Assessing patient withdrawal in cancer clinical trials: A systematic evaluation of reasons and transparency in reporting.

Abstract

e23003 Background: Patient withdrawal in oncology clinical trials ranges from 5% to 45%, posing significant threats to trial integrity. High attrition rates can lead to underpowered studies, selection bias, diminished generalizability, and wasted resources. Despite the importance of mitigating withdrawal, the reasons and extent of withdrawal are often poorly documented. The lack of standardized practices for documenting withdrawal reasons, coupled with infrequent use of written consent withdrawal forms, introduces ambiguity in interpreting patient intentions and limits on further participation (Cleland et al. 2004). To the best of our knowledge, this is the first study to systematically assess trial-reported reasons for withdrawal in cancer clinical trials, aiming to improve trial design and data reliability. Methods: We reviewed all therapeutic phase III oncology clinical trials conducted between August 20th, 2014, and August 20th, 2024, from the ClinicalTrials.gov database. Supportive care trials and trials without a consort diagram or descriptive equivalent were excluded. Withdrawal without defined reason was considered a patient or investigator-initiated withdrawal without specified cause (e.g., noncompliance, patient request, loss to follow up). Results: 300 phase III therapeutic oncology trials with 165,674 patients across 14 primary cancer types and a median follow-up of 35.6 months (25th–75th percentile: 4.4–66.8 months) were included, while 40 trials (11.8%) were excluded due to unavailable attrition data. Included trials were predominantly unblinded (185 trials, 61.6%) and randomized (263 trials, 87.7%) designs, with control arms utilizing placebo treatment alone (92 trials, 30.7%) or standard care (173 trials, 57.7%) therapies. Documented reasons for patient withdrawal are recorded in the Table. Withdrawal without a defined reason ranged from 0-64.4% per trial (25th-75th percentile: 4.5-10.5%). Of the patients who withdrew without a defined reason, 60.1% (N = 10,638) were patient-initiated, 25.7% (N = 4,435) were investigator-initiated, and 10.0% (N = 1,769) were labeled as lost to follow-up. Conclusions: 14.9% of patient discontinuations in phase III oncology trials were initiated by patients or investigators without defined reasons. To enhance data integrity and transparency, future efforts should prioritize key stakeholders developing and requiring standardized written withdrawal of consent documentation to more clearly assess the reason and extent of withdrawal. Reasons for trial discontinuation. Reason # of patients (% of trial discontinuations) Progression of disease 41,656 (36.8%) Death 30,620 (27.1%) Withdrawal without defined reason 16,842 (14.9%) Treatment-related adverse effects 10,742 (9.5%) Sponsor termination 5,420 (4.8%) Other 4,008 (3.5%) Ineligibility 2,952 (2.6%) Max clinical benefit 879 (0.8%)