Radium-223 therapy in metastatic castration-resistant prostate cancer: baseline predictors of treatment completion and survival in a reimbursement-regulated Taiwanese cohort.

objectiveReal-world outcomes of Ra-223 in Asian populations vary and are influenced by reimbursement policies that restrict use to symptomatic patients. We aimed to identify baseline predictors of treatment completion and overall survival (OS) in metastatic castration-resistant prostate cancer (mCRPC) within a reimbursement-regulated Taiwanese cohort and to derive an exploratory composite prognostic classification.

methodsWe retrospectively studied 170 men with symptomatic, bone-predominant mCRPC who received Ra-223 under Taiwan National Health Insurance criteria (April 2019-December 2024; follow-up through April 2026). Baseline variables were evaluated as predictors of treatment completion (all six cycles) by logistic regression and of OS by Kaplan-Meier and Cox proportional-hazards models. To avoid immortal time bias, treatment completion was excluded from the OS model. A 0-5 classification was constructed from the independent baseline risk factors and internally validated by bootstrap resampling.

resultsThe completion rate was 61% (104/170) and median OS was 15.2 months. Baseline alkaline phosphatase (ALP ≤ 102 U/L; OR 0.20) and hemoglobin (Hb ≥ 10 g/dL; OR 0.35) and absence of prior radiotherapy (OR 0.46) were independently associated with treatment completion (modeled as odds of early discontinuation). Five baseline factors were independently associated with OS: no prior chemotherapy (HR 0.63), no prior androgen-receptor-pathway-inhibitor (ARPI) therapy (HR 0.47), prostate-specific antigen (PSA) ≤ 19.4 ng/mL (HR 0.49), ALP ≤ 102 U/L (HR 0.36), and Hb ≥ 10 g/dL (HR 0.45); the ARPI association is exploratory (18 ARPI-naive patients). The 0-5 composite classification stratified three risk groups (log-rank p < 0.001): restricted mean survival 58.8 months for low-risk versus median OS of 15.6 and 6.7 months for intermediate- and high-risk groups; the optimism-corrected C-index was 0.69.

conclusionsIn a symptomatic, reimbursement-regulated mCRPC cohort, baseline ALP and Hb are associated with Ra-223 completion, whereas survival is associated with tumor activity (ALP, PSA), marrow reserve (Hb), and prior systemic therapy. The classification provides prognostic stratification within a Ra-223-treated population but does not establish treatment-specific benefit and requires external validation.