Does Rofecoxib Cardiovascular Risk Persist After Discontinuation?

To The Editor

Interim results from the Adenomatous Polyp PRevention On Vioxx (APPROVe) trial demonstrated that rofecoxib was associated with an increased risk of cardiovascular events,¹ leading to Merck’s withdrawal of rofecoxib from the worldwide market in September 2004. In October 2008, long-term follow-up findings were published;² during a one-year period of post-drug observation, increased cardiovascular risk associated with rofecoxib persisted after stopping treatment.

These findings raise concerns about long-term risks incurred by individuals who had taken the drug. Data from only one other rofecoxib clinical trial, study 078,³ provides information about long-term off-drug cardiovascular risk after clinical trial drug discontinuation. Accordingly, our objective was to examine whether increased cardiovascular risk of rofecoxib persisted after stopping treatment in study 078.

Methods

We used data from study 078 made available through Vioxx litigation by Merck. Details of the randomized, double-blind trial, which examined whether rofecoxib delayed progression of Alzheimer’s disease among adults with mild cognitive impairment, have been published elsewhere.³ We only included subjects who took at least one dose of the study drug (rofecoxib 25 mg daily or placebo). Our pre-specified outcome measure was incidence of any investigator-reported death or cardiovascular thromboembolic (CVT) adverse event, identified from the adverse event reporting terms, as in prior research.⁴

Data were retrospectively analyzed in an intention-to-treat (ITT) fashion, including all randomized individuals and adverse events as of the study termination date. Time at risk for events after stopping treatment began 14 days after the last dose, consistent with Merck’s protocol defining off-drug events; subjects who were alive and had not previously had a cardiovascular event were deemed at risk. In all survival analyses, time of last contact or study termination, whichever was earlier, was used to censor time at risk. Kaplan–Meier estimates of cumulative event rates over time were used to construct event curves, which were compared with logrank tests. Risk ratios [RR] and 95% Confidence Intervals [CI] were calculated with Cox proportional hazards models. All statistical tests were 2-tailed and analyses were performed using the R statistical software environment, Version 2.8.0 (http://www.r-project.org).

Results

Among 1451 subjects, 723 were assigned to rofecoxib and 728 to placebo. Mean age was 75.0 and 33% were female. Median duration of on-drug treatment was 623 days among rofecoxib subjects and 724 days among placebo. Forty-three percent (n=617) of subjects were at risk for a first event and had off-drug follow-up data available; median duration was 189 days among rofecoxib subjects and 125 days among placebo.

During off-drug follow-up, 22 investigator-reported CVT adverse events occurred among rofecoxib subjects, 6 among placebo subjects (RR=2.97, 95% CI, 1.20–7.32; p=0.02). Similarly, 23 deaths occurred among rofecoxib subjects, 9 among placebo subjects (RR=2.10, 95% CI, 0.97–4.34; p=0.06). Overall, an investigator-reported CVT adverse event or death occurred during off-drug follow-up for 45 subjects, 32 among 287 patient-years of rofecoxib use and 13 among 234 patient-years of placebo use (RR=2.01, 95% CI, 1.05–3.82; p=0.03).

Discussion

There was 100% increased cardiovascular risk associated with rofecoxib during off-drug follow-up among study 078 subjects, confirming the 79% increased cardiovascular risk associated with rofecoxib observed in APPROVe.² Furthermore, estimates of off-drug cardiovascular risk from study 078 are likely conservative for two reasons. First, cardiovascular safety was not a primary or secondary outcome of the study and may not have been systematically collected or reported, raising issues of ascertainment bias. Second, study discontinuation rates were high and there was not a planned, rigorously-conducted period of off-drug follow-up, made more challenging because all patients had cognitive impairment. In addition, our analysis updates the published findings from study 078, as subsequent investigation suggested that these were incomplete.⁵

Implications for individuals who used rofecoxib are unclear, as it was withdrawn from the market nearly 5 years ago. No studies have been conducted to determine for how long the increased cardiovascular risk associated with rofecoxib persists, to elucidate the mechanism, or to examine other non-steroidal anti-inflammatory drugs, selective or non-selective. To improve drug safety evaluation within clinical trials, periods of off-drug surveillance should be used when appropriate to ensure observation of long-term effects.

Figure.

Investigator-reported cardiovascular thrombotic events and all-cause deaths observed during off-drug follow-up within study 078.