PSA-based intermediate endpoints to compare radiotherapy and radical prostatectomy for the treatment of localized prostate cancer.

366 Background: There is no validated PSA-based endpoint to compare oncological outcomes after radical prostatectomy (RP) and radiation therapy (RT). We aimed to define a novel PSA-based intermediate endpoint that leads to similar prostate cancer-specific mortality (PCSM) for primary treatment comparison. Methods: This population-based study included all men with cT1-3, cN0M0 prostate cancer (PCa) in Stockholm who underwent RP or RT with PSA follow-up from 2003 to 2021. Competing risk regression, accounting for non-PCa mortality, was used to compare the cumulative incidence of PCSM in patients meeting different PSA-based endpoints after RP and RT. Endpoints were predefined by systematically increasing PSA cut-offs and PSA doubling times (PSA-DT) at recurrence. The analysis was adjusted for age at diagnosis, Charlson, PSA, ISUP Gleason at biopsy, cT stage, and treatment year, and was repeated to identify endpoint combinations that resulted in similar PCSM for RT and RP (subdistribution hazard ratio [SHR] of RT vs RP close to 1). Results: 12,010 (67.7%) man underwent RP, and 5,743 (32.3%) underwent RT. The median follow-up for survivors was 91 months (IQR 70–154) after treatment. A total of 428 men died from PCa. The adjusted SHR for patients meeting the current definition of biochemical recurrence (BCR) after RT (PSA ≥ nadir + 2) was 3.35 (95% CI 2.67, 4.23) compared to RP patients with a rising PSA≥0.2. Similar results were found when comparing PSA nadir + 2 with patients reaching higher PSA cut-offs after RP (≥ 0.5 and ≥ 1 ng/ml). Further comparisons were therefore performed by lowering the PSA cut-offs for the RT endpoint (nadir + 1-0.5 ng/ml). The endpoint definition leading to similar survival rates was PSA nadir + 0.5 ng/ml after RT and PSA ≥ 0.5 ng/ml with PSA DT &lt; 9 months after RP (SHR RT vs RP 1.00, 95% CI 0.78, 1.29). The adjusted cumulative incidence of the endpoint was higher after RT. Conclusions: We performed a systematic analysis of PSA-based early endpoints following primary treatment for PCa and found that patients who underwent RT and reached a PSA ≥ nadir + 0.5 ng/ml had similar survival rates to patients with PSA ≥ 0.5 ng/ml and PSA DT &lt; 9 months after RP. Our results highlight that current BCR definitions are suboptimal for primary treatment comparison and present, for the first time, promising novel early endpoints suitable for further validation. Competing risk regression comparing cumulative incidence of prostate cancer specific mortality after meeting the predefined endpoints after radiotherapy or radical prostatectomy (reference group). Endpoint compared Radical Prostatectomy(reference) Radiotherapy Adjusted SHR for PCa Mortality (RT vs RP) Adjusted Coefficient for PCa Mortality (RT vs RP) SHR 95% CI P Coeff 95% CI P 1 &gt;0.2 nadir+2 ng/ml N of Pt 2501 836 3.35 2.67, 4.23 &lt;0.001 1.21 0.98,1.44 &lt;0.001 2 &gt;0.5 nadir+2 ng/ml N of Pt 1365 836 2.08 1.66, 2.60 &lt;0.001 0.73 0.51,0.96 &lt;0.001 3 &gt;0.5+ DT&lt;12 nadir+2 ng/ml N of Pt 805 836 1.68 1.32, 2.12 &lt;0.001 0.52 0.28,0.75 &lt;0.001 4 &gt;0.5+ DT&lt;9 nadir+2 ng/ml N of Pt 667</j