Considerations in specifying the duration of followup of antiretroviral trials

The duration of followup of antiretroviral trials should consider the possibility that treatments have time-limited efficacy and will not be taken by all as prescribed. A short-term study, by definition, does not provide information concerning long-term persistence of clinical benefit and tolerability. Thus a trial may give misleading answers if, in practice, the experimental treatment regimen is to be used over a longer time period than studied. On the other hand, long-term trials of each new agent or combination are not feasible and may not be relevant if, in practice, drugs are changed frequently. A solution to this conundrum is to formulate research questions as strategy questions with nested studies of specific agents or combination of agents. Such trials would specify not only the first-step treatments (initial random assignments) but also second and third-line treatments (possibly also randomly assigned). In the short-term nested studies, new agents would be compared with one another and a control using biologic markers and a combination of clinical endpoints. Different strategies (stepped-care approaches) would be compared over longer time periods using survival. This approach has the potential for both quantifying the relative short-term efficacy of specific drugs or combinations and providing data to develop better practice guidelines for using antiretrovirals over long time periods.