Endpoints in lung cancer trials: today’s challenges for clinical statistics

Abstract

The identification of potential molecular targets in lung cancer has stimulated the today’s search for antitumor agents to be tested in clinical trials with an appropriate endpoint. Three main categories of classical endpoints are applied: survival time endpoints, symptom endpoints, and endpoints relying on patients’ reporting, with the gold standard overall survival (OS). Efforts have been taken to substitute OS by surrogates. As a surrogate for OS, progression-free survival (PFS) should have the inherent considerable advantage, that it can detect subpopulations with longer PFS intervals early. OS and PFS was linked directly by splitting OS into PFS and the rest as survival time past progression (SPP). Variation in SPP is a quantitative factor in validating PFS as a surrogate of OS. If accompanied by some independent measures like quality of life (QoL) or treatment toxicity, PFS should be able to cover the clinical benefit achieved by treatment. In an adaptive trial design, the impact of SPP can be modeled by adjustments of the covariate structure by factors as QoL. OS as the gold standard is easy to measure and is precise but results are available late. PFS in contrast is complex to measure but will become attractive, because its results are available earlier and are not influenced by subsequent therapies. Therefore, PFS, as an endpoint with some extra measures, has become an acceptable alternative to OS in lung cancer trials recently. Extra measures under discussion to enrich PFS are QoL and information of SPP, or adaptive designs.