Patterns of Recurrence Following Radiation and ADT for Pathologic Lymph Node Positive Prostate Cancer: A Multi-Institutional Study

Abstract

Retrospective studies have suggested that post-operative radiation therapy (RT) can reduce the risk of cause-specific mortality in men with pathologic nodal involvement (pN1) after radical prostatectomy (RP). We evaluated prognostic factors and patterns of recurrence in patients who received post-operative RT +/- androgen deprivation therapy (ADT) for pN1 disease in a multicentric cohort of 4 academic centers. Data from patients with pN1 prostate cancer after RP who subsequently received RT with short term (< = 6 mo) or long term (>6 mo) ADT were obtained from 4 academic institutions. Patterns of recurrence, biochemical progression free survival (bPFS) and distant metastasis free survival (DMFS) were evaluated. A total of 270 patients with a median follow-up of 48 months were included. Gleason grade group (GG) 2 was present in 20 patients (7.5%), GG 3 in 81 (30%), GG4 in 36 (13.5%), GG5 in 130 (49%) patients. 256 (95%) patients had extracapsular extension, 70% had seminal vesicle invasion, 59% had positive surgical margins, and 66% had a detectable post-operative PSA. The number of positive nodes at surgery were 1 in 59%, 2 in 19% and >2 in 22% of patients. Of the 83 patients that had pre-RT imaging, 46 (55%) had a PET scan (PSMA, or fluciclovine); 25 (30%) of those had lymph nodes detected on imaging prior to RT. Median time from RP to RT was 6 mo (IQR 4.5-9.1 mo). 96% received radiation to both the prostate bed (median dose 68.4Gy) and pelvic lymph nodes (median dose 46Gy). ADT was prescribed short-term (20%) or long-term (68%), while 26 (10%) received no ADT, and 7 (3%) had an unknown duration. Biochemical failure (bF) was observed in 29% of men, with 5% having pelvic nodal failure and 11.5% having distant metastases (majority in bones, followed by paraaortic nodes) at last follow up. Of 59 patients who had normal baseline testosterone levels, 37% recovered their testosterone by last follow-up. The 4-year bRFS was 72% for all patients and was 83% for those with a pre-RT PSA of <0.1 ng/mL, 76% for PSA 0.1-<0.5 ng/mL, 60% for PSA 0.5-2 ng/mL, and 35% for PSA >2 ng/mL. On multivariable analysis, maximum pre-RT PSA ≥0.5 (0.5 to 2.0 vs <0.1 HR = 3.19; >2.0 vs <0.1 HR = 9.00), use of LTADT (HR = 0.47), and percent pN1 (HR = 1.03) were significantly associated with bF. Pre-RT PSA >2 (HR = 4.10), use of LT ADT (HR = 0.33) and percent pN1 (HR = 1.03) were also significantly associated with DM. In pN1 patients, pelvic RT and ADT at low PSA levels is associated with improved oncologic outcomes compared to treatment at higher levels, suggesting that PSA may have prognostic value for pN1 prostate cancer.