By Nicholas Alp, M.D., Ph.D. and Didier Saur, M.D., ICON plc
The goal of a cardiovascular outcomes trial is to provide an acceptable base of evidence regarding safety and/or efficacy for regulatory and commercial acceptance. Sponsors aim to design studies that are operationally feasible, cost effective, and that produce results applicable to the target population.
With the global implementation of therapies that modify cardiovascular risk, which now constitute standard of care (including statins, ACE inhibitors and anti-platelet drugs), the background CV event rate even in higher risk populations has fallen and is now in the range of 2-3 percent per year. It is, therefore, becoming increasingly difficult to demonstrate further incremental absolute therapeutic benefits for new drugs in cardiovascular disease. As a result, studies designed to demonstrate superiority of a new drug versus placebo on a background standard of care need very large sample sizes. An alternative approach is to design studies with noninferiority statistical comparisons, which in general require smaller sample sizes. Nevertheless, in a population with a background event rate of 2 percent per year, 610 adjudicated events would be required for the upper bound of the two-sided 95 percent confidence interval of the risk ratio to be less than 1.3 – equivalent to 30,500 patient years of exposure (in practice, 10,167 patients followed for an average of three years).
A key strategic approach is to define pragmatic inclusion and exclusion criteria, balancing the need to generalize the results with the desire to enrich the study population with higher-risk subjects, as well as to increase the event rate and the likelihood of detecting either benefit or harm. The balance of secondary prevention versus primary prevention populations in a trial will depend on factors such as the respective event rates, target indication for the drug, and feasibility considerations driven by the relative availability of higherversus lower-risk subjects.