Background: Common endpoints of many cancer clinical trials are overall survival, progression-free survival or relapse. Despite their given importance, these measures do not account for quality of life (QOL). The Q-TWiST (Quality adjusted Time Without Symptoms and Toxicity) method has been proposed by Gelber in the 1990s to enable a statistical comparison between two treatment groups by incorporating both benefit (survival) and risk (toxicity, QOL). Q-TWiST partitions lifespan of patients from beginning of treatment until death into three time segments corresponding to three distinct health states: toxicity (TOX), TWiST (Time Without Symptoms of Treatment or Diseases) and relapse (REP). Aim: The purpose of this systematic review was to evaluate the utilizations of Q-TWiST method in cancer clinical trials and how these results were translated to clinical practice in oncology field. Methods: We searched PubMed, Embase, Cochrane library and the bibliographies of relevant articles to locate additional publications in Jan 2018. Two evaluators independently reviewed and selected eligible studies based on keyword “Q-TWiST” and predetermined selection criteria. Clinical trials in oncology field quantifying Q-TWiST in the analysis were included. We focused how health states were defined and utility coefficients were determined. Other relevant contents were collected and summarized. Results: Out of 79 studies, 36 oncology clinical trials utilizing Q-TWiST method were identified. We included 20 studies published within last decade (2008-2018) in final review. Trials ranged from 182 to 1110 subjects. Breast cancer was most common cancer type. All studies were post hoc analysis which used data from previous publications and applied Q-TWiST in analysis. Methods in determining QOL, also called utility coefficients, varied among included trials. Among 20 studies, six directly measured QOL in their trials and used it for Q-TWiST method, two obtained QOL data from other studies with similar study participants. Most commonly used questionnaires were EORTC QLQ-C30 and EuroQOL EQ-5D. Utility coefficients were arbitrarily set in 12 studies. Most common values were 0.5, 1 and 0.5 for TOX, TWiST and REP respectively. Definition of these health states, especially toxicity, were consistent among included studies. Few studies discussed both statistical and clinical significance of their findings. Conclusion: Q-TWiST method has been increasingly used as a post hoc analysis in many oncology clinical trials within last decade. However, some current limitations should be compromised to make Q-TWiST become a clinically useful method for physicians and patients. We suggest that definition of health states must be standardized, methods to determine utility coefficients must be validated and universal in the future analysis.
Older adult participation in cancer clinical trials: A systematic review of barriers and interventions