Using pre-clinical information to establish a safe dose in first-in-men studies

Before a novel compound is evaluated in a series of clinical trials, a number of preclinical, in-vitro and in-vivo studies are typically undertaken. A crucial step in the transition between pre-clinical and clinical studies is to extrapolate a dose that has been established as safe (and efficacious) in animal studies to a safe dose to be used in first-in-men studies. This is usually done using allometric scaling, a process that adjusts the dose found in a particular animal species to other species and ultimately humans based on the typical weight. This process is, however, often poorly understood and often applied incorrectly.

The project will begin by investigating current practice when determining the starting dose for first-in-men studies and a review of existing methods for dose extrapolation. A model which dynamically updates the extrapolated best dose based on accumulating data in the first-in-men study such that the weight of the preclinical evidence is gradually reduced for direct information in humans is then developed. A Bayesian framework that synthesizes the pre-clinical evidence which is subsequently updated with observations in the study in humans will be utilised.