Prevalence and characteristics of prematurely terminated cancer clinical trials.

Abstract

6613 Background: Cancer clinical trials require a substantial investment of patient and financial resources. Trials that are terminated prematurely contribute little to the scientific knowledge base and divert resources from answering other critical questions. To our knowledge, the prevalence of prematurely terminated trials, and reasons for termination, has not been comprehensively assessed. Methods: The ClinicalTrials.gov database was queried to identify all phase II and phase III interventional trials for any neoplasm registered between 9/2005 and 11/2012. The disease type, enrollment status, phase, funding source, and number of study sites were collected for each trial. A ”prematurely terminated” trial was any trial with a status of “terminated,” “withdrawn,” or “suspended.” The reason for termination was collected for each prematurely terminated trial. Results: We identified 11,483 interventional cancer trials, of which 1,326 (11.5%) were terminated prematurely. The reasons for termination are shown in the Table. The proportion of cancelled trials was similar among the various disease types, but was slightly higher in lung cancer (14.4% vs. 11.2%, p = 0.001), and slightly lower in skin cancer (8.2% vs. 11.7%, p = 0.022). The most common reason for premature termination was poor accrual (33.1%) though no reason was provided for 23% of trials. Trials sponsored by industry were more likely to be canceled than those with NIH or “other” sponsorship (13.4% vs. 9.9%, p = 0.001). Single center trials were more likely to be canceled than trials with multiple sites (13.6% vs. 9.4%, p < 0.001). Conclusions: More than one in ten cancer clinical trials is terminated prematurely. Full reporting of the reasons for premature termination is necessary for further evaluation of the cancer clinical trial enterprise and identification of predictors of study early closure. [Table: see text]