Cure models for the analysis of time-to-event data in cancer studies

Abstract

In settings when it is biologically plausible that some patients are cured after definitive treatment, cure models present an alternative to conventional survival analysis. Cure models can inform on the group of patients cured, by estimating the probability of cure, and identifying factors that influence it; while simultaneously focusing on time to recurrence and associated factors for the remaining patients.

Keywords: cure models; soft tissue sarcoma; survival analysis.