• Mallorie H Fiero, Jessica K Roydhouse, Vishal Bhatnagar, Ting-Yu Chen, Bellinda L King-Kallimanis, Shenghui Tang, and Paul G Kluetz.
• Office of Biostatistics, Center for Drug Evaluation and Research, US Food and Drug Administration, Silver Spring, MD, USA. Electronic address: mallorie.fiero@fda.hhs.gov.
• Lancet Oncol. 2022 May 1; 23 (5): e229-e234.

AbstractTime-to-event endpoints for patient-reported outcomes, such as time to deterioration of symptoms or function, are frequently used in cancer clinical trials. Although time-to-deterioration endpoints might seem familiar to cancer researchers for being similar to survival or disease-progression endpoints, there are unique considerations associated with their use. The complexity of time-to-deterioration endpoints should be weighed against the information that they add to the tumour, survival, and safety data used to inform the risks and benefits of an investigational drug. Here we use the estimand framework to show how analytical decisions answer different clinical questions of interest, some of which might be uninformative. Challenges including the consideration of intercurrent events, the difficulty in maintaining adequate completion rates, and considerable patient and trial burden from long-term, serial, patient-reported outcome measurements render time to deterioration a problematic approach for widespread use. For trials in which a comparative benefit in symptoms or function is an objective, an analysis at pre-specified relevant timepoints could be a better approach.Copyright © 2022 Elsevier Ltd. All rights reserved.

13 keywords...

hide…