Beyond expanded access, clinical trials for gene therapies often present unique challenges. Conventional clinical trials for gene therapies, Dr. Byrne explains, often are designed to have an initial phase I/II dose escalation study with 2-3 cohorts, followed by dose selection and a phase IIb confirmatory study. In contrast with more conventional drugs, the phase IIb study for gene therapies is often designed as a delayed start placebo-controlled study. This means that participants who initially receive the placebo will later cross over and receive the treatment. For the types of serious diseases that many of these products are intended to treat, there are risks to this design. For example, patients may seroconvert or become ineligible due to the loss of functional capacities that are required to participate in the study. However, there are a variety of reasons to choose this design rather than a concurrent placebo-controlled study, like small patient populations and ethical considerations. Regardless, depending on the characteristics of the disease, the population, or another consideration, alternative controls may be considered.
The small size of many of the patient populations that gene therapies are designed to treat often presents problems for clinical study design, recruitment, and analysis. Demonstrating efficacy or a change from controls is difficult. Very recently, the FDA has signaled they might grant accelerated approval for a surrogate endpoint that could be expected to predict favorable clinical outcomes; however, these outcomes tend to be messy. Alternative clinical trial designs could address some of these challenges. For example, Dr. Byrne suggests that adaptive designs, which allow changes to the design based on accumulating data from the trial might better leverage small patient populations. FDA guidance does address such adaptive designs. These designs can allow access to treatment for populations that are so small that the evidence cannot reach the threshold for marketing approval.
Scientific concerns have also limited potential clinical trial design. For example, there is a belief that once a patient has received an AAV gene therapy, the presence of neutralizing antibodies would preclude ever dosing that patient again. However, significant progress has been made in this regard, and it is now possible in some cases to dose patients repeatedly, allowing sponsors to perform within subject dose escalation. As we accumulate knowledge and listen to patient input, clinical study designs will likely continue to evolve.