Funding source, conflict of interest and positive conclusions in neuro-oncology clinical trials

We aimed to test any association between authors' conclusions and self-reported COI or funding sources in central nervous system (CNS) studies. A review was performed for CNS malignancy clinical trials published in the last 5 years. Two investigators independently classified study conclusions according to authors' endorsement of the experimental therapy. Statistical models were used to test for associations between positive conclusions and trials characteristics. From February 2010 to February 2015, 1256 articles were retrieved; 319 were considered eligible trials. Positive conclusions were reported in 56.8% of trials with industry-only, 55.6% with academia-only, 44.1% with academia and industry, 77.8% with none, and 76.4% with not described funding source (p = 0.011). Positive conclusions were reported in 60.4% of trials with unrelated COI, 60% with related COI, and 60% with no COI reported (p = 0.997). Factors that were significantly associated with the presence of positive conclusion included trials design (phase 1) [OR 11.64 (95 CI 4.66-29.09), p < 0.001], geographic location (outside North America or Europe) [OR 1.96 (95 CI 1.05-3.79), P = 0.025], primary outcomes (non-overall or progression free survival) [OR 3.74 (95 CI 2.27-6.18), p < 0.001], and failure to disclose funding source [OR 2.45 (95 CI 1.22-5.22), p = 0.011]. In a multivariable regression model, all these factors remained significantly associated with trial's positive conclusion. Funding source and self-reported COI did not appear to influence the CNS trials conclusion. Funding source information and COI disclosure were under-reported in 14.1 and 17.2% of the CNS trials. Continued efforts are needed to increase rates of both COI and funding source reporting.