Beyond the Proportional Hazards Model: Analysis of Treatment-effects on Time-to-Event Endpoints in Cancer Development
26 August 2020 | 11:00 UTC / GMT
Time Zone Converter
In cancer clinical trials, time-to-event endpoints, such as overall survival or progression-free survival, are used widely as primary endpoints and are of great interest. Currently, treatment effects on such endpoints are almost exclusively summarized by using the hazard ratio (HR) estimated based on a proportional hazards (PH) model. However, the PH assumption is quite restrictive. Moreover, there is an increasing body of evidence that it may not be tenable for, e.g., cancer immunotherapy. For some specific cancer types, a violation of the PH assumption can be due to a fraction of long-term survivors. In the latter a situation, the proportion of “cured” patients becomes a crucial component of the assessment of patient benefit, and being able to distinguish a curative effect from a life-prolonging one conveys important additional information in the evaluation of a new treatment.
For all these reasons, there is an increasing need and interest in developing and applying alternative metrics and inference tools to assess the benefit-risk profile more efficiently and in a timely fashion. In the session, we present alternatives to the HR and PH model that are gaining in relevance for evaluation of treatment effects in oncology. In particular, we focus on the restricted mean survival‑time (RMST), semiparametric accelerated failure‑time (AFT) model, and cure_fraction (CF) models. For all these methods, important methodological developments have taken place in the recent years. We review the developments, illustrate the application of the methods, and discuss their merits and limitations. While we focus on oncology as the application domain, the methodological content of the session should be accessible to any applied statistician.
Geert molenberghs, Catholic University of Leuven & Hasselt University
Chen Hu, Johns Hopkins University
Utility of restricted mean survival time in oncology clinical trials
Tomasz Burzykowski, Hasselt University & International Drug Development Institute (IDDI)
Time for a broader use of accelerated failure-time models in cancel trials
Catherine Legrand, Catholic University of Louvain
To use a cure model or not, is that the question?