Treatment-Free Survival and the Pattern of Follow-Up Treatments After Curative Prostate Cancer Treatment, a Real-World Analysis of Big Data from Electronic Health Records from a Tertiary Center.

Prospective trials provide robust evidence for prostate cancer (PCa) treatment but often include highly selective populations, limiting generalizability. Real-world data (RWD) can address these gaps and inform personalized care. : This study aimed to evaluate treatment-free survival (TFS) and secondary treatment sequences after initial curative therapy for PCa using electronic health record (EHR) data and to analyze associated medication profiles. : We studied 3024 patients treated with radical prostatectomy (RP), brachytherapy (BT), or curative radiotherapy (RT) at Erasmus MC (2009-2023), the Netherlands. Outcomes included TFS, treatment sequences, and medication patterns across treatment lines. : Median age at diagnosis was 65 years (IQR 61-69) for RP, 68 (62-73) for BT, and 72 (68-76) for RT. At 10 years, TFS was 89% (95% CI: 84.9-94.1) for BT, 85% (95% CI: 83-87) for RT, and 71% (95% CI: 65.7-75.8) for RP. Most patients remained treatment-free, but up to five treatment lines occurred, mainly in patients with low comorbidity scores. Medication profiles reflected treatment-related morbidity: alpha-blocker use increased after BT and RT, while bladder relaxants were common after RP. Comorbidity-related medication use remained low among patients undergoing multiple sequenced treatments. : These findings highlight the real-world application of multiple secondary treatments after different primary curative therapy options for PCa and associated comorbidity and medication use patterns. They confirm the durability and long-term effectiveness of curative treatments in real-world PCa care. By combining treatment trajectories and medication profiles, RWD provides insights for personalized counseling, helping clinicians and patients anticipate long-term treatment needs, and enabling informed decisions aligned with health status and preferences.