Estimating and communicating survival times to patients with incurable cancer who are treated with targeted therapies

Abstract

The aim of this thesis was to help oncologists estimate and communicate survival times to patients with incurable cancer receiving targeted therapies. For patients with incurable cancer receiving chemotherapy, simple multiples of an overall survival (OS) curve’s median accurately estimate its percentiles corresponding to worst case (90th percentile, 1/4 x median OS), typical (25th to 75th percentile, 1/2 to 2 x median OS) and best case scenarios (10th percentile, >3 x median OS) for survival. We aimed to determine if this principle can be applied to patients with incurable cancer receiving targeted therapies. In both clinical trial participants and non-trial patients with HER2 positive metastatic breast cancer (MBC) we demonstrated that the same simple multiples accurately estimated worst case and typical scenarios. Follow up duration was insufficient to quantify the best case scenario. We determined the accuracy of oncologists’ estimates of expected survival time (EST) in patients receiving the targeted therapy regorafenib, using data from a randomised trial. 66 oncologists estimated the EST for each of 152 patients. Only one third of oncologists’ estimates of EST were precise. Simple multiples of the oncologists’ estimates were accurate for calculating scenarios for survival with 9% of patients living ≤1/4 of their EST, 57% living between half to double their EST, and 12% living ≥3 times their EST. Our survey examined medical oncologists’ practices of prognosis communication. It showed that most oncologists reported providing quantitative estimates of EST to patients. Oncologists reported difficulty estimating survival time was a significant barrier to discussing prognosis. This thesis provides oncologists with a framework for estimating and explaining survival times to patients with incurable cancer treated with targeted therapies. It aims to improve patient understanding of prognosis thereby leading to better treatment decisions and end of life outcomes.