The selection of endpoints is critical to the interpretation of clinical trials results. In early phase clinical trials investigators seek evidence of activity, then predictors of clinical benefit, and in most randomized trials, accepted surrogates of meaningful benefit. At one extreme we may seek improvements in overall survival while in another we may be satisfied simply with evidence of tumor stabilization or changes in circulating biomarkers. In between these extremes are intermediate and composite endpoints including “partial responses”, “clinical benefit rate”, and those that are time-dependent endpoints such as disease-free or progression free survival. Many of the discussions and disagreements we hear in the context of drug development, benefit, and approval can be linked to a fundamental lack of agreement on the value and meaning of many of these endpoints. Activity of a given novel agent may, or may not, persuade some investigators, clinicians, and patients, that it is a worthwhile addition to our treatment options. On the other hand, a small but statistically significant improvement in overall survival in advanced cancer may, or may not, convince this same audience that the tested treatment is worthy of approval and use. Again, the intermediate endpoints may be unreliable or only weak predictors of more profound benefits. We will review and explore these issues considering not only the endpoints themselves but also their meaning in specific clinical settings and the impact of endpoint selection on further drug development.
Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr ES3-3.