Methodological aspects of lung cancer clinical trials in the era of targeted agents

Methodology of clinical trials conducted in lung cancer, similarly to other tumours, has been recently challenged by the particular characteristics of new targeted agents. Traditional methodology of phase II trials has been questioned, both for the choice of the endpoint and for the study design itself. Due to the mechanism of action of new drugs, cytostatic more than cytotoxic at least in principle, the usual endpoint of phase II trials, objective response rate, is now often replaced by alternative event-related endpoints, such as progression-free survival or progression-free rate at a fixed time-point. Randomized phase II trials, considered in the past the exception rather than the rule, have been encouraged, as the only design useful to give clear information on the activity of experimental treatments. Conduction of phase III trials remains mandatory to demonstrate treatment efficacy, but their endpoints and design are currently object of discussion. With targeted agents, great efforts have been made to identify predictive factors of treatment efficacy, but this aspect appears to be more complicated than hypothesized in principle. The history of clinical trials with epidermal growth factor receptor inhibitors in advanced NSCLC is a good example of the uncertainty about predictive factors and selection criteria. Moreover, non-inferiority design has been used for several phase III trials comparing targeted agents with chemotherapy. In this review, recent aspects of clinical trials methodology in lung cancer are described, and examples of their application are discussed.