1365PD - Do contemporary randomized controlled trials meet ESMO thresholds for clinically meaningful benefit?

Date 08 October 2016
Event ESMO 2016 Congress
Session Public health and health economics
Topics Bioethics, Legal, and Economic Issues
Presenter Joseph Del Paggio
Citation Annals of Oncology (2016) 27 (6): 474-482. 10.1093/annonc/mdw387
Authors J.C. Del Paggio1, B. Azariah2, R. Sullivan3, W. Hopman1, F.V. James2, S. Roshni2, I. Tannock4, C. Booth1
  • 1Medicine, NCIC Clinical Trials Group, Cancer Research Institute, K7L 3N6 - Kingston/CA
  • 2Clinical Oncology, Regional Cancer Centre Thiruvananthapuram/Trivandrum, Thiruvananthapuram (Trivandrum)/IN
  • 3Cancer Epidemiology & Population Health, King's College Hospital, London/GB
  • 4Medical Oncology, Princess Margaret Hospital, Toronto/CA

Abstract

Background

ESMO has developed a framework for evaluating the magnitude of clinical benefit (ESMO-MCBS) of new cancer therapies. We evaluate the extent to which contemporary randomized controlled trials (RCTs) are designed to detect differences in outcome that meet the proposed ESMO thresholds for clinically meaningful benefit (CMB).

Methods

All RCTs evaluating systemic therapy for breast, non-small cell (NSCLC), colorectal (CRC), and pancreas cancer published 2011-2015 were reviewed. Two authors abstracted data regarding trial characteristics and applied the ESMO-MCBS to study results. Data from the statistical methods section were used to determine if the RCT was powered to detect an effect that would meet CMB.

Results

277 eligible RCTs were included (40% breast, 31% NSCLC, 22% CRC, 6% pancreas). Median sample size was 532 and 83% were funded by industry. Among the 225 RCTs in which an ESMO design score could be assigned, 69% were powered to detect an effect size that would not meet the threshold for CMB. Factors associated with being powered for small effect size included disease (79% lung, 71% GI, 61% breast, p = 0.038), treatment intent (82% palliative, 37% curative, p 

Conclusions

Less than one-third of RCTs with statistically significant results meet ESMO thresholds for clinically meaningful benefit, and this represents only 15% of all published trials. Investigators and funding agencies should adopt more stringent thresholds for meaningful benefit in the design of future RCTs.

Clinical trial identification

Legal entity responsible for the study

Queen's University

Funding

Kingston General Hospital

Disclosure

All authors have declared no conflicts of interest.