagen alpha-1(III) chain precursor G-protein coupled receptor protein signaling <0.0001 1.27
168542 pathway
Blood circulation
ENSG00000 TSHZ3 Teashirt homolog 3 Transcription 0.0001 1.19
121297
ENSG00000 HMBOX1 Homeobox-containing protein 1. Regulation of transcription 0.0003 1.29
147421
ENSG00000 FOXO1 Forkhead box protein O1A Transcription 0.0003 1.29
150907 G-protein coupled receptor protein signaling
pathway
Anti-apoptosis
Blood vessel development
Regulation of cell proliferation
ENSG00000 GABRA2 GABA(A) receptor subunit alpha-2 Ion transport 0.0005 1.002
151834 Regulation of neurotransmitter levels
ENSG00000 PDE4A cAMP-specific 3',5'-cyclic Signal transduction 0.0006 1.16
065989 phosphodiesterase 4A
ENSG00000 DYRK2 Dual specificity tyrosine- Protein amino acid phosphorylation 0.0008 1.15
127334 phosphorylation-regulated kinase 2 Apoptosis
DNA damage respons
OTTHUMG0 TNXB tenascin XB isoform 2 Signal transduction 0.0008 1.16
0000031088
ENSG00000 KCND3 Potassium voltage-gated channel Potassium ion transport 0.0035 1.30
171385 subfamily D member 3
ENSG00000 PRKAG2 5'-AMP-activated protein kinase subunit Glycogen metabolic process 0.0044 1.16
106617 gamma-2
ENSG00000 MME Neprilysin Proteolysis 0.0044 1.26
196549 Cell-cell signalling
ENSG00000 NDRG4 Vascular smooth muscle cell-associated Cell differentiation 0.0044 1.16
103034 protein 8 Response to stress
Cell growth
ENSG00000 TMTC1 Transmembrane and TPR repeat- RNA processing 0.0044 1.10
133687 containing protein 1
Genomic location:1-56359985- Unknown 0.0051 1.20
56360053
ENSG00000 ODAM odontogenic ameloblast-associated Odontogenesis of dentine-containing teeth 0.0051 1.31
109205 protein
ENSG00000 FLVCR2 Feline leukemia virus subgroup C Transport 0.0051 1.12
119686 receptor-related protein 2
ENSG00000 FUT4 Alpha-(1,3)-fucosyltransferase Carbohydrate metabolic process 0.0051 1.18
196371 Protein aminoacind glycosylation
ENSG00000 GARNL3 GTPase activating Rap/RanGAP Regulation of small GTPase mediated signal 0.0074 1.15
136895 domain-like 3 transduction
ENSG00000 REV3L DNA polymerase zeta catalytic subunit DNA replication 0.0074 1.16
009413
ENSG00000 RFTN1 Raftlin Unknown 0.0116 1.04
131378
ENSG00000 KYNU Kynureninase Metabolic process 0.0116 1.19
115919
ENSG00000 SYNCRIP Unknown 0.0133 1.088
135316
TUG-1 taurine upregulated gene 1 miscRNA 0.0135 1.13
ENSG00000 Q96L18_ Unknown 0.0135 1.13

212853 HUMAN
ENSG00000 RASAL2 Ras GTPase-activating protein nGAP Regulation of small GTPase mediated signal 0.0137 1.14
075391 transduction
Signal transduction
OTTHUMG0 XIST X (inactive)-specific transcript Unknown 0.0140 1.14
0000021839
ENSG00000 SOCS2 Suppressor of cytokine signaling 2 Regulation of cell growth 0.0142 1.16
120833 Intracellular signalling cascade
Anti-apoptosis
ENSG00000 PLCL1 phospholipase C-like 1 Signal transduction 0.0143 1.19

115896 Intracellular signalling cascade
ENSG00000 Q9HAI8_ CDNA FLJ11561 fis, clone Unknown 0.0157 1.11
170015 HUMAN HEMBA1003142.
ENSG00000 RORA Nuclear receptor ROR-alpha Regulation of transcription, DNA dependent 0.0165 1.09
069667 Signal transduction
ENSG00000 CMTM2 Chemokine- like factor superfamily Chemotaxis 0.0165 1.13
140932 member 2
ENSG00000 HTRA3 Pregnancy-related serine protease Proteolysis 0.0181 1.13
170801 Regulation of cell growth
ENSG00000 C9orf33 Protein HFSE-1. Unknown 0.0181 1.12
204423
ENSG00000 GRHL2 Grainyhead-like protein 2 homolog Transcription 0.0181 1.17
083307
ENSG00000 LMCD1 LIM and cysteine-rich domains protein Negative regulation of transcription from 0.0186 1.36
071282 1 RNA polymerase II promoter
ENSG00000 PBEF1 Pre-B-cell colony-enhancing factor 1 Signal transduction 0.0214 1.19
105835 Cell proliferation
ENSG00000 NP_001030 Unknown 0.0227 1.14
205730 013.1
ENSG00000 DDX21 Nucleolar RNA helicase 2 Protein binding 0.0230 1.13
165732
ENSG00000 CD248 Endosialin precursor Ca ion/ sugar binding 0.0254 1.15
174807
ENSG00000 SULF2 Extracellular sulfatase Sulf-2 precursor Metabolic process 0.0281 1.13
196562
ENSG00000 TLR4 Toll-like receptor 4 precursor Innate immune respons 0.0281 1.16
136869 Signal transduction
Actvation of I/NF-kappaB kinase cascade
Inflammatory respons
ENSG00000 ANXA1 Annexin A1 Cytoskeleton organization and biogenesis 0.0281 1.13
135046 Cell surface receptor signal transduction
Cell proliferation regulation
ENSG00000 KIF1B Kinesin-like protein KIF1B Microtubule-based movement 0.0285 1.02
054523
ENSG00000 SDC2 Syndecan-2 precursor Cell proliferation 0.0306 1.008
169439
ENSG00000 FAM135A KIAA1411 Unknown 0.0353 1.14
082269
ENSG00000 PAQR8 Membrane progestin receptor beta Steroid binding 0.0364 1.14
170915
ENSG00000 N4BP2L2 phosphonoformate immuno-associated ATP binding 0.0364 1.13
139617 protein 5.
ENSG00000 LRRC8B Unknown 0.0364 1.12
197147
OTTHUMG0 HNRPR heterogeneous nuclear ribonucleoprotein Unknown 0.0404 1.11
0000003224 R
ENSG00000 BIRC3 Inhibitor of apoptosis protein 1 Anti-apoptosis 0.0411 1.36
023445 Cell surface receptor signal transduction
ENSG00000 PDGFRA Alpha platelet-derived growth factor Protein amino acid phosphorylation 0.0411 1.16
134853 receptor precursor Cell proliferation
Cell surface receptor signal transduction
ENSG00000 AKAP13 A-kinase anchor protein 13 Intracellular signaling cascade 0.0421 1.13
170776
ENSG00000 MAT2A S-adenosylmethionine synthetase One-carbon compound metabolic process 0.0421 1.13
168906 isoform type-2 Response to hormone stimulus
ENSG00000 RPGR X-linked retinitis pigmentosa GTPase Visual perception 0.0456 1.09
156313 regulator.
ENSG00000 TCEAL3 Transcription elongation factor A Transcription 0.0456 1.14
196507 protein-like 3
ENSG00000 YTHDC1 YTH domain-containing protein 1 mRNA processing 0.0456 1.11
083896
ENSG00000 RTF1 RNA polymerase-associated protein Transcription 0.0456 1.11
137815 RTF1 homolog
ENSG00000 ZNF142 Zinc finger protein 142 Transcription 0.0486 1.13
115568 DNA repair
ENSG00000 Unknown 0.0492 1.089
180715
ENSG00000 CD36 Platelet glycoprotein 4 Cell adhesion 0.0493 1.032
135218 Blood coagulation
ENSG00000 TMEM37 Voltage-dependent calcium channel Ion transport 0.0497 1.19
171227 gamma-like subunit
Gene symbols in bold are also significantly differentially expressed in the subgroup analysis.
Chapter 4 | 139
Table 3: Genes significantly down-regulated in women with implantation failure versus

controls, identified in the primary analysis of 24 patients and 24 controls.
Ensembl Gene Description (n=12) Function/ grouping p- Fold
Symbol value change↓
ENSG00000 TRA2A_ Transformer-2 protein homolog mRNA processing and splicing 0.0013 1.19
164548 HUMAN
ENSG00000 PAFAH1B Platelet-activating factor Nervous system development 0.0133 1.11
079462 3 acetylhydrolase IB subunit gamma Spermatogenesis
ENSG00000 NP_065915 CDNA FLJ30664 fis, clone Unknown 0.0135 1.18
144460 .1 FCBBF1000604
ENSG00000 NCAPD2 Condensin complex subunit 1 Mitosis 0.0140 1.12
010292 Chromosome condensation
Cell division
ENSG00000 CPE Carboxypeptidase E precursor Proteolysis 0.0152 1.22
109472 Neuropeptide signalling pathway
ENSG00000 C17orf81 Dermal papilla-derived protein 6 Transcription 0.0182 1.11
170291
ENSG00000 PDIA3 Protein disulfide-isomerase A3 Signal transduction 0.0281 1.15
167004 precursor Activation of apoptosis
ENSG00000 CA12 Carbonic anhydrase 12 precursor One-carbon compound metabolic process 0.0316 1.19
074410
ENSG00000 PGD 6-phosphogluconate dehydrogenase, Electron transport 0.0316 1.10
142657 decarboxylating Penthose biosynthesis
ENSG00000 MAL T-lymphocyte maturation-associated Cell differentiation 0.0353 1.14
172005 protein Signal transduction
Induction of apoptosis
ENSG00000 CBX1 Chromobox protein homolog 1 Chromatin assembly or disassembly 0.0387 1.10
108468
ENSG00000 FABP3 Fatty acid-binding protein, heart Transport 0.0456 1.20
121769 Negative regulation of cell proliferation
Gene symbols in bold are also significantly differentially expressed in the subgroup analysis.
Table 4: Genes significantly differentially expressed with an average fold change > 1.5 in the
subgroup analysis of 7 patients and 16 controls meeting a stricter criterion.
Ensembl Gene Description Function/ grouping p- Fold
Symbol (n=16) value change↑
ENSG00000 CYP26A1 Cytochrome p450, family 26, subfamily Metabolic process 0.038 2.67
095596 A, polypeptide 1
ENSG00000 SLC46A2 Thymic stromal cotransporter homolog Transport 0.0001 2.37
119457
ENSG00000 SULT1E1 Estrogen sulfotransferase Estrogen metabolic process 0.0001 2.32
109193
ENSG00000 PKHD1L1 Fibrocystin L 0.0246 2.21
205038
ENSG00000 MEGF10 multiple EGF-like-domains 10 Myelination 0.0096 2.07
145794
ENSG00000 SYT2 Synaptotagmin-2 Transport 0.0034 1.93
143858
ENSG00000 OLFM4 Olfactomedin 4 precursor 0.0207 1.71
102837
ENSG00000 BIRC3 Inhibitor of apoptosis protein 1 Anti-apoptosis 0.0330 1.69
023445
ENSG00000 ADAMTS8 ADAMTS-8 precursor Negative regulation of cell proliferation 0.0158 1.62
134917
ENSG00000 PTGS2 Prostaglandin G/H synthase 2 precursor Electron transport 0.0000 1.62
073756
ENSG00000 TTYH2 Tweety 2 isoform 1 Unknown 0.0432 1.56
141540
ENSG00000 CNTD2 Cyclin N-terminal domain-containing Regulation of cell cycle 0.0055 1.55
105219 protein 2
ENSG00000 CEP70 Centrosomal protein of 70 kDa Signal transduction 0.0029 1.54
114107
ENSG00000 TMEM56 Transmembrane protein 56 0.0001 1.53
152078
ENSG00000 NFIB Nuclear factor 1 B-type Transcription 0.0403 1.53
147862
ENSG00000 SOX17 Transcription factor SOX-17 Transcription 0.0451 1.52
164736
Gene symbols in bold are also significantly differentially expressed in the primary analysis.
Figure 1: Histograms of fold changes in gene expression of each patient and control
compared to the commercial reference of 4 genes of interest.
Vertical interrupted lines: average fold change of controls/patients versus commercial reference.
Arrows: average fold change between patients and controls.
The statistically differentially expressed transcripts were compared to those reported in

previous studies investigating endometrial gene expression from pre-receptive and
receptive phase endometrium (Mirkin et al., 2005; Carson et al., 2002; Ponnampalam
et al., 2004; Riesewijk et al., 2003; Talbi et al., 2006), endometrial gene expression in
women with endometriosis (Kao et al., 2003) or an intrauterine device (IUD)
(Horcajadas et al., 2006), or after ovarian stimulation (Macklon et al., 2008) versus
Chapter 4 | 141
controls, and also endometrial gene expression from women with implantation failure
(Tapia et al., 2008). The overlap in identified genes is presented in Table 5, 3 of these
9 studies showed no overlap (Tapia et al., 2008; Carson et al., 2002; Kao et al., 2003).
Table 5: Statistically differentially expressed genes in the primary analysis or subgroup

analyses contrasted with previous reported gene lists.
Gene Description RIF WOI WOI WOI WOI IUD Stim

Symbol (n=9) LH+8 vs. ESE-MSE LH+7 vs MSE vs IUD vs Stim vs nat
LH+3 vs. rest cycle LH+2 ESE no IUD cycle
Present Mirkin Ponnam Riesewijk Talbi Horcajadas Macklon
study 2005 palan 2004 2003 2006 2006 2008
FOXO1 Forkhead box protein ↑ ↑ ↑
O1A
GABRA2 GABA(A) receptor ↑ ↑ ↓
subunit alpha-2
FUT4 Alpha-(1,3)- ↑ ↑
fucosyltransferase
TLR4 Toll-like receptor 4 ↑ ↑
precursor
BIRC3 Inhibitor of apoptosis ↑ ↑
protein 1
CD36 Platelet glycoprotein 4 ↑ ↓ ↓
ADAMTS8* ADAMTS-8 precursor ↑ ↓
OPRK1* Kappa-type opioid ↑ ↓
receptor
CA12 Carbonic anhydrase 12 ↓ ↑
precursor
AMD1* S-adenosylmethionine ↑ ↑
decarboxylase
proenzyme
CYP26A1* Cytochrome P450 ↑ ↑ ↑
26A1
FAM13A1* Protein FAM13A1 ↑ ↑
HPGD* Prostaglandin ↑ ↓
dehydrogenase 1
IFNGR1* Interferon-γ receptor α ↑ ↑
chain precursor
MUC16* Mucin-16 ↑ ↑
NFIB* Nuclear factor 1 B- ↑ ↑
type
RIF: recurrent implantation failure; WOI: window of implantation; LH: luteinizing hormone; ESE: early-
secretory endometrium; MSE: mid-secretory endometrium; IUD: intra-uterine device; NAT: during natural
cycle; STIM: during ovarian stimulated cycle.
* Only significantly differentially expressed in the subgroup analysis.
Gene ontologies
To investigate whether the genes observed to be differentially expressed in women
with recurrent implantation failure indicate the dysregulation of specific biological
processes, these genes were queried for their gene ontology (GO) classifications. In
the primary analysis of women with implantation failure versus controls, 63 genes
were significantly upregulated. One biological processes was significantly enriched in
the list of significantly upregulated genes compared to the reference gene list, namely
‘response to cytokine stimulus’ (p = 0.008, ratio of observed to expected [R] 193). In

the subgroup analysis of women with > 10 embryos transferred versus 16 controls, 76
genes were significantly upregulated in women with implantation failure. Three
biological processes were significantly enriched in the list of significantly upregulated
genes in the subgroup analysis compared to the reference gene list. In addition to
response to cytokine stimulus (p = 0.009, R 194), arachidonic acid metabolic process
(p = 0.009, R 194), and organic acid biosynthetic process (carboxylic acid biosynthetic
process) (p = 0.012, R 17) were identified.
Clustering and classification of samples

Unsupervised hierarchical clustering of all 48 endometrial samples from women with
implantation failure and controls based on the expression profile of all genes included
in the array could not discriminate between the gene expression profiles of a receptive
or less receptive endometrium (data not shown). In addition, we used supervised
classification as described in the methods to classify samples. The performance of the
resulting predictor was independently validated on a test set of samples, which was not
used to build the predictor. The full procedure was repeated 100 times to obtain an
accurate estimation of classification performance. However, the 100 individual
predictors were unable to correctly determine whether the endometrial sample is taken
from a normal receptive endometrium or from women with recurrent implantation
failure in on average 45.3% of cases (95% CI: 43.4- 47.2%). After random assignment
of labels (patient and control), another 100 predictors were obtained, which showed a
false prediction rate of 52.1% (95% CI: 49.6- 54.6%). A two-sample t-test showed that
the ability to classify samples as a normal (control) or less receptive endometrium
(patient) by their gene expression profile was significantly better compared to the
ability to classify samples with random assigned labels (p= 0.000013). Although this is
a significant difference, the performance of the classifier leaves little scope for
prediction of samples of unknown class in clinical practice.
Validation of the microarray data by PCR

In order to validate the data generated by microarray analysis, a selected set of genes
was submitted to real time reverse transcriptase polymerase chain reaction (RT-PCR),
namely FOXO1, MUC16, Cyp26A1, and SULT1E1. The results obtained by this real
time PCR analysis showed a good correlation with the results obtained by the
Chapter 4 | 143
microarray experiment, p <0.0001 for all four genes. The RNA expression was
normalised for housekeeping gene ribosomal protein L19 (RPL19) for microarray and
PCR data. The significant differences in expression of these genes in patients and
controls were corroborated for FOXO1, and SULT1E1. CYP26A1 expression was
higher in women with implantation failure versus controls of borderline significance
(p<0.06). MUC16 was not significantly differentially expressed in patients and
controls (p=0.8) (Figure 2). However, this may have been expected since a 1.19 fold
change in expression is small to confirm by PCR.
Figure 2: Boxplots from results obtained by real time RT-PCR analysis of 4 genes.
Y-axis: relative expression normalised to RPL19 expression

Discussion
The present study reports the largest set of microarray data on endometrial gene
expression in women with recurrent implantation failure to date. 77 genes were
observed to be significantly differentially expressed in these women compared to
control subjects. Of these genes, 65 were upregulated and 12 genes were
downregulated compared to women who did achieve pregnancy with ICSI for severe
male factor infertility only.
Genes involved in a change in activity of a cell in terms of movement, secretion,

enzyme production or gene expression as a result of a cytokine stimulus were
significantly enriched in women with implantation failure. The endometrium becomes
receptive under the influence of endocrine, paracrine and autocrine stimuli.
Subsequently, the fetal-endometrial cross-talk is critical for a successful implantation.
Genes upregulated in endometrial stromal cells in response to the trophoblast included
those for inflammatory response and immune response (Popovici et al., 2006).
Therefore, it may be hypothesized that an aberrant strong response by the
endometrium on stimuli may result in rejection of the embryo.
In a subgroup analysis of women with implantation failure (> 10 embryos transferred),

three biological processes were significantly enriched in the list of significantly
upregulated genes, namely the response to cytokine stimuli, and two metabolic process
involved in the chemical reactions and pathways resulting in the formation of organic
acids, carboxylic acids, and arachidonic acid. An upregulation of arachidonic acid
metabolism is an intriguing observation. Cyclo-oxygenase-2 (COX2), the rate-limiting
enzyme for the conversion of arachidonic acid into prostaglandin, is essential for
blastocyst implantation and decidualization (Lim et al., 1997). There is also extensive
evidence for a positive involvement of prostaglandins in implantation (Kennedy et al.,
2007). An increase in arachidonic acid metabolism in women with implantation failure
may be a sign of compensation for a decreased endometrial receptivity, such as
shortcoming vascular permeability.
Several genes that are significantly differentially expressed in women with

implantation failure versus controls warrant further discussion. Forkhead box O
Chapter 4 | 145
(FOXO1, synonyms FKH1, FKHR, FOXO1A), which was significantly higher

expressed in women with implantation failure versus controls, is a member of the
FOXO sub-family of forkhead/ winged-helix family of transcription factors and a
critical mediator in cell fate decision involved in the control of cell cycle arrest and the
induction of apoptosis (Brosens and Gellersen, 2006). FOXO1 interacts with the
progesterone receptor, which acquires control of the diverse gene families involved in
decidualization. An in-vitro study showed that among the 3405 significantly regulated
genes upon decidualization of human endometrial stromal cells, 507 (15.3 %) genes
were aberrantly expressed upon FOXO1 knockdown (Takano et al., 2007). The study
revealed that FOXO1 plays an essential role in coordinating different aspects of this
process, including cell proliferation, differentiation, immune modulation, and defences
against environmental or oxidative stress. Previously, FOXO1 was shown to be
significantly upregulated in the receptive versus the pre-receptive endometrium
(Mirkin et al., 2005; Talbi et al., 2006). Subjects with endometriosis were previously
shown to present with reduced FOXO1 expression in the endometrium compared to
controls (Shazand et al., 2004; Burney et al., 2007). Since FOXO1 plays an essential
coordinating role, dysregulation of FOXO1 expression may be detrimental to
endometrial receptivity.
Prior to the receptive phase, the glycocalyx present on the uterine luminal epithelial
surface of the endometrium provides a barrier to embryo adherence. It has been
hypothesized that loss of the anti-adhesive barrier is required for the attachment of the
embryo. To date, most research has focused on cell surface associated mucin 1
(MUC1), encoding a membrane bound glycosylated phosphoprotein (Aplin, 1997).
Pinopodes (Bentin-Ley et al., 1999) or uterodomes (Murphy, 2000), which are
cyclically appearing epithelial protrusions present during the receptive phase and
possibly a marker of endometrial receptivity, render the cell surface adhesive to
trophectoderm cells. Recently, MUC16 expression (also known as the ovarian tumor
cell marker CA125) was shown to be lost on pinopode surfaces during the receptive
phase (LH+6) more consistently then MUC1. Moreover, short interfering RNA
knockdown of MUC16 showed an increased trophoblast adherence in vitro, whereas
MUC1 knockdown did not affect trophoblast adherence (Gipson et al., 2008). These
data suggest that MUC16 contributes to the formation of a barrier and that loss of
MUC16 is necessary to facilitate adherence and subsequent implantation of the
embryo. The present study showed a significant increased MUC16 expression in

women with implantation failure (> 10 embryos transferred). On the contrary, others
showed an increased MUC16 expression during the receptive phase compared to the
pre-receptive phase (Talbi et al., 2006). The increased expression of MUC16 in
women with recurrent implantation failure could not be confirmed by PCR, possibly
because the microarray findings on MUC16 were false or as a result of the relatively
low fold change. The human trophoblast expresses high levels of L-selectin after
hatching, which binds to uterine epithelial oligosaccharide ligands. Fucosyltransferase
4 (FUT4, synonyms: CD15, ELFT, FCT3A, FUC-TIV) is an enzyme responsible for
the final catalyzing step in the production of many L-selectin ligands. It has been
shown that FUT4 (mRNA and protein) is upregulated in human endometrium during
the period of implantation and is regulated by progesterone (Ponnampalam et al.,
2004; Ponnampalam, 2008). In the present study FUT4 expression was significantly
higher in women with implantation failure versus controls. Although speculative, this
may be a result from an abnormal or a deficit L-selectin expression.
Another interesting significantly upregulated gene in women with implantation failure

is ADAM metallopeptidase with thrombospondin type 1 motif 8, also known as
ADAMTS8 (synonym: METH2). Previously, this gene has been shown to be
downregulated in endometrial biopsies during the window of implantation compared
to the pre-receptive phase (Talbi et al., 2006). This gene encodes a member of the
ADAMTS protein family, which share a metalloproteinase domain, a disintegrin-like
domain, and a thrombospondin type 1 (TS) motif. The enzyme encoded by this gene
disrupts angiogenesis in vivo (Vazquez et al., 1999). Both ADAMTS1 and ADAMTS8
can inhibit vascular endothelial growth factor (VEGF) induced angiogenesis.
ADAMTS1 and -8 can be categorized into the same family on the basis of sequence
similarities and targeted substrates (Porter et al., 2005). To our knowledge, targeted
disruption of the mouse ADAMTS8 gene has not been investigated. However,
disruption of the ADAMTS1 gene resulted in an impaired development of follicles,
implantation, and intrauterine development (Shindo et al., 2000).
Cytochrome p450, family 26, subfamily A, polypeptide 1 (Cyp26A1) showed the

largest increase in expression among the subgroup of women (> 10 embryos
transferred) versus controls (2.67 fold change). The gene expression of this retionoic
Chapter 4 | 147
acid catabolic enzyme markedly increases in the secretory phase (Deng et al., 2003)
and in response to progesterone (Jeong et al., 2005). On the contrary, endometrial
Cyp26A1 expression is significantly lower in women with endometriosis (Burney et
al., 2007).
Endometrial estrogen-specific sulfotransferase (SULT1E1) may be an important

enzyme in endometrial receptivity because of its role in estrogen inactivation. Sulfated
estrogens do not bind to the estrogen receptor and are therefore hormonally inactive
(Strott, 1996). SULT1E1 expression is highest in the secretory phase compared to the
proliferative phase (Rubin et al., 1999). SULT1E1 knock out female mice showed
reduced fertility caused by placental thrombosis and spontaneous fetal loss, which was
prevented by treatment with either an anticoagulant or antiestrogen (Tong et al., 2005).
The present study showed a 2.32 fold increased expression of SULT1E1 in women
with implantation failure (> 10 embryos transferred) versus controls. A high
expression may diminish endometrial receptivity, since a delicate balance exists
between the cyclically oestrogen and progesterone mediated preparation of the
endometrium.
A transcriptomic signature which distinguishes receptive and refractory endometrium

was not discerned in the present study and fold changes in expression were relatively
small. This suggests that aberrant endometrial gene expression may be of aetiological
importance in only a proportion of patients with idiopathic recurrent implantation
failure. It may also reflect the heterogeneity in causes for implantation failure,
different molecular processes in the endometrium may be affected, an embryo factor
may have been present, and a number of women may have met the criteria for
recurrent implantation failure simply due to chance events in the absence of
underlying fertility diminishing factors. Although differences in expression were
relatively small, it is unknown what fold change in expression is of biological
significance. Transcription factors, such as FOXO1, have the ability to regulate the
transcription of hundreds of genes (Takano et al., 2007) and a change in gene
expression may result in a larger effect on secreted proteins, which can be regulated
downstream.
To our knowledge, only one study also investigated gene expression profiles in
women with implantation failure to date (Tapia et al., 2008), including 3 patients with
ovarian failure. In that study, genes >2 fold change in expression are reported with no
statistical analysis performed. The present study revealed no corresponding genes.
This is not surprising however, given the degree of variation in gene expression
determined in the present study on 24 patients. Moreover, in the study by Tapia et al.
biopsies were taken in artificial cycles induced by exogenous sex steroids rather than
natural cycles (Tapia et al., 2008), which have been shown to affect the endometrial
transcriptome (Macklon et al., 2008; Mirkin et al., 2004).
A number of limitations of the presented study should be taken into consideration

when interpreting the results. An endometrial biopsy contains different types of cells
present in the endometrium. Therefore, genes expressed in one of these type of cells,
e.g. epithelial cells, may be differentially expressed in women with implantation
failure, but may not be of sufficient amplitude to be detected in the entire biopsy
(Carson et al., 2002). Cell-type specific gene expression has been demonstrated for
genes upregulated during the window of implantation (Franchi et al., 2008). Laser
capture microdissection has been used to separate both stromal and epithelial
compartments (Horcajadas et al., 2007). Endometrial biopsies were performed without
the presence of an embryo in the uterine cavity, which is clearly impracticable for
ethical reasons. Therefore we cannot exclude that the endometrial-embryonic dialogue
is disturbed in women with implantation failure, beyond the maternally endocrine
driven changes in the endometrium. This has been investigated in mice where gene
expression profiles of implantation and inter-implantation sites have been compared
(Reese et al., 2001). An in vivo experiment of human endometrial stromal cells
cultured in the presence of trophoblast conditioned media was recently published
(Hess et al., 2007). A total of 4817 genes were found to be differentially expressed in
response to the trophoblast derived factors. Moreover, the uteri of women with
implantation failure have never been exposed to pregnancy, whereas the controls have
given birth. We cannot exclude that pregnancy itself induced an imprinting in the
endometrium as has been proposed by experimental data from animals and also
addressed by Tapia et al. (Ginger et al., 2001). However, there is no evidence for such
endometrial epigenetic modifications in humans (Tapia et al., 2008). Lastly, we
included women with implantation failure during (mainly) ovarian stimulated cycles,
Chapter 4 | 149
whereas the biopsy was derived from a natural cycle. It can be hypothesized that these
women react aberrantly to ovarian stimulation, rather than a permanent intrinsic
problem in the endometrium as a cause of implantation failure.
In conclusion, a number of potentially interesting genes have been identified,
providing novel clues for understanding endometrial dysfunction in infertility and
starting points for further investigation. However, an aberrant endometrial gene
expression profile appears to be present only in a limited number of patients with
recurrent implantation failure. Future studies could lead to new insights in the
biological basis of the roles of these genes in the implantation process, which might
ultimately reveal novel leads for therapeutic intervention.
Chapter 5
____________________________________________
General Discussion
General Discussion
The studies presented in this thesis address one of the major rate-limiting steps for
success in fertility treatment. A better understanding of the molecular signals that
regulate endometrial receptivity and implantation competency is of clinical relevance,
because unravelling these signals may lead to strategies to correct implantation
failure. Moreover, at present in many couples no aetiology for their infertility is
identified. A diminished endometrial receptivity may account for a number of women
with unexplained infertility. A reliable clinical test for endometrial receptivity would
help to identify these women and may help to counsel these couples. However, the
clinical implications of understanding the molecular signals regulating endometrial
receptivity go beyond improving fertility treatment. It may also have implications for
the development of contraceptives, and potentially it may improve clinician’s ability
to treat disorders related to implantation and placentation disorders, including pre-
eclampsia and fetal growth restriction.
As outlined in the introduction, for ethical and practical reasons it is difficult to study
human embryo implantation in vivo, since many approaches will risk disturbing the
process itself.

The endometrium is receptive to blastocyst implantation during a spatially restricted
‘window’, when the luminal epithelium is favourable for blastocyst implantation. It is
well established that embryos cannot implant outside this window. An ideal
molecular marker of receptivity should be obtained non-invasively, enabling
investigation during conception cycles during this ‘window of implantation’.
However, the complexity and molecular redundancy observed in human implantation
indicates the need to identify molecular profiles conductive to implantation.
Endometrial secretion aspiration and analysis by a multiplex immunoassay provides
these advantages and may therefore offer a clinically useful approach to study the
endometrial factor in human embryo implantation (chapter 3.1). Our results confirm
earlier observations by our group (van der Gaast et al. 2003), which indicate that this
technique can be safely carried out prior to embryo transfer during IVF/ ICSI
treatment. Endometrial secretion aspiration immediately prior to embryo transfer is
Chapter 5 General Discussion | 153
well tolerated, easy to perform, and the aspiration provides sufficient material for
analysis in 99.5% of cases. Moreover, significant differences in the cytokine and
chemokine profile in cervical mucus were observed compared with endometrial
secretions, which implies that the aspirate is indeed endometrial secretion and does
not represent cervical mucus or a significant contamination during the procedure. In
chapter 3.2 we have shown that endometrial secretion cytokine profiling identifies a
profile conducive to clinical pregnancy, and the ability to predict clinical pregnancy
by IL-1β and TNF-α levels in endometrial secretions was similar and additive to the
predictive value of embryo quality.
Interestingly, IL-1β and TNF-α are both pro-inflammatory cytokines. Pregnancy was
postulated to be an event requiring a predominance of anti-inflammatory mediators,
such as IL-10 and IL-5. It was noted that pathologies associated with humoral
responses, such as lupus erythematosus, are known to flare up during pregnancy,
whereas cell mediated ones, such as rheumatoid arthritis, often enter in remission.
These observations formed the basis for the T helper cell type 1/ 2 (Th1/Th2)
paradigm. The model was developed in mice in 1995 (Chaouat et al. 1995). It was
shown that placentae from abortion prone mice were deficient in their production of
anti-inflammatory cytokines, which was companied by increased levels of local
inflammatory cytokines. rIL-10, anti-IFN-γ or pentoxifillin (an anti-TNF agent)
reduced the resorption, while anti-IL-10 increased the resorption rate in the abortion
prone matings. Moreover, induction of an inflammatory response by
lipopolysaccharides was associated with foetal demise and premature labour (Chaouat
2007). These results indicated that local anti-inflammatory cytokines can play a vital
role in the survival of the foetal allograft by counteracting deleterious inflammatory
cytokines. These findings in mice were later extended to humans. Excess IFN-γ and
TNF-α production have been associated with IVF failure and miscarriage (Hill et al.,
1995). Another study showed that maternal peripheral blood mononuclear cells
(PBMCs) from women with a history of recurrent miscarriage responded differently
on placental antigens in co-culture than PBMCs from controls, showing a bias
towards Th1-type reactivity in women with a history of recurrent miscarriage
(Raghupathy et al. 1999).
However, the Th1/Th2 theory was soon challenged by a variety of data. Anti-IL-10
did not have an effect on the resorption rate in non-abortion prone mice (Chaouat et
al. 1995). Much of our understanding of endometrial receptivity is derived from
mouse knock-out studies. Relatively few of these cytokine gene knock-out studies
have shown effects on reproduction and they were mainly Th1 cytokines, whereas IL-
4 and IL-10 knock-out mice did not show an effect on their fertility (White et al.
2004). IFN-γ producing uNK cells were seen as ‘bad guys’ in the endometrium.
However, an increasing body of evidence showed that uNK cells accumulate in the
endometrium during the implantation period and uNK cell knock-out mice showed
reproductive defects (Barber and Pollard, 2003). It became clear that it is not as
straightforward as previously thought. Although the established pregnancy may
indeed be a Th2 phenomenon, Th1 activity is required during the early implantation
period. Implantation is a process of aggression in which maternal and embryonic
signals induce a state of controlled inflammation, which increases the production of
both pro-inflammatory cytokines, inducing angiogenesis and increased vascular
permeability, but also anti-inflammatory cytokines (van Mourik et al. 2008). The
results presented in chapter 3.2, with a positive association between the endometrial
secretion concentration of TNF-α and a negative association with IL-β and
pregnancy, strengthen the enunciation that the Th1/Th2 paradigm is obsolete.
A second aim was to study the effect of ovarian stimulation on the intrauterine
cytokine profile. As described in chapter 3.3, the in vivo milieu encountered by the
embryo after transfer is significantly altered by ovarian stimulation. Whether these
changes in the intra-uterine milieu affect endometrial receptivity is unknown. In
stimulated cycles both TNF-α and IL-β increased approximately 2.7 fold. Therefore,
the ratio of both cytokines was not different in natural and ovarian stimulated cycles.
However, since we have shown that subtle differences in the endometrial secretion
cytokine profile are associated with the likelihood of conceiving, the marked
differences in cytokine profiles in ovarian stimulated cycles compared to the golden
standard of a receptive endometrium, namely the endometrium in a natural cycle,
may be detrimental. Increasing awareness of these possible effects on endometrial
receptivity and their burden on the patient, stimulate development of milder ovarian
stimulation regimens (Macklon et al. 2000; Fauser et al. 1999; Edwards et al. 1996).
Although less embryos are obtained after mild stimulation, an increased percentage of
euploid embryos per number of oocytes retrieved has been reported (Baart et al.
2006). High estradiol levels on the day of hCG administration, which coincide with a
higher ovarian response, have been shown to reduce implantation rates independent
of embryo quality (Simon et al. 1995; Yu and Ng, 2000). The study described in
chapter 3.3 also suggests an increased disruption of the intrauterine cytokine profile
with an increased number of oocytes retrieved. Secondly, concerns have been raised
by the authors of a meta-analysis on a possible negative effect direct on the
endometrium by the use of GnRH antagonists compared to agonists (Al-Inany et al.
2006). We have shown only minor differences in cytokine profiles after GnRH
antagonist or agonist treatment in comparison to the differences observed in natural
and stimulated cycles. We could not confirm findings of a previous study on
endometrial gene expression, which concluded that endometrial development after a
GnRH antagonist mimics the natural endometrium more closely than after the use of
a GnRH agonist (Simon et al. 2005). Employing endometrial secretion analysis to
investigate the effect of ovarian stimulation on the intra-uterine environment during
the window of implantation may help to develop less disruptive ovarian stimulation
protocols for IVF.
There is evidence that embryo implantation rates may be impaired by the pathogenic
effects of bacterial vaginosis. In chapter 3.4, the association between bacterial
vaginosis and the intra-uterine cytokine profile was assessed. Previously it has been
shown that almost half of patients with symptomatic BV showed histopathological
evidence of plasma cell endometritis (Korn et al. 1995). Our data of women with
mainly asymptomatic BV, showed cytokine concentrations which were within the
normal range compared to controls, with a tendency to a more pro-inflammatory
profile in women with a higher Nugent score. Screening for bacterial vaginosis in
women undergoing ART is not recommended on the basis of these results.
Future ‘promise’ of endometrial secretion analysis

The studies presented in chapter 3 have generated a number of new research
questions. Although endometrial secretion analysis may have considerable potential
implications for future IVF treatment, at present, the technique remains primarily a
promising and novel tool to investigate human embryo implantation in vivo.

However, it is tempting to hypothesize the clinical implications endometrial secretion
analysis may have in future IVF treatment. It is important to refine the array by
including novel potential markers for endometrial receptivity. As yet, we have not
been able to include certain proposed markers of the window of implantation in the
multiplex immunoassay, either because appropriate antibodies were not available
(Glycodelin, Osteopontin), or because of problems arising from cross-interference
(IL-11, LIF and M-CSF) (Lessey et al. 1995; Chryssikopoulos et al. 1996; Ledee-
Bataille et al. 2002). Novel potential markers for endometrial receptivity should be
included in future studies, whereas others should be eliminated. Although unknown,
refining the array may improve the sensitivity and specificity of predicting pregnancy
and could lead to future tests used to determine if the physiological state of the intra-
uterine milieu is appropriate for embryo transfer. Considerable differences in the
intra-uterine milieu were observed in natural and stimulated cycles. Therefore,
cryopreservation of the embryo and transfer in a natural cycle may be an alternative
when the intra-uterine milieu is less favourable in the IVF cycle. Further studies are
required to assess the intercycle variability of the intra-uterine cytokine expression
and to assess how the intrauterine cytokine milieu varies through the cycle and in
particular the luteal phase.
Endometrial secretions are produced by the epithelial cells from endometrial glands.
It is unknown how secreted proteins present in the secretions reflect protein
expression in the endometrium. Therefore we aim to study the correlation between
the protein expression in endometrial secretions and endometrial biopsies. It addition
to a complex array of growth factors and cytokines, endometrial secretions provide
nutrition for the conceptus. The glycoprotein rich secretion is considered to support
the embryo during the pre-implantation period (Burton et al. 2002). Epidemiological
and animal studies reveal that risk of adult disease can be set in part in utero by
nutritional factors operating from the preimplantation period onwards. Endocrine-
metabolic pathologies contribute significantly to this disease risk. Rodent embryos
exposed to maternal protein restriction exclusively during the pre-implantation period
produce offspring that are overweight, hypertensive, have cardiovascular function
defects and abnormal stress-related behaviour (Fleming et al. 2004). However, little
is known about how maternal nutrition affects the environment of embryos in

humans. Studies on the effect of lifestyle and nutritional factors on the intra-uterine
milieu (such as aminoacids and folic acid) are currently designed.
Obviously, our findings presented in chapter 3 should be confirmed by other

laboratories. It should be realized that different IVF clinics use different protocols,
involving many different stimulation protocols, different day of embryo transfer and
different criteria for oocyte retrieval and also the patient population may differ. This
may affect the endometrial secretion cytokine profile present prior to embryo transfer.
The clinical problem of recurrent implantation failure

Clinicians have sought to develop clinical interventions aimed at enhancing
implantation rates from IVF. The competitive commercial context in which IVF is
usually practised and the frustration caused by recurrent implantation failure has often
led to the early adoption of promising new interventions before clear evidence for
their efficacy and safety has been established. As described in chapter 2.1, there is no
evidence supporting the empirical use of most pharmaceutical interventions, such as
aspirin, nitric oxide, aromatase inhibitors, ascorbic acid, prolonged progesterone
treatment, glucocorticoids, or insulin sensitizing drugs, in an unselected IVF
population. Adjuvant glucocorticoids resulted in higher pregnancy rates in women
undergoing IVF rather than ICSI, compared to controls. However, data to support
their use is not convincing, since results were of borderline significance. Given that
the safety is also uncertain, we would at present not recommend their use in an
unselected IVF population. These results do not exclude the use of pharmaceutical
interventions in certain subgroups of patients. Aspirin has been shown to be effective,
either alone or in combination with heparin, in the treatment of recurrent miscarriage
in women with antiphospholipid antibody syndrome (Empson et al. 2005; Rai et al.
2002). Aromatase inhibitors in women with PCOS (Mitwally et al. 2004) and
glucocorticoids in women with auto-antibodies (Geva et al. 2000; Ando et al. 1996)
may be promising and warrant further investigation. Although evidence is limited,
one RCT showed that the use of glucocorticoids significantly increased pregnancy
rates in certain subgroups of women with undergoing assisted hatching (Cohen et al.
1990; Primi et al. 2004). Outwith specific indications, adjuvant pharmaceutical

interventions should be restricted to well-designed studies.
On the contrary, significant improvements in clinical pregnancy rates can be achieved

by giving due attention to the type of catheter used and embryo transfer position. An
ultrasound guided transfer with (if possible) a soft embryo transfer catheter (Buckett
et al., 2006), positioned close to the middle of the endometrial cavity (Oliveira et al.,
2004; Coroleu et al., 2002), may optimise the embryo transfer procedure. However,
the causes for implantation failure are multifactorial. It is unlikely that additional
medical interventions, such as those addressed in this chapter, will be shown to have
anything but a marginal effect on IVF outcomes, until the factors which determine the
ability of an embryo to successfully implant are better understood.
Previous approaches to investigate implantation have generally relied on the analysis

of individual candidate genes. To date no single marker has been identified that is
specific and sensitive in identifying a receptive endometrium though. Therefore it is
assumed that reproductive mechanisms, including implantation, depend on redundant
pathways. Microarray analysis enables an approach from a global genomic
perspective. The study described in chapter 4 reports the biggest set of microarray
data on endometrial gene expression in women with implantation failure to date. The
establishment of a predictor set of genes after DNA micro-array has been successful
in other areas of research. A set of 102 genes successfully predicted the presence of
lymph node metasteses for primary head and neck squamous cell carcinomas
(Roepman et al. 2005). Similarly, DNA micro-array gene expression profiling has
been shown to predict clinical outcome in women with breast cancer (van ‘t Veer et
al. 2002). One of the study aims addressed in chapter 4 was to identify a molecular
endometrial signature to differentiate between a normal or less-receptive
endometrium. However, we have shown that from our results it is not possible to
construct a more limited gene array, which might be used for prediction of recurrent
implantation failure. This finding probably reflects the heterogeneous aetiology of
recurrent implantation failure.
Global gene expression studies employing micro-arrays are essentially hypothesis

generating. Several interesting genes identified by this study warrant further
investigation on the protein level in the endometrium by histochemical studies, such
as FOXO1, ADAMTS8, MUC16, FUT4, SULT1E1 and Cyp26A1. The endometrial
expression of these proteins throughout the menstrual cycle and in women with
recurrent miscarriage, endometriosis, unexplained infertility or recurrent implantation
failure, will be interesting studies to undertake and could lead to new insights in the
biological basis of the roles of these genes in these disorders and the implantation
process.
General clinical context

The studies presented in this thesis aimed to explore the role of the endometrium in
the implantation process. However, it takes two to tango. The embryo interacts with
the endometrium, which receives signals from the embryo (van Mourik et al., 2008).
Clearly for ethical and practical reasons an endometrial biopsy or endometrial
secretion aspiration cannot be taken with an embryo in the uterine cavity. Moreover,
effects brought about by the embryo may only be present locally rather than
throughout the endometrium. Therefore, in the present thesis investigations are
limited to the maternally endocrine driven changes in the endometrium. The
endometrial-embryonic dialogue may be decisive in the successfulness of the
implantation though. The biggest challenge is to investigate this cross-talk between
the implanting embryo and maternal endometrium during the early stages of
implantation. Animal models offer a way around this problem. Much of our
understanding of early human development is indeed inferred from studies in knock-
out mice. However as discussed previously, caution is required when data from mice
are extrapolated to humans. Although both species have hemochorial placentas, there
are important differences in the implantation process (Carson et al. 2000).
Alternatively, in-vitro experiments may be better suited for deducing these
mechanisms. Studies on interactions between the embryo and the endometrium, or the
effect of soluble factors expressed by the embryo on the endometrial transcriptome
are currently undertaken in our centre. However, of course the controlled conditions
present in the in vitro system differ significantly from those in vivo.
The most effective way to prevent implantation failure after IVF is by reducing the
number of women attending IVF therapy. Women who put off trying to have children
until their mid-thirties or later worsen their chances of becoming pregnant and risk
losing out on motherhood altogether. Women should be enabled by society to delay
their careers instead of their childbearing. Paid maternity and parental leaves and
safe, affordable day care should be available. Secondly, increasing evidence suggests
that pre-conceptional interventions designed to optimize factors related to lifestyle
and nutrition may improve outcomes from fertility treatment. For example, obesity
has become a disease of epidemic proportions. In the Netherlands approximately 40%
of adults is overweight and 10% is obese. The role of preconceptional counselling in
order to optimise reproductive health has been largely ignored and increased attention
could contribute to improving implantation rates and pregnancy outcomes after IVF.
Concerning the endometrial factor in the human implantation process in the general
IVF population, we believe that we can conclude from the studies presented in this
thesis that the endometrium is an important determinant of success and is not simply
facilitory. However, a diminished endometrial receptivity as the primary cause of
recurrent implantation failure may only be present in a limited number of patients,
who cannot be easily identified by known clinical parameters or an endometrial gene
expression profile. A test for endometrial receptivity remains elusive.
References | 161
References
____________________________________________
Akoum A, Metz CN, Al Akoum M et al. 2006 Macrophage migration inhibitory factor expression in the
intrauterine endometrium of women with endometriosis varies with disease stage, infertility status, and pelvic
pain. Fertility and Sterility 85, 1379-85.
Al-Inany GH, Abou-Setta AM, Garzo G 2006 Embryo transfer catheters for ART cycles. (Protocol). The
Cochrane Database of Systematic Reviews.
Al-Inany HG, bou-Setta AM, Aboulghar M. 2006 Gonadotrophin-releasing hormone antagonists for assisted
conception. Cochrane Database of Systematic Reviews Issue 3, CD001750
Amsel R, Totten PA, Spiegel CA, 1983 Nonspecific vaginitis. Diagnostic criteria and microbial and
epidemiologic associations. American Journal of Medicine 74, 14-22.
Anahara R, Toyama Y, Koda M et al. 2006 Deletion of macrophage migration inhibitory factor gene induces
down regulation of sex hormones and ultrastructural abnormalities in mouse testes. Reproductive Toxicology
21, 167-70.
Andersen CY, Hornnes P. 1994 Intrafollicular concentrations of free cortisol close to follicular rupture. Human
Reproduction 9, 1944-9.
Ando T, Suganuma N, Furuhashi M, et al. 1996 Successful glucocorticoid treatment for patients with abnormal
autoimmunity on in vitro fertilization and embryo transfer. Journal of Assisted Reproduction and Genetics 13,
776-81.
Aplin JD 1997 Adhesion molecules in implantation. Rev Reprod 2, 84-93
Aplin JD 1999 MUC-1 glycosylation in endometrium: possible roles of the apical glycocalyx at implantation.
Human Reproduction 14 Suppl 2, 17-25.
Azem F, Many A, Ben A, et al. 2004 Increased rates of thrombophilia in women with repeated IVF failures.
Hum Reprod 19, 368-70
Arck PC, Troutt AB, Clark DA 1997 Soluble receptors neutralizing TNF-alpha and IL-1 block stress-triggered
murine abortion. American Journal of Reproductive Immunology 37, 262-6.
Arcuri F, Cintorino M, Vatti R, et al. 1999 Expression of Macrophage Migration Inhibitory Factor Transcript
and Protein by First-Trimester Human Trophoblasts. Biology of Reproduction 60, 1299-1303.
Asimakopoulos B, Koster F, Felberbaum R et al. 2006 Cytokine and hormonal profile in blood serum and
follicular fluids during ovarian stimulation with the multidose antagonist or the long agonist protocol. Human
Baart EB, Van Opstal D, Eijkemans MJ, et al. 2007 Does the magnitude of ovarian stimulation for IVF affect
chromosomal competence of embryos as assessed by PGS? Human Reproduction 22, 980-8.
Barber EM and Pollard JW 2003 The uterine NK cell population requires IL-15 but these cells are not required
for pregnancy nor the resolution of a Listeria monocytogenes infection. The Journal of Immunology 171, 37-46.
Basir GS, WS O, Ng EH, et al. 2001 Morphometric analysis of peri-implantation endometrium in patients
having excessively high oestradiol concentrations after ovarian stimulation. Human Reproduction 16, 435-40.
Battaglia C, Salvatori M, Maxia N, et al. 1999 Adjuvant L-arginine treatment for in-vitro fertilization in poor
responder patients. Human Reproduction 14, 1690-7.
Battaglia C, Regnani G, Marsella T, et al. 2002 Adjuvant L-arginine treatment in controlled ovarian
hyperstimulation: a double-blind, randomized study. Human Reproduction 17, 659-65.
Beitins IZ, Bayard F, Ances IG, et al. 1972 The transplacental passage of prednisone and prednisolone in
pregnancy near term. Journal of Pediatrics 81, 936-45.
Beklen A, Ainola M, Hukkanen M et al. 2007 MMPs, IL-1, and TNF are regulated by IL-17 in periodontitis.
Journal of Dental Research 86, 347-51.
Bentin-Ley U, Sjogren A, Nilsson L, et al. 1999 Presence of uterine pinopodes at the embryo-endometrial
interface during human implantation in vitro Human Reproduction 14, 515-20
Berkkanoglu M, Isikoglu M, Seleker M et al. 2006 Flushing the endometrium prior to the embryo transfer does
not affect the pregnancy rate. Reproductive Biomedicine Online 13, 268-71.
Bider D, Amoday I, Tur-Kaspa I, et al. 1996 The addition of a glucocorticoid to the protocol of programmed
oocyte retrieval for in-vitro fertilization--a randomized study. Human Reproduction 11, 1606-8.
Bider D, Amoday I, Yonesh M, et al. 1996 Glucocorticoid administration during transfer of frozen-thawed
embryos: a prospective, randomized study. Fertility and Sterility 66, 154-6.
Bider D, Blankstein J, Levron J, et al. 1997 Gonadotropins and glucocorticoid therapy for "low responders"--a
controlled study. Journal of Assisted Reproduction and Genetics 14, 328-31.
Blois SM, Joachim R, Kandil J, et al. 2004 Depletion of CD8+ Cells Abolishes the Pregnancy Protective Effect
of Progesterone Substitution with Dydrogesterone in Mice by Altering the Th1/Th2 Cytokine Profile. The
Journal of Immunology 172, 5893-9.
Boomsma CM, Macklon NS 2006 What can the clinician do to improve implantation? Reproductive
Biomedicine Online 13, 845-55.
Boomsma CM, Keay SD, Macklon NS 2007 Peri-implantation glucocorticoid administration for assisted
reproductive technology cycles. Cochrane Database of Systematic Reviews .
Boomsma CM, Kavelaars A, Eijkemans MJ et al. 2009 Endometrial secretion analysis identifies a cytokine
profile predictive of pregnancy in in-vitro fertilization. Human Reproduction 1, 1-9.
Boomsma CM, Kavelaars A, Eijkemans MJC et al. 2009 Cytokine profiling in endometrial secretions: a non
invasive window on endometrial receptivity. Reproductive Biomedicine Online 18, 85-94.
Borthwick JM, Charnock-Jones DS, Tom BD et al. 2003 Determination of the transcript profile in human
endometrium. Molecular Human Reproduction 9, 19-33.
Botti G, Fugueroa Casas PR, Tozzini I, et al. 1998 Effect of glucocorticoid therapy on embryo implantation
(Abstract). Fertility and Sterility Abstract Book, s423.
Bourgain C, Ubaldi F, Tavaniotou A et al. 2002 Endometrial hormone receptors and proliferation index in the
periovulatory phase of stimulated embryo transfer cycles in comparison with natural cycles and relation to
clinical pregnancy outcome. Fertility and Sterility 78, 237-244.
Bourgain C, Devroey P 2003 The endometrium in stimulated cycles for IVF. Human Reproduction Update 9,
515-22.
References | 163
Boxmeer JC, Brouns MM, Lindemans J, et al. 2005 Folic acid treatment affects oocyte environment. Journal of
Social Gynecological Investigation 12, 24A.
Boxmeer JC, Smit M, Weber RFA, et al. 2006 Seminal plasma cobalamin significantly correlates with sperm
concentration in men undergoing IVF or ICSI procedures. Journal of Andrology 28, 521-7.
Brazma A, Hingamp P, Quackenbush J, et al. 2001 Minimum information about a microarray experiment
(MIAME)-toward standards for microarray data. Nat Genet 29, 365-71
Brison DR, Houghton FD, Falconer D, et al. 2004. Identification of viable embryos in IVF by non-invasive
measurement of amino acid turnover. Human Reproduction 19, 2319-24.
Brown SE, Mandelin E, Oehninger S et al. 2000 Endometrial glycodelin-A expression in the luteal phase of
stimulated ovarian cycles. Fertiity and Sterility 74, 130-133
Brosens JJ, Gellersen B 2006 Death or survival - progesterone-dependent cell fate decisions in the human
endometrial stroma. J Mol Endocrinol 36, 389-98
Buckett WM 2003 A meta-analysis of ultrasound-guided versus clinical touch embryo transfer. Fertility and
Sterility 80, 1037-41.
Buckett WM 2006 A review and meta-analysis of prospective trials comparing different catheters used for
embryo transfer. Fertility and Sterility 85, 728-34.
Bulmer JN, Morrison L, Longfellow M, et al. 1991. Granulated lymphocytes in human endometrium:
histochemical and immunohistochemical studies. Human Reproduction 6, 791-8.
Burney RO, Talbi S, Hamilton AE, Vo et al. 2007 Gene expression analysis of endometrium reveals
progesterone resistance and candidate susceptibility genes in women with endometriosis. Endocrinology 148,
3814-26
Burton GJ, Watson AL, Hempstock J, et al. 2002 Uterine glands provide histiotrophic nutrition for the human
fetus during the first trimester of pregnancy Journal of Clinical Endocrinology and Metabolism 87, 2954-9.
Caballero-Campo P, Dominguez F, Coloma J, et al. 2002 Hormonal and embryonic regulation of chemokines
IL-8, MCP-1 and RANTES in the human endometrium during the window of implantation. Molecular Human
Carson DD, Bagchi I, Dey SK, et al. 2000. Embryo implantation. Developmental Biology 223, 217-237.
Carson DD, Lagow E, Thathiah A et al. 2002 Changes in gene expression during the early to mid-luteal
(receptive phase) transition in human endometrium detected by high-density microarray screening. Molecular
Human Reproduction 8, 871-879
Catt JW, Ryan J P, Saunders DM, et al. 1994 Short-term corticosteroid treatment does not improve implantation
for embryos derived from subzonal insertion of sperm. Fertility and Sterility 61, 565-6.
Cauci S, Guaschino S, de Aloysio D, et al. 2003 Interrelationships of interleukin-8 with interleukin-1{beta} and
neutrophils in vaginal fluid of healthy and bacterial vaginosis positive women. Molecular Human Reproduction
9, 53-8.
Cauci SAB, Driussi SILV, Guaschino SECO, et al. 2002 Correlation of Local Interleukin-1beta Levels with
Specific IgA Response Against Gardnerella vaginalis Cytolysin in Women with Bacterial Vaginosis. American
Journal of Reproductive Immunology 47, 257-64.
Chaouat G, Menu E, Clark DA, et al. 1990 Control of fetal survival in CBA x DBA/2 mice by lymphokine
therapy. Journal of Reproductive Medicine 89, 447-58.
Chaouat G, Assal MA, Martal J et al. 1995 IL-10 prevents naturally occurring fetal loss in the CBA x DBA/2
mating combination, and local defect in IL-10 production in this abortion-prone combination is corrected by in
vivo injection of IFN-tau. The Journal of Immunology 154, 4261-8.
Chaouat G, Zourbas S, Ostoijc S et al. 2001 New insights into maternal-fetal interactions at implantation.
Reproductive Biomedicine Online 2, 198-203.
Chaouat G, Ledee-Bataille N, Dubanchet S, et al. 2004 TH1/TH2 paradigm in pregnancy: paradigm lost?
Cytokines in pregnancy/early abortion: reexamining the TH1/TH2 paradigm. International Archives of Allergy
and Immunology 134, 93-119.
Chaouat G 2007. The Th1/Th2 paradigm: still important in pregnancy? Seminars in Immunopathology 29, 95-
113.
Chaouat G, Dubanchet S, Ledee N 2007 Cytokines: Important for implantation? Journal of Assisted
Reproduction and Genetics 24, 491-505.
Chard T 1995. Cytokines in implantation. Human Reproduction Update 1, 385-396.
Chobotova K, Spyropoulou I, Carver J et al. 2002 Heparin-binding epidermal growth factor and its receptor
ErbB4 mediate implantation of the human blastocyst. Mechanisms of Development 119, 137-44.
Choudhury SR and Knapp LA 2001 Human reproductive failure II: Immunogenetic and interacting factors.
Human Reproduction Update 7, 135-160.
Christian M, Marangos P, Mak I, et al. 2001. Interferon-gamma modulates prolactin and tissue factor
expression in differentiating human endometrial stromal cells. Endocrinology 142, 3142-51.
Chryssikopoulos A, Mantzavinos T, Kanakas N, et al. 1996 Correlation of serum and follicular fluid
concentrations of placental protein 14 and CA-125 in in vitro fertilization-embryo transfer patients. Fertility
and Sterility 66, 599-603.
Chwalisz K & Garfield RE 2000 Role of nitric oxide in implantation and menstruation. Human Reproduction
15 Suppl 3, 96-111.
Clark EA, Golub TR, Lander ES, et al. 2000 Genomic analysis of metastasis reveals an essential role for RhoC.
Nature 406, 532-5
Cohen J, Inge KL, Suzman M, et al. 1989 Videocinematography of fresh and cryopreserved embryos: a
retrospective analysis of embryonic morphology and implantation. Fertility and Sterility 51, 820-7.
Cohen J, Malter H, Elsner C, et al. 1990 Immunosuppression supports implantation of zona pellucida dissected
human embryos. Fertility and Sterility 53, 662-5.
Cohen M, Meisser A, Haenggeli L, et al. 2006 Involvement of MAPK pathway in TNF-{alpha}-induced MMP-
9 expression in human trophoblastic cells Molecular Human Reproduction 12, 225-32.
Cole PA & Robinson CH 1990 Mechanism and inhibition of cytochrome P-450 aromatase. Journal of
Medicinal Chemistry 33, 2933-42.
Coroleu B, Barri PN, Carreras O, et al. 2002 The influence of the depth of embryo replacement into the uterine
cavity on implantation rates after IVF: a controlled, ultrasound-guided study. Human Reproduction 17, 341-6.
Costea DM, Gunn LK, Hargreaves C, et al. 2000 Delayed luteo-placental shift of progesterone production in
IVF pregnancy. International Journal of Gynaecology and Obstetrics 68, 123-9.
References | 165
Costello MF, Chapman M & Conway U 2006 A systematic review and meta-analysis of randomized controlled
trials on metformin co-administration during gonadotrophin ovulation induction or IVF in women with
polycystic ovary syndrome. Human Reproduction 21, 1387-99.
Coulam CB, Stern JJ, Soenksen DM, et al. 1995 Comparison of pulsatility indices on the day of oocyte retrieval
and embryo transfer. Human Reproduction 10, 82-4.
Coulam CB, Jeyendran RS, Fishel LA, et al. 2006 Multiple thrombophilic gene mutations are risk factors for
implantation failure. Reproductive Biomedicine Online 12, 322-7
Coutifaris C, Myers ER, Guzick DS, et al. 2004. Histological dating of timed endometrial biopsy tissue is not
related to fertility status. Fertility and Sterility 82, 1264-72.
Creus M, Ordi J, Fabregues F et al. 2003 The effect of different hormone therapies on integrin expression and
pinopode formation in the human endometrium: a controlled study. Human Reproduction 18, 683-693
Croucher CA, Lass A, Margara R, et al.1998 Predictive value of the results of a first in-vitro fertilization cycle
on the outcome of subsequent cycles. Human Reproduction 13, 403-8
Croy BA, Ashkar AA, Foster RA et al. 1997 Histological studies of gene-ablated mice support important
functional roles for natural killer cells in the uterus during pregnancy. Journal of Reproductive Immunology 35,
111-33.
Croy BA, Chantakru S, Esadeg S, et al. 2002 Decidual natural killer cells: key regulators of placental
development (a review). Journal of Reproductive Immunology 57, 151-68.
Damario MA, Lesnick TG, Lessey BA 2001 Endometrial markers of uterine receptivity utilizing the donor
oocyte model. Human Reproduction 16, 1893-9.
Daya S, Gunby J 2004 Luteal phase support in assisted reproductive cycles. Cochrane Database of Systematic
Reviews Issue 3, CD004830.
Daya S 2006 Is there a benefit of low-dose aspirin in assisted reproduction? Current Opinions in Obstetrics and
Gynecology 18, 313-8.
de Jager W, te Velthuis VH, Prakken BJ, et al. 2003 Simultaneous detection of 15 human cytokines in a single
sample of stimulated peripheral blood mononuclear cells. Clinical and diagnostic laboratory immunology 10,
133-9.
Deb K, Chatturvedi MM, Jaiswal YK. 2004 Gram-negative bacterial endotoxin- induced infertility: a birds eye
view. Gynecol Obstet Invest. 57, 224-32.
Delage G, Moreau JF, Taupin JL, et al. 1995. In-vitro endometrial secretion of human interleukin for DA
cells/leukaemia inhibitory factor by explant cultures from fertile and infertile women. Human Reproduction 10,
2483-2488.
DeLoia JA, Krasnow JS, Brekosky J, et al. 1998. Regional specialization of the cell membrane-associated,
polymorphic mucin (MUC1) in human uterine epithelia. Human Reproduction 13, 2902-2909.
DeLoia JA, Stewart-Akers AM, Brekosky J, et al. 2002 Effects of exogenous estrogen on uterine leukocyte
recruitment. Fertility and Sterility 77, 548-554
Deng L, Shipley GL, Loose-Mitchell DS, et al. 2003 Coordinate regulation of the production and signaling of
retinoic acid by estrogen in the human endometrium. J Clin Endocrinol Metab 88, 2157-63
Devroey P, Bourgain C, Macklon NS, et al. 2004 Reproductive biology and IVF: ovarian stimulation and
endometrial receptivity. Trends in Endocrinology and Metabolism 15, 84-90.
Dey SK, Lim H, Das SK et al. 2004 Molecular cues to implantation. Endocrine Reviews 25, 341-73.
Dicker D, Ashkenazi J, Feldberg D, et al.1992 The value of repeat hysteroscopic evaluation in patients with
failed in vitro fertilization transfer cycles. Fertility and Sterility 58, 833-5
Di Nisio M, Peters L & Middeldorp S 2005 Anticoagulants for the treatment of recurrent pregnancy loss in
women without antiphospholipid syndrome. Cochrane Databse of Systematic Reviews.
Diedrich K, Fauser BCJM, Devroey P et al. On behalf of the Evian Annual Reproduction (EVAR) Workshop
Group 2007: The role of the endometrium and embryo in human implantation. Human Reproduction Update
13, 365-77.
Dimitriadis E, White CA, Jones RL et al. 2005 Cytokines, chemokines and growth factors in endometrium
related to implantation. Human Reproduction Update 11, 613-30.
Dmowski WP, Rana N, Michalowska J, et al. 1995 The effect of endometriosis, its stage and activity, and of
autoantibodies on in vitro fertilization and embryo transfer success rates. Fertility and Sterility 63, 555-2.
Dominguez F, Pellicer A, Simon C 2003 The Chemokine Connection: Hormonal and Embryonic Regulation at
the Human Maternal-Embryonic Interface--A Review Placenta 24, S48- S55
Dominguez F, Gadea B, Esteban FJ et al. 2008 Comparative protein-profile analysis of implanted versus non-
implanted human blastocysts. Human Reproduction 23, 1993-2000
Donoso P, Staessen C, Fauser BCJM, et al. 2007. Current value of preimplantation genetic aneuploidy
screening in IVF. Human Reproduction Update 13, 15-25.
Dosiou C and Giudice LC 2005 Natural Killer Cells in Pregnancy and Recurrent Pregnancy Loss: Endocrine
and Immunologic Perspectives. Endocrine Reviews 26, 44-62.
Drake PM, Red-Horse K, Fisher SJ 2004 Reciprocal chemokine receptor and ligand expression in the human
placenta: implications for cytotrophoblast differentiation Developmental Dynamics 229, 877-85.
Duvan CI, Ozmen B, Satiroglu H, et al. 2006 Does addition of low-dose aspirin and/or steroid as a standard
treatment in nonselected intracytoplasmic sperm injection cycles improve in vitro fertilization success? A
randomized, prospective, placebo-controlled study. Journal of Assisted Reproduction and Genetics 23, 15-21.
Ebner T, Moser M, Sommergruber M, et al. 2003. Selection based on morphological assessment of oocytes and
embryos at different stages of preimplantation development: a review. Human Reproduction Update 9, 251-62.
Eckert LO, Moore DE, Patton DL, et al. 2003 Relationship of vaginal bacteria and inflammation with
conception and early pregnancy loss following in-vitro fertilization. Infectious Diseases and Obstetrics and
Gynecology11.11-7.
Edwards RG, Lobo R & Bouchard P 1996 Time to revolutionize ovarian stimulation. Human Reproduction 11,
917-9.
Egbase PE, al-Sharhan M, al-Othman S, et al. 1996 Incidence of microbial growth from the tip of the embryo
transfer catheter after embryo transfer in relation to clinical pregnancy rate following in-vitro fertilization and
embryo transfer. Human Reproduction 11, 1687-9.
Eijkelkamp N, Heijnen CJ, Lucas A et al. 2007 G-protein-coupled receptor kinase 6 controls chronicity and
severity of dextran sodium sulphate-induced colitis in mice. Gut 56, 847-54.
Empson M, Lassere M, Craig JC, et al. 2002 Recurrent pregnancy loss with antiphospholipid antibody: a
systematic review of therapeutic trials. Obstetrics and Gynecology 99, 135-44.
References | 167
Empson M, Lassere M, Craig J, et al. 2005 Prevention of recurrent miscarriage for women with
antiphospholipid antibody or lupus anticoagulant. Cochrane Databse of Systematic Reviews.
ESHRE PGD Consortium Steering Committee 2002 ESHRE Preimplantation Genetic Diagnosis Consortium:
data collection III (May 2001). Human Reproduction 17, 233-46
Ezzeldin F, Rahman AF, Agameya AF, et al. 2003 Prednisolone administration in the luteal phase in ICSI
(Abstract). Human Reproduction Abstract book 18, 116.
Fanchin R, Harmas A, Benaoudia F, et al. 1998 Lundkvist U, Olivennes F, Frydman R. Microbial flora of the
cervix assessed at the time of embryo transfer adversely affects in vitro fertilization outcome. Fertility and
Fanchin R, Righini C, de Ziegler D, et al. 2001 Effects of vaginal progesterone administration on uterine
contractility at the time of embryo transfer. Fertility and Sterility 75, 1136-40.
Fanchin R, Righini C, Olivennes F, et al. 1998 Uterine contractions at the time of embryo transfer alter
pregnancy rates after in-vitro fertilization. Human Reproduction 13, 1968-74.
Farhi J, Weissman A, Steinfeld Z, et al. 2000 Estradiol supplementation during the luteal phase may improve
the pregnancy rate in patients undergoing in vitro fertilization-embryo transfer cycles. Fertility and Sterility 73,
761-6.
Fauser BC, Devroey P, Yen SS, et al. 1999 Minimal ovarian stimulation for IVF: appraisal of potential benefits
and drawbacks. Human Reproduction 14, 2681-6.
Fauser BC & Macklon NS 2004 Chapter: Medical Approaches to Ovarian Stimulation for Infertility. Book: Yen
and Yaffe's Reproductive Endocrinology. Elsevier Saunders. Editors: Strauss JF, Barbieri R.
Fauser BC, Devroey P & Macklon NS 2005 Multiple birth resulting from ovarian stimulation for subfertility
treatment. Lancet 365, 1807-16.
Fleming TP, Kwong WY, Porter R et al. 2004 The embryo and its future (Review) Biology of Reproduction 71,
1046-54.
Forges T, Monnier-Barbarino P, Guillet-May F, et al. 2006 Corticosteroids in patients with antiovarian

antibodies undergoing in vitro fertilization: a prospective pilot study. European Journal of Clinical
Pharmacology 62, 699-705.
Franchi A, Zaret J, Zhang X, et al. 2008 Expression of immunomodulatory genes, their protein products and
specific ligands/receptors during the window of implantation in the human endometrium. Molecular Human
Reproduction 14, 413-21
Fridstrom M, Carlstrom K, Sjoblom P, et al. 1999 Effect of prednisolone on serum and follicular fluid androgen
concentrations in women with polycystic ovary syndrome undergoing in-vitro fertilization. Human
Gaast van der MH., Eijkemans MJC, Net van der JB, et al. 2006 The optimum number of oocytes for a
successful first IVF treatment cycle. Reproductive Biomedicine Online 13, 476-80.
Gargett CE, Lederman FL, Lau TM et al. 1999 Lack of correlation between vascular endothelial growth factor
production and endothelial cell proliferation in the human endometrium. Human Reproduction 14, 2080-8.
Gaudoin M, Rekha P, Morris A, et al. 1999 Bacterial vaginosis and past chlamydial infection are strongly and
independently associated with tubal infertility but do not affect in vitro fertilization success rates. Fertility and
Genbacev OD, Prakobphol A, Foulk RA, et al. 2003. Trophoblast L-selectin-mediated adhesion at the maternal-
fetal interface. Science 299, 405-8.
Geva E, Fait G, Lerner-Geva L, et al. 1999 The possible role of antiovary antibodies in repeated in vitro
fertilization failures. American Journal of Reproductive Immunology 42, 292-6.
Geva E, Amit A, Lerner-Geva L, et al. 2000 Prednisone and aspirin improve pregnancy rate in patients with
reproductive failure and autoimmune antibodies: a prospective study. American Journal of Reproductive
Immunology 43, 36-40.
Ghosh D, De P & Sengupta J 1994 Luteal phase ovarian oestrogen is not essential for implantation and
maintenance of pregnancy from surrogate embryo transfer in the rhesus monkey. Human Reproduction 9, 629-
37.
Gianaroli L, Magli MC, Munne S, et al. 1997 Will preimplantation genetic diagnosis assist patients with a poor
prognosis to achieve pregnancy? Human Reproduction 12, 1762-7
Ginger MR, Gonzalez-Rimbau MF, Gay JP, et al 2001 Persistent changes in gene expression induced by
estrogen and progesterone in the rat mammary gland. Mol Endocrinol 15, 1993-2009
Gipson IK, Blalock T, Tisdale A, et al. 2008 MUC16 is lost from the uterodome (pinopode) surface of the
receptive human endometrium: in vitro evidence that MUC16 is a barrier to trophoblast adherence. Biology of
Reproduction 78, 134-42
Giudice LC 1994. Growth factors and growth modulators in human uterine endometrium: their potential
relevance to reproductive medicine. Fertility and Sterility 61, 1-17.
Gleicher N and Barad D 2006 The relative myth of elective single embryo transfer. Human Reproduction 21,
1337-1344.
Gobert B, Barabarino-Monnier P, Guillet-Rosso F, et al. 1990 Ovary antibodies after IVF. Lancet 335, 723.
Golub TR, Slonim DK, Tamayo P, et al. 1999 Molecular classification of cancer: class discovery and class
prediction by gene expression monitoring. Science 286, 531-7
Goswami SK, Das T, Chattopadhyay R, et al. 2004 A randomized single-blind controlled trial of letrozole as a
low-cost IVF protocol in women with poor ovarian response: a preliminary report. Human Reproduction 19,
2031-5.
Gott AL, Hardy K, Winston RM et al. 1990 Non-invasive measurement of pyruvate and glucose uptake and
lactate production by single human preimplantation embryos. Human Reproduction 5, 104-108.
Graham CH, Lysiak JJ, McCrae KR, et al. 1992 Localization of transforming growth factor-beta at the human
fetal-maternal interface: role in trophoblast growth and differentiation. Biology of Reproduction 46, 561-72.
Grandone E, Colaizzo D, Lo BA, et al. 2001 Inherited thrombophilia and in vitro fertilization implantation
failure. Fertility and Sterility 76, 201-2
Gray CA, Taylor KM, Ramsey WS et al. 2001 Endometrial glands are required for preimplantation conceptus
elongation and survival. Biology of Reproduction 64, 1608-1613
Greb RR., Behre HM & Simoni M 2005 Pharmacogenetics in ovarian stimulation - current concepts and future
options. Reproductive Biomedicine Online 11, 589.
Greene JF, III, Kuehl TJ, Allen SR. 2000 The papanicolaou smear: inadequate screening test for bacterial
vaginosis during pregnancy. American Journal of Obstetrics and Gynecology 182, 1048-9.
References | 169
Griesinger G, Franke K, Kinast C, et al. 2002 Ascorbic Acid Supplement During Luteal Phase in IVF. Journal
of Assisted Reproduction and Genetics 19, 164-8.
Guimond MJ, Wang B & Croy BA 1998 Engraftment of bone marrow from severe combined immunodeficient
(SCID) mice reverses the reproductive deficits in natural killer cell-deficient tg epsilon 26 mice. Journal of
Experimental Medicine 187, 217-23.
Haas GG Jr, Manganiello P. 1987 A double-blind, placebo-controlled study of the use of methylprednisolone in
infertile men with sperm-associated immunoglobulins. Fertility and Sterility 47, 295-301.
Hanna J, Goldman-Wohl D, Hamani Y, et al. 2006 Decidual NK cells regulate key developmental processes at
the human fetal-maternal interface Nature Medicine 12, 1065-74.
Hannan NJ, Jones RL, White CA, et al. 2006. The Chemokines, CX3CL1, CCL14, and CCL4, Promote Human
Trophoblast Migration at the Feto-Maternal Interface. Biology of Reproduction 74, 896-904.
Harlow CR, Jenkins JM, Winston RM. 1997 Increased follicular fluid total and free cortisol levels during the
luteinizing hormone surge. Fertility and Sterility 68, 48-53.
Harrell FE 2001 Missing data. In: Regression Modeling Strategies With Applications to Linear Models,
Logistic Regression, and Survival Analysis. Springer-Verlag New York, Inc. pp. 41-51
Hasegawa I, Yamanoto Y, Suzuki M, et al. 1998 Prednisolone plus low-dose aspirin improves the implantation
rate in women with autoimmune conditions who are undergoing in vitro fertilization. Fertility and Sterility 70,
1044-8.
Healey S, Tan SL, Tulandi T, et al. 2003 Effects of letrozole on superovulation with gonadotropins in women
undergoing intrauterine insemination. Fertlity and Sterility 80, 1325-9.
Hedges S, Barrientes F, Desmond R, et al. 2006 Local and Systemic Cytokine Levels in Relation to Changes in
Vaginal Flora. Journal of Infectious Diseases. 193, 556-62.
Heffner LJ 2004. Advanced Maternal Age -- How Old Is Too Old? New England Journal of Medicine 351,
1927-9.
Hefler LA & Gregg AR 2002 Inducible and endothelial nitric oxide synthase: genetic background affects
ovulation in mice. Fertility and Sterility 77, 147-51.
Hendry WF, Hughes L, Scammell G, et al. 1990 Comparison of prednisolone and placebo in subfertile men
with antibodies to spermatozoa. Lancet 335, 85-8.
Hess AP, Hamilton AE, Talbi S, et al. 2007 Decidual Stromal Cell Response to Paracrine Signals from the
Trophoblast: Amplification of Immune and Angiogenic Modulators. Biology of Reproduction 76, 102-17
Hill JA, Polgar K, and Anderson DJ 1995 T-helper 1-type immunity to trophoblast in women with recurrent
spontaneous abortion. JAMA 273, 1933-1936.
Hill JA 1995 T-helper 1-type immunity to trophoblast: evidence for a new immunological mechanism for
recurrent abortion in women. Human Reproduction 10 Suppl 2, 114-20.
Hohmann FP, Macklon NS & Fauser BC 2003 A randomized comparison of two ovarian stimulation protocols
with gonadotropin-releasing hormone (GnRH) antagonist cotreatment for in vitro fertilization commencing
recombinant follicle-stimulating hormone on cycle day 2 or 5 with the standard long GnRH agonist protocol.
Journal of Clinical Endocrinology Metabolism 88, 166-73.
Hoozemans DA, Schats R, Lambalk CB et al. 2004 Human embryo implantation: current knowledge and
clinical implications in assisted reproductive technology. Reproductive Biomedicine Online 9, 692-712.
Horai R, Asano M, Sudo K et al. 1998 Production of mice deficient in genes for interleukin (IL)-1alpha, IL-
1beta, IL-1alpha/beta, and IL-1 receptor antagonist shows that IL-1beta is crucial in turpentine-induced fever
development and glucocorticoid secretion. Journal of Experimental Medicine 187, 1463-75.
Horcajadas JA, Riesewijk A, Martin J et al. 2004 Global gene expresion profiling of human endometrial
receptivity. Journal of Reproductive Immunology 63, 41-9.
Horcajadas JA, Riesewijk A, Polman J et al. 2005 Effect of controlled ovarian hyperstimulation in IVF on
endometrial gene expression profiles Molecular Human Reproduction 11, 195-205.
Horcajadas JA, Sharkey AM, Catalano RD, et al. 2006 Effect of an intrauterine device on the gene expression
profile of the endometrium. J Clin Endocrinol Metab 91, 3199-207
Horcajadas JA, Pellicer A and Simon C 2007 Wide genomic analysis of human endometrial receptivity: new
times, new opportunities. Human Reproduction Update 13, 77-86.
Hornstein MD, Davis OK, Massey JB, et al. 2000 Antiphospholipid antibodies and in vitro fertilization success:
a meta-analysis. Fertility and Sterility 73, 330-3
Hornung D, Dohrn K, Sotlar K et al. 2000 Localization in Tissues and Secretion of Eotaxin by Cells from
Normal Endometrium and Endometriosis. Journal of Clinical Endocrinology Metabolism 85, 2604-8.
Houghton FD, Hawkhead JA, Humpherson PG, et al. 2002. Non-invasive amino acid turnover predicts human
embryo developmental capacity. Human Reproduction 17, 999-1005.
Hugues JN, Cedrin-Durnerin I, Bstandig B, et al. 2006 Administration of gonadotropin-releasing hormone

agonist during the luteal phase of GnRH-antagonist IVF cycles. Human Reproduction Abstract Book: 22nd
Annual meeting of the ESHRE, O-007 21, i3.
Hunault CC, Laven JSE, van Rooij IAJ, et al. 2005 Prospective validation of two models predicting pregnancy
leading to live birth among untreated subfertile couples. Human Reproduction 20, 1636-41.
Hunt JS, Fishback JL, Andrews GK, et al. 1988. Expression of class I HLA genes by trophoblast cells. Analysis
by in situ hybridization. Journal of Immunology 140, 1293-9.
Igarashi M 1977 Augmentative effect of ascorbic acid upon induction of human ovulation in clomiphene-
ineffective anovulatory women. International Journal of Infertility 22, 168-73.
Imai T, Kurachi H, Adachi K et al. 1995 Changes in epidermal growth factor receptor and the levels of its
ligands during menstrual cycle in human endometrium. Biology of Reproduction 52, 928-38.
Inagaki N, Stern C, McBain J, et al. 2003 Analysis of intra-uterine cytokine concentration and matrix-
metalloproteinase activity in women with recurrent failed embryo transfer Human Reproduction 18, 608-15.
Iwamoto R, Yamazaki S, Asakura M et al. 2003. Heparin-binding EGF-like growth factor and ErbB signaling is
essential for heart function. Proceedings of the National Academy of Sciences of the United States of America
100, 3221-6.
Jeong JW, Lee KY, Kwak I, et al. 2005 Identification of murine uterine genes regulated in a ligand-dependent
manner by the progesterone receptor. Endocrinology 146, 3490-505
Jeziorska M, Salamonsen LA, Woolley DE. 1995 Mast cell and eosinophil distribution and activation in human
endometrium throughout the menstrual cycle. Biology of Reproduction 53, 312-20.
Joesoef MR, Schmid GP, Hillier SL. 1999 Bacterial vaginosis: review of treatment options and potential
clinical indications for therapy. Clinical Infectious Diseases 28 Suppl 1:S57-S65.
References | 171
Jones HW Jr 1996 What has happened? Where are we? Human Reproduction (Abstract). 11, 7.
Jones RL, Hannan NJ, Kaitu'u TJ, et al. 2004 Identification of chemokines important for leukocyte recruitment
to the human endometrium at the times of embryo implantation and menstruation Journal of Clinical
Endocrinology and Metabolism 89, 6155-67.
Kamiyama S, Teruya Y, Nohara M, et al. 2004 Impact of detection of bacterial endotoxin in menstrual effluent
on the pregnancy rate in in vitro fertilization and embryo transfer. Fertility Sterility 82, 788-92.
Kao LC, Tulac S, Lobo S et al. 2002 Global gene profiling in human endometrium during the window of
implantation. Endocrinolog,y 143, 2119-38.
Kao LC, Germeyer A, Tulac S, et al. 2003 Expression profiling of endometrium from women with
endometriosis reveals candidate genes for disease-based implantation failure and infertility. Endocrinology 144,
2870-81
Karmakar S, Das C 2002 Regulation of Trophoblast Invasion by IL-1beta and TGF- beta1 American Journal of
Reproductive Immunology 48, 210-9.
Kats R, Al Akoum M, Guay S et al. 2005 Cycle-dependent expression of macrophage migration inhibitory
factor in the human endometrium. Human Reproduction 20, 3518-25.
Kauma SW, Aukerman SL, Eierman D, et al. 1991 Colony-stimulating factor-1 and c-fms expression in human
endometrial tissues and placenta during the menstrual cycle and early pregnancy. Journal of Clinical
Endocrinology and Metabolism 73, 746-51.
Keay SD, Harlow CR, Wood PJ, et al. 2002 Higher cortisol:cortisone ratios in the preovulatory follicle of
completely unstimulated IVF cycles indicate oocytes with increased pregnancy potential. Human Reproduction
17, 2410-4.
Kemeter P, Feichtinger W. 1986 Prednisolone supplementation to Clomid and/or gonadotrophin stimulation for
in-vitro fertilization--a prospective randomized trial. Human Reproduction 1, 441-4.
Kennedy TG, Gillio-Meina C, Phang SH 2007 Prostaglandins and the initiation of blastocyst implantation and
decidualization. Journal of Reproduction and Fertility 134, 635-43
Khong TY and Robertson WB 1987 Placenta creta and placenta praevia creta. Placenta 8, 399-409.
Kim CH, Cho YK, Mok JE. 1996 The efficacy of immunotherapy in patients who underwent superovulation
with intrauterine insemination. Fertility and Sterility 65, 133-8.
Kim CH, Chae HD, Kang BM, et al. 1997 The immunotherapy during in vitro fertilization and embryo transfer
cycles in infertile patients with endometriosis. The Journal of Obstetrics and Gynaecology Research 23, 463-
70.
Kitaya K, Nakayama T, Daikoku N, et al. 2004 Spatial and Temporal Expression of Ligands for CXCR3 and
CXCR4 in Human Endometrium Journal of Clinical Endocrinology and Metabolism 89, 2470-6.
Klebanoff MA, Schwebke JR, Zhang J, et al. 2004 Vulvovaginal symptoms in women with bacterial vaginosis.
Obstetrics and Gynecology 104, 267-72.
Klonoff-Cohen H, Bleha J & Lam-Kruglick P 2002 A prospective study of the effects of female and male
caffeine consumption on the reproductive endpoints of IVF and gamete intra-Fallopian transfer. Human
Kolibianakis E, Bourgain C, Albano C, et al. 2002 Effect of ovarian stimulation with recombinant follicle-
stimulating hormone, gonadotropin releasing hormone antagonists, and human chorionic gonadotropin on
endometrial maturation on the day of oocyte pick-up. Fertility and Sterility 78, 1025-9.
Korn AP, Bolan G, Padian N, et al. 1995 Plasma cell endometritis in women with symptomatic bacterial
vaginosis. Obstetrics and Gynecology 85, 387-90.
Kowalik A, Liu HC, He ZY, et al. 1999 Expression of the insulin-like growth factor-1 gene and its receptor in
preimplantation mouse embryos; is it a marker of embryo viability? Molecular Human Reproduction 5, 861-5.
Krohn MA, Hillier SL, Eschenbach DA. 1989 Comparison of methods for diagnosing bacterial vaginosis
among pregnant women. Journal of Clinical Microbiology 27, 1266-71.
Kuhn R, Lohler D, Rennick K, et al. 1993 Interleukin-10-deficient mice develop chronic enterocolitis. Cell. 75,
263-74.
Kuo HC, Hsu CC, Wang ST, et al. 1997 Aspirin improves uterine blood flow in the peri-implantation period.
Journal of the Formosan Medical Association 96, 253-7.
Kuo RC, Baxter GT, Thompson SH, et al. 2000 NO is necessary and sufficient for egg activation at
fertilization. Nature 406, 633-6.
Lai TH, Zhao Y, Shih I et al. 2006 Expression of L-selectin ligands in human endometrium during the
implantation window after controlled ovarian stimulation for oocyte donation. Fertility and Sterility. 85, 761-
763
Laird SM, Tuckerman EM, Dalton CF et al. 1997 The production of leukaemia inhibitory factor by human
endometrium: presence in uterine flushings and production by cells in culture. Human Reproduction 12, 569-
74.
Laird SM, Tuckerman EM, Li TC 2006 Cytokine expression in the endometrium of women with implantation
failure or recurrent miscarriage. Reproductive Biomedicine Online 13, 13-23.
Larsson PG, Platz-Christensen JJ, Sundstrom E. 1991 Is bacterial vaginosis a sexually transmitted disease?
International Journal of Sexual Transmitted Diseases and AIDS. 2, 362-4.
Lass A, Peat D, Avery S, Brinsden P. 1998 Histological evaluation of endometrium on the day of oocyte
retrieval after gonadotrophin-releasing hormone agonist-follicle stimulating hormone ovulation induction for
in-vitro fertilization. Human Reproduction 13, 3203-3205
Lédée-Bataille N, Laprée-Delage G, Taupin JL et al. 2002 Concentration of leukaemia inhibitory factor (LIF) in
uterine flushing fluid is highly predictive of embryo implantation. Human Reproduction 17, 213-8.
Ledee-Bataille N, Olivennes F, Kadoch J et al. 2004a Detectable levels of interleukin-18 in uterine luminal
secretions at oocyte retrieval predict failure of the embryo transfer. Human Reproduction 19, 1968-73.
Ledee-Bataille N, Dubanchet S, Coulomb-L'hermine A et al. 2004b A new role for natural killer cells,
interleukin (IL)-12, and IL-18 in repeated implantation failure after in vitro fertilization. Fertility and Sterility
81, 59-65.
Ledee-Bataille N, Bonnet-Chea K, Hosny G et al. 2005 Role of the endometrial tripod interleukin-18, -15, and -
12 in inadequate uterine receptivity in patients with a history of repeated in vitro fertilization-embryo transfer
failure. Fertility and Sterility 83, 598-605.
Lee TH, Wu MY, Chen MJ, et al. 2004 Nitric oxide is associated with poor embryo quality and pregnancy
outcome in in vitro fertilization cycles. Fertility and Sterility 82, 126-31.
Lee YL, Liu Y, Ng PY et al. 2008 Aberrant expression of angiopoietins-1 and -2 and vascular endothelial
growth factor-A in peri-implantation endometrium after gonadotrophin stimulation. Human Reproduction 23,
894-903
References | 173
Leese HJ, Sturmey RG, Baumann CG, et al. 2007. Embryo viability and metabolism: obeying the quiet rules.
Human Reproduction 22, 3047-3050.
Leitich H, Kiss H. 2007 Asymptomatic bacterial vaginosis and intermediate flora as risk factors for adverse
pregnancy outcome. Best practice & research. Clinical obstetrics & gynaecology 21, 375-90.
Lessey BA, Killam AP, Metzger DA et al. 1988 Immunohistochemical analysis of human uterine estrogen and
progesterone receptors throughout the menstrual cycle. Journal of Clinical Endocrinology and Metabolism 67,
334-340
Lessey BA, Damjanovich L, Coutifaris C, et al. 1992 Integrin adhesion molecules in the human endometrium.
Correlation with the normal and abnormal menstrual cycle. Journal of Clinical Investigation 90, 188-95.
Lessey BA, Castelbaum AJ, Sawin SW, et al. 1995 Integrins as markers of uterine receptivity in women with
primary unexplained infertility. Fertility and Sterility 63, 535-42.
Lessey BA 1998 Endometrial integrins and the establishment of uterine receptivity. Human Reproduction 13
Suppl 3, 247-58.
Lessey BA, Gui Y, Apparao KB et al. 2002 Regulated expression of heparin-binding EGF-like growth factor
(HB-EGF) in the human endometrium: a potential paracrine role during implantation. Molecular Reproduction
and Development 62, 446-55.
Lewin A, Benshushan A, Mezker E, et al. 1994 The role of estrogen support during the luteal phase of in vitro
fertilization-embryo transplant cycles: a comparative study between progesterone alone and estrogen and
progesterone support. Fertility and Sterility 62, 121-5.
Li TC, MacKenna A and Roberts R 1993 The techniques and complications of out-patient uterine washing in
the assessment of endometrial function. Human Reproduction 8, 343-6.
Lila N, Carpentier A, Amrein C, et al. 2000 Implication of HLA-G molecule in heart-graft acceptance. Lancet
355, 2138.
Lim H, Paria BC, Das SK, et al. 1997 Multiple Female Reproductive Failures in Cyclooxygenase 2-Deficient
Mice. Cell 91, 197-208
Lintsen AM, Pasker-de Jong PC, de Boer EJ, et al. 2005 Effects of subfertility cause, smoking and body weight
on the success rate of IVF. Human Reproduction 20, 1867-75.
Liversedge NH, Turner A, Horner PJ, et al. 1999 The influence of bacterial vaginosis on in-vitro fertilization
and embryo implantation during assisted reproduction treatment. Human Reproduction 14, 2411-5.
Lockwood CJ, Reece EA, Romero R, et al. 1986 Anti-phospholipid antibody and pregnancy wastage. Lancet 2,
742-3.
Loetscher P, Seitz M, Clark-Lewis I, et al. 1994 Monocyte chemotactic proteins MCP-1, MCP-2, and MCP-3
are major attractants for human CD4+ and CD8+ T lymphocytes The FASEB Journal 8, 1055-60.
Long EO 1999 Regulation of immune responses through inhibitory receptors. Annual Review of Immunology
17, 875-904.
Lord JM, Flight IH & Norman RJ 2003 Insulin-sensitising drugs (metformin, troglitazone, rosiglitazone,
pioglitazone, D-chiro-inositol) for polycystic ovary syndrome. Cochrane Databse of Systematic Reviews.
Luck MR & Zhao Y 1993 Identification and measurement of collagen in the bovine corpus luteum and its
relationship with ascorbic acid and tissue development. Journal of Reproduction and Fertility 99, 647-52.
Luck MR, Jeyaseelan I & Scholes RA 1995 Ascorbic acid and fertility. Biology of Reproduction 52, 262-6.
Lukassen HG, Joosten I, van Cranenbroek B et al. 2004 Hormonal stimulation for IVF treatment positively
affects the CD56bright/CD56dim NK cell ratio of the endometrium during the window of implantation.
Molecular Human Reproduction 10, 513-520
Lukaszuk K, Liss J, Lukaszuk M, et al. 2005 Optimization of estradiol supplementation during the luteal phase
improves the pregnancy rate in women undergoing in vitro fertilization-embryo transfer cycles. Fertility and
Ma Wg, Song H, Das SK, et al. 2003 Estrogen is a critical determinant that specifies the duration of the window
of uterine receptivity for implantation. Proceedings of the National Academy of Sciences 100, 2963-8.
Macklon NS & Fauser BC 2000 Impact of ovarian hyperstimulation on the luteal phase. Journal of
Reproduction and Fertility 55, 101-8.
Macklon NS, Geraedts JP, Fauser BC 2002 Conception to ongoing pregnancy: the 'black box' of early
pregnancy loss Human Reproduction Update 8, 333-43.
Macklon NS, Pieters MHEC, Hassan MA, et al. 2002 A prospective randomized comparison of sequential
versus monoculture systems for in-vitro human blastocyst development. Human Reproduction 17, 2700-5.
Macklon NS, Stouffer RL, Giudice LC, 2006 The science behind 25 years of ovarian stimulation for in vitro
fertilization. Endocrine Reviews. 27,170-207.
Macklon NS, Van der Gaast MH, Hamilton AE et al. 2008 The effect of stimulation with recombinant FSH and
GnRH antagonists on endometrial gene expression in the absence of luteal support. Reproductive Sciences 15,
357-365
Magli MC, Jones GM, Gras L, et al. 2000 Chromosome mosaicism in day 3 aneuploid embryos that develop to
morphologically normal blastocysts in vitro. Human Reproduction 15, 1781-6.
Martin KL, Barlow DH, Sargent IL 1998 Heparin-binding epidermal growth factor significantly improves
human blastocyst development and hatching in serum-free medium. Human Reproduction 13, 1645-52.
Martinelli I, Taioli E, Ragni G, et al. 2003 Embryo implantation after assisted reproductive procedures and
maternal thrombophilia. Haematologica 88, 789-93
Mastenbroek S, Twisk M, van Echten-Arends J, et al. 2007 In vitro fertilization with preimplantation genetic
screening. New England Journal of Medicine 357, 9-17.
Masuda A, Yoshikai Y, Aiba K, et al. 2002 Matsuguchi T. Th2 Cytokine Production from Mast Cells Is
Directly Induced by Lipopolysaccharide and Distinctly Regulated by c-Jun N-Terminal Kinase and p38
Pathways. Journal of Immunology 169, 3801-10.
Matsubayashi H, Arai T, Izumi S, et al. 2001 Anti-annexin V antibodies in patients with early pregnancy loss or
implantation failures. Fertility and Sterility 76, 694-9
Matteo MG, Greco P, Rosenberg P, et al. 2007 Normal percentage of CD56bright natural killer cells in young
patients with a history of repeated unexplained implantation failure after in vitro fertilization cycles. Fertility
Mattsby-Baltzer I, Platz-Christensen JJ, Hosseini N, et al. 1998 IL-1beta, IL-6, TNFalpha, fetal fibronectin, and
endotoxin in the lower genital tract of pregnant women with bacterial vaginosis. Acta Obstet Gynecol Scand.
77, 701-6.
References | 175
McDonald HM, Brocklehurst P, Gordon A. 2007 Antibiotics for treating bacterial vaginosis in pregnancy.
Cochrane Database Syst Rev. Issue 1, CD000262.
Medawar PB 1953 Some immunological and endocrinological problems raised by the evolution of viviparity in
vertebrates. In Society for Experimental Biology, (New York: Academic Press), 320-38.
Meekins JW, Pijnenborg R, Hanssens M, et al. 1994 A study of placental bed spiral arteries and trophoblast
invasion in normal and severe pre-eclamptic pregnancies. British Journal of Obstetric and Gynaecology 101,
669-674.
Meisser A, Chardonnens D, Campana A, et al. 1999 Effects of tumour necrosis factor-{alpha}, interleukin-1
{alpha}, macrophage colony stimulating factor and transforming growth factor on trophoblastic matrix
metalloproteinases Molecular Human Reproduction 5, 252-60.
Menezo Y, Elder K, and Viville S 2006 Soluble HLA-G release by the human embryo: an interesting artefact?
Mercader A, Garcia-Velasco JA, Escudero E, et al. 2003 Clinical experience and perinatal outcome of
blastocyst transfer after coculture of human embryos with human endometrial epithelial cells: a 5-year follow-
up study. Fertility and Sterility 80, 1162-8.
Mericskay M, Kitajewski J, Sassoon D 2004 Wnt5a is required for proper epithelial-mesenchymal interactions
in the uterus. Development 131, 2061-72.
Mikkelsen AL 2005 Strategies in human in-vitro maturation and their clinical outcome. Reproductive
Biomedicine Online 10, 593-9.
Miller C and Sassoon DA 1998 Wnt-7a maintains appropriate uterine patterning during the development of the
mouse female reproductive tract. Development 125, 3201-11.
Mirkin S, Nikas G, Hsiu JG et al. 2004 Gene expression profiles and structural/functional features of the peri-
implantation endometrium in natural and gonadotropin stimulated cycles. Journal of Clinical Endocrinology
Metabolism 89, 5742-52.
Mirkin S, Arslan M, Churikov D, et al. 2005 In search of candidate genes critically expressed in the human
endometrium during the window of implantation. Human Reproduction 20, 2104-17
Mitwally MFM & Casper RF 2001 Use of an aromatase inhibitor for induction of ovulation in patients with an
inadequate response to clomiphene citrate. Fertility and Sterility 75, 305-9.
Mitwally MF & Casper RF 2003 Aromatase inhibition reduces gonadotrophin dose required for controlled
ovarian stimulation in women with unexplained infertility. Human Reproduction 18, 1588-97.
Mitwally MF & Casper RF 2004 Aromatase inhibition reduces the dose of gonadotropin required for controlled
ovarian hyperstimulation. Journal of Social Gynecological Investigation 11, 406-15.
Miyazaki S, Tanebe K, Sakai M, et al. 2002 Interleukin 2 receptor gamma chain (gamma(c)) knockout mice
show less regularity in estrous cycle but achieve normal pregnancy without fetal compromise. American
Moffett-King A. 2002 Natural killer cells and pregnancy. Nature Reviews. Immunology 2, 656-663
Moffett A, Regan L, Braude P 2004 Natural killer cells, miscarriage, and infertility British Medical Journal
329, 1283-5.
Moffitt D, Queenan JT, Jr., Veeck LL, et al. 1995 Low-dose glucocorticoids after in vitro fertilization and
embryo transfer have no significant effect on pregnancy rate. Fertility and Sterility 63, 571-7.
Mollo A, Strina I, Alviggi C, et al. 2002 Effect of low dose aspirin and prednisone administered after ovum
pick-up on implantation and abortion rates in women with no auto-immune disorders undergoing ICSI.
Giornale Italiane Obst. Gin. 14, 261-5.
Moore DE, Soules MR, Klein NA, et al. 2000 Fujimoto VY, Agnew KJ, Eschenbach DA. Bacteria in the
transfer catheter tip influence the live-birth rate after in vitro fertilization. Fertility and Sterility 74, 1118-24.
Morris M, Nicoll A, Simms I, et al. 2001 Bacterial vaginosis: a public health review. British Journal of
Obstetrics and Gynaecology 108, 439-50.
Mottla GL, Smotrich DB, Gindoff PR, et al. 1996 Increasing clinical pregnancy rates after IVF/ET. Can
immunosuppression help? Journal of Reproductive Medicine 41, 889-91.
Munne S 2006 Chromosome abnormalities and their relationship to morphology and development of human
embryos. Reproductive Biomedicine Online 12, 234-53.
Murphy CR 2000 Understanding the apical surface markers of uterine receptivity: Pinopods--or uterodomes?
Murphy SP, Fast LD, Hanna NN, et al. 2005 Uterine NK Cells Mediate Inflammation-Induced Fetal Demise in
IL-10-Null Mice. Journal of Immunology 175, 4084-90.
Murray MJ, Meyer WR, Zaino RJ et al. 2004 A critical analysis of the accuracy, reproducibility, and clinical
utility of histologic endometrial dating in fertile women. Fertility and Sterility 81, 1333-43.
Nagaoka K, Nojima H, Watanabe F, et al. 2003 Regulation of blastocyst migration, apposition and initial
adhesion by a chemokine, IFN-gamma -inducible protein 10 kDa (IP-10), during early gestation The Journal of
biological chemistry 278, 29048-56.
Nagata Y, Yoshimitsu K, Shirakawa K.1993 [A quality control for the culture system using endotoxin assays in
human in vitro fertilization and embryo transfer]. Nippon Sanka Fujinka Gakkai Zasshi. 45, 220-6. Japanese.
Navot D, Scott RT, Droesch K, et al. 1991 The window of embryo transfer and the efficiency of human
conception in vitro. Fertility and Sterility 55, 114-118.
Narayanan M, Murthy PS, Munaf SA, et al. 1995 Antiovarian antibodies and their effect on the outcome of
assisted reproduction. Journal of Assisted Reproduction and Genetics 12, 599-605
.
Naz RK. 2004 Modalities for treatment of antisperm antibody mediated infertility: novel perspectives.
American Journal of Reproductive Immunology 51, 390-7.
Ng EHY, Chan CCW, Tang OS, et al. 2004 Comparison of endometrial and subendometrial blood flow
measured by three-dimensional power Doppler ultrasound between stimulated and natural cycles in the same
patients. Human Reproduction 19, 2385-2390
Ng EHY, Chan CCW, Tang OS, et al. 2006 The role of endometrial and subendometrial vascularity measured
by three-dimensional power Doppler ultrasound in the prediction of pregnancy during frozen-thawed embryo
transfer cycles. Human Reproduction 21, 1612-7.
Nishino E, Matsuzaki N, Masuhiro K et al. 1990 Trophoblast-derived interleukin-6 (IL-6) regulates human
chorionic gonadotropin release through IL-6 receptor on human trophoblasts. Journal of Clinical
Endocrinology Metabolism 71, 436-41.
Noyes N, Hertig AT and Rock J 1950 Daring the endometrial biopsy. Fertility and Sterility 1, 3-25.
Nugent RP, Krohn MA, Hillier SL. 1991 Reliability of diagnosing bacterial vaginosis is improved by a
standardized method of gram stain interpretation. Journal of Clinical Microbiology 29, 297-301.
References | 177
Nyboe AA, Popovic-Todorovic B, Schmidt KT, et al. 2002 Progesterone supplementation during early
gestations after IVF or ICSI has no effect on the delivery rates: a randomized controlled trial. Human
Ohl J, Lefebvre-Maunoury C, Wittemer C, et al. 2002 Nitric oxide donors for patients undergoing IVF. A
prospective, double-blind, randomized, placebo-controlled trial. Human Reproduction 17, 2615-20.
Oliveira JB, Martins AM, Baruffi RL, et al. 2004 Increased implantation and pregnancy rates obtained by
placing the tip of the transfer catheter in the central area of the endometrial cavity. Reproductive Biomedicine
Online 9, 435-41.
Olivennes F, Ledee-Bataille N, Samama M et al. 2003 Assessment of leukemia inhibitory factor levels by
uterine flushing at the time of egg retrieval does not adversely affect pregnancy rates with in vitro fertilization.
Fertility and Sterility 79, 900-4.
Ombelet W, Cadron I, Gerris J, et al. 2005 Obstetric and perinatal outcome of 1655 ICSI and 3974 IVF
singleton and 1102 ICSI and 2901 IVF twin births: a comparative analysis. Reproductive Biomedicine Online
11, 76-85.
Ombelet W, Martens G, De SP, et al. 2006 Perinatal outcome of 12,021 singleton and 3108 twin births after
non-IVF-assisted reproduction: a cohort study. Human Reproduction 21, 1025-32.
Oum K, Nakhuda GS, Ferin MJ et al. 2006 Detection of secreted intact hCG in conditioned media from human
blastocysts 5 days after oocyte retrieval (Abstract). Fertility and Sterility 86 suppl 1, s219.
Pakkila M, Rasanen J, Heinonen S, et al. 2005 Low-dose aspirin does not improve ovarian responsiveness or
pregnancy rate in IVF and ICSI patients: a randomized, placebo-controlled double-blind study. Human
Pandian Z, Templeton A, Serour G, et al. 2005 Number of embryos for transfer after IVF and ICSI: a Cochrane
review. Human Reproduction 20, 2681-2687.
Park-Wyllie L, Mazzotta P, Pastuszak A, et al. 2000 Birth defects after maternal exposure to corticosteroids:
prospective cohort study and meta-analysis of epidemiological studies. Teratology 62, 385-92.
Parr B A, McMahon AP 1998. Sexually dimorphic development of the mammalian reproductive tract requires
Wnt-7a. Nature 395, 707-10.
Peppel van de J, Kemmeren P, van BH, Radonjic M, et al. 2003 Monitoring global messenger RNA changes in
externally controlled microarray experiments. EMBO Rep 4, 387-93
Piccinni MP, Maggi E and Romagnani S 2000 Role of hormone-controlled T-cell cytokines in the maintenance
of pregnancy. Biochemical Society Transactions 28, 212-15.
Pijnenborg R, Vercruysse L, and Hanssens M 2006 The uterine spiral arteries in human pregnancy: facts and
controversies Placenta 27, 939-958.
Pinsky L, Digeorge A.M. 1965 Cleft palate in the mouse: a teratogenic index of glucocorticoid potency. Science
147, 402-3.
Piot P. 1984 Bacterial vaginosis. An evaluation of treatment. Scandinavian journal of urology and nephrology.
Supplementum 86, 229-35.
Pirard C, Donnez J & Loumaye E 2006 GnRH agonist as luteal phase support in assisted reproduction
technique cycles: results of a pilot study. Human Reproduction 21, 1894-1900.
Platz-Christensen JJ, Mattsby-Baltzer I, Thomsen P,et al. 1993 Endotoxin and interleukin-1 alpha in the
cervical mucus and vaginal fluid of pregnant women with bacterial vaginosis. American Journal of Obstetrics
and Gynecology 169,1161-6.
Pollard JW, Hunt JS, Wiktor-Jedrzejczak W, et al. 1991. A pregnancy defect in the osteopetrotic (op/op) mouse
demonstrates the requirement for CSF-1 in female fertility. Developmental Biology 148, 273-283.
Poli V, Balena R, Fattori E et al. 1994. Interleukin-6 deficient mice are protected from bone loss caused by
estrogen depletion. The EMBO Journal 13, 1189-96.
Pongcharoen S, Somran J, Sritippayawan S et al. 2007 Interleukin-17 Expression in the Human Placenta.
Placenta 28, 59-63.
Ponnampalam AP, Rogers PAW 2008 Expression and regulation of fucosyltransferase 4 in human
endometrium. Journal of Reproduction and Fertility 136, 117-23
Ponnampalam AP, Weston GC, Trajstman AC, et al. 2004 Molecular classification of human endometrial cycle
stages by transcriptional profiling. Molecular Human Reproduction 10, 879-93
Popovic-Todorovic B, Loft A, Bredkjaeer HE, et al. 2003a A prospective randomized clinical trial comparing
an individual dose of recombinant FSH based on predictive factors versus a 'standard' dose of 150 IU/day in
'standard' patients undergoing IVF/ICSI treatment. Human Reproduction 18, 2275-82.
Popovic-Todorovic B, Loft A, Lindhard A, et al. 2003b A prospective study of predictive factors of ovarian
response in 'standard' IVF/ICSI patients treated with recombinant FSH. A suggestion for a recombinant FSH
dosage normogram. Human Reproduction 18, 781-7.
Popovici RM, Betzler NK, Krause MS, et al. 2006 Gene expression profiling of human endometrial-trophoblast
interaction in a coculture model. Endocrinology 147, 5662-75
Porter S, Clark IM, Kevorkian L, et al. 2005 The ADAMTS metalloproteinases. Biochem J 386, 15-27
Pope CS, Cook EKD, Arny M, et al. 2004 Influence of embryo transfer depth on in vitro fertilization and
embryo transfer outcomes. Fertility and Sterility 81, 51-8.
Primi MP, Senn A, Montag M, et al. 2004 A European multicentre prospective randomized study to assess the
use of assisted hatching with a diode laser and the benefit of an immunosuppressive/antibiotic treatment in
different patient populations. Human Reproduction 19, 2325-33.
Pritts EA & Atwood AK 2002 Luteal phase support in infertility treatment: a meta-analysis of the randomized
trials. Human Reproduction 17, 2287-99.
Qublan HS, Eid SS, Ababneh HA, Amarin et al. 2006 Acquired and inherited thrombophilia: implication in
recurrent IVF and embryo transfer failure. Human Reproduction 21, 2694-8
Quenby S, Bates M, Doig T, et al. 1999. Pre-implantation endometrial leukocytes in women with recurrent
miscarriage. Human Reproduction 14, 2386-91
Quenby S, Vince G, Farquharson R, et al. 2002 Recurrent miscarriage: a defect in nature’s quality control?
Quenby S, Kalumbi C, Bates M, Farquharson R et al. 2005 Prednisolone reduces preconceptual endometrial
natural killer cells in women with recurrent miscarriage. Fertility an Sterility 84, 980-4.
Quenby S, Farquharson R. 2006 Uterine natural killer cells, implantation failure and recurrent miscarriage.
Reproductive Biomedicine Online 13, 24-28
References | 179
Quinn C, Ryan E, Claessens EA, et al. 2007 The presence of pinopodes in the human endometrium does not
delineate the implantation window. Fertility and Sterility 87, 1015-21.
Raghupathy R, Makhseed M, Azizieh F et al. 1999. Maternal Th1 and Th2 type reactivity antigens in normal
human pregnancy and unexplained recurrent spontaneous abortions. Celullar Immunology 196, 122-30.
Rai R, Regan L, Kutteh WH, et al. 2002 Antiphospholipid Syndrome in Pregnancy: A Randomized, Controlled
Trial of Treatment. Obstetrics and Gynecology 100, 1354-6.
Ralph SG, Rutherford AJ, Wilson JD. 1999 Influence of bacterial vaginosis on conception and miscarriage in
the first trimester: cohort study. British Medical Journal. 319,:220-3.
Rand JH, Wu XX, Andree HA, et al. 1997 Pregnancy loss in the antiphospholipid-antibody syndrome--a
possible thrombogenic mechanism. New England Journal of Medicine 337, 154-60.
Rebmann V, Switala M, Eue I, et al. 2007 Rapid evaluation of soluble HLA-G levels in supernatants of in vitro
fertilized embryos. Human Immunology 68, 251-8.
Reimand K, Talja I, Metskula K, et al. 2001 Autoantibody studies of female patients with reproductive failure.
Rein MS, Jackson KV, Sable DB, et al. 1996 Dexamethasone during ovulation induction for in-vitro
fertilization: a pilot study. Human Reproduction 11, 253-5.
Reis FM, Lhullier C, Edelweiss MI et al. 2005 In vivo assessment of the regulation of transforming growth
factor alpha, epidermal growth factor (EGF), and EGF receptor in the human endometrium by
medroxyprogesterone acetate. Journal of Assisted Reproduction and Genetics 22, 19-24.
Reese J, Das SK, Paria BC, et al. 2001 Global gene expression analysis to identify molecular markers of uterine
receptivity and embryo implantation. J Biol Chem 276, 44137-45
Riesewijk A, Martin J, van Os R et al. 2003 Gene expression profiling of human endometrial receptivity on
days LH+2 versus LH+7 by microarray technology. Molecular Human Reproduction 9, 253-64.
Robb L, Li R, Hartley L, et al 1998 Infertility in female mice lacking the receptor for interleukin 11 is due to a
defective uterine response to implantation. Nature Medicine 4, 303-8.
Robertson SA, Mau VJ, Young IG et al. 2000a Uterine eosinophils and reproductive performance in interleukin
5-deficient mice. Reproduction 120, 423-32.
Robertson SA, O'Connell AC and Ramsey A 2000b The effect of interleukin-6 deficiency on implantation, fetal
development and parturition in mice. Proceedings of the Australian Society for Reproductive Biology 31, 97.
Roepman P, Wessels LF, Kettelarij N et al. 2005 An expression profile for diagnosis of lymph node metastases
from primary head and neck squamous cell carcinomas Nature Genetics 37, 182-6.
Romero R, Wu YK, Brody DT, et al. 1989 Oyarzun E, Duff GW, Durum SK. Human decidua: a source of
interleukin-1. Obstetrics and Gynecology 73, 31-4.
Romero R, Mazor M, Manogue K, et al. 1991 Oyarzun E, Cerami A. Human decidua: a source of cachectin-
tumor necrosis factor. European Journal of Obstetrics and Reproductive Biology. 41, 123-7.
Romundstad LB, Romundstad PR, Sunde A, et al. 2006 Increased risk of placenta previa in pregnancies
following IVF/ICSI; a comparison of ART and non-ART pregnancies in the same mother. Human
Reproduction del153.
Roudebush WE, Wininger JD, Jones AE, et al. 2002 Embryonic platelet-activating factor: an indicator of
embryo viability. Human Reproduction 17, 1306-10.
Rubin GL, Harrold AJ, Mills JA, et al. 1999 Regulation of sulphotransferase expression in the endometrium
during the menstrual cycle, by oral contraceptives and during early pregnancy. Molecular Human Reproduction
5, 995-1002
Rubinstein M, Marazzi A & Polak dF 1999 Low-dose aspirin treatment improves ovarian responsiveness,
uterine and ovarian blood flow velocity, implantation, and pregnancy rates in patients undergoing in vitro
fertilization: a prospective, randomized, double-blind placebo-controlled assay. Fertility and Sterility 71, 825-9.
Saadat P, Boostanfar R, Slater CC et al. 2004 Accelerated endometrial maturation in the luteal phase of cycles
utilizing controlled ovarian hyperstimulation: impact of gonadotropin-releasing hormone agonists versus
antagonists. Fertility and Sterility 82, 167-171
Salim R, Ben-Shlomo I, Colodner R, et al. 2002 Bacterial colonization of the uterine cervix and success rate in
assisted reproduction: results of a prospective survey. Human Reproduction. 17, 337-40.
Schatz F, Krikun G, Runic R, et al. 1999 Implications of decidualization-associated protease expression in

implantation and menstruation. Seminars in Reproductive Endocrinology 17, 3-12.
Schisterman EF, Vexler A, Whitcomb BW et al. 2006 The limitations due to exposure detection limits for
regression models. American Journal of Epidemiology 163, 374-83.
Schoolcraft W 2001 Textbook of assisted reproductive techniques; Laboratory and clinical perspectives.
Chapter: Embryo Transfer. Taylor and Francis.
Schroder K, Herrtzog PJ, Ravasi T et al. 2004 Interferon-{gamma}: an overview of signals, mechanisms and
functions. Journal of Leukocyte Biology 75, 163-89.
Seckl J. 2004 Prenatal glucocorticoids and long-term programming (Abstract). European Journal of
Endocrinology 151, U49-U62.
Seif M, Edi-Osagie E, Farquhar C, et al. 2005 Assisted hatching on assisted conception (IVF & ICSI).
Cochrane Database of Systematic Reviews.
Seli E, Sakkas D, Scott R, et al. 2007 Noninvasive metabolomic profiling of embryo culture media using
Raman and near-infrared spectroscopy correlates with reproductive potential of embryos in women undergoing
in vitro fertilization. Fertility and Sterility 88, 1350-7.
Sermon K, Moutou C, Harper J, et al. 2005 ESHRE PGD Consortium data collection IV: May-December 2001.
Shazand K, Baban S, Prive C, et al. 2004 FOXO1 and c-jun transcription factors mRNA are modulated in
endometriosis. Molecular Human Reproduction 10, 871-7
Sher G, Keskintepe L, Batzofin J, et al. 2005 Influence of early ICSI-derived embryo sHLA-G expression on
pregnancy and implantation rates: a prospective study. Human Reproduction 20, 1359-63.
Sherwin JR, Smith SK, Wilson A et al. 2002 Soluble gp130 is up-regulated in the implantation window and
shows altered secretion in patients with primary unexplained infertility. Journal of Clinical Endocrinology
Metabolism 87, 3953-60.
Shindo T, Kurihara H, Kuno K, et al. 2000 ADAMTS-1: a metalloproteinase-disintegrin essential for normal
growth, fertility, and organ morphology and function. J Clin Invest 105, 1345-52
Shohayeb A, Demerdash T. 2005 Effects of corticosteroids in hypothyroid patients with antithyroid antibodies
on IVF outcome (Abstract). Human Reproduction Abstractbook.
Sibug RM, Helmerhorst FM, Tijssen AMI, et al. 2002 Gonadotrophin stimulation reduces VEGF120 expression
in the mouse uterus during the peri-implantation period. Human Reproduction 17, 1643-1648
References | 181
Simon C, Frances A, Piquette GN et al. 1994 Embryonic implantation in mice is blocked by interleukin-1
receptor antagonist. Endocrinology 134, 521-8.
Simon C, Cano F, Valbuena D, et al. 1995 Clinical evidence for a detrimental effect on uterine receptivity of
high serum oestradiol concentrations in high and normal responder patients. Human Reproduction 10, 2432-7.
Simon C, Gimeno MJ, Mercader A et al. 1997 Embryonic regulation of integrins beta 3, alpha 4, and alpha 1 in
human endometrial epithelial cells in vitro. Journal of Clinical Endocrinology Metabolism 82, 2607-16.
Simon C, Landeras J, Zuzuarregui JL, et al. 1999 Early pregnancy losses in in vitro fertilization and oocyte
donation Fertility and Sterility 72, 1061-5.
Simon C, Oberyé J, Bellver J, et al. 2005 Similar endometrial development in oocyte donors treated with either
high- or standard-dose GnRH antagonist compared to treatment with a GnRH agonist or in natural cycles.
Smarr SC, Wing R, Hammond MG. 1988 Effect of therapy on infertile couples with antisperm antibodies.
American Journal of Obstetrics and Gynecology 158, 969-73.
Smith SK 2001 Angiogenesis and reproduction. British Journal of Obstetrics and Gynaecology 108, 777-83.
Smitz J, Bourgain C, van Waesberghe L, et al. 1993 A prospective randomized study on oestradiol valerate
supplementation in addition to intravaginal micronized progesterone in buserelin and HMG induced
superovulation. Human Reproduction 8, 40-5.
Spandorfer SD, Neuer A, Giraldo PC, et al. 2001 Relationship of abnormal vaginal flora, proinflammatory
cytokines and idiopathic infertility in women undergoing IVF. Journal of Reproductive Medicine 46, 806-10.
Staun-Ram E, Goldman S, Gabarin D, et al. 2004 Expression and importance of matrix metalloproteinase 2 and
9 (MMP-2 and -9) in human trophoblast invasion Reproductive Biology and Endocrinology 2, 59.
Steptoe PC and Edwards RG 1978 Birth after the reimplantation of a human embryo Lancet 2, 366.
Stern C, Chamley L, Hale L, et al. 1998 Antibodies to [beta]2 glycoprotein I are associated with in vitro
fertilization implantation failure as well as recurrent miscarriage: results of a prevalence study. Fertility and
Stern C, Chamley L, Norris H, et al. 2003 A randomized, double-blind, placebo-controlled trial of heparin and
aspirin for women with in vitro fertilization implantation failure and antiphospholipid or antinuclear antibodies.
Fertility and Sterility 80, 376-83.
Stewart CL, Kaspar P, Brunet LJ, et al. 1992 Blastocyst implantation depends on maternal expression of
leukaemia inhibitory factor. Nature 359, 76-9.
Stewart CL, Cullinan EB 1997 Preimplantation development of the mammalian embryo and its regulation by
growth factors. Developmental Genetics 21, 91-101.
Steyerberg EW, Eijkemans MJ, Harrell FE, Jr, et al. 2000 Prognostic modelling with logistic regression
analysis: a comparison of selection and estimation methods in small data sets Staistics in Medicine 19, 1059-79.
Strickler RC, Christianson C, Crane JP, et al. 1985 Ultrasound guidance for human embryo transfer. Fertility
Strott CA 1996 Steroid sulfotransferases. Endocr Rev 17, 670-97
Strowitzki T, Germeyer A, Popovici R, et al. 2006 The human endometrium as a fertility-determining factor.
Human Reproduction Update 12, 617-30.
Sturm-Ramirez K, Gaye-Diallo A, Eisen G, et al. 2000 High Levels of Tumor Necrosis Factor alfa and
Interleukin beta in Bacterial Vaginosis May Increase Susceptibility to Human Immunodeficiency Virus.
Journal of Infectious Diseases 182,467-73.
Sukhikh GT, Kasabulatov NM, Van'ko LV, et al. 2005 Ratio between the number of Th1 and Th2 lymphocytes
in the peripheral blood and concentration of proinflammatory cytokines in lochia of women with postpartum
endometritis. Advances in experimental medicine and biology 140, 672-4.
Szereday L, Varga P and Szekeres-Bartho J 1997 Cytokine production by lymphocytes in pregnancy. American
Talbi S, Hamilton AE, Vo KC, et al. 2006 Molecular Phenotyping of Human Endometrium Distinguishes
Menstrual Cycle Phases and Underlying Biological Processes in Normo-Ovulatory Women. Endocrinology
147, 1097-1121.
Tan SL, Balen A el HE, Campbell S, et al. 1992 The administration of glucocorticoids for the prevention of
ovarian hyperstimulation syndrome in in vitro fertilization: a prospective randomized study. Fertility and
Tan BK, Vandekerckhove P, Keay SD. 2004 Glucocorticoid supplementation during ovarian stimulation for
IVF or ICSI (Protocol). Cochrane Databse of Systematic Reviews.
Tang T, Glanville J, Orsi N, et al. 2006 The use of metformin for women with PCOS undergoing IVF
treatment. Human Reproduction 21, 1416-25.
Taniguchi F. 2005 Results of prednisolone given to improve the outcome of in vitro fertilization-embryo
transfer in women with antinuclear antibodies. Journal of Reproductive Medicine 50, 383-8.
Taylor PV, Campbell JM, Scott JS 1989 Presence of autoantibodies in women with unexplained infertility.
American Journal of Obstetrics and Gynecology 161, 377-9.
Takano M, Lu Z, Goto T, Fusi et al. 2007 Transcriptional Cross Talk between the Forkhead Transcription
Factor Forkhead Box O1A and the Progesterone Receptor Coordinates Cell Cycle Regulation and
Differentiation in Human Endometrial Stromal Cells. Mol Endocrinol 21, 2334-49
Tapia A, Gangi LM, Zegers-Hochschild F, et al. 2008 Differences in the endometrial transcript profile during
the receptive period between women who were refractory to implantation and those who achieved pregnancy.
Human Reproduction 23, 340-51
Templeton A, Morris JK 1998 Reducing the Risk of Multiple Births by Transfer of Two Embryos after in Vitro
Fertilization. N Engl J Med 339, 573-7
Tong MH, Jiang H, Liu P, et al. 2005 Spontaneous fetal loss caused by placental thrombosis in estrogen
sulfotransferase-deficient mice. Nat Med 11, 153-9
The European IVF-monitoring program 2005 Assisted reproductive technology in Europe, 2001. Results
generated from European registers by ESHRE. Human Reproduction 20, 1158-76.
Thomas K, Thomson AJ, Sephton V et al. 2002 The effect of gonadotrophic stimulation on integrin expression
in the endometrium. Human Reproduction 17, 63-68
Thomas K, Thomson A, Wood S, et al. 2003 Endometrial integrin expression in women undergoing in vitro
fertilization and the association with subsequent treatment outcome. Fertility and Sterility 80, 502-7.
Threadgill DW, Dlugosz AA, Hansen LA et al. 1995 Targeted disruption of mouse EGF receptor: effect of
genetic background on mutant phenotype. Science 269, 230-4.
References | 183
Toder V, Fein A, Carp H, et al. 2003 TNF-alpha in pregnancy loss and embryo maldevelopment: a mediator of
detrimental stimuli or a protector of the fetoplacental unit? Journal of Assisted Reproduction and Genetics 20,
73-81.
Tolstrup JS, Kjaer SK, Munk C, et al. 2003 Does caffeine and alcohol intake before pregnancy predict the
occurrence of spontaneous abortion? Human Reproduction 18, 2704-10.
Tranguch S, Daikoku T, Guo Y et al. 2005 Molecular complexity in establishing uterine receptivity and
implantation. Cellular and Molecular Life Sciences 62, 1964-73.
Tseng L and Mazella J 1999 Prolactin and its receptor in human endometrium. Seminars in Reproductive
Endocrinology 17, 23-7.
Tuckerman E, Laird SM, Prakash A, et al. 2007 Prognostic value of the measurement of uterine natural killer
cells in the endometrium of women with recurrent miscarriage. Human Reproduction 22, 2208-13.
Tulac S, Overgaard MT, Hamilton AE et al. 2006 Dickkopf-1, an inhibitor of Wnt signaling, is regulated by
progesterone in human endometrial stromal cells. Journal of Clinical Endocrinology Metabolism 91, 1453-61.
Twisk M, Mastenbroek S, Bossuyt PM, et al. 2005 The effectiveness of preimplantation genetic screening.
Ubaldi F, Bourgain C, Tournaye H, et al. 1997 Endometrial evaluation by aspiration biopsy on the day of
oocyte retrieval in the embryo transfer cycles in patients with serum progesterone rise during the follicular
phase. Fertility and Sterility 67, 521-6.
Ubaldi F, Rienzi L, Ferrero S, et al. 2002 Low dose prednisolone administration in routine ICSI patients does
not improve pregnancy and implantation rates. Human Reproduction 17, 1544-7.
Urman B, Mercan R, Alatas C, et al. 2000 Low-Dose Aspirin Does Not Increase Implantation Rates in Patients
Undergoing Intracytoplasmic Sperm Injection: A Prospective Randomized Study. Journal of Assisted
Reproduction and Genetics 17, 586-90.
van der Gaast MH, Beier-Hellwig K, Fauser BC, et al. 2003 Endometrial secretion aspiration prior to embryo
transfer does not reduce implantation rates. Reproductive Biomedicine Online 7, 105-9.
van der Gaast MH, Classen-Linke I, Krusche CA et al. 2008 Impact of ovarian stimulation on mid-luteal
endometrial tissue and secretion markers of receptivity. Reproductive Biomedicine Online 17, 553-563.
van 't Veer LJ, Dai H, van de Vijver MJ et al. 2002 Gene expression profiling predicts clinical outcome of
breast cancer Nature 415, 530-6.
van Mourik M, Heijnen CJ, Macklon NS 2008 Embryonic implantation, cytokines, adhesion molecules and
immune cells in establishing an implantation environment. Journal of Leukocyte Biology 85, 4-19.
Vazquez F, Hastings G, Ortega MA, et al. 1999 METH-1, a Human Ortholog of ADAMTS-1, and METH-2
Are Members of a New Family of Proteins with Angio-inhibitory Activity. Journal of Biology and Chemistry
274, 23349-57
Verberg MF, Macklon NS, Heijnen EM et al. 2007 ART: iatrogenic multiple pregnancy? Best Practice and
Research. Clinical Obsterics and Gynaecology 21, 129-43.
Verlinsky Y, Cohen J, Munne S, et al. 2004 Over a decade of experience with preimplantation genetic
diagnosis. Fertility and Sterility 82, 302-3.
von Wolff M, Thaler CJ, Strowitzki T, et al. 2000 Regulated expression of cytokines in human endometrium
throughout the menstrual cycle: dysregulation in habitual abortion Molecular Human Reproduction 6, 627-34.
Von Wolff M, Stieger S, Lumpp K et al. 2002 Endometrial interleukin-6 in vitro is not regulated directly by
female steroid hormones, but by pro-inflammatory cytokines and hypoxia. Molecular Human Reproduction 8,
1096-102.
Waldenstrom U, Hellberg D & Nilsson S 2004 Low-dose aspirin in a short regimen as standard treatment in in
vitro fertilization: a randomized, prospective study. Fertility and Sterility 81, 1560-4.
Wang H, Dey SK 2006 Roadmap to embryo implantation: clues from mouse models. Nature Reviews. Genetics.
7, 185-99.
Wang J, Mayernik L, Schultz JF et al. 2000 Acceleration of trophoblast differentiation by heparin-binding

EGF-like growth factor is dependent on the stage-specific activation of calcium influx by ErbB receptors in
developing mouse blastocysts. Development 127, 33-44.
Wang X, Hagberg H, Mallard C et al. 2006 Disruption of interleukin-18, but not interleukin-1, increases
vulnerability to preterm delivery and fetal mortality after intrauterine inflammation. American Journal of
Pathology 169, 967-76.
Wegmann TG, Lin H, Guilbert L, et al. 1993 Bidirectional cytokine interactions in the maternal-fetal
relationship: is successful pregnancy a TH2 phenomenon? ImmunologyToday 14, 353-6.
White CA, Johansson M, Roberts CT et al. 2004. Effect of Interleukin-10 Null Mutation on Maternal Immune
Response and Reproductive Outcome in Mice. Biology of Reproduction 70, 123-31.
Wilcox AJ, Weinberg CR, O'Connor JF, et al. 1988 Incidence of early loss of pregnancy New England Journal
of Medicine 319, 189-94.
Wilcox AJ, Baird DD, Weinberg CR 1999 Time of implantation of the conceptus and loss of pregnancy. New
England Journal of Medicine 340, 1796-9.
Wilson JD, Ralph SG, Rutherford AJ. 2002 Rates of bacterial vaginosis in women undergoing in vitro
fertilisation for different types of infertility. British Journal of Obstetrics and Gynaecology 109, 714-7.
Wright VC, Schieve LA, Reynolds MA, et al. 2005 Assisted reproductive technology surveillance--United
States, 2002. Morbidity and mortality weekly report. Surveillance summaries 54, 1-24.
Wu H, Kerr MK, Cui X, et al. 2003 MAANOVA: a software package for the analysis of spotted cDNA
microarray experiments. In: The analysis of gene expression data: methods and software.; 313-341
Wuu YD, Pampfer S, Becquet P et al. 1999 Tumor necrosis factor alpha decreases the viability of mouse
blastocysts in vitro and in vivo. Biology of Reproduction 60, 479-83.
Yang YH, Dudoit S, Luu P, et al. 2002 Normalization for cDNA microarray data: a robust composite method
addressing single and multiple slide systematic variation. Nucleic Acids Res 30:e15
Yie SM, Balakier H, Motamedi G, et al. 2005 Secretion of human leukocyte antigen-G by human embryos is
associated with a higher in vitro fertilization pregnancy rate. Fertility and Sterility 83, 30-6.
Yoo H J, Barlow DH, and Mardon HJ 1997 Temporal and spatial regulation of expression of heparin-binding
epidermal growth factor-like growth factor in the human endometrium: a possible role in blastocyst
implantation. Developmental Genetics 21, 102-8.
Younis JS, Ezra Y, Sherman Y, et al. 1994 The effect of estradiol depletion during the luteal phase on
endometrial development. Fertility and Sterility 62, 103-7.
Zhang X, Xu H, Lin J et al. 2005 Peritoneal fluid concentrations of interleukin-17 correlate with the severity of
endometriosis and infertility of this disorder. British Journal of Obstetrics and Gynaecology 112, 1153-5.
Nederlandse Samenvatting | 185
Nederlandse Samenvatting
____________________________________________
Hoofdstuk 1
In de introductie wordt de achtergrond en de reikwijdte van dit proefschrift
besproken. Ondanks de terugplaatsing van embryo’s met een goede morfologie is de
kans op zwangerschap 25-35 procent per embryo terugplaatsing. De kans op
innesteling is de ‘bottleneck’ in de slagingskans van de IVF behandeling. Het
endometrium is receptief voor de innesteling van een embryo gedurende de 19de tot
de 24ste cyclusdag in een cyclus van 28 dagen. Een cascade van lokaal werkende
chemokinen, cytokinen en groeifactoren is van belang bij de dialoog tussen het
endometrium en het embryo intra-uterien. Het werd voorheen aangenomen dat een
cytokine profiel in het endometrium dat het Th-2 type T cellen ondersteunt de kans
op conceptie vergroot. Echter, recentere literatuur heeft aangetoond dat dit zeer
waarschijnlijk een over-simplificatie van de werkelijkheid betreft. Een aantal markers
voor endometrium receptiviteit zijn voorgesteld, zoals integrinen, glycodelin, Hb-
EGF, CSF-1 en LIF. Echter geen individuele factor is specifiek en sensitief gebleken
in het voorspellen van een zwangerschap. Tegenwoordig is er een ontwikkeling
gaande richting technologieën die tientallen tot duizenden genen of gen producten
tegelijk kunnen kwantificeren.
Om een aantal redenen is er de rol van het endometrium binnen de humane embryo
implantatie moeilijk te onderzoeken. Morfologische criteria om het endometrium te
beschrijven worden veel gebruikt. Sinds de jaren ’50 zijn de Noyes criteria, welke het
endometrium volgens histologische criteria beoordelen om de fase gedurende de
menstruatiecyclus te bepalen, de gouden standaard. Echter, deze criteria zijn niet
geschikt gebleken om endometrium van fertiele en infertiele vrouwen te
onderscheiden. Daarnaast is het nut van het gebruik van de Noyes criteria in de
kliniek beperkt als gevolg van het subjectieve karakter en de hoge intra- en
interobserver variabiliteit. De cyclische aanwezigheid van de zogenaamde
‘pinopodes’, uitstulpingen in het oppervlakte epitheel van het endometrium, is ook
geen geschikte marker voor endometrium receptiviteit gebleken, omdat deze niet
goed gecorreleerd zijn met de zogenaamde ‘window of implantation’. Veel van onze
kennis op het gebied van implantatie is afkomstig van studies in dieren. Echter,
relatief weinig gen knock-out studies hebben een effect op de vruchtbaarheid laten
zien. Daarnaast kan informatie afkomstig van muizen studies niet met zekerheid
geëxtrapoleerd worden naar de mens.
De meeste technieken om het endometrium te onderzoeken vereisen een

endometrium biopt. Aangezien een biopt niet tijdens de ‘window of implantation’
afgenomen kan worden in conceptie cycli zonder de kans op zwangerschap hiermee
negatief te beïnvloeden, wordt gezocht naar non-invasieve onderzoekstechnieken.
Een voorbeeld hiervan is het spoelen van de uterus holte. Waarschijnlijk als gevolg
van een hoge mate van verdunning en variabiliteit in het verkregen volume, zijn de
resultaten van deze onderzoeken echter inconsistent. Perifere serum monsters bieden
ook geen uitkomst, omdat deze niet representatief zijn voor het intra-uteriene milieu.
Een andere mogelijkheid is het aspireren van endometrium secreet uit de baarmoeder
voorafgaande aan de embryo terugplaatsing binnen een IVF behandeling. Dit biedt
een nieuwe mogelijkheid om het intra-uteriene milieu waarin het embryo wordt
teruggeplaatst te analyseren en te correleren aan de kans op conceptie.
Hoofdstuk 2
In hoofdstuk 2.1 worden de huidige klinische therapieën om de kans op embryo
implantatie na een IVF behandeling te vergroten besproken. In hoofdstuk 2.2 wordt
dieper in gegaan op het gebruik van glucocorticoïden in de peri-implantatie fase en
wordt tevens een meta-analyse verricht van de beschikbare gerandomiseerde studies.
Hoofdstuk 2.1
De embryo transfer techniek is een kritieke handeling tijdens de IVF behandeling.
Studies hebben significante betere resultaten van de behandeling laten zien
afhankelijk van de methode van terugplaatsing, het type catheter, en de plaats van
terugplaatsing ten opzichte van de fundus. In toenemende mate worden verscheidene
farmaceutische therapieën toegepast om de kans op innesteling te vergroten. Echter,
het bewijs voor de effectiviteit en veiligheid van deze middelen voor dit doeleinde is
vaak beperkt. De beschikbare literatuur voor het gebruik van aspirine, nitraten,
aromatase remmers, vitamine C, progesteron, glucocorticoïden, insuline

sensitiserende medicatie, en GnRH agonisten wordt in dit hoofdstuk besproken.
Daarnaast wordt het effect van mildere ovariële simulatie schema’s op de kans van
innesteling, het endometrium en de embryo kwaliteit beschreven.
Hoofdstuk 2.2
Glucocorticoïden beïnvloeden mogelijk de endometrium receptiviteit door immuun
modulerende effecten, zoals een reductie van het aantal uteriene Natural Killer cellen,
normaliseren van het cytokine profiel in het endometrium en suppressie van
inflammatie. Een meta-analyse van 13 gerandomiseerde studies naar het empirisch
gebruik van glucocorticoïden tijdens de peri-implantatie-fase in IVF of ICSI
behandelingen liet geen significant verschil in de kans op zwangerschap zien (OR
1.16, CI 0.94 tot 1.44). Echter, een subgroep analyse van 650 vrouwen die IVF
ondergingen in plaats van ICSI (6 RCTs) liet wel een kleine maar significant hogere
kans op zwangerschap zien na het gebruik van glucocorticoïden (OR 1.50, CI 1.05 tot
2.13). Enkele studies in specifieke subgroepen, zoals vrouwen met autoantilichamen
of in combinatie met assisted hatching lieten ook een verbetering in
zwangerschapskans zien. Conflicterende resultaten werden verkregen in studies
welke het effect van adjuvante glucocorticoïden tijdens de stimulatiefase op de
ovariële respons onderzochten. Concluderend, het gebruik van adjuvante
glucocorticoïden in een IVF of ICSI behandeling in een ongeselecteerde
patiëntengroep lijkt niet zinvol. Echter, er is insufficiënt bewijs in bepaalde
patiëntengroepen, zoals vrouwen met autoantilichamen, endometriose, of in
combinatie met ‘assisted hatching’. Voor deze vrouwen of paren zouden
glucocorticoïden wel van nut kunnen zijn. Verder onderzoek hiernaar en naar de
veiligheid van glucocorticoïden tijdens de implantatiefase en vroege zwangerschap is
vereist. Het empirisch gebruik van glucocorticoïden is thans alleen verantwoord in de
context van gerandomiseerde gecontroleerde studies.
Hoofdstuk 3
In hoofdstuk 3 worden 4 studies besproken welke het cytokine profiel in
endometrium secreet hebben onderzocht, in relatie tot zwangerschap, implantatie,
ovariële stimulatie en de aanwezigheid van een bacteriële vaginose.
Hoofdstuk 3.1
Het doel van dit hoofdstuk was het ontwikkelen van een nieuwe techniek om het
intra-uteriene milieu te onderzoeken tijdens de implantatiefase en de veiligheid en
betrouwbaarheid hiervan vast te stellen. Endometrium secreet werd geaspireerd
voorafgaande aan de embryo terugplaatsing bij 210 vrouwen die een IVF of ICSI
behandeling ondergingen. Het cytokine profiel in het aspiraat werd onderzocht
middels een multiplex immunoassay. Tien van de zeventien factoren, IL-1ß, IL-6, IL-
12, IL-18, TNF-α, MIF, Eotaxin, MCP-1, IP-10, en VEGF, waren detecteerbaar in
90-100% van de monsters. Hb-EGF, IL-5, IL-17, IL-10, Dkk-1 en IL-15 waren
detecteerbaar in 23-76% van de monsters, terwijl de concentratie van IFN- γ beneden
de detectielimiet van de assay was in alle gevallen. Aspiratie van cervicaal slijm in
plaats van endometrium secreet werd voorkomen. Echter, om een mogelijke
contaminatie te onderzoeken, werd bij 22 vrouwen zowel endometrium secreet als
cervicaal slijm geaspireerd ter vergelijking van de cytokine profielen. Het cytokine
profiel in cervicaal slijm verschilde significant van endometrium secreet. De
veiligheid van endometrium secreet aspiratie voorafgaande aan de embryo
terugplaatsing werd onderzocht middels een ‘matched-control’ studie naar de kans op
zwangerschap in het cohort van 210 participanten en 210 controles, ‘gematched’
voor belangrijke prognostische factoren. De controle patiënten ondergingen hun
behandeling gedurende dezelfde tijdsperiode als de studie participanten. Er werden
geen significante verschillen in de kans op zwangerschap gevonden. Concluderend,
het bepalen van een cytokine profiel in het endometrium secreet biedt een objectieve,
niet invasieve veilige manier om de endometrium factor in humane embryo
implantatie te onderzoeken.
Hoofdstuk 3.2
In dit hoofdstuk worden de resultaten beschreven van een onderzoek naar de relatie
tussen het cytokine profiel in het endometrium secreet en de kans op een klinisch
herkenbare zwangerschap of embryo implantatie. Endometrium secreet werd
geaspireerd voorafgaande aan de embryo terugplaatsing bij 210 vrouwen. Om naast
zwangerschap ook de embryo implantatie te kunnen onderzoeken, hebben vrouwen
gedurende 10 dagen dagelijks een urine monster gespaard en ingevroren voor latere
hCG analyse. Multivariate analyse liet een significante associatie zien tussen de
concentratie van MCP-1 (negatief) en IP-10 (positief) en de kans op een embryo

innesteling. Daarnaast werd een significante associatie gezien tussen de concentratie
TNF-α (positief) en IL-1β (negatief) en de kans op een klinisch herkenbare
zwangerschap. De voorspellende waarde van TNF-α en IL-1β voor het ontstaan van
een zwangerschap was equivalent en additief aan de voorspellende waarde voor het
optreden van een zwangerschap aan de hand van de embryo kwaliteit. Concluderend,
endometrium secreet analyse biedt perspectief in het onderzoek naar de endometrium
receptiviteit.
Hoofdstuk 3.3
Suprafysiologische sex steroïden concentraties als gevolg van ovariële stimulatie in
het kader van IVF behandelingen worden beschouwd de endometrium receptiviteit
negatief te beïnvloeden. In dit hoofdstuk wordt de invloed van ovariële stimulatie op
het intra-uterien cytokine profiel onderzocht. 203 patiënten ondergingen ovariële
stimulatie middels een GnRH antagonist of agonist protocol in het kader van een IVF
of ICSI behandeling en ondergingen endometrium secreet aspiratie voorafgaande aan
de embryo terugplaatsing. Een subgroep van 41 vrouwen participeerde zowel in de
gestimuleerde cyclus op de dag van de embryo terugplaatsing, als in een normale
ovulatoire cyclus zes dagen na de spontane LH piek. Een Bonferroni correctie voor
‘multiple testing’ werd toegepast. Significant hogere concentraties van IL-1ß, IL-5,
IL-10, IL-12, IL-17, TNF-α, Hb-EGF, Eotaxin, en Dkk-1 werden gevonden in
endometrium secreet verkregen in de gestimuleerde cyclus vergeleken met de
normale cyclus. Multivariate analyse in 203 patiënten liet een significante associatie
zien tussen het intra-uterien cytokine profiel en het aantal verkregen oöcyten en het
type GnRH analoog. Concluderend, het in-vivo milieu in welke het embryo wordt
teruggeplaatst is significant anders in een gestimuleerde cyclus in vergelijking tot een
normale spontane cyclus. Dit effect lijkt toe te nemen wanneer een groter aantal
oöcyten verkregen wordt.
Hoofdstuk 3.4
In dit hoofdstuk wordt gekeken naar de associatie tussen de aanwezigheid van een
bacteriële vaginose (BV) en het intra-uterien cytokine profiel. De hypothese is dat
bacteriële vaginose gepaard gaat met een meer pro-inflammatoir cytokine profiel.
Voorafgaande aan de embryo terugplaatsing werd bij 198 vrouwen endometrium

secreet geaspireerd en een cervicale smear werd afgenomen voor Gram kleuring en
screening voor BV middels de Nugent criteria. Bij 8.6% van de vrouwen bleek er
sprake van een BV. Multivariate analyse liet een significant positieve associatie zien
tussen IL-1β en de Nugent score, en significant negatieve associatie tussen IL-10 en
Eotaxin, en de Nugent score. Concluderend, een hogere Nugent score is geassocieerd
met een meer pro-inflammatoir cytokine profiel in het endometrium secreet. Echter,
het effect op de endometrium receptiviteit is niet bekend.
Hoofdstuk 4
Herhaaldelijk implantatie falen wordt gedefinieerd als drie of meer onsuccesvolle
IVF behandelingen met een embryo terugplaatsing of het falen van conceptie na de
terugplaatsing van 10 embryo’s of meer. Een verminderde endometrium receptiviteit
zou de oorzaak van implantatie falen kunnen zijn in een subgroep van vrouwen met
dit probleem. Er is veel onderzoek gedaan naar de rol van cytokinen, chemokinen en
groeifactoren en hun rol bij de humane embryo implantatie. Echter, er is geen
individuele factor geïdentificeerd welke cruciaal lijkt te zijn bij de humane embryo
innesteling. DNA microarray gen expressie onderzoek maakt het mogelijk het
endometrium genoom breed te onderzoeken. Specifieke veranderingen in
endometrium gen expressie profielen werden geïdentificeerd gedurende de
menstruatiecyclus en ook tijdens niet-fysiologische of pathologische condities zoals
ovarieel gestimuleerde cycli, of bij de aanwezigheid van een spiraal of endometriose.
In dit hoofdstuk wordt een studie beschreven met als doel het identificeren van
verschillen in het endometrium transcriptoom van 24 vrouwen, 38 jaar of jonger, met
idiopatisch herhaaldelijk implantatie falen na IVF versus een controle groep. De
controle groep bestaat uit vrouwen die bevallen zijn van een gezond kind na hun
eerste of tweede ICSI poging, met enkel een ernstige andrologische factor als
verklaring voor hun infertiliteit. Endometrium biopsieën werden 7 dagen na de LH
piek afgenomen in een natuurlijke cyclus en RNA werd geanalyseerd middels DNA
microarray analyse. 75 genen kwamen significant anders tot expressie in beide studie
groepen. qRT-PCR voor 4 genen liet een goede correlatie zien in microarray en PCR
resultaten. Een aantal interessante genen met een rol in endometrium decidualisatie,
embryo adhesie en angiogenese werd geïdentificeerd, waaronder FOXO-1, MUC16
en ADAMTS8. Hierna zal verder onderzoek gedaan worden op eiwit niveau.

Gesuperviseerde hiërarchische clustering werd toegepast op een deel van de monsters
om een set genen te identificeren welke de monsters van patiënten en controles zou
kunnen onderscheiden. De voorspellende waarde van deze set van genen werd
vervolgens getoetst op de overige (en dus onafhankelijke) monsters. Deze methode
bleek helaas niet goed in staat het endometrium als normaal of verminderd receptief
correct te classificeren. De resultaten suggereren dat in een groep vrouwen met
idiopatisch herhaaldelijk falen van de embryo innesteling de oorzaken heterogeen
zijn.
Hoofdstuk 5
In dit hoofdstuk worden de conclusies die uit de studies in dit proefschrift getrokken
konden worden bediscussieerd. Dit proefschrift laat zien dat endometrium secreet
analyse een veilige manier biedt om de endometrium factor in humane embryo te
onderzoeken. De resultaten beschreven in hoofdstuk 3.2, met een positieve associatie
tussen de concentratie van TNF-α en een negatieve associatie met IL-β en
zwangerschap, benadrukken het vermoeden dat het Th1/Th2 paradigma obsoleet is.
De voorspellende waarde van het intra-uterien cytokine profiel voor het ontstaan van
een zwangerschap was even groot als de voorspellende waarde van de embryo
kwaliteit. Hieruit concluderen wij dat het endometrium een belangrijke rol heeft in de
embryo innesteling en niet alleen een faciliterende functie. Echter, bij vrouwen met
herhaaldelijk falen van de embryo innesteling is er geen sprake van een duidelijk
afwijkend endometrium gen expressie profiel en een verminderde endometrium
receptiviteit is waarschijnlijk slechts bij een kleine groep aanwezig. De endometrium-
embryo dialoog is mogelijk doorslaggevend in het slagen van de implantatie. In-vitro
experimenten bieden de mogelijkheid de vroege fasen van de humane embryo
innesteling onder gecontroleerde omstandigheden te onderzoeken. Momenteel lopen
er studies binnen onze onderzoeksgroep op het gebied van embryo-endometrium
interactie, zoals het effect van oplosbare factoren uitgescheiden door het embryo op
het endometrium transcriptoom.
List of publications | 193
List of publications
____________________________________________
Boomsma CM, Kavelaars A, Eijkemans MJC, Fauser BCJM, Heijnen CJ, Macklon NS. Ovarian stimulation for
in-vitro fertilization alters the intra-uterine cytokine, chemokine and growth factor milieu encountered
by the embryo. Submitted.
Boomsma CM, Kavelaars A, Eijkemans MJC, Fauser BCJM, Heijnen CJ, Macklon NS. Is bacterial vaginosis
associated with a pro-inflammatory cytokine profile in endometrial secretions of women undergoing
IVF? Submitted.
Boomsma CM, Groot-Koerkamp MJA, Brabers NHCH, Leenen van D, Eijkemans MJC, Hooff van SR, Fauser
BCJM, Kavelaars A, Heijnen CJ, Holstege FCP, Macklon NS. Determination of the peri-implantation
endometrial transcript profile in women with recurrent implantation failure versus controls. Submitted.
Boomsma CM, Kavelaars A, Eijkemans MJC, Lentjes EG, Fauser BCJM, Heijnen CJ, Macklon NS. Analysis
of the endometrial secretion cytokine profile to predict pregnancy in in-vitro fertilization. Human
Reproduction 2009; 1:1-9.
Boomsma CM, Kavelaars A, Eijkemans MJC, Teklenburg G, Amarouchi K, Fauser BCJM, Heijnen CJ,
Macklon NS. Cytokine profiling in endometrial secretions: a non-invasive window on endometrial
receptivity. Reprod Biomed Online 2009; 18:85-94.
Macklon NS, Boomsma CM. ‘Omics’ and the molecular fingerprint of a receptive endometrium. Focus on
Reproduction. January 2009.
Boomsma CM, Macklon NS. Does glucocorticoid therapy in the peri-implantation period have an impact on
IVF outcomes? Curr Opin Obstet Gynecol 2008;20:249-56.
Boomsma CM, Fauser BCJM, Macklon NS. Pregnancy complications in women with polycystic ovary
syndrome. Semin Reprod Med 2008;26:72-84.
Boomsma CM, Heineman MJ, Cohlen BJ, Farquahar C. Semen preparation techniques for intrauterine
insemination. Cochrane Database Syst Rev 2007 (Updated review 2004); Issue 4.
Boomsma CM, Keay SD, Macklon NS. Peri-implantation glucocorticoid administration for assisted
reproductive technology cycles. Cochrane Database Syst Rev 2007; Issue 1.
Boomsma CM, Macklon NS. What can the clinician do to improve implantation? Reprod Biomed Online
2006;13:845-55.
Boomsma CM, Eijkemans MJ, Hughes EG, Visser GH, Fauser BC, Macklon NS. A meta-analysis of pregnancy
outcomes in women with polycystic ovarian syndrome versus controls. Hum Reprod Update
2006;12:673-83.
Leij van der FR, Huijkman NC, Boomsma C, Kuipers JR, Bartelds B. Genomics of the Human Carnitine
Acyltransferase Genes, Mol Genet Metab 2000;71:139-53.
Boomsma CM, Macklon NS. Chapter 38: medical strategies to improve AT outcome: current evidence. In:
Textbook of Infertility and Assisted Reproduction. Edited by Drs. Rizk, JA. Garcia-Velasco, H Sallam
and A Makrigiannakis. Cambrigde University Press.
List of auhors and their affiliations | 195
List of authors and their affiliations

____________________________________________
Amarouchi K Laboratory of Psychoneuroimmunology
University Medical Center Utrecht, The Netherlands
Bozkurt N Departments of Obstetrics and Gynecology
Gazi University Ankara, Turkey
Brabers NHCH Department for Physiological Chemistry
Eijkemans MJC Department of Public Health
Erasmus Medical Center Rotterdam, The Netherlands
Fauser BCJM Department of Reproductive Medicine and Gynaecology
Groot Koerkamp MJA Department for Physiological Chemistry
Heijnen CJ Laboratory of Psychoneuroimmunology
Hooff van SR Department for Physiological Chemistry
Kavelaars A Laboratory of Psychoneuroimmunology
Leenen van D Department for Physiological Chemistry
Lentjes EG Laboratory of Endocrinology
Macklon NS Department of Reproductive Medicine and Gynaecology
Teklenburg G Department of Reproductive Medicine and Gynaecology
Dankwoord | 197
Dankwoord
____________________________________________
Het boekje is klaar! Heel veel mensen zijn van belang geweest voor het tot stand komen van
dit proefschrift. Ik wil iedereen bedanken die hieraan wetenschappelijk of anderzins heeft
bijgedragen. Ik wil dit boekje echter niet sluiten zonder een aantal van hen persoonlijk te
bedanken.
Professor Macklon, beste Nick. Hij is af! Dankzij jou motiverende woorden (No really, it’ll
be the talk of the town!), en een deur die eigenlijk altijd open stond voor vragen, is het
gelukt. Je hebt me de kans gegeven 3 jaar lang aan iets moois te werken en alle
mogelijkheden hierbij gegeven. Ik ben je dankbaar voor het in mij gestelde vertrouwen. Ik
hoop in de toekomst nog vaak met je te mogen samenwerken!
Professor Heijnen, beste Cobi. Voor de één een première bij de immunologie en voor de
ander bij de fertiliteit. Andere aandachtsgebieden en een andere manier van werken. Naar
mijn mening heeft de middenweg hierin tot het beste resultaat geleid! Ik heb bewondering
voor je kennis en inzicht en ben blij dat ik hier veel van heb kunnen leren. Bedankt daarvoor!
Beste Annemieke. Al vond ik versie 1 meestal klaar voor submissie, ik kan niet anders dan
toegeven dat een tiental versies later het manuscript toch echt naar een hoger niveau getild
was. Ik heb veel geleerd van je analytische benadering. Je hulp bij de experimenten de
afgelopen jaren is voor mij van grote waarde geweest. Bedankt!
Professor Fauser, beste Bart. Jij zorgt ervoor dat wij ons als onderzoekers op de afdeling
onderdeel voelen van een groter geheel. Bewonderenswaardig was het dat je vaak bij overleg
binnen korte tijd de vinger op de zere plek wist te leggen. Soms mindere motiverende
woorden (Je zal niet de eerste zijn wiens carrière op het endometrium is stuk gelopen…),
maar toch vooral een bevlogen enthousiasme voor onderzoek welke aanstekelijk werkt!
Beste René, jouw statistische kennis is onmisbaar geweest. Ik sluit me graag aan bij een
eerdere stelling: “Bij iedere zichzelf respecterende onderzoeksgroep zou een statisticus
betrokken moeten zijn”. De data lieten ons nog wel eens in de steek, bedankt dat je zo
ontzettend je best hebt gedaan om de juiste conclusies te trekken. Daarnaast was het erg
gezellig dat je er op vrijdag altijd bij was en ook de vrijdagmiddagborrel vaak niet afsloeg!
Professor FCP Holstege, Professor TAE Stout, Professor MJ Heineman en Professor FM

Helmerhorst wil ik bedanken voor het plaatsnemen in de beoordelingscommissie. Professor
Simon, I am honoured that you are a member of the examination panel and one of my
opponents.
Beste Professor Heineman en dr. Cohlen, jullie stonden aan het begin van de
wetenschappelijke ‘carrière’. Dankzij jullie zijn er deuren voor me geopend, bedankt
daarvoor!
Beste Professor Holstege, Sander, Patrick, Dik, Nathalie en vooral natuurlijk Marian van het
Department for Physiological Chemistry. Bedankt voor jullie hulp en vooral éindeloze
geduld, er is een wereld voor me open gegaan!
Beste Karima, Jitske, en Hanneke van de PNI, bedankt voor de ontzettende nauwgezette en
altijd snelle hulp bij de experimenten. Jullie aandeel in dit boekje is zeer groot geweest!
Sjerp, Peter, Dagmar, en Esther, de embryologen van het IVF-laboratorium, dank voor de
antwoorden op vele vragen en de hulp bij het in de praktijk brengen van de studies. Caroline,
Els, Joke, Jolanda, Linde, Marleen, Renata, Erna en vooral ook Truus, Michael en Margot
van het IVF lab, dank allemaal voor ál het extra werk dat jullie in het kader van de
onderzoeken hebben verricht en de hulp als ik afwezig was! Lieneke, Nijske, Anna,
Frederieka, Yvonne, Jeroen, Monique en Marieke, IVF artsen, ik wil jullie vooral bedanken
voor de gezelligheid! Daarnaast hebben jullie je altijd erg flexibel getoond bij het starten van
nieuwe studies en kon ik altijd op jullie hulp rekenen. Bedankt! Marieke, bedankt voor je
“gezonde controles”. Yvonne, fijn dat jij mijn opvolger bent en we hopelijk nog veel samen
gaan werken.
Beste Inge, Eef, en Ad van de endocrinologie. Ik heb jullie waarschijnlijk tot het uiterste
gedreven met ál die buisjes… Het einde is in zicht. Ze hebben het boekje helaas niet meer
Dankwoord | 199
gehaald, maar er moet nog wat te onderzoeken overblijven! Bedankt voor de prettige
samenwerking.
Teun en Tjerk, een begrip. Bedankt voor jullie hulp bij alle stress-veroorzakende ICT
problemen.
Alle paren die meegedaan hebben aan een van de onderzoeken wil ik graag bedanken voor
hun inzet en tijd. Ik hoop dat de resultaten zoals beschreven in dit proefschrift van betekenis
zullen zijn voor de IVF behandeling in de toekomst.
Tessa en Ingrid, bedankt voor de gezelligheid en jullie hulp en tips en trucs! Van fax-
apparaten tot trouwjurken.
Henk, Eleonora, Piet, Marjan, Jan, en Ineke. Bedankt voor jullie interesse in de voortgang
van mijn onderzoek de afgelopen jaren. Frank, bedankt voor je aanstekelijke enthousiasme.
Angelique, bedankt voor de gezelligheid en leerzame momenten. Ik hoop nog veel van jullie
te leren.
Alle betrokkenen van de afdeling voortplantingsgeneeskunde van het Meander Medisch

Centrum, het Diakonessenziekenhuis Utrecht en het Rijnstate ziekenhuis in Arnhem wil ik
graag bedanken voor jullie inzet bij het includeren van patiënten in de verschillende cohort
studies.
Collega-uitvinders van het AZU en de overkant, een lach…een traan…. Naast onderzoek
veel geleerd over het maken van cappuccino’s, Sittard’s dialect, body-bags in the US of A,
en broodjes van de week. Fijn vooruitzicht dat we komende jaren collega’s zullen zijn.
Kamergenoten Wenche en Gijs, succes met de laatste loodjes. Wenche, ik heb jou meer
gezien dan Pieter afgelopen jaren, maar mis het zeker! Gijs, erg gelachen maar ook leuk
gediscussieerd over “the embryo” of “the endometrium” as a primary regulator
of….Dynamite! Hopelijk blijft DE LIJST voorlopig leeg en klinkt vooral de submissie-bel.
Marian Kosterman, moeder van de onderzoekers, bedankt voor de gezelligheid!
Ook wil ik graag mijn nieuwe collega’s en de gynaecologen in het Diakonessenhuis

bedanken voor het begrip tijdens de laatste maanden onderzoek waarin het boekje nog
gemaakt moest worden. Dr. van Haaften, Maarten, bedankt voor het vertrouwen drie jaar
geleden en de hulp deze onderzoeksplaats te bemachtigen!
Lieve vrienden. Zonder ski-hut feestjes, weekendjes Istanbul (oh nee Brussel), Barcelona,
pre- en after parties, carnavál, kerstdiners, hockey-toernooien of het park-café, was het
boekje misschien wel veel eerder af geweest. Toch prijs ik mij gelukkig!
Gijs en Anandi, ontzettend leuk dat jullie mij 14 april bijstaan als paranimf!
Familie Smits, ik bof met zo’n lieve schoonfamilie. Het zou nog wel eens druk kunnen
worden in Oswald, want de verhalen over de skivakanties zijn inmiddels ook onder mijn
collega’s beroemd (of berucht) geworden. Gebroeders Smits, buurmannen, dat er nog maar
veel gezellige avonden mogen volgen! Lieve Laura, ben erg blij met de ‘versterking’.
Lieve papa, mama, Nicolien en Jeroen, en Emily en Maarten, en kleine Fleur. Wat fijn om uit
zo’n warm nest te komen! Papa, ik vind het erg leuk en bijzonder dat ik 14 april sta waar jij
27 jaar geleden al stond. Papa en mama, bedankt voor alles.
Laatst en liefst, Pieter! Samen zijn met jou maakt al het andere heerlijk onbelangrijk!
Curriculum Vitae | 201
Curriculum Vitae
____________________________________________
Carolien Boomsma werd op 13 februari 1979 geboren te Amersfoort als jongste van 3
kinderen. Al gauw verhuisde de familie Boomsma naar Groningen. Na het behalen van haar
middelbare school diploma aan het Maartens college ging zij in 1997 geneeskunde studeren
aan de Universiteit Groningen. Tijdens deze periode ontmoet Carolien Pieter. Het eerste half
jaar van haar co-schappen in 2001 heeft zij gelopen in het Sint Elisabeth Hospitaal te
Curaçao. Na het afronden van haar co-schappen in Groningen weet Carolien Pieter te
overtuigen van het nut van een literatuurstudie aan de andere kant van de aardbol. Zij loopt
haar wetenschappelijke stage bij de ‘Cochrane Menstrual Disorders and Subfertility Group’
in Auckland, Nieuw-Zeeland, onder begeleiding van Prof. MJ Heineman, Prof. C Farquhar
en dr BJ Cohlen. Na het behalen van haar artsenbul in 2004 heeft zij ervaring opgedaan met
verloskunde in een district hospitaal in Lukulu, Zambia. Hierna was de keuze voor de
gynaecologie én een ontwikkeld land definitief. Eenmaal terug op Nederlandse bodem begint
zij aan haar eerste baan als IVF-arts/ ANIOS Gynaecologie en Obstetrie in het
Diakonessenziekenhuis te Utrecht (opleider dr M van Haaften). Na een jaar start zij in
augustus 2005 met haar promotie-onderzoek onder begeleiding van Prof. NS Macklon en
Prof. CJ Heijnen. Het project kwam tot stand door samenwerking tussen de afdeling
Voortplantingsgeneeskunde en het research laboratorium van de PsychoNeuroImmunologie.
Het nodige doorzettingsvermogen wordt gevraagd, maar het eindresultaat ligt nu voor u. Zij
begon de klinische opleiding tot gynaecoloog in oktober 2008 in het Diakonessenziekenhuis
te Utrecht (opleider dr PC Scholten). Carolien trouwde op 5 september 2008 met Pieter
Smits.
List of abbreviations | 203
List of abbreviations
____________________________________________
AA Ascorbic Acid
ANCOVA analysis of covariance
AOA Anti Ovarian Antibodies
ART Assisted Reproductive Techniques
APS Antiphospholipid Antibody Syndrome
APC activated protein C
ASA Anti Sperm Antibodies
AUC Area under the curve
BMI Body Mass Index
BSA Bovine Serum Albumin
BV Bacterial Vaginosis
CBAVD Congenital bilateral absence of the vas deferens
CI Confidence Interval
COH Controlled Ovarian Hyperstimulation
COX2 Cyclo-oxygenase-2
DHEAS dehydroepiandrosterone
Dkk-1 Dickkopf homolog 1
ELISA Enzyme Linked Immunosorbent Assay
EST expressed sequence tag
FOXO1 Forkhead Box O 1
FUT4 Fucosyltransferase 4
GnRH Gonadotropin Releasing Hormone
GM-CSF Granulocyte macrophage colony stimulating factor
GO Gene Ontology
Hb-GF Heparin-binding Epidermal Growth Factor
hCG human Chorionic Gonadotropin
HLA Human Leukocyte Antigens
hMG human Menopausal Gonadotropin
ICSI Intra Cytoplasmic Sperm Injection
IFN- γ Interferon-γ
IGFBP-1 Insulin-like Growth Factor Binding Protein-1
IL Interleukin
IP-10 Interferon-γ inducible 10 kD protein
IUI Intrauterine Insemination
IUD Intrauterine Device
IVF In-Vitro Fertilization
KO Knock-out
k-nn k-nearest neighbour
LIF Leukemia Inhibitory Factor
LH Luteinizing Hormone
MCP-1 Monocyte Chemotactic Protein-1
MIAME Minimum information about a microarray experiment
MIF Macrophage migration Inhibitory Factor
MMP Matrix Metalloproteinases
MUC1 Mucin1
NO Nitric Oxide
OHSS Ovarian Hyperstimulation Syndrome
OR Odds Ratio
LPS Lipopolysaccharides
LIF Leukemia Inhibitory Factor
MHC Major Histocompatibility Complex
MMP Matrix Metalloproteinase
p probability
PAF Platelet Activating Factor
PCOS Polycystic Ovary Syndrome
PBS Phosphate-Buffered Saline
PBMC Peripheral Blood Mononulear Cells
PGS Preimplantation Genetic Screening
PT-INR prothrombin time
rFSH recombinant Follicle Stimulating Hormone
RCT Randomized Controlled Study
RIF Recurrent Implantation Failure
RIN RNA integrity numbers
ROC Receiver Operating Characteristics
RT-PCR reverse transcriptase polymerase chain reaction
SD standard deviation
SET Single Embryo Transfer
SNR signal-to-noise-ratio
TGF-β Tumor Growth Factor β
Th1/2 T helper cell type 1/ 2
TNF-α Tumor Necrosis Factor-α
TSH thyroid-stimulating hormone
uNK uterine Natural Killer
VEGF Vascular Endothelial Growth Factor

Booms Ma

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

Booms Ma

Uploaded by

Copyright:

Available Formats

The endometrial factor in human

Thesis, Utrecht University, with a summary in Dutch.

© 2009 Carolien Boomsma

De endometrium factor in humane embryo

ter verkrijging van de graad van doctor aan de Universiteit

Chapter 2 Current clinical therapies to improve endometrial

Chapter 3 Studying the endometrial factor: cytokine profiling in

Chapter 4 The endometrial factor in recurrent implantation failure

Figure 1: An overview of the outcome of human conceptions

Adapted from Macklon et al., 2002.

Human embryo implantation

Figure 2. Different stages of the embryo implantation

Adapted from Chaouat et al 2003.

The invasion of cytotrophoblasts requires enzymatic digestion of the extracellular

the embryo on morphological criteria (fragmentation, number of blastomeres,

Non-invasive methods to assess embryo quality in order to improve embryo selection

profiles (Dominguez et al. 2008). However, identification of mediators secreted by

Uterine receptivity is defined as a restricted period when the uterus supports

Techniques to study endometrial receptivity

Morphological changes in the endometrium

The use of animal models to study human embryo implantation

Addressing molecular redundancy

Figure 3: Implantation strategies among species

Another possible approach is the analysis of endometrial secretions aspirated from

Aims and outline of this thesis

described and the available evidence is reviewed in order to clarify the

In order to improve implantation in IVF, the ability to investigate endometrial

Microarray analysis enables an approach from a global genomic perspective.

What can the clinician do to improve implantation?

(Reproductive BioMedicine Online 2006; 13: 845-55)

Embryo transfer technique

Adjuvant pharmaceutical therapies

RCTs investigating the use of aspirin as an empirical therapy in non-selected IVF

Study n Dose Timing Pregnancy rate p-

Nitric oxide donors

The optimal duration of progesterone administration remains to be clarified. Many

Although providing an appealing rationale for extended progesterone

Although supplementation of progesterone is widely used to improve implantation

Levels expressed as mean ± standard error.

There is therefore evidence to support a possible role for glucocorticoids in

Insulin sensitizing drugs

Ovarian stimulation regimens

Ovarian hyperstimulation and the resultant supra-physiological estradiol levels have

maturation and defective induction of progesterone receptors (Devroey et al. 2004).

Increasing awareness of the possible detrimental effects of standard ovarian

The optimal number of growing follicles or oocytes retrieved in order to maximize

Selecting the optimal embryo, endometrium and patient

maturation (Mikkelsen et al. 2005) and the use of embryo-endometrial co-culture

Parallel to these techniques, preconceptional assessment of endometrial receptivity

embryo transfer itself has been shown to be critical in assisted reproductive

Increasing awareness of the possible detrimental effects of ovarian hyperstimulation

and nutritional factors in order to optimise reproductive health could contribute to

Embryo implantation failure can be caused by multiple factors; as a result, no single

Table 2: Summary of possible methods for a clinician to improve implantation rates

Peri-implantation glucocorticoid administration for

(Curr Opin Obstet Gynecol. 2008; 20: 249-56)

Glucocorticoids in the general IVF population

Recently, a meta-analysis has been performed of randomized controlled trials (RCTs)

Study Glucocorticoids Control OR (fixed) Weight OR (fixed)

01 Pregnancy rate after IVF

02 Pregnancy rate after ICSI

0.1 0.2 0.5 1 2 5 10

Overall there is no clear evidence to support empirical administration of

Glucocorticoids for couples with autoantibodies