Non-Publication of Interventional Studies for Uveal Melanoma

Introduction Evidence-based medicine relies on publicly available research findings, and study non-publication may predispose to incomplete evidence on intervention efficacy or redundant funding for re-attempting unpublished studies. This study investigated non-publication rates among United States (US)-registered interventional studies for uveal melanoma. Methods In April 2025, ClinicalTrials.gov was searched for interventional uveal melanoma studies. Interventional studies were included if completed at least three years prior to the database search, excluding completed trials awaiting results publication. Studies were labeled as having publicly available results if data were posted through ClinicalTrials.gov or journal publication. For published studies, we determined whether their primary study endpoint result was statistically significant. For all records, we collected study characteristics, including phase, sponsor, country of study, year of study initiation, enrollment, melanoma type, treatment type, and randomization status. The primary outcome was the overall rate of non-publication among all records. Secondarily, we provided descriptive characteristics of included records and compared rates of non-publication between sublevels of primary endpoint significance status and study characteristics. Where applicable, the Fisher-exact, Wilcox rank-sum, and binomial exact tests analyzed these comparisons. Analysis was completed in RStudio (v2022.02+433). Results From 102 records, we identified 98 trials (N=4669) for final inclusion. Overall, 37.76% of records did not have identifiable online publication on any platform. Whether published or unpublished, 22.95% and 16.22% of trials were terminated, respectively, and the reasons for termination were heterogeneous. Trials with industry sponsoring were significantly more often published than those without industry backing (p=0.024). No significant association was found between publication status and any one of study phase, enrollment size, study duration, or randomization status. Among the 60 studies published online, 63.93% had reported significant primary study endpoints. Given this, the calculated probability of publication to either ClinicalTrials.gov or journal publication was 65% [95% CI: 51.60–76.87%; p=0.03]. No significant association was found between significant positive primary outcome results and any one of randomization status or sponsor type. Conclusion About one-third of US-registered uveal melanoma interventional studies did not reach publication. Future studies should prioritize reporting transparency for results or reasons for non-publication.