Screen failure and treatment dropout in oncology clinical trials: A multicenter analysis from Türki̇ye.

Abstract

e23009 Background: Screen failure and treatment dropout remain major barriers to successful conduct of oncology clinical trials, particularly in real-world, middle-income settings. Data evaluating both patient- and center-related factors influencing screen failure, dropout, and time to dropout are limited. Methods: This multicenter retrospective analysis included 1,229 oncology patients screened for clinical trials across 13 centers in Türkiye. Causes of screen failure and dropout were descriptively analyzed. Time to dropout was assessed using median values and compared across subgroups using the Mann–Whitney U and Kruskal–Wallis tests. Results: Overall, 1,155 patients experienced screen failure, and 74 patients dropped out after trial initiation. Patients with screen failure had a median age of 64 years, and 67.3% were male. The most frequent cancer types among screen failure patients were lung (57.0%), breast (13.7%), and prostate cancer (8.6%). The primary causes of screen failure in the overall cohort were non-fulfillment of inclusion criteria (69.4%), presence of exclusion criteria (21.4%), and patient non-compliance (4.1%). Among the 74 patients who dropped out, 63.5% were male, with a median age of 62 years; the most common diagnoses were lung (44.6%), gastric (27.0%), and breast cancer (23.0%). The leading causes of dropout were adverse events (33.8%), disease progression (28.4%), and patient dissatisfaction (14.9%). Median time (months) to dropout was significantly shorter in female patients (3 vs 7, p = 0.014), patients without comorbidities (7 vs 8, p = 0.036), those treated in neoadjuvant or adjuvant compared with metastatic settings (7.5 vs 6.5 vs 8, p < 0.001), patients receiving parenteral compared with oral or oral plus parenteral regimens (7 vs 9.5 vs 13, p = 0.035), and in centers with a clinical research unit or a clinical research center compared with centers without an administratively organized unit (7 vs 6.5 vs 10, p = 0.010). Finally, time to dropout decreased with increasing sub-investigator research experience, defined as the total number of clinical trials in which the sub-investigator had previously participated (0–5, 6–10 and > 10 trials, respectively; median 9.5 vs 7 vs 5 months; p = 0.022). Principal investigator experience and total trial volume per center were not associated with time to dropout. Conclusions: In this multicenter study, screen failure in clinical trials was predominantly driven by eligibility criteria, while treatment dropout was mainly related to toxicity and disease progression. Time to dropout varied according to patient characteristics, treatment setting, route of administration, and center-level factors, underscoring the influence of both clinical and operational elements on trial retention. These findings highlight the importance of pragmatic trial design and strengthened research infrastructure to improve feasibility and retention.