| |
Risk of myocardial infarction and nucleoside analogues
Comment
|
| |
| |
The Lancet Early Online Publication, 2 April 2008
James H Stein a and Judith S Currier b
a. Division of Cardiovascular Medicine, University of Wisconsin School of Medicine and Public Health, Madison, WI 53717, USA
b. Division of Infectious Diseases, UCLA Center for Clinical AIDS Research and Education, University of California David Geffen School of Medicine, Los Angeles, CA, USA
In today's Lancet, the D:A:D Study Group analyses whether the nucleoside reverse transcriptase inhibitors zidovudine, didanosine, stavudine, lamivudine, and abacavir are associated with increased risk of myocardial infarction.1 Their analysis included 517 patients with myocardial infarction, and 157 912 person-years of follow-up of a cohort with a myocardial infarction rate of 3·3 per 1000 patient-years. Their unexpected finding was that ongoing or recent exposure to abacavir (relative risk 1·90, 95% CI 1·47-2·45) or didanosine (1·49, 1·14-1·95), but not the thymidine analogues zidovudine or stavudine, increased risk of myocardial infarction.1 Because randomised trials show that abacavir improves the lipid profile when substituted for a protease inhibitor, patients with poor lipid profiles or at high risk of coronary heart disease might have been more likely to receive abacavir.2-4
Participants in the D:A:D study who were receiving abacavir and didanosine had greater risk of coronary heart disease than those prescribed other nucleoside reverse-transcriptase inhibitors.1 However the authors make a series of compelling arguments suggesting that channelling bias did not explain their findings, on the basis of statistical adjustments for factors that contribute to risk of coronary heart disease and the observation that the excess risk of myocardial infarction with abacavir and didanosine was present only while participants were taking these medications. If channelling of high-risk patients explained their observations, the excess risk should not have disappeared after cessation of abacavir and didanosine. A better understanding of why patients stopped these medications and if other factors related to discontinuation might have led to increased risk of myocardial infarction would be helpful.
What are the clinical implications of these findings? In general, findings from observational studies should not lead to changes in clinical practice, especially without confirmation. In this case, however, the magnitude of the increased risk of myocardial infarction among the subset of individuals at high risk of coronary heart disease cannot be ignored. High risk of coronary heart disease was defined as greater than 20% predicted risk of angina pectoris, unstable angina, myocardial infarction, or sudden death over a decade.5 In these individuals (about 6% of the D:A:D cohort), one additional myocardial infarction would be expected for every 11 treated with abacavir or every 20 treated with didanosine for 5 years. On the basis of this risk, alternatives to abacavir and didanosine in high-risk patients should be considered; although there are no similar data for the alternative, tenofovir. However, the decision to switch antiretroviral therapy must be made cautiously given the known toxicities of inadequate viral suppression, including increased cardiovascular risk.6 Modification of risk factors for coronary heart disease is crucial: over 5 years, one additional myocardial infarction is expected for every six high-risk patients who smoke cigarettes.7
Patients at low or moderate risk of coronary heart disease should not be alarmed by the estimated 1·5-2·0 fold increased relative risk of myocardial infarction, because the added absolute risk is modest. They should carefully manage their risk factors while awaiting confirmatory research. Because the associations between the use of abacavir or didanosine and myocardial infarction were unexpected and observational studies have recognised limitations for quantifying the risk of coronary heart disease related to medical therapies, this association must be confirmed and a mechanism should be sought. Given its time course, mechanisms that affect atherosclerotic plaque stability, endothelial function, or thrombosis are likely to be involved. Compared with other nucleoside reverse transcriptase inhibitors, abacavir and didanosine might have adverse antiretroviral or immunological effects that influence systemic and arterial inflammation.
By contrast with the D:A:D report, Amy Cutrell and colleagues8 from GlaxoSmithKline (the makers of abacavir) report, in a Correspondence letter in today's Lancet, that myocardial infarction did not increase in a summary of 54 pooled studies, 13 of which randomly assigned patients to receive abacavir. Although the low overall rates of myocardial infarction are somewhat reassuring, Cutrell and colleagues' analysis is not powered to detect meaningful differences: it was based on only 18 myocardial infarctions and the limitations of summaries of pooled data for uncommon events in studies not designed to detect them are well known. Because coronary events were not adjudicated formally in these antiretroviral therapy efficacy studies, interpretation of the rates of "coronary artery disorders" is difficult.8 Available data on coronary heart disease from clinical trials, such as those included in the Cutrell report, should be submitted for peer review so their design and analyses can be described in detail and their conclusions fully interpreted.
The benefits of antiretroviral therapy were gleaned from a strong tradition of randomised trials. By contrast, most of the data on risks of coronary heart disease associated with antiretroviral therapy come from observational and short-term efficacy studies. Because patients are living longer with HIV and coronary heart disease is becoming a real risk to survival, studies lasting 24-48 weeks cannot guide what usually is lifelong therapy. Studies of antiretroviral efficacy should be of longer duration so their toxicities can be better understood. At the very least, long-term follow-up registries should be set up. Coronary events should be adjudicated in all randomised trials of antiretroviral efficacy, and well-validated surrogates of vascular outcomes should regularly be incorporated into efficacy studies. Recent randomised studies of antiretroviral therapy that prospectively investigated risk of coronary heart disease have challenged previously held beliefs and improved our understanding of the effects of antiretroviral therapy and HIV on vascular disease.6,9 The power of the randomised efficacy trials should be used to investigate the coronary risks associated with antiretroviral therapy, so our dependence on observational cohorts and short-term efficacy studies as the only sources of information about the long-term risks of antiretroviral therapy can be reduced.
|
|
| |
| |
|
|
|
