The Safety of COX-2

Inhibitors
John J. Cush, MD
Presbyterian Hospital of Dallas
 GI Outcomes Trials: Design
VIGOR (n=8076) CLASS (n=7982)
Drug Rofecoxib 50 mg QD Celecoxib 400 mg BID
(2x max chronic dose) (2x max chronic dose)

Patients RA OA (72 %), RA (28 %)

Comparator Naproxen 500 mg BID Ibuprofen 800 mg TID

Diclofenac 75 mg BID

Low dose ASA No Yes (21 %)

Duration Median 9 months Median 9 months

Maximum 13 months Maximum 13 months
6 months reported

Bombardier et al. N Engl J Med. 2000;343:1520-1528 Silverstein et al. JAMA. 2000; 284:1247-
 NSAID Impact/Cost
 13 million chronic users (70 million Rx/yr)
 Dyspepsia 20% > Gastric ulcers 10%
 Serious GI complications
 1.3 % RA pts
 0.73% OA pts
 Hospitalization (UGI bleed) 103,000/year
 5-10% mortality (est 16,500 deaths/year)
 NSAID deaths 15th most common cause in USA
 Cost > $2 billion/year
 NSAID Concerns
 13 million chronic Mortality for 7 Selected
users (U.S.) Disorders (1997)
 Dyspepsia in 10-20%
 Serious GI
complications in 1.3%
 Fatalities in 5-10% of
these
 Direct cost/year > $2
billion
 ~ 15th most common Wolfe, et al: NEJM 1999
cause of death (U.S.)
 Mechanism of Action of NSAIDs
CO2H

Arachidonic acid
COX-1 Non-specific COX-2
“Constitutive” NSAIDs “Inducible”
COX-2 NSAIDs

GI Mucosa Platelet
Prostaglandins
Prostaglandins Thromboxane

Mediate pain,
GI mucosal
Hemostasis inflammation, and fever
Protection
Bakhle et al. Med Inflamm. 1996;5:305-323.
Vane et al. Inflamm Res. 1995;44:1-10.
 GI Outcomes Trials: Design
VIGOR (n=8076) CLASS (n=7982)
Drug Rofecoxib 50 mg QD Celecoxib 400 mg BID
(2x max chronic dose) (2x max chronic dose)

Patients RA OA (72 %), RA (28 %)

Comparator Naproxen 500 mg BID Ibuprofen 800 mg TID

Diclofenac 75 mg BID

Low dose ASA No Yes (21 %)

Duration Median 9 months Median 9 months

Maximum 13 months Maximum 13 months
6 months reported

Bombardier et al. N Engl J Med. 2000;343:1520-1528 Silverstein et al. JAMA. 2000; 284:1247-
 CLASS Trial: Upper GI Complications
Alone and With Symptomatic Ulcers
= celecoxib
6

= NSAIDs (ibuprofen + 5
p = 0.02
diclofenac) 4
p = 0.09
49 / 1384
All Patients 3

20 / 1384
30 / 1441

Annualized Incidence %
2

1
11 / 1441
0
6

5 p = 0.02
4

Patients Not Taking Aspirin 3

p = 0.04 32 / 1101
2
14 / 1101 16 / 1143
1
5 / 1143
0

p = 0.49
6 17 / 283
5 14/ 298
4 p = 0.92
Patients Taking Aspirin 3
6 / 283
2
6 / 298
1

0
Ulcer ComplicationsSymptomatic Ulcers and
Silverstein et al. JAMA 2000; 284:1247-1255 Ulcer Complications
 Events Leading up to FDA Mtg
 Results of VIGOR and CLASS
 9/30/04 Merck voluntary w/d of Vioxx (based
on APPOVE trial)
 Reanalyses of all COX2 data
 12/9/04 FDA warning of CV events in Bextra
CABG trial
 12/17/04 NCI stop APC trial (concerns over
celecoxib data)
 12/20/04 NIH stops ADAPT trial (concerns
over naproxen)
 FDA: COX-2 Safety
February 16-18, 2005
 Arthritis Advisory Committee
 Drug Safety and Risk Management
 Other speakers, SGEs
 8 rheums, 19 physicians, 8 statisticians, 1 ethicist,
patient and industry representatives
 ISSUES:
1. Does the agent pose a risk for CV events?;
2. Does the risk versus benefit profile of the drug support its
marketing in the U.S.?
3. If continued marketing is supported, what actions are
recommended to ensure its safe use?
 FDA COX-2 Hearing
 Limitations of available data
 Most trials of short duration, using low dose
 Most trials done with efficacy endpoints
 Few designed with CVS safety endpoint
 Most trials were active comparator and NOT
placebo controlled trials
 Observational studies (presented) are for
generating signals and hypotheses
 Safety signals drawn from other indications are
valid, but not conclusive unless repeated or of
sufficiently great magnitude
Table 1A. Placebo-controlled trials: Prospective CV Endpoint*
Populatio COX-2 dose/d
Trial N Control Results
n (Rx duration)
APPROV Polyp Rofecoxib 25 Cardiac evnt 31 v 12;
2586 Placebo
E** prevention mg (3 yr) HR 2.8 (CI, 1.4-5.4)
Valdecoxib 40
CV/thromboembolic
CABG- High risk mg
1671 Placebo 11 v 3;
II** CABG Parecoxib 40
RR 3.7 (CI, 1-13.5)
mg (10 d)
Parecoxib IV No increase in MI;
CABG-I 80mg Increased CVA in
462 CABG Placebo
(035)** Valdecoxib 80 parecoxib/ valdecoxib
mg (10 days) (2.9% v 0.7%)
Table 1B. Placebo-controlled trials: Sporadic, Observed CV Adverse Events
COX-2, dose/day
Trial N Population Control Results
(Rx duration)
Celecoxib 400
HR 400mg RR 3.0
mg
203 Polyp (0.3-28.6)
APC Celecoxib 800 Placebo
5 prevention 800 mg RR 6.1 (0.7-
mg
50.3)
(>2.8 yrs)
Alzheimer Alzheimer’s Celecoxib 400 MI 2 v 0;
425 Placebo
s-001 prevention mg (1 yr) CV deaths 2.4 v 1.4%
156 Polyp Celecoxib 400
Pre-SAP Placebo No CV increase
1 prevention mg
246 Alzheimer’s Celecoxib 400 Placebo
ADAPT+ No CV increase
3 prevention mg (~2 yrs) Naproxen
Alzheimer 209 Alzheimer’s Rofecoxib 25 mg
Placebo No CV increase
/MCI 1 prevention (1 yr)
197 Polyp
VICTOR$ Rofecoxib (3 yr) Placebo RR 3.14 (1-9.75)
6 PY prevention
Table 2A. Naproxen active comparator trials: Observed Adverse events
Trial N Populati COX-2,dose Compared Results
Rofecoxib MI: 20 v 4 @ 8mos
VIGOR 8076 RA Naproxen
50mg/10 mo RR 2.38 (1.39-4.0)
Metaan
Etoricoxib 3457 Etoricoxib Naproxen RR 1.70 (0.9-3.18)
alyses
TARGET Lumiracoxib
9471 OA Naproxen HR 1.77 (0.82-3.84)
(0117)** 400 mg 1 yr)
Table 2B. NSAID active comparator trials: Observed Adverse events
Celebrex800 Diclofenac
CLASS** 5968 OA, RA No CV increase
mg (9-5mos) Ibuprofen
Celebrex Naproxen
Success-1 13194 OA No CV increase
200, 400 mg Diclofenac
OA/GI Etoricoxib No CV increase
EDGE** 7111 Diclofenac
tolerab (9-16 mos) Small ↑ HTN
TARGET Lumiracoxib
8773 OA Ibuprofen No CV increase
(2332)** 400 mg/1 yr
 Serious CV/T1 Events
Different Endpoints (VIGOR)
Vioxx 50 mg
64
Naproxen
45
32 35
19 18

Investigator CV/ T Adjudicated 2 APTC 3

1 Cardiovascular Thrombotic. 2Confirmed by CV adjudication committee.

3 Antiplatelet Trialist Collaboration composite endpoint: CV & unknown

cause of death, non-fatal myocardial infarction and non-fatal stroke.

Vioxx - APPROVe APTC Events
Time to Event Plot

Rofecoxib 25

Placebo
Vioxx – APPROVe: Effect of ASA on APTC *

Vioxx 25 Placebo RR, p-value

(95% CI)
All patients 1287 1299
n/PYR 33/3053 16/3322
2.25
Rate/100 PYR P=0.008
1.08 0.48
Non-ASA user 1074 1095
28/2564 12/2817 2.57
n/PYR
(1.31,5.06)
Rate/100 PYR 1.09 0.43
ASA user 213 204
n/PYR 5/489 4/505 1.29
Rate/100 PYR 1.02 0.79 (0.28,6.50)

* Based on October 2004, submission (no final dataset).

 CV Signal in Vioxx Databases
AAC Labeling changes

1998 2000 2001-2002

NDA VIGOR 102 RAe SURs/AD
----Epi and re-analyses-----
Versus Plac+NSAIDs Napr Napr Napr Placebo

APTC N Y t t N
CV Death N N t t t
NF MI N Y t t N
NF Stroke N N N N N
N= no signal. Y= Clear signal. t = Trend. AAC: February 8, 2001 FDA Arthritis Advisory Committee
meeting. RAe: Rheumatoid Arthritis efficacy supplement. SURs/AD: Safety Update Reports including
Alzheimer’s disease studies. Epi= epidemiologic studies. NF= non-fatal.
 CLASS - APTC-like Events

Event Celebrex Diclofenac Ibuprofen

(n=3987) (n=1996) (n=1985)

CV death 11 5 5

MI 19 4 9

Stroke 4 6 6

TOTAL 34 (0.9%) 15 (0.8%) 20 (1.0%)

 FDA COX-2 Hearing
 Vote: Retain on the market
 Celecoxib 32-0
 Valdecoxib 17-13 (2 abstentions)
 Rofecoxib 17-15
 Unanimously in favor of:
  “Black box” warning for CV risk & COX-2 drugs
  Education measures for pts and physicians
  Restrictions on direct-to-consumer advertising
  “Warning” added to Current NSAIDs
?  Compassionate use liquid rofecoxib for kids
 Summary
 COX-2 inhibitors equipotent to NSAIDs
 GI Toxicity lower with COX-2, but negated
by low dose ASA 81 mg/d
 Would this be off-set by use of PPI
 CV risk modestly higher w/ COX-2 inhibitor
 Same for NSAIDs?
 Not affected by use of low dose ASA
 FDA: COX-2 Safety
 Risk/benefit of COX-2 inhibitors favors continued use in U.S
 Patients should be counseled about cardiovascular risks.
 While all NSAIDs may impose similar cardiovascular risks,
some agents (naproxen 500 mg bid, celecoxib 200 mg qd)
appear to be safer than others (e.g., rofecoxib, valdecoxib,
diclofenac, and ibuprofen).
 Cardiovascular risks of the COX-2 inhibitors may be greater
with higher doses, longer durations of therapy, and when
used in high risk individuals.
 The use of low dose aspirin does not consistently abrogate
the potential CV risk of a COX-2 inhibitor. Patients who
require the cardioprotective effects of aspirin may not be
ideal candidates for COX-2 inhibitor or NSAID therapy.
 The use of COX-2 inhibitors (and other NSAIDs thought to
increase CV risk) should be avoided in high risk individuals
 Bextra
FDA Action: 4/7/05
We have concluded that the overall risk versus benefit profile for this product is
unfavorable and we have requested Pfizer to voluntarily withdraw the product from
the market. They have agreed to suspend sales and marketing in the U.S., pending
further discussions with the Agency. To resume marketing of the drug, the sponsor
would have to demonstrate that Bextra has a clear benefit over existing therapies or
a lower risk compared to other COX-2 selective inhibitors.
 Celebrex
We have concluded that Celebrex should remain available as a prescription drug, but
with changes to the labeling, a MedGuide, and commitments from Pfizer for
additional studies to better define the cardiovascular effects of the drug.
 Vioxx
A proposal by Merck to reintroduce Vioxx to the market would require a supplemental
new drug application. Such an application would be reviewed with consideration of
the risk to benefit balance of the proposed indications and populations for use,
warnings in the label, and all relevant data. We expect that the proposal would also
be reviewed at a public Advisory Committee meeting.
 Prescription Non-Selective NSAIDs- Based upon the available data, we have
concluded that an increased risk of CV events may be a class effect for
NSAIDs. Therefore, at this time, changes to the prescribing information for all of
these drugs are warranted, until the risk profile of the individual agents can be better
assessed.
 Non-prescription Non-Selective NSAIDs- The labeling for low dose, non-prescription
products that contain ibuprofen, naproxen and ketoprofen will be revised to include
warnings about potential CV and GI risks, advisories recommending certain patients
seek physician input before use and stronger reminders to follow the instructions of
the label concerning dose and duration of treatment
 A New Standard for Drug
Development
 Any new NSAID, must be clinically better
than current alternatives
 Any new NSAID, must be safer
 Usual phase IV trials must be done before
approval, not after.
 NSAIDs Available by Prescription
NSAIDs
Diclofenac (Voltaren)
Diclofenac/Misoprostol (Arthrotec)b
Fenoprofen (Nalfon)
Flurbiprofen (Ansaid)
Ibuprofen (Motrin)a
Indomethacin (Indocin)
Ketoprofen (Orudis)a
Meclofenamate
Mefenamic acid (Ponstel)
Nabumetone (Relafen)
Naproxen (Naprosyn, Anaprox)a
Oxaprozin (Daypro)
Piroxicam (Feldene)
Sulindac (Clinoril)
Tolmetin (Tolectin)
2004 Physician’s Desk Reference
SALICYLATES COX-2 INHIBITORS
Aspirina (Zorprin, Easprin) Celecoxib (Celebrex)
Diflunisal (Dolobid) Valdecoxib (Bextra)
Salsalate (Disalcid, Salflex)
Choline salicylate (Trilisate)
Magnesium salicylate (Magan)
In Development
Etoricoxib
Parecoxibc
Lumiracoxib
Previously Available
Rofecoxib (Vioxx)
a Also available as over-the-counter preparations in the U.S.
b Combination tablet of NSAID/synthetic prostaglandin E1
c Parenterally administered
 In Vitro Selectivity: COX-2/COX-1 Ratio
lumiracoxib
etoricoxib
rofecoxib
valdecoxib > 50-fold COX-2 selective
etodolac
nimesulide
diclofenac 5- 50-fold COX-2
celecoxib selective
meloxicam

fenoprofen < 5-fold COX-2

ibuprofenselective
tolmetin
naproxen
aspirin
indomethacin
ketoprofen
Warner et al. FASEB J. 2004:18:790-804 flurbiprofen
ketorolac
-3 -2 -1 0 1 2 3
Increasingly COX-2 Increasingly COX-1
Selective Selective
Range of COX Selectivity for COX-1 and COX-2
(log10 IC50 COX-2/COX-1)
 Figure 1. Pooled and study specific relative risks and 95% confidence intervals for the
RR OF MYOCARDIAL INFARCTION
risk of myocardial infarction associated with use of NSAIDs as a class, naproxen and
ibuprofen.

Garcia-Rodriguez I

Ray I

Solomon Garcia Rodriguez 2004

Watson

Schlienger

Ray II

Mamdani

Kurth

Kimme
l
Garcia-Rodriguez II

Pooled NSAIDs
NSAIDs RR=1.04 [1.00-1.08
Garcia-Rodriguez I

Ray I

Solomon

Watson

Schlienger

Ray II

Mamdani

Kimmel

NAPROXEN Garcia-Rodriguez II RR=0.88 [0.8-0.95]

Pooled Naproxen

Garcia-Rodriguez I

Ray I

Solomon

Watson

Ray II

Kimmel

IBUPROFEN Garcia-Rodriguez II

Pooled Ibuprofen RR = 1.03 [.96-1.1]

0 1 2
? : Study specific, adjusted RRs; ? Pooled RR; — 95% C

NAPROXEN : HALF AS GOOD AS ASPIRIN?

Long-term NSAID Use May Increase Risks of
Cardiovascular Death-Norwegian Study, 2005
 Population-based, nested case-control study of 454
Scandinavian patients diagnosed with oral cancer
between 1975 and 2003, and 454 gender- and age-
matched controls. (Sudbo J, et al. 2005 - Manuscript submitted)
CV deaths
2.06
Any NSAID 42
1.16
Aspirin 2
2.86
Ibuprofen 12
1.70
Naproxen 7
2.26
Indomethacin 10
1.84
Piroxicam 7
1.90
Ketoprofen 4

0 1 2 3 4 5 6
Hazard Ratio for CV Death in Long-term NSAID Users Compared to Non-users
(with 95% Confidence Intervals)
 On Safety…
“The desire for safety stands against every
great and noble enterprise”
- Tacitus

“ I think we too often make choices based on

the safety of cynicism, and what we’re lead
to is a life not fully lived”
- Ken Burns