Loading...
Nonsteroidal anti-inflammatory drugs (NSAIDs) impair the progression of colonic neoplasia; however, the gastroduodenal adverse effects of nonselective NSAIDs have limited their use for this purpose. Because cyclooxygenase-2–selective NSAIDs (coxibs) are associated with fewer gastrointestinal side effects, they have been evaluated in several studies of patients with colon polyps. One such study, APPROVe, was discontinued early because of increased risk for adverse cardiovascular events in patients who received rofecoxib (Journal Watch Gastroenterology Apr 12 2005); rofecoxib was withdrawn from the market promptly. Other studies have suggested that celecoxib is associated with similar adverse events.
Prior to the withdrawal of rofecoxib, investigators in the U.K. initiated a study to evaluate the effect of rofecoxib on cancer recurrence in 2327 patients with stage II or III colorectal cancer who underwent curative surgical resection plus chemotherapy, radiotherapy, both, or neither. Patients were randomized to receive daily rofecoxib (25 mg) or placebo. Although the study was discontinued after the withdrawal of rofecoxib, the enrolled patients were followed for another 2 years to determine the incidence of adverse cardiovascular events.
The median duration of active treatment was 7.4 months, and the median follow-up was 33 months. Twenty-three adjudicated cardiovascular events occurred during therapy or within 14 days of drug termination: 7 in the placebo group and 16 in the rofecoxib group (relative risk, 2.66; 95% CI, 1.03–6.86; P=0.04). Fourteen additional events occurred during the 2 years of follow-up: 8 in the placebo group and 6 in the rofecoxib group (2-year RR, 1.50; 95% CI, 0.76–2.94; P=0.24). When the stricter Antiplatelet Trialists’ Collaboration (APTC) endpoints were used, higher risks were noted in the rofecoxib group, but none achieved statistical significance. The authors concluded that coxibs, specifically rofecoxib, are associated with increased risk for adverse cardiovascular events in colorectal cancer patients who take the drug for 7 months but that this higher risk is attenuated within 2 years after drug discontinuation.
Kerr DJ et al. Rofecoxib and cardiovascular adverse events in adjuvant treatment of colorectal cancer. N Engl J Med 2007 Jul 26; 357:360-9.
Comment
These results are similar to those of other trials and meta-analyses of rofecoxib and celecoxib. Cardiovascular risk was identified after 7.4 months of therapy in this study, whereas this effect did not became apparent until 18 months in the APPROVe trial. The reason for the difference in findings is unclear, but perhaps risk for cardiovascular disease increases sooner after coxib initiation than we thought, or perhaps patients who already have had cancer are prone to cardiovascular problems. The failure to demonstrate a significant difference using APTC endpoints could have been due to the short duration of therapy. The marginal statistical significance of the results and the lack of significance for differences in APTC endpoints illustrate the difficulties inherent in identifying incidence differences for rare adverse events, even in large prospective trials.
This study was terminated prematurely, and survival results were not reported in this paper. However, the cardiovascular risks of coxib therapy might exceed the antineoplastic benefits of these drugs in patients with colonic neoplasia. The good news is that cardiovascular risk associated with coxibs does not appear to persist for longer than 2 years after therapy ends.