The Vioxx Withdrawal What Happened John A Baron

The Vioxx Withdrawal What Happened? John A. Baron Dartmouth Medical School for the APPROVe Investigators

Vioxx A Recent Quote The licensing of Vioxx and its continued use …. have been public-health catastrophes. (Lancet, Nov 5)

COX-2 Inhibitors What are They? Should we Care?

Cyclooxygenase & Prostaglandins Membrane phospholipids Arachidonate PGG 2 COX PGH 2 PGE 2 PGF 2 PGD 2 PGI 2 TXA 2

Prostanoids Phospholipids, Arachidonic Acid Phospholipase A 2 PGG 2 Peroxidase NSAIDs PGH 2 Isomerase Prostaglandins Thromboxanes Angiogenesis Platelet function Apoptosis Vascular Reactivity Invasiveness Inflammation COX Cell Signaling Molecules

Cyclooxygenase 2 Isoforms Cox-1 consitutive house-keeping Cox-2 inducible inflammation, cancer Cox-2 w/o Cox-1 inhibition may offer: · Anti-inflammatory effects · Cancer prevention AND · Protection of stomach · No bleeding

Vioxx Early History • Vioxx (Rofecoxib) released in 1999 • An “early” COX-2 inhibitor more selective than celecoxib • Premise of COX-2 inhibitors: greater safety than traditional NSAIDs at least equal efficacy

Thrombotic CV Events: the VIGOR Study Overall RR: 2. 38 (1. 39, 4. 00)

Thrombotic CV Events: Phase II OA Overall RR 1. 09 (0. 69, 1. 73)

Cardiovascular Background Summary In randomized trials prior to APPROVe cardiovascular risk for rofecoxib was: • Higher than for naproxen • Similar to non-naproxen NSAIDs • Similar to Placebo (limited data beyond 2 years)

APPROVe Study (Adenomatous Polyp Prevention with VIOXX) Standard Adenoma Prevention Study Subjects with recent adenoma 3 -year adenoma endpoint 1 -year research colonoscopy Rofecoxib 25 mg vs. placebo 107(!) sites, 39 in U. S.

APPROVe Study overseen by External Steering Committee External Safety Monitoring Board (ESMB) Adjudication of Serious CV Events Prespecified Protocol for CV effects

APPROVe Study Design Rofecoxib 25 mg (N~1214) Randomization Colo* Study Visit Month -4. 5 -1. 5 0 (N~1214) Placebo Colo 12 24 *non-study, within 3 months prior to screening 36

APPROVe Eligibility Inclusion Criteria ≥ 40 years old histologically confirmed large bowel adenoma Prior MI, PTCA, CABG OK if > 1 year prior T 0 Exclusion Criteria Uncontrolled hypertension (>165/95 mm Hg), angina at rest or minimal activity, CHF at rest

APPROVe Baseline Characteristics Rofecoxib N=1287 Male (%) 62 Mean age 59 years Aspirin use (%) 19 Hypertension (%) 36 CV risk* (%) 29 Current Smoker (%) 22 Placebo N=1299 62 59 years 18 34 26 22 * CV hx, or ≥ 2 of: hx of DM, cholesterol, HTN, smoker

APPROVe CV Events, as of 8/16/2004 118 Investigator-reported events 70 Confirmed Thrombotic Events MI, Unstable Angina, Sudden Death Stroke, TIA DVT, PE, Arterial Thrombosis 49 Confirmed APTC events* Death: CV or unknown cause MI Stroke *APTC = Antiplatelet Trialists’ Collaboration BMJ. 1994

APPROVe CV Events Rate per 100 (N/ P-Yrs) Placebo N=1299 Thrombotic (70 Events) APTC (49 Events) Rofecoxib N=1287 0. 75 1. 48 (25/3315) (45/3041) 0. 48 1. 08 (16/3322) (33/3053) Relative Risk (95%CI) 1. 96 (1. 20, 3. 19) 2. 25 (1. 24, 4. 08) RR for CHF/PE/Cardiac Failure: 4. 29 (1. 43, 12. 82)

APPROVe Confirmed Thrombotic Events Placebo (N=1299) Rofecoxib (N=1287) Cardiac Events 11 30 Cerebrovascular Events 7 15 Peripheral Vascular Events 7 3

APPROVe Confirmed Thrombotic Endpoints Overall RR: 1. 96 (1. 20, 3. 19)

Thrombotic CV Events Alzheimer’s Disease Studies Overall RR 1. 01 (0. 67, 1. 53)

APPROVe Thrombotic Events Subgroup analyses Age • Aspirin use Hypertension • Cigarette smoking Diabetes • CV risk* Hypercholesterolemia No treatment by Subgroup Interactions *CV hx, or ≥ 2 of: hx of DM, cholesterol, HTN, smoker

APPROVe: Blood Pressure Preliminary analyses not suggestive of a relationship between blood pressure rise and risk

APPROVe CV Events Summary risk of thrombotic CV events after 18 months of Tx 1 st 18 months consistent with prior placebo-controlled and nonnaproxen controlled data On the basis of these data, VIOXX was withdrawn

APPROVe: The Future Mechanism of CV toxicity uncertain Analyses ongoing Patients will be followed for one year per protocol Adenoma data will be analyzed Study within 3 months of completion anyway ~75% of subjects had completed treatment

APPROVe Research Team John Baron†, Robert S Bresalier††, Robert Sandler‡, Robert Riddell§, Angel Lanas║, Dion Morton¶, Alise Reicin#, Bettina Oxenius#, Kevin Horgan#, Hui Quan# †Dartmouth Medical School ††University of Texas MD Anderson Cancer Center; ‡University of North Carolina at Chapel Hill; §Mount Sinai Hospital, Toronto; ║University Clinic Hospital, Zaragoza, Spain; ¶University of Birmingham, UK; #Merck Research Laboratories

Cumulative Metaanalysis MI (not Total CVD) Combined RR = 2. 24 (1. 24 -4. 02 Year 1997 1998 Juni et al, 2004 # Patients # Events 5193 13, 269 16 40 1999 2000 VIGOR Study 2002 21, 432 64

Naproxen & MI Risk Observational Data Jick (2000) Rahme (2002) Ray (2002) Schlienger (2002) Solomon (2002) Watson (2002) Mamdani (2003) Kimmel (2004) Juni et al, 2004 Graham (2004) Garcia Rdoriguez (2004) Combined RR (0. 86 0. 75 -0. 99)

COX-2 Inhibitors & CVD What are the Possible Mechanisms?

Aspirin COX-2 Inhibition COX-1 Prostacyclin Thromboxane Decreased CV events COX-2 Prostacyclin Increased CV events

Atherosclerosis An Inflammatory Process • Cox-2 over expressed in atheroma • Cox-2 inhibitors might be beneficial? ?

NSAIDs and Blood Pressure Frishman, Am J Cardiol, 2002

Change from baseline (mm Hg) Baseline mean = 136 mm Hg Frishman, 2002 Baseline mean = 136 mm Hg Baseline mean = 81 mm Hg

NSAIDs & GFR Harris, Am J Cardiol, 2002

NSAIDs and CHF “An Underrecognized Public Health Problem” • NSAIDs can CHF • Stronger effect with Hx of CVD (Especially drugs w/ long half life) Heerdink et al, Arch Intern Med, 1998 Page & Henry, Arch Intern Med, 2000 Feenstra et al, Arch Intern Med, 2002

COX-2 & Cardiovascular Disease Background • COX-2 vascular prostacyclin • COX-2 inflammation Net effect on atherosclerotic disease? • COX-2 involved in renal tubular function • COX-2 inhibition may lead to: fluid retention HTN

Vioxx Summary • Increases in CVD • Probably delayed • Probably rare Published commentary uninformed