Use of placebo in oncology randomized controlled trials (RCT) is ethically controversial. Placebo may introduce bias, as toxicity profiles of treatment arms can inadvertently unblind subjects and investigators.
We investigated the use of placebo in urologic oncology RCTs, hypothesizing that most placebo-controlled trials are effectively unblinded, either explicitly with open-label design or implicitly due to large differences in adverse events (AE) or oncologic outcomes.
Urologic oncology RCTs utilizing placebo were identified via ClinicalTrials.gov. Interventional prostate, bladder/urothelial, and renal cancer trials from 2014 to 2024 were included. Subject incompletion, all-cause mortality, AE rates, and serious AE (SAE) rates were identified and compared between placebo and active arms using χ2 and Fisher's exact tests.
Sixty studies met inclusion criteria and included 66 placebo arms with 12,918 subjects and 81 active arms with 16,098 subjects. There was no significant difference in incompletion rates between placebo and active arms. Subjects enrolled in active arms reported statistically significant higher SAE and AE rates compared to those in placebo arms across the majority of physiological domains, including 18/24 domains for SAEs and 13/24 for AEs. This relationship persisted in sensitivity analyses where unblinded trials were excluded.
In urologic oncology placebo-controlled RCTs, active arms are associated with significantly higher rates of AEs and SAEs compared with placebo arms. These findings indicate a strong possibility that true blinding is not possible in oncology RCTs, even with optimal study design, and serve to better inform future clinical trial design and implementation challenges in employing placebo control.
Urology practice. 2024 Jun 26 [Epub]
Zijing Cheng, Timothy Campbell, Trevor C Hunt, Ashley Li, Karen Doersch, Jathin Bandari
Department of Urology, University of Rochester Medical Center, Rochester, New York.
PubMed http://www.ncbi.nlm.nih.gov/pubmed/39196661