Regression modelling of overall survival and progression-free survival

Abstract

There are three endpoints commonly used in oncology clinical trials, which are known as overall survival (OS), time to progression (TTP) and progression-free survival (PFS). Recently, PFS has become an important alternative endpoint to OS. In this thesis, both exponential and Weibull distributions are used to investigate the joint model of OS and PFS. Regression modelling will be introduced to investigate the effect of a treatment indicator on the distribution parameters for OS, TTP, and PFS. Both simulated data and real data will be used to investigate and demonstrate methods. The parameters of the models will be estimated by the maximum likelihood estimation. Furthermore, Wald tests will be performed to investigate covariate effects.