Abstract
IMPORTANCE: Research biopsies are frequently incorporated within clinical trials in oncology and are often a mandatory requirement for trial enrollment. However, limited information is available regarding the extent and completeness of research biopsy reporting. OBJECTIVES: To determine the rate of research biopsy reporting for clinical trials registered in ClinicalTrials.gov and determine the clinical trial factors that correlated with research biopsy reporting. DESIGN, SETTING, AND PARTICIPANTS: ClinicalTrials.gov (CTG) was searched for all oncologic therapeutic clinical trials with completion dates between January 1, 2000, and January 1, 2015, with end point category terms including biopsy, biopsies, or tissue. The date of the final publication search was March 12, 2018. Trials conducting only diagnostic biopsies or trials using bone marrow biopsies or liquid biopsies were excluded. Credit for biopsy reporting was given for any mention of performing or results from tissue biopsies in publications. Clinical trials were compared with the highest level of corresponding publication or registry report. Fisher exact test was used for analysis. RESULTS: A total of 301 clinical trials were identified, with a median of 37 patients (range, 1-1310 patients) enrolled per trial. After a median follow-up time of 5.8 years from trial completion, 244 of 301 trials (81.1%) reported results: publications in 195 (64.8%) and CTG registry in 49 (16.3%). Reporting of trial results was associated with later-stage trials (phase 2/3) (137 of 153 [89.5%] for phase 2/3 vs 107 of 148 [72.3%] for phase 1 or 1/2 trials; P < .001). Results from research biopsies were reported in 153 of 301 (50.8%) trials or in 153 of 244 (62.7%) trials with published results. Rates varied by type of presentation: 142 of 195 publications (72.8%) vs 11 of 49 CTG reports (22.4%) (P < .001). Conducting mandatory biopsies (82.1% [101 of 123] vs 43.0% [52 of 121]; P < .001), early-phase clinical trials (70.1% [75 of 107] vs 56.9% [78 of 137]; P = .03), and listing the biopsy as a primary objective in CTG (76.3% [45 of 59] vs 58.4% [108 of 185]; P = .01) was associated with improved biopsy reporting. Trials that met their primary end point (71.9% [115 of 160] vs 45.2% [38 of 84]; P < .001) and those published in higher-impact journals (81.1% [77 of 95] vs 65.0% [65 of 100]; P = .01) had improved biopsy reporting. Mandatory biopsies and biopsy reporting increased over time with similar slopes (P = .58). CONCLUSIONS AND RELEVANCE: Despite ethical obligations to report research biopsies, only 50.8% of all trials that included a research biopsy-related end point in CTG reported on these biopsy-related results. Improved efforts are needed to report results obtained from research biopsies.