Abstract
The introduction of prostate-specific PET imaging to detect radiorecurrent prostate cancer (RRPCa) and the establishment of HDR brachytherapy salvage at our institution have resulted in an increased case volume of RRPCa being treated with HDR salvage. We sought to compare the patient make-up and outcomes of this new cohort in comparison to the last two decades of RRPCa patients salvaged with brachytherapy at our institution.
We retrospectively reviewed 20 years of RRPCa patients salvaged with whole-gland brachytherapy (2000-2021) with at least 3 years of follow-up. We identified 30 patients treated with MRI-guided LDR (Cohort 1: 2000-10), 39 patients treated with trans-rectal ultrasound(TRUS)-guided LDR (Cohort 2: 2011-20), and 32 patients treated with (TRUS guided: 22, CT planned: 10) HDR (Cohort 3: 2017-2021). All patients were biopsy-proven, locally-recurrent, without evidence of metastatic progression and received pre-salvage MRI. Percentages are actuarial and calculated using cumulative incidence. Survival analysis was conducted first via Log-Rank test, and Cox proportional hazards for univariate and multivariate analyses.
The percentage of patients with EPE on MRI for cohort 1, cohort 2, and cohort 3 were 17%, 8.1%, and 26%, respectively. Pre-salvage, Axumin PET rates were 0%, 18%, and 100%. Rates of concurrent and adjuvant ADT at salvage were 3.3%, 59%, and 84%. At median follow-ups of 20.5, 6.9, and 3.4 years, the percentage of patients without BCR at 3 y were (cohort 1- 79.3%, cohort 2- 78.6%, cohort 3- 75.1%). The three salvage strategies were not statistically different with respect to BCR (Log-Rank p = 0.4). As the MRI patients (cohort 1) have been previously reported [Nguyen PL, Cancer 2007] and TRUS LDR is utilized as an HDR alternative at many other practices, we then reported outcomes between TRUS and HDR patients. There was no statistical difference in OS, DMFS, Local (Intraprostatic or SV) recurrence, or LRR by Log-Rank test (p > 0.05) (Fig 1). Long-term GU grade 3 toxicity (90d to 3 y) was not statistically different (Cohort 2: 4/39 (10%), Cohort 3: 4/32 (13%)). There was similarly no difference for long-term GI grade 3 toxicity (Cohort 2: 1/39 (2.6%), Cohort 3: 1/32 (3.1%)). PSA DT < 12 mo trended towards an association with BCR (p = 0.12) and LRR (p = 0.088) on UVA (Cox proportional hazards). SVI on pre-salvage MRI was negatively prognostic, UVA (p = 0.045 and p = 0.027) and MVA when accounting for cohort and PSA DT (p= 0.038 and p = 0.022) for BCR and LRR, respectively. Of the 6 local-regional recurrences after HDR salvage the LRR pattern of failure was: 1 in the prostate and subsequently R inguinal nodes and bilateral seminal vesicles; 2 regional only in the pelvic sidewall and LNs; 1 in the prostate and LNs, 1 in the SV and LNs, and 1 in the prostate and SV only (1 local recurrence only, 2 regional recurrence only, 3 local and regional recurrence)
This single-institution retrospective comparison study demonstrates that modern, salvage HDR brachytherapy for RRPCa is equally effective (BCR, LRR, OS) when compared to historical methods (MRI guided LDR) and contemporary alternatives (TRUS LDR). Additionally, long-term GU and GI toxicities are rare for both HDR and TRUS LDR. As the majority of LRR failures included intraprostatic or SV failure (4/6) for HDR therapy, prospective studies involving further dose-escalation to PSMA-avid lesions may be warranted.