Parallel anticancer drug development and molecular stratification to qualify predictive biomarkers: dealing with obstacles hindering progress

Abstract

Current anticancer drug development still largely follows the classic designs developed for chemotherapeutic agents over the past 4 to 5 decades, remaining slow, costly, and inefficient, with continuing high risks of costly late drug attrition. A Pharmacologic Audit Trail has been described to decrease these risks, incorporating pharmacokinetic, pharmacodynamic, intermediate efficacy endpoints, as well as patient stratification molecular biomarkers. Molecular biomarker-based patient selection in hypothesis-testing early clinical trials is critical to clinically qualify putative predictive biomarkers for rationally designed, molecularly targeted drugs as early as possible. Nevertheless, major concerns have been raised about the impact of using such biomarkers in early trials, in view of the costs and time involved to develop multiple certified assays for clinical use. The rapid evolution of novel technologies of utility to this field, such as next-generation sequencing and circulating tumor-cell isolation, makes these valid concerns of critical importance. We therefore propose a more efficient parallel predictive biomarker and clinical anticancer drug development process to deal with the obstacles hindering progress.