Salvage radical prostatectomy: baseline prostate cancer characteristics and survival across SEER registries

Abstract Objective To test for baseline prostate cancer characteristics and survival differences after salvage radical prostatectomy (SRP) across 18 Surveillance, Epidemiology and End Results (SEER) registries from 2004 and 2016. Material and Methods We tabulated PSA, cT-stage, age and SRP rates across individual SEER registries. Kaplan-Meier and competing risks regression methodologies depicted cancer-specific mortality (CSM) and other cause mortality (OCM). Finally, overall mortality (OM) was compared to predicted life expectancy. Results Overall, 428 SRP patients (2004-2016) were identified in the SEER database. Median follow-up duration was 74 months (IQR 31-114). Their median age at diagnosis was 68 (IQR 61-73) with a median PSA at diagnosis of 8.8 ng/ml (IQR 5.4-18.6 ng/ml) and 10% cT3-4-stage (range 0-23.8%). Variability existed across individual SEER registries regarding age, PSA, cT-stage and annual number of SRPs (range: 0-17), as well as cumulative numbers of SRPs (range: 7-73) between 2004 and 2016. At ten years, CSM was 23.2% vs. OCM 19.3%. Finally, SRP patients exhibited higher ten-year OM (43.3%) than that predicted by life tables (31.8%). Conclusion SRP is rarely performed. In most SEER registries SRP use is very occasional: More than two average annual SRPs were reported in only five of all registries. Nonetheless, across all registries, SRP patients showed marginal to moderate differences in PSA, cT-stage and age at diagnosis. However, at ten years of follow-up, one out of five SRP patients died of other causes and observed OM was higher than expected (36%).