Clinical Cardiology: New Frontiers
Cardiovascular medicine is undergoing extraordinary and rapid transformation. Circulation intends to keep its readership at the forefront of advances in our knowledge. Hence, in this issue, we introduce a section entitled “Clinical Cardiology: New Frontiers.” Approximately once per month, we will engage topics of a clinically relevant nature to integrate new discoveries and knowledge into the armamentarium of the clinical practitioner. We intend that these articles will be helpful in influencing the thinking and clinical practice of physicians. Our primary goal is to provide useful information and practical advice on the management of contemporary clinical problems. Eligible topics will be drawn from a wide range of areas, including pathophysiology, diagnostic methods, epidemiology, natural history, and therapy.
Because we aim to provide mechanistic insights and evaluate areas of controversy and uncertainty, we will generally invite two key opinion leaders to write articles in this section. Readers will note, however, that they will be from different institutions so as to achieve both diversity and balance. We welcome feedback on this new initiative.
Cardiovascular and Renal Advisory Panel of the FDA
The Cardiovascular and Renal Advisory Panel of the FDA met on October 23 and 24, 1997, to discuss the broad issue of evaluation of active-control trials and to consider the platelet antagonist clopidogrel.
Although the traditional basis for drug approval has included demonstration of superiority to placebo in two trials, it is increasingly recognized that under some circumstances, a placebo-controlled trial is not feasible. In a half-day workshop, issues related to conduct and analysis of active-control trials were discussed. Placebo-controlled trials are usually designed to demonstrate superiority over placebo, and lack of rigor in a placebo-controlled trial generally reduces the chance of finding a significant difference between treatment and placebo. Therefore, there is a great incentive to conduct placebo-controlled trials rigorously. By contrast, active-control trials may be designed to demonstrate superiority or equivalence of a new treatment against a standard treatment; demonstration of equivalence in an active-control trial may arise because the two drugs are in fact equally effective or because the trial was not designed or executed in a sufficiently rigorous fashion to detect differences. Implicit in the design of active-control trials are assumptions about the efficacy (and confidence intervals [CIs] around that efficacy) of the comparator active drug. Calculation of these values allows the investigators to estimate the outcome that would be expected with a putative placebo in an active-control trial. Under some circumstances, this estimate of the outcome with a putative placebo may be used to estimate the efficacy of the new treatment. However, the CIs for that estimate may be quite wide. For these and other reasons, active-control trials may require very large numbers of patients to demonstrate efficacy or superiority to standard therapies. Although in theory it would be possible with this methodology to find a treatment better than placebo but worse than standard therapy, such a finding is unlikely, given the CIs around the point estimates of the various outcomes.
The discussion of clopidogrel centered around many of the same issues, because the primary basis for the new drug application was the results of CAPRIE, a single trial involving more than 19 000 patients with various forms of vascular disease randomly assigned to clopidogrel (75 mg/d) or to aspirin (325 mg/d).1 In CAPRIE, there was an 8.7% reduction in the incidence of the combined end point of ischemic stroke, myocardial infarction, or vascular death (P=.043). Patients who entered the trial because of myocardial infarction did better with aspirin, whereas those entering with peripheral vascular disease and stroke did better with clopidogrel. However, despite the size of the trial, there was not enough power to be certain about the reliability or generalizability of a differential treatment effect in these subgroups. After discussion related to these issues, to the questions of active-control trials outlined above, and to concerns regarding the exact status of some patients at the end of the trial, the committee decided, in a split vote, to recommend approval for the use of clopidogrel in patients with vascular disease. The committee further voted that despite the marginally significant P value for the benefit of clopidogrel over aspirin, the evidence that clopidogrel is actually superior to aspirin is not sufficient to be confident. Conversely, the recommendation for approval was based on the certainty that clopidogrel is better than placebo and meets all reasonable criteria for being at least equivalent to aspirin.
- Copyright © 1998 by American Heart Association