Persistent Prostate-Specific Antigen Following Radical Prostatectomy for Prostate Cancer and Mortality Risk

Importance Whether the conventional 1.5-month to 2.0-month time interval following radical prostatectomy (RP) for prostate cancer (PC) is sufficient to accurately document a persistent prostate-specific antigen (PSA) remains unanswered. Objective To evaluate the time necessary to accurately document a persistent PSA level after RP. Design, Setting, and Participants This cohort study evaluated whether a significant interaction existed between (1) a pre-RP PSA level greater than 20 ng/mL vs 20 ng/mL or less and (2) persistent PSA vs undetectable PSA after RP on PC-specific mortality (PCSM) risk and all-cause mortality (ACM) risk, adjusting for known PC prognostic factors, age at RP, year of RP, and the time-dependent use of post-RP radiation therapy (RT) and androgen deprivation therapy (ADT). Whether an increasing persistent PSA level was associated with a worse prognosis was also investigated. Patients with T1N0M0 to T3N0M0 PC treated with RP between 1992 and 2020 at 2 academic centers were included. Follow-up data were collected until November 2023. Data were analyzed from July 2024 to January 2025. Exposure RP. Main Outcomes and Measures Adjusted hazard ratio (aHR) of ACM and PCSM risk. Results Of 30 461 patients included in the discovery cohort, the median (IQR) age was 64 (59-68) years; of 12 837 patients included in the validation cohort, the median (IQR) age was 59 (54-64) years. Compared with patients with undetectable PSA, among patients with persistent PSA, a pre-RP PSA level greater than 20 ng/mL vs 20 ng/mL or less was significantly associated with reduced ACM risk (aHR, 0.69; 95% CI, 0.51-0.91; P = .01; P for interaction < .001) and PCSM risk (aHR, 0.41; 95% CI, 0.25-0.66; P < .001; P for interaction = .02). This result remained after adjustment for prostate volume and was confirmed in the validation cohort for PCSM risk and may represent a higher proportion of patients with a pre-RP PSA greater than 20 ng/mL vs 20 ng/mL or less who could have reached an undetectable PSA level if additional time for PSA assessment occurred before initiating post-RP therapy for presumed persistent PSA. Notably, there was more frequent and a shorter median time to post-RP RT plus ADT or ADT use in patients with a pre-RP PSA greater than 20 ng/mL (244 of 446 [54.7%] at a median [IQR] of 2.68 [1.51-4.40] months) vs 20 ng/mL or less (338 of 972 [34.8%] at a median [IQR] of 3.30 [2.00-5.39] months). These treatment times were shorter than the times to an undetectable PSA in observed patients (median [IQR] of 2.96 [1.84-3.29] months vs 3.37 [2.35-4.09] months, respectively). Increasing persistent PSA level was associated with an increased ACM risk (aHR, 1.14; 95% CI, 1.04-1.24; P = .004) and PCSM risk (aHR, 1.27; 95% CI, 1.12-1.45; P < .001). Conclusions and Relevance PSA level assessed for at least 3 months after RP may minimize overtreatment, and in this study, a higher persistent PSA level was associated with a worse prognosis.