Characteristics of stereotactic radiotherapy clinical trials registered on ClinicalTrials.gov

BACKGROUND: We aim to analyze the characteristics of stereotactic radiotherapy trials registered on ClinicalTrials.gov, and to compare completed and stopped early trials to identify predictors of trial failure. METHODS: All interventional stereotactic radiotherapy trials registered on ClinicalTrials.gov before Dec 31, 2019 were downloaded. Trial characteristics over time and between different regions were compared by Chi-square test. Binary logistic regression was used to explore characteristics associated with trials stopped early. RESULTS: A total of 760 trials were included. A higher proportion of trials were about lung cancer (20.4%), prostate cancer (14.7%) and central nervous system (14.6%). Most trials were phase1 and/or 2 trials (63.0%), single group (62.4%), nonrandomized (71.7%), open-label (95.0%) and single center (75.8%). The median sample size was 40 (0–1,716). Only 15.1% and 13.5% were funded by industry and National Institutes of Health (NIH), respectively. 15.4% stopped early with status includes “suspended”, “terminated” and “withdrawn”. Of the 113 “completed” trials, only 28 were published on PubMed. Compared with 2010 to 2014, trials from 2015 to 2019 were more likely to be randomized (20.0% vs. 34.4%; P=0.001), with 2 study arms (27.1% vs. 42.1%; P=0.002), industry-funded (11.0% vs. 19.1%; P=0.028) and conducted in Asia (7.6% to 15.8%; P=0.002). Trials from North America were more oriented toward phase 1 research (24.4% vs. 6.1% for Europe and 6.5% for Asia, P<0.001), nonrandomized (77.7% vs. 56.8% for Europe and 64.1% for Asia, P<0.001). Trials from Asia were more likely to have recruiting status (56.5% vs. 45.5% for Europe and 43.6% for North America, P<0.001). Multivariate regression analysis showed that randomized (OR 8.090, P=0.001), and enrollment patients ≤50 (OR 3.813, P<0.001) were associated with trials stopped early. CONCLUSIONS: Stereotactic radiotherapy trials are predominantly early-phase, small, single arm, nonrandomized and open label. Trials with randomized allocation or enrollment patients ≤50 were more likely to stop early.