Lourdes Villalba, M.D.

1

Vioxx Cardiovascular Safety

• Lourdes Villalba, M.D.
• DAAODP, ODE V
• February 16, 2005

2
Overview of Presentation

• Goals
• Brief background/chronology
• Overview of Vioxx data sources
• Classification of CV events
• CV safety in different Vioxx databases
• Summary of Vioxx CV safety
• Challenges in interpreting Vioxx CV safety

3
Vioxx Milestones
First IND

• 4/02 Label Update
• 10/02 PCPS Prostate Ca. prevention
• 10/02 Merck Pooled CV analysis
• 12/02 Advice letter to Merck

NDA submit Acute Pain/DYSM/OA

• 6/00 VIGOR results

• 9/04 APPROVe results
• Ongoing Studies Stopped
• Product Withdrawn

• 4/99 AAC meeting
• 5/99 NDA approved

95
97
98
99
00
01
02
03
04
05
94
96
Arthritis/DSaRM COX-II Safety Meeting

• IND for APPROVe

• 2/01 AAC Meeting
• Review of additional information
• Internal Briefings COX-II Safety

• 3/04 Updated Alzh CV Data
• 3/04 Migraine approval
• 8/04 JRA approval
• 8/04 Kaiser Study Abstract

• IND for Alzheimers
• 3/98 COX-II Safety AC

• GIDAC Meeting (APPROVe)

4
Overview of VioxxData Sources Before APPROVe
1998 2000 2001-2004 2001-2004 2001-2004 2001-2004 2001-2004
NDA VIGOR 102 RAe1 SUR2 AER3 Epi
Indication OA RA OA RA Various Re-analysis
Vioxx 12.5
25
50
Ibuprofen
Diclofenac
Nabumetone
Placebo 4 5 6
Naproxen
1RAe RA efficacy supplement. 2 SURs Safety
Update Reports, include Alzheimers studies.
3FDA Adverse Event Reporting. 4up to 18 weeks.
5up to 12 weeks. 6Alzheimers studies up to 3
years. Epi epidemiologic studies.

5
Cardiovascular Endpoints

• All events reported under CV system
• CV Deaths
• Discontinuations due to CV adverse events
• Serious CV adverse events
• CV/Thrombotic (pre-specified definition) referred
  for blinded CV adjudication
• Confirmed CV/Thrombotic
• APTC (Anti-platelet Trialist Collaboration)
• CV and unknown cause of death
• Non-fatal MI
• Non-fatal stroke (ischemic hemorrhagic)
• (does not include unstable angina, transient
  ischemic attack peripheral events)

1
2
1. Routine FDA review 2. VIGOR and all
subsequent Vioxx studies
6
Serious CV/T1 Events Different Endpoints (VIGOR)
64
45
35
32

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.
7
Vioxx NDA Database (Nov. 1998)

• Large NDA, approx. 5,400 patients on rofecoxib
• Approx. 3,300 patients in multiple dose OA trials
  (6-week, 6-month and one-year studies, some with
  extension up to 21 months).
• 371 patients exposed for 1 year at 12.5 mg
• 381 patients exposed for 1 year at 25 mg doses
• 270 patients exposed to 50 mg for 6 months
• (Greater exposure than International Conference
  on Harmonisation Guidances)

8
Vioxx NDA All CV EventsPotentially Thrombotic,
Serious and Non-serious1 (Nov 1998)
6-week Studies 6-week Studies 6-week Studies
Vioxx 12.5 Vioxx 25 Vioxx 50 Ibuprofen Nabum Placebo
N 725 N 735 N 97 N 470 N 115 ______ N 412
5 (0.7) 5 (0.7) 1 (1.1) 2 (0.4) 0 1 (0.2)
24-week Studies 24-week Studies 24-week Studies 24-week Studies 24-week Studies 24-week Studies
Vioxx 12.5 Vioxx 25 Vioxx 50 Ibuprofen Diclofenac Placebo
N 490 N 879 N 379 N 377 _____ N 498 N 371
7 (1.4) 10 (1.1) 4 (1.1 ) 2 (0.5 ) 10 (2.0 ) 3 (0.8 )
CV safety Vioxx appears between ibuprofen and
diclofenac. 1
Includes terms such as coronary artery disease,
angina pectoris, myocardial infarction,
cerebrovascular accident (These are not confirmed
events) Placebo only 18 weeks exposure
9
Vioxx NDA Approved May, 1999

• Adequate evidence of efficacy for
• Acute pain (50 mg) for five days
• Dysmenorrhea (50 mg)
• Osteoarthritis (12.5 and 25 mg dose)
• Adequate evidence of safety for the intended uses
  as indicated in the label
• Safety profile similar to comparators NSAIDs
• CV safety profile in between ibuprofen
  diclofenac
• HTN Dose response, 50 mggt 12.5 and 25 mg
• Endoscopic data suggest Vioxx better than
  ibuprofen
• Liver Vioxx better than diclofenac

10
Vioxx - Patients with Confirmed
APTC Events VIGOR (June 2000)
Vioxx 50 mg Vioxx 50 mg Naproxen Naproxen
N 4047 N 4047 N 4029 N 4029
n Rate2 n Rate2 RR3
APTC1 35 1.30 18 0.67 1.94
CV Deaths 6 0.22 6 0.22 1.00
NF4 MI 18 0.67 4 0.15 4.514
NF Isch. Stroke 9 0.33 8 0.30 1.10
NF Hem. Stroke 2 0.07 1 0.04 2.00
1Patients may have more than one event. n number
of events 2 Rate Rate per 100 patient years
based on PYR Vioxx 2697, naproxen 2698. 3
Overall relative Risk Vioxx vs naproxen. 4 NF
Non fatal. Date of submission to FDA.
11
VIGOR Confirmed CV/ Thrombotic Events Time to
Event Plot
Hazard Ratio Vioxx vs. Naproxen
Overall 2.4 gt 8 months 4.0
12
Vioxx- VIGOR CV/Thrombotic Events Over Time
Included in April, 2002 Vioxx label

13
CV Death, Non-fatal MI, Stroke Vioxx Studies
090 0851 (June 2000)
Vioxx 12.5 mg Nabum Placebo
090 N390 N392 N196
CV Death 0 0 0
Non fatal MI 3 1 0
Non fatal stroke 1 0 0
085 N424 N410 N208
CV Death 0 0 0
Non-fatal MI 1 0 0
Non-fatal stroke 0 0 0
1 6-week OA studies. 12 Low dose ASA. Date
of submission to FDA, along with VIGOR.
14
February 8, 2001 Arthritis
Advisory Committee Conclusions

• Vioxx GI safety profile superior to naproxen
• CV signal of concern
• Given study design, unclear how it applies to
• 12.5 and 25 mg doses (approved for chronic use)
• populations other than RA
• NSAIDs other than naproxen
• population at CV risk using low dose aspirin
• Labeling changes should reflect both, benefits
  and potential harms
• Additional data needed to clarify these issues

15
Vioxx - APTC Events RA Efficacy Database
(February, 2001)

• Complex database. Vioxx 25, 50 and naproxen
• - Studies 068, 096 097, 098 103
• 12-week base studies (Part I) with blinded
  extensions up to 3
  years. Re-randomization of low doses and placebo
  to Vioxx 25, 50 or naproxen
  after initial 12 weeks.
• Vioxx 25 50
  Placebo Naproxen
• Part I(N) 823 729
  989 516
  
• Extensions
  

Date of submission to FDA
16
APTC events - VioxxRA Efficacy Database

• Treatment Events Pt-years
  Risk per
  
  
• at
  risk 100 pt-years
• Placebo 1 160
  0.6
• Vioxx 12.5 2 29
  6.9
• Vioxx 25 5 501
  1.0
• VIoxx 50 6 430
  1.4
• Naproxen 1 406
  0.3

Increased CV risk for Vioxx 25 and 50 mg as
compared to naproxen.
Studies 096, 097, 098 103 (068 not included).
12 week base studies with extensions up to 3
years. ASA not allowed.
17
Vioxx - APTC Events ADVANTAGE (March and
April, 2001)
Vioxx 25 mg N2785 n Rate2 Vioxx 25 mg N2785 n Rate2 Naproxen N2771 n Rate2 Naproxen N2771 n Rate2 RR3
APTC1 10 1.56 7 1.11 1.41
CV Deaths 4 0.63 0 - -
NF MI 5 0.78 1 0.16 -
NF Isch. Stroke 1 0.16 6 - -
1 Patients may have more than one event. 12 week,
OA, 12 low dose ASA. n events. crude
rate. Include 3 sudden deaths (one
re-adjudicated by FDA reviewer) and one ruptured
aortic aneurism on Vioxx. 2 Rate per 100 patient
years at risk. PYR Vioxx 640 and naproxen
629. Date of submission to FDA. Also known as
Study 102.
18
Vioxx - Safety Update Reports (July 2001)

• Studies in the original NDA (follow up)
• New studies
• Short term (4 - 6 week)
• Long-term (gt1 year)
• Study 083 (bone density) Vioxx 25 vs. ibuprofen
• Alzheimers Vioxx 25 vs. placebo
• Updated meta-analysis of CV events
• No differences in confirmed CV/thrombotic or
  APTC events

Date of submission to FDA Does not include
Sponsors responses to FDA information requests.
19
Vioxx in Alzheimers Disease (AD)1 Long-term
Vioxx 25 mg vs. Placebo (July 2001)

• Established AD
• 091 - 15 months, 350 patients per arm, age 65 y
  
• Completed showed no efficacy
• Median exposure 13 months
• 026 - Similar design to 091 terminated early
• Median exposure 6 months
• Prevention of AD in Mild Cognitive Impairment age
  50 y
• 078 - 2 years or until 220 AD events, extended
  to
• 4 years 730 patients per arm,
  ongoing at the time
  
• of the SURs median exposure 18
  months

1 60 male mean age 75 years amended to allow
low dose aspirin (7)
20
APTC events in Alzheimers Studies 091 078
(July 2001 February 2002)

Vioxx 25 mg N 1069 n Rate2 Vioxx 25 mg N 1069 n Rate2 Placebo N 1069 n Rate Placebo N 1069 n Rate RR3
APTC1 17 1.34 27 1.84 0.37
CV Deaths 8 0.63 5 0.34 1.84
NF MI 6 0.47 11 0.75 0.63
NF Isch Stroke 2 0.16 12 0.82 0.19
NF Hem Stroke 1 0.08 1 0.07 1.15
Date of submission to FDA. 1 One patient may
have more than one event. n events. 2 Rate
per 100 patient years. PYR exposure Vioxx 1267,
placebo 1464. 3Overall relative risk of Vioxx
vs. placebo. Median duration 14 months.
21
Vioxx - Safety Update Reports (July 2001)
Meta-analysis of APTC endpoints Vioxx vs. Placebo
and vs. non-naproxen NSAIDs
Vioxx all doses Placebo RR (95 CI)
Case/PYR1 Rate2 Case/PYR1 Rate2
Totals Vioxx vs Placebo 38/2518 (1.51) 34/2099 (1.62) 0.93 (0.57,1.53)
Totals Vioxx vs non-naproxen NSAIDs 29/2648 (1.10) 17/1303 (1.30) 0.84 (0.45,1.63)
PYR patient years at risk. Rate rate per 100
patient years.
22
CV Signal in Vioxx Databases
AAC
1998 2000
2001-2002
NDA
VIGOR 102 RAe SURs/AD Versus
PlacNSAIDs Napr Napr Napr
Placebo


APTC NA NA Y t t N
CV Death N N N t t t
NF MI N N Y t t N
NF Stroke 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 Alzheimers disease studies.
NF non-fatal
23
CV Signal in Vioxx Databases
AAC
Labeling changes
1998 2000
2001-2002
NDA
VIGOR 102 RAe SURs/AD
----Epi and re-analyses----- Versus
PlacNSAIDs Napr Napr Napr Placebo

APTC N N Y t t N
CV Death N N N t t t
NF MI N N Y t t N
NF Stroke 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 Alzheimers disease studies. Epi
epidemiologic studies. NF non-fatal.
24
Vioxx CV Labeling Changes April 2002

• Clinical Studies, Special studies, VIGOR
• Cardiovascular findings
• Description of results
• Two tables - Table 3 CV/T events over time
• - Table 4 CV/T events by
  category
• Precautions
• Indications
• Adverse reactions
• HTN in patients with RA with 25 mg dose
• Dosage and administration

25
Vioxx CV Labeling Changes April 2002 (contd)

• Precautions - Cardiovascular Effects
• The information below should be taken into
  consideration and caution should be exercised
  when VIOXX is used in patients with a medical
  history of ischemic heart disease.
• Description CV results in VIGOR
• Reference to a placebo-controlled database
  derived from 2 studies with a total of 2142
  elderly patients (Alzheimers data) including CV
  mortality (8 and 3 cases).
• The significance of the cardiovascular findings
  from these 3 studies (VIGOR and 2
  placebo-controlled studies) is unknown.
• Prospective studies specifically designed to
  compare the incidence of serious CV events in
  patients taking VIOXX versus NSAID comparators or
  placebo have not been performed.
• Prospective, long-term studies on concomitant
  administration of VIOXX and aspirin evaluating
  cardiovascular outcomes have not been conducted.

26
Vioxx CV Labeling ChangesApril 2002 (contd)

• Indications Rheumatoid Arthritis
• Adverse Reactions
• Rheumatoid Arthritis. In studies of at least
  three months, the incidence of hypertension in RA
  patients receiving the 25 mg dose once daily was
  10 and the incidence of hypertension in patients
  receiving naproxen 500 mg twice daily was 4.7.
• Dosage and Administration
• Rheumatoid Arthritis
• The recommended dose is 25 mg once daily. The
  maximum recommended daily dose is 25 mg.
• Acute Pain and Primary Dysmenorrhea
• Chronic use of Vioxx 50 mg daily is not
  recommended.

27
Study 203 (October 2002)

• Prospective analysis of Cardiovascular Thrombotic
  events in three long-term placebo-controlled
  studies
• APPROVe (colon polyp prevention)
• VICTOR (Oxford study)
• PCPS (prostate cancer prevention)

Data of submission of proposed analysis to FDA
28
APTC Events in Alzheimers Studies 091 078
(Updated March 04)

Vioxx 25mg N 1069 n Rate2 Vioxx 25mg N 1069 n Rate2 Placebo N 1074 n Rate2 Placebo N 1074 n Rate2 RR3
APTC1 32 1.88 40 2.07 0.91
CV Deaths 11 0.65 10 0.52 1.25
NF MI 14 0.82 14 0.73 1.14
NF Isch Stroke 6 0.35 17 0.88 0.40
NF Hem. Stroke 1 0.06 1 0.05 1.13
Date of submission to FDA. 1 One patient may
have more than one event. n events. 2 Rate
per 100 patient years at risk (PYR). PYR Vioxx
1699, placebo 1930. 3Overall relative risk of
Vioxx vs. placebo, based on PYR.
29
CV/T events, Risk Over Time Alzheimers 091
078 (Updated March 2004)
30
APTC Events KM Estimates (95 CI) Vioxx
Alzheimers Disease (091 078) (Updated March
2004)
Overall RR Vioxx 25 mg vs naproxen 1.0

31
Vioxx - All Cause Mortality (On-Drug)
Alzheimers Disease Studies
091 078.
All cause mortality 36/19 (V 25 /Placebo)
32
Vioxx NaproxenEpidemiologic Studies (2001-2004)

• Suggest increased CV/T risk for Vioxx 50 mg dose
• No clear evidence of
• - CV/T risk with Vioxx 12.5 mg and 25 mg doses
• - Cardioprotective effect of naproxen
• FDA awaiting results of long-term placebo
  controlled studies (ongoing Pooled Analysis of
  CV/T events)

33
Vioxx - APTC events in APPROVe1 (as per January,
2005)

Vioxx 25 mg N1287 n Rate3 Vioxx 25 mg N1287 n Rate3 Placebo N1299 n Rate Placebo N1299 n Rate RR4
APTC2 34 1.11 18 0.54 2.055
CV Deaths 6 0.20 5 0.15 1.30
MI (all) 21 0.68 9 0.27 2.535
NF Isch Stroke 10 0.33 6 0.18 1.81
NF Hem. Stroke 1 0.03 1 0.03 1.09
13 year on-treatment, 1 year F.U. history of
colonic polyps mean age 59 years 62 male
19 low dose aspirin. 2 Patients with events
(patients with multiple events may be counted
more than once). 3 Rate per 100 patient-years at
risk. PYR patient years at risk
Vioxx 3070 and placebo 3334. 4 RR Overall
relative risk Vioxx to placebo. 5 Stat
significant.

34
Vioxx - APPROVe APTC Events Time to Event Plot

• Rofecoxib 25

Placebo
35
Vioxx APPROVe Effect of ASA on APTC
Vioxx 25 Placebo RR, p-value (95 CI)
All patients n/PYR 1287 33/3053 1299 16/3322 2.25 P0.008
Rate/100 PYR 1.08 0.48 2.25 P0.008
Non-ASA user n/PYR 1074 28/2564 1095 12/2817 2.57 (1.31,5.06)
Rate/100 PYR 1.09 0.43 2.57 (1.31,5.06)
ASA user n/PYR 213 5/489 204 4/505 1.29 (0.28,6.50)
Rate/100 PYR 1.02 0.79 1.29 (0.28,6.50)
Based on October 2004, submission (no final
dataset).
36
Vioxx APPROVeEffect of BP on CV/T Events
Patients with investigator reported
thromboembolic or APTC events (excluding non CV
deaths) by on-treatment Hypertension
On treatment HTN Rofecoxib 25 mg n/PYR Rate x 100 pt years Placebo n/PYR Rate x 100 pt years
Patients with no HTN 35/2058 (1.7) 32/2510 (1.3)
Patients with HTN once 20/449 (4.5) 4/413 (1.0)
Patients with HTN (twice or more) 18/520 (3.5) 8/381 (2.1)
nevents. Hypertension defined as patients with
DBP 100 or SBP 160 mm Hg.
37
Summary - Cardiovascular Signal
in Vioxx Databases
1998 2000
2001-2004 2004 NDA
VIGOR 102/RA Alzheimer Epi APPROVe
Studies PlacNSAIDs Napr
Napr Placebo Placebo


APTC NA Y t N Y
CV Death2 N N t Y N
NF MI N Y t N Y
NF Stroke N N N N Y

N no signal. Y Clear signal. t Trend. Epi
epidemiologic studies and meta-analyses.
38
Summary Selected CV Endpoints Large Vioxx
Databases
PTY risk 1 VIGOR Vioxx 50 Naprox N2697 N2698 VIGOR Vioxx 50 Naprox N2697 N2698 Alzheimers Vioxx 25 Placebo N1699 N1930 Alzheimers Vioxx 25 Placebo N1699 N1930 APPROVe Vioxx 25 Placebo N3070 N3334 APPROVe Vioxx 25 Placebo N3070 N3334
APTC events APTC events APTC events APTC events APTC events APTC events APTC events
n PY-rate 2 35 1.30 18 0.77 32 1.88 40 2.07 34 1.11 18 0.54
Myocardial Infarction (fatal and non-fatal) Myocardial Infarction (fatal and non-fatal) Myocardial Infarction (fatal and non-fatal) Myocardial Infarction (fatal and non-fatal) Myocardial Infarction (fatal and non-fatal) Myocardial Infarction (fatal and non-fatal) Myocardial Infarction (fatal and non-fatal)
n PY-rate 2 20 0.74 4 0.14 15 0.88 15 0.77 21 0.68 9 0.27
All cause mortality (on drug) All cause mortality (on drug) All cause mortality (on drug) All cause mortality (on drug) All cause mortality (on drug) All cause mortality (on drug) All cause mortality (on drug)
n PY-rate 2 22 0.82 15 0.56 36 2.12 19 0.98 10 0.36 10 0.30
n events. 1 PYR patient years at risk, assuming
constant overall risk. 2 PY-rate Overall rate
per 100 patient years.
39
Challenges in Interpreting Vioxx CV Safety

• Vioxx CV signal was clear when compared to
  naproxen
• CV findings inconsistent when compared to placebo
  (APPROVe different from Alzheimers)
• No adequate comparative CV safety data for Vioxx
  non-naproxen NSAIDs
• Until recently, no long-term placebo-controlled
  data for any of the traditional NSAIDs
• Extent of role of BP on CV/T events with Vioxx is
  unclear
• Role of low dose ASA in protecting for CV/T
  events with Vioxx unknown

40
Issues to Consider in Trial Design for COX-2
inhibitors

• Different ways of looking at CV endpoints
• All CV events APTC (which is
  best?)
• In addition to overall risk, need to consider
  risk over time (long time)?
• Different rates of CV/T events in different
  populations. Which one is the best to address CV
  safety?
• Need to control for BP effects
• Need to include more than one traditional NSAID
  comparator