Tamoxifen‐treated breast carcinoma patients and the risk of acute myocardial infarction and newly diagnosed angina

The report by Bradbury et al.1 of lower myocardial infarction (MI) and angina risk in current tamoxifen users has important implications for clinical decisions. Women with and those at high risk for breast cancer must balance tamoxifen's benefits, primarily a reduced risk of breast cancer occurrence or recurrence, against adverse effects like increased risk for endometrial cancer and thromboembolism,2, 3 as well as weigh the relative risk–benefit profiles of tamoxifen and aromatase inhibitors.2 Given that heart disease is the leading cause of death in women, convincing evidence that tamoxifen prevents MI may considerably alter clinical decisions. However, the Bradbury et al. study1 has two important methodologic limitations that suggest caution in concluding that tamoxifen prevents MI. First, because nearly all the breast cancer patients available for study received tamoxifen, the reference group comprised women with bladder, colorectal, or nonmelanoma skin cancer. Thus, the reduced MI risk could be an artifact of the choice of reference group. For example, many breast cancers are responsive to endogenous hormones, and one could speculate that women with levels of endogenous hormones supporting breast cancer development also are less likely to develop MI. Another possibility is that the other cancers have risk factors in common with MI, such as high blood pressure (bladder cancer) and diabetes (colon cancer), placing women in the reference group at higher MI risk for reasons unrelated to tamoxifen. Second, including angina as an outcome equivalent to MI is questionable. Whereas specific criteria exist for MI diagnosis,4 angina is a nonspecific patient-reported symptom. Although angina is associated with increased MI risk, it is neither a sufficient nor necessary marker of heart disease.4 Patients may report angina-like symptoms for other conditions such as gastrointestinal distress or panic attacks. In a case–control study of incident MI following a first primary breast cancer diagnosis, in which all patients had breast cancer, we found no association between MI and tamoxifen (odds ratio 1.2, 95% confidence interval 0.7–1.9; controlled for smoking, hypertension, diabetes, and radiotherapy).5 Existing data such as those available to Bradbury et al.1 provide a rich resource for epidemiologic and outcomes research, but the inherent limitations of data collected primarily for reasons other than research must be carefully considered when studies are designed and reported. Given our findings5 and the limitations of the Bradbury et al. study,1 we believe women and clinicians should not base their treatment decisions on a potential cardioprotective effect of tamoxifen. Ann M. Geiger Ph.D.*, Leslie Bernstein Ph.D. , * Research and Evaluation Department, Kaiser Permanente Southern California, Pasadena, California, Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, California.