Abstract Body

In general, active-controlled noninferiority (NI) trials are considered when superiority trials, to an active control or placebo, are not possible due to ethical or other considerations. NI trials share some of the same biases as historically controlled trials because they rely on information external to the clinical trial. Food and Drug Administration (FDA) guidance states that NI designs are credible and appropriate only in situations in which the active control has shown a consistent effect (generally compared to placebo) in prior superiority trials conducted in a patient population similar to the population in the clinical investigation being planned. This is called the constancy assumption and allows for assay sensitivity in an NI trial. NI is met if the new intervention is 'not unacceptably worse' than the active control by a specified amount, the NI margin. The NI margin should be no larger than the effect the active control had in previous trials. Unless a placebo group is also included, NI trials depend on the assumption that the active control had its expected effect in the trial. From a regulatory perspective knowing the active control had its expected effect is necessary to ensure that a trial that concludes NI has identified a treatment that is superior to placebo. HIV treatment trials have successfully used NI trials for antiretroviral (ARV) drug development for many years; however, quantifying the treatment effect of each component of an ARV regimen has been challenging as drug regimens evolve, which can have consequences when designing an NI trial.  HIV prevention research also illustrates the limitations of NI trial designs. Although collective data show that emtricitabine/tenofovir disoproxil fumarate (FTC/TDF) can be highly efficacious at preventing HIV infection when taken as prescribed in uninfected individuals, the prophylactic effect has been highly variable over time and by population.  Two trials that included FTC/TDF arms in cisgender women in Africa showed a lack of pre-exposure prophylaxis (PrEP) efficacy due to poor adherence. The lack of a consistent PrEP effect across trials in all populations invalidates a constancy assumption of FTC/TDF as an active control in an NI trial for some populations. Other types of trial designs using external controls might be more credible and appropriate than NI trials and are currently being explored.