Correspondence

Risks and Benefits of Celecoxib to Prevent Colorectal Adenomas

N Engl J Med 2006; 355:2371-2373November 30, 2006

Article

To the Editor:

In the article by Arber et al. (Aug. 31 issue)1 about the Prevention of Colorectal Sporadic Adenomatous Polyps (PreSAP) trial, the percentage of patients with the combined end point of adjudicated serious cardiovascular events did not differ significantly between the celecoxib and placebo groups. Using data from Supplementary Appendix 2, we calculate that the rate ratio for “cardiac disorders” (including myocardial infarction) is 1.6 (95% confidence interval [CI], 0.9 to 2.8), whereas for “nervous system disorders” (including stroke), it is 0.7 (95% CI, 0.4 to 1.3). Similar results were reported by Bertagnolli et al.2 in the Supplementary Appendix of the Adenoma Prevention with Celecoxib (APC) trial for patients treated with 200 mg of celecoxib twice daily — for cardiac disorders, the rate ratio was 1.7 (95% CI, 1.0 to 2.9), and for nervous system disorders, it was 1.1 (95% CI, 0.5 to 2.4). In a meta-analysis of randomized, placebo-controlled trials of celecoxib lasting at least 6 weeks,3 the rate ratio was 2.3 (95% CI, 1.0 to 5.1) for myocardial infarction and 1.0 (95% CI, 0.5 to 1.8) for stroke. Recently, we also found in two observational studies that celecoxib was associated with acute myocardial infarction but not with ischemic stroke.4,5 These data strongly suggest that a more informative analysis of cardiovascular risks associated with celecoxib should distinguish between myocardial infarction and stroke, rather than use a composite end point.

Frank Andersohn, M.D.
Charité–Universitätsmedizin Berlin, 10117 Berlin, Germany
frank.andersohn@charite.de

Samy Suissa, Ph.D.
McGill University Health Centre, Montreal, QC H3A1A1, Canada

Edeltraut Garbe, M.D., Ph.D.
Charité–Universitätsmedizin Berlin, 10117 Berlin, Germany

Dr. Suissa reports serving as an external consultant for Merck in preparation for the Food and Drug Administration COX-2 Advisory Committee Meeting.

5 References

1
Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the prevention of colorectal adenomatous polyps. N Engl J Med 2006;355:885-895
Full Text | Web of Science | Medline

2
Bertagnolli MM, Eagle CJ, Zauber AG, et al. Celecoxib for the prevention of sporadic colorectal adenomas. N Engl J Med 2006;355:873-884
Full Text | Web of Science | Medline

3
Caldwell B, Aldington S, Weatherall M, Shirtcliffe P, Beasley R. Risk of cardiovascular events and celecoxib: a systematic review and meta-analysis. J R Soc Med 2006;99:132-140
CrossRef | Web of Science | Medline

4
Andersohn F, Suissa S, Garbe E. Use of first- and second-generation cyclooxygenase-2-selective nonsteroidal antiinflammatory drugs and risk of acute myocardial infarction. Circulation 2006;113:1950-1957
CrossRef | Web of Science | Medline

5
Andersohn F, Schade R, Suissa S, Garbe E. Cyclooxygenase-2 selective nonsteroidal anti-inflammatory drugs and the risk of ischemic stroke: a nested case-control study. Stroke 2006;37:1725-1730
CrossRef | Web of Science | Medline

Author/Editor Response

Distinguishing differences in the rates of individual components of the cardiovascular risk associated with celecoxib has the potential to clarify further the precise pathophysiological mechanisms responsible for this increased risk. Nevertheless, the very low absolute number of fatal and nonfatal strokes and myocardial infarctions, presented here in Table 1Table 1Patients with Individual Adjudicated Cardiovascular Events Occurring before 37 Months after Randomization or on January 31, 2005, Whichever Came First. and published simultaneously with the efficacy reports,1 reinforces the fact that, when one is drawing conclusions regarding relative rates of individual end points, extreme caution is necessary to avoid overinterpretation. We did not calculate hazard ratios for the individual end points precisely because the very small number of events precludes accurate assessment of risk associated with them. Rather, combining events that have common pathophysiological bases seems preferable when dealing with such small numbers. The data do not support the conclusion that celecoxib is associated with an increased risk of myocardial infarction but not of stroke, since even a calculation of relative risk with the use of the numbers in Table 1 would demonstrate widely overlapping confidence intervals.

Scott Solomon, M.D.
Brigham and Women's Hospital, Boston, MA 02115

Janet Wittes, Ph.D.
Statistics Collaborative, Washington, DC 20036

Nadir Arber, M.D.
Tel Aviv Sourasky Medical Center, 64239 Tel Aviv, Israel

Monica Bertagnolli, M.D.
Brigham and Women's Hospital, Boston, MA 02115

Ernest Hawk, M.D., M.P.H.
National Cancer Institute, Bethesda, MD 20892

Bernard Levin, M.D.
University of Texas M.D. Anderson Cancer Center, Houston, TX 77030

for the APC and PreSAP Trial Investigators

1 References

1
Solomon SD, Pfeffer MA, McMurray JJV, et al. Effect of celecoxib on cardiovascular events and blood pressure for the prevention of colorectal adenomas. Circulation 2006;114:1028-1035
CrossRef | Web of Science | Medline

Author/Editor Response

Andersohn and colleagues raise the question of whether stroke and myocardial infarction should be part of the same composite outcome, which is described as prespecified. The post hoc hypothesis offered by Andersohn and colleagues used data from the tables in the supplementary appendixes of the original articles, in which adverse events were grouped into categories by body system, such as “cardiac disorders” and “nervous system disorders.” These categories are heterogeneous collections of adverse events. For instance, 37 different types were included in the cardiac category. It is not clear that a comparison of the associations of celecoxib with “cardiac disorders” and “nervous system disorders” is informative. The optimal approach, used by investigators in both studies, is to define events by standard criteria and have them adjudicated by an independent committee. Separate analyses of the risks of adjudicated myocardial infarction and stroke would be of interest, but there will be too few myocardial infarction and stroke events to draw firm conclusions from these data. In our editorial, the risk–benefit analysis, which used the adjudicated data, accurately represents the differences in risk among cardiovascular outcomes that included myocardial infarction, stroke, heart failure, and death from cardiovascular disease.

Bruce M. Psaty, M.D., Ph.D.
University of Washington, Seattle, WA 98101-1448

John D. Potter, M.B., B.S., Ph.D.
Fred Hutchinson Cancer Research Center, Seattle, WA 98109-1024

Citing Articles (1)