Investigating documented reasons for withdrawal from clinical trials for patients with advanced prostate cancer (PCa).

Abstract

105 Background: In clinical trials, treatment efficacy rates are based on participants who can tolerate treatment through study endpoints. Attrition from clinical trials can potentially lead to biased results (PMID: 30379822). Many oncology clinical trials censor the subject data of participants because of early trial intervention discontinuation due to adverse events (AEs), withdrawal of consent, or loss to follow-up. Despite high attrition rates reported across various oncology clinical trials (10-38%), there is inconsistency regarding investigating the reasons for withdrawal (PMID: 23132290). We aim to identify the causes of the withdrawal of participants with advanced PCa from clinical trials. Methods: In this IRB-approved single-center retrospective study (Sept 2014 to Sept 2022), the OnCore Research Database was used to capture data on the participants with advanced PCa enrolled in clinical trials. The included trials were specific to prostate cancer, and mixed solid tumor trials were excluded. Endpoint reports from all included trials were assessed. Participants who discontinued treatment for any reason were included. The primary objective was to obtain reasons for clinical trial withdrawal before primary endpoints, identify potential clinical intervention areas, and determine potential areas of adjustment to future study designs to improve retention rates. Eligibility criteria: diagnosis of metastatic prostate cancer, receipt of protocol treatment until discontinuation, and documentation of a reason for discontinuation of the protocol treatment. Results: 272 pts from the 18 clinical trials met eligibility criteria and were included in the study. The most common reasons documented for discontinuation were: the progression of disease (46.69%), other reasons (15.44%), adverse events (5.15%), personal preference (15.81%), and completion of protocol treatment (11.76%) of the total number of enrolled participants. We noted a statistically significant difference between participants who withdrew for other reasons in the castration-resistant versus castration-sensitive setting (6.08% vs. 32.38%, p value=0.03). Notably, declining performance status and symptom progression were listed under the “other reasons” category. The documentation of patient withdrawal due to clinical symptoms was reported across multiple withdrawal categories. Conclusions: In approximately one-third of participants, the reason for protocol treatment discontinuation was not apparent (documented as patient preference and other reasons with no further specifications). Further exploration of the exact reasons contributing to therapy discontinuation in these patients will be critical to improving patient retention in clinical trials.