VIOXX® Information Center

Since Merck announced the voluntary withdrawal of Vioxx^® (rofecoxib), questions have been raised about what we knew before and after we obtained the results of the VIGOR study, and what actions we took before and after we knew those results. This article addresses these questions and reviews the facts and the decision to withdraw Vioxx^® based on the APPROVe trial.

We became aware of the initial VIGOR results in March 2000. VIGOR evaluated the gastrointestinal safety profile of 50 mg of Vioxx^® compared to 1500 mg of naproxen daily in patients with rheumatoid arthritis. In this study, where Vioxx^® was used at twice the recommended dose for rheumatoid arthritis and four times the recommended dose for osteoarthritis, there was a higher incidence of cardiovascular (CV) events in patients receiving Vioxx^® and that difference was statistically significant. However, all data from the previous studies demonstrated no difference in the cardiovascular event rate between Vioxx^® and placebo or between Vioxx^® and non-naproxen NSAID. Therefore, we promptly unblinded key safety data from two ongoing placebo-controlled trials in Alzheimer’s disease (one trial for prevention of the disease, and one trial in patients with a mild form of the disease). This additional safety data did not show any tendency for a higher CV risk in patients taking Vioxx^® when compared to placebo.

Because of a potential mechanism based reason to explain the higher CV risk in patients taking Vioxx^®, we continued our efforts to further characterize the cardiovascular safety profile of Vioxx^®. We developed a prospective plan to analyze the cardiovascular event rates in three large, placebo-controlled studies:

All three studies were overseen by a steering committee of external medical experts. In addition, each of these trials had an external safety monitoring board that was unblinded to the safety data during the trial and hence was responsible for monitoring patient safety in the trials. Importantly, the external safety monitoring board was required to meet at least twice a year to evaluate and report on the ongoing safety data from the studies.

While these trials were being designed or ongoing, we submitted, in May 2000, the initial VIGOR results to the New England Journal of Medicine for publication and presented the data at a major scientific meeting. Later that same year, the data was published in the New England Journal of Medicine.

Merck submitted a supplemental New Drug Application (sNDA) to the FDA in June 2000. This NDA included extensive data on the VIGOR trial and new proposed prescribing information for Vioxx^®. Merck and Merck Frosst Canada worked diligently with the FDA and Health Canada, respectively, to review these data and develop revised prescribing information.

When we received the data from our APPROVe trial, in which there was an increased CV risk in patients using Vioxx^® beginning after 18 months of continuous use, Merck acted immediately to voluntarily withdraw Vioxx^® from the market. Based on preliminary analyses, there were 7.5 confirmed CV events per 1,000 patient years on placebo and 14.8 confirmed CV events per 1,000 patient-years on Vioxx^®.

Therefore, it was not until September 2004—4.5 years after the APPROVe trial was initiated—that the safety monitoring board determined that there was an issue and recommended that the APPROVe trial be stopped. Within one week of becoming aware of the information, after extensive reflection and consultation, both within the company and with external consultants, we decided to withdraw Vioxx^® from the market. We believe this decision reflects Merck’s fundamental and unwavering commitment to patient safety.