Serum [‐2]proPSA as a Candidate Phenotypic Correlate of Molecular Alterations and Prognostic Marker in Metastatic Castration‐Resistant Prostate Cancer

Background: Serum biomarkers that complement molecular profiling and provide prognostic information are of interest in metastatic castration-resistant prostate cancer (mCRPC). Methods: We performed a single-centre retrospective exploratory analysis of 31 patients with mCRPC receiving systemic therapy. Baseline serum total PSA, free PSA, [-2]proPSA (p2PSA) and prostate health index (PHI) were evaluated together with tumour somatic molecular alterations and overall survival (OS). Baseline p2PSA was analysed both continuously after log2 transformation and using an exploratory 300 pg/mL threshold. On-treatment p2PSA change was reanalysed as a time-dependent covariate because follow-up sampling occurred at variable times. Results: Median age was 75.1 years (range 60.2-83.2). Median follow-up was 38.3 months (95% CI 22.5-42.8). Each doubling of baseline p2PSA was associated with shorter OS (HR 1.51, 95% CI 1.16-1.96; p = 0.0019). Patients with p2PSA > 300 pg/mL (n = 8; 6 deaths) had shorter OS than those with p2PSA ≤ 300 pg/mL (n = 23; 8 deaths; HR 4.61, 95% CI 1.46-14.56; log-rank p = 0.0043). Total PSA, free PSA and PHI were also associated with OS, and adding p2PSA to total PSA did not significantly improve model fit (likelihood-ratio p = 0.156). Baseline p2PSA was higher in patients with a pathogenic/likely pathogenic somatic alteration detected (n = 9) than in those with no pathogenic alteration detected (n = 22; Mann-Whitney p = 0.0348; Cliff's delta 0.49), but this association attenuated after exclusion of extreme values. On-treatment p2PSA change was not associated with OS after time-dependent modelling (failure to achieve ≥ 50% decline: HR 1.44, 95% CI 0.42-4.90; p = 0.557). Conclusions: In this small hypothesis-generating mCRPC cohort, higher baseline p2PSA was associated with poorer OS and with the presence of pathogenic/likely pathogenic somatic tumour alterations. These findings support prospective validation of p2PSA as an exploratory prognostic marker and candidate phenotypic correlate, rather than a surrogate for tumour genomics.