Development and Validation of a Treatment-Specific Prognostic Model for Overall Survival in Patients with Metastatic Castration-Sensitive Prostate Cancer Treated with First-Line Apalutamide: A Multicenter Real-World Study

<title>Abstract</title> <p> <bold>Background</bold> Apalutamide plus androgen deprivation therapy (ADT) is a standard first-line treatment for metastatic castration-sensitive prostate cancer (mCSPC). However, outcomes remain heterogeneous, and prognostic factors specific to apalutamide-treated populations are not well established. We aimed to evaluate real-world outcomes and develop a treatment-specific prognostic model. <bold>Methods</bold> We retrospectively analyzed 208 patients with mCSPC treated with first-line apalutamide at multiple Japanese institutions. Overall survival (OS) was evaluated using Kaplan–Meier analysis and Cox proportional hazards models. Independent prognostic factors were used to construct a risk model, and performance was assessed using Harrell’s concordance index (c-index) and compared with LATITUDE and CHAARTED classifications. <bold>Results</bold> The median baseline prostate-specific antigen level was 129 ng/mL. A deep PSA response (&lt;0.2 ng/mL) was achieved in 68.8% of patients, and grade ≥3 adverse events occurred in 17.3%. Multivariable analysis identified primary Gleason pattern 5 (hazard ratio [HR] 4.32, P &lt; 0.001), extent of disease ≥2 (HR 3.89, P = 0.01), and lactate dehydrogenase ≥216 U/L (HR 3.95, P = 0.002) as independent predictors of poor OS. Patients with ≥2 risk factors had significantly worse OS (HR 6.83, 95% confidence interval 2.73–17.1; P &lt; 0.001). The proposed model showed better discrimination (c-index 0.762) than LATITUDE (0.593) and CHAARTED (0.615). <bold>Conclusions</bold> Clinical outcomes with apalutamide vary according to baseline disease characteristics. A simple three-factor model may improve prognostic stratification and support individualized treatment decisions in mCSPC. </p>