Future Developments: Innovative Trial Design

The phase 1 followed by phase 2 followed by phase 3 trial approach to new drug development in acute myeloid leukemia (AML) has been followed for many years. However, although the purpose of a phase 1 trial is often said to determine the maximum tolerated dose, patients enter phase 1 trials in the hope of response, which is not considered in phase 1. Likewise phase 2 trials focus on a single outcome, although several outcomes may be of clinical interest. The decision to move from a single-arm phase 2 study to phase 3 inherently involves comparison of the new therapy with an older one, but the effects of important prognostic covariates on this comparison are often omitted from this comparison. Although randomized trials are essential, phase 3 trials can be problematic if the control arm is very likely to be unsatisfactory, such that patients would be reluctant to be randomized to it. Below, the author describes several new designs to address these problems.