Salvage prostate therapies for local pelvic recurrence after primary radiation for prostate cancer

Abstract Objectives The purpose of this study is to evaluate salvage treatment after primary radiation therapy (RT) for prostate cancer within our hospital system. Specifically, we aim to evaluate treatment patterns, oncologic outcomes and complications. Materials and methods We retrospectively identified men with post‐RT recurrence treated with salvage radical prostatectomy (sRP), seminal vesiculectomy, cryoablation, high‐dose‐rate brachytherapy (HDR) or irreversible electroporation (IRE). Salvage failure was defined as biochemical recurrence (BCR; Phoenix criteria, nadir + 2 ng/mL for cryoablation, HDR and IRE or prostate‐specific antigen (PSA) > 0.2 ng/mL for sRP and seminal vesiculectomy) or need for subsequent therapy. Cox regression identified factors associated with time to failure. Results Seventy‐one patients underwent salvage therapy after RT; 58 (81.7%) cryoablation, 7 (9.9%) sRP, 2 (2.8%) robotic seminal vesiculectomies, and 4 (5.6%) other modalities. Failure occurred in 30 (45%) patients at a median of 12.5 months (IQR 6, 23). Grade group (GG) 4–5 disease prior to RT was associated with time to failure (HR 2.18, 95% CI 0.96, 4.97, p = 0.063) on multivariable analysis; however, the difference was not significant. Treatment type was not significantly associated with failure‐free survival ( p = 0.7). Four patients experienced Clavien–Dindo grade III or higher complications: three following sRP (two anastomosis dehiscence, clot retention) and one following cryoablation (stricture). Ten patients (14%) underwent local salvage for prostate recurrences in the setting of N1/M1 disease with concurrent systemic therapy, with no treatment escalation or complications at median 16 months follow‐up time. Conclusions In this institutional cohort, salvage cryoablation demonstrated an acceptable short‐term failure‐free survival with fewer observed high‐grade complications than salvage prostatectomy. Salvage cryoablation also appeared feasible in select patients with low‐volume nodal or metastatic disease. These findings are hypothesis‐generating and require prospective validation.