Association Between Biopsy PD-L1 Combined Positive Score and Pathological Upgrading at Radical Prostatectomy in Grade Group 1 Prostate Cancer: A Retrospective Study

Abstract

Programmed death-ligand 1 (PD-L1) expression in Grade Group 1 (GG1) prostate carcinoma biopsies has not been established as a marker of pathological upgrading at radical prostatectomy (RP). This two-center retrospective case–control study included 172 men with GG1 carcinoma in the biopsy closest to RP: all 86 patients upgraded to GG2 and 86 randomly selected non-upgraded controls from 126 eligible GG1 patients. PD-L1 was assessed with the SP263 clone using the combined positive score (CPS), with CPS ≥ 1 predefined before outcome comparison. CPS ≥ 1 occurred in 36/86 upgraded cases (41.9%) and 18/86 controls (20.9%; unadjusted odds ratio [OR], 2.72; 95% confidence interval [CI], 1.39–5.33; p = 0.005). After adjustment for serum prostate-specific antigen, positive-core count, Prostate Imaging Reporting and Data System category, and center, CPS ≥ 1 remained associated with upgrading (adjusted OR, 2.91; 95% CI, 1.47–5.97; p = 0.003). A Firth bias-reduced sensitivity model including digital rectal examination yielded a similar CPS estimate. Biopsy CPS ≥ 1 was associated with biopsy–RP grade discordance, but the study does not establish an optimal cutoff, predictive performance, or clinical utility. Prospective multicenter validation, including active-surveillance outcomes, is required.