Strategies for optimizing combinations of molecularly targeted anticancer agents

Key Points The use of drug combinations to circumvent tumour resistance is a well-established principle of cancer therapy. There are now a large number of molecular targeted anticancer drugs available to test in clinical trials. Establishing strategies to choose from the almost limitless possibilities and focus on prioritizing the most promising combinations is therefore paramount. The variables to consider in the clinical development of combinations of targeted agents include: mechanisms of action and resistance of single agents and their mechanisms of interaction; rational and robust preclinical demonstration of additivity or preferably synergy for a combination; and a strategy for testing proof-of-principle in the clinical setting that reflects the preclinical models that showed activity. Combination strategies for targeted agents that have been proposed to date can be divided into three broad categories: combinations to maximize the inhibition of a specific target; to maximize inhibition of a pathway by targeting multiple components; and to expand inhibition of multiple cellular mechanisms. In general, strategies to rationally evaluate the activity of combination regimens should take into consideration both knowledge of the single agents — including mechanisms of action, clinical pharmacology, toxicity profile and antitumour activity — and non-clinical evaluation of the combination. To maximize translatability of preclinical study results, the proposed regimen should be tested in a variety of tumour models at clinically achievable doses/exposures, and, if possible, the target status and molecular context of the tumour models that correspond to treatment outcomes should be defined. Our current ability to move beyond empirical selection and evaluation of molecularly targeted agents and their combinations is often limited by inadequate knowledge of the biology and molecular features of specific tumours; inadequate understanding of the mechanisms of action/resistance for individual agents and their combinations, as well as inadequate tools to measure and compare treatment effects in laboratory models that can be predictably used in clinical development. To overcome these barriers, future studies should continue to focus on several essential tasks: a systematic approach to non-clinical studies of targeted agents and combinations; correlative studies in both preclinical and clinical settings; and the development of reliable assays or tests for clinical use f