Author: Stella Flampouri, Heng Li, Sarah Quirk, Timothy Ritter, Mihaela Rosu-Bubulac, Toni M. Roth, Michael B. Roumeliotis, Koren Smith, Wade P. Smith 👨🔬
Affiliation: Johns Hopkins University, Virginia Commonwealth University, University of Washington, VCU Health System, UMass Chan Medical School/IROC RI, Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins University, Emory University, Brigham and Women's Hospital; Harvard Medical School, Memorial Sloan Kettering Cancer Center 🌍
Purpose: To assess clinical trial reproducibility for translation to standard clinical care in radiation therapy.
Methods: The systematic review adheres to the Preferred Reported Items for Systematic Reviews and Meta-Analyses protocol (PRISMA). For this study, the review consisted of a template produced by the study authors created to assess trials for radiotherapy reproducibility. The domains of evaluation included prescription and target dosimetry, normal tissue dosimetry, volume definition, treatment planning and delivery, post-trial reporting, and data sharing reporting for a total of 47 categories. The clinical trials included for analysis were arbitrarily selected among available CCTG, TROG and NCTN studies closed to accrual between 2016 and 2019. For each trial meeting these criteria, a PUBMED search was conducted to identify peer-reviewed manuscripts, conference proceedings, or conference submissions (i.e. abstracts). To be considered eligible for review, the report must meet three criteria: directly reference the trial name, trial identifier, or NCT number; and include the listed principal investigator as an author. Each trial was reviewed independently by three study authors, where any discrepancy in interpretation was resolved through directed discussion. Analysis included a summary of compliance to the reproducibility template.
Results:
A total of 13 trials were included in the analysis. In each of the domains, consistent reporting in prescription and target dosimetry was 59.8% (9 categories), normal tissue dosimetry was 63.5% (4 categories), volume definition was 58.2% (7 categories), treatment planning and delivery was 44.4% (13 categories), post-trial dosimetry reporting was 0.3% (5 categories), and data sharing was 73.8% (5 categories). Post-trial dosimetry reporting demonstrates a substantial deficit in translating knowledge that is required to emulate clinical trials in a standard clinical setting.
Conclusion:
In many domains of trial reporting, the evaluation for reproducibility is low (less than 60%). Post-trial dosimetry reporting was identified as a significant domain of concern.