Overcoming the "Valley of Death" in Medications Development for Alcohol Use Disorder.

As the development of novel pharmacotherapies for alcohol use disorder (AUD) has been slow, the discovery and testing of more efficacious pharmacotherapies for AUD represent a high priority research area. In fact, the transition from preclinical to clinical testing of novel compounds has been termed the "valley of death" in medications development. One key obstacle consists of the lack of an articulated set of goals for each stage of medications development. Specifically, the knowledge outputs required to make the transition from safety testing, to early efficacy detection, to confirming clinical efficacy remain unclear, and this is despite a great deal of interest and substantial financial investment in developing novel therapeutics for AUD. This qualitative critical review seeks to draw parallels and lessons from the well-established stage model for behavioral therapies research with alcohol and other substance use disorders and to apply these insights into AUD pharmacotherapy development. We argue that human laboratory models and/or pilot randomized controlled trials should serve as intermediaries in the transition from preclinical studies to large, and costly, randomized controlled efficacy trials. The relative strengths and weaknesses of pilot clinical trials versus human laboratory studies for bridging the "valley of death" are discussed and explored via a Monte Carlo data simulation study. Multiple permutations of suitable research designs informed by the behavioral therapies development model are discussed with the overall goal of promoting consilience and maximizing efficiency across all phases of clinical testing of novel AUD pharmacotherapies.