Objective: To evaluate outcomes in prostate cancer patients classified as high-risk (HR) or very high-risk (VHR) who were treated with conformal radiation therapy (CRT) and androgen deprivation therapy (ADT).
Methods: Between 11/2001 and 3/2012, 203 patients with HR disease received CRT to the prostate (78-82 Gy) and pelvic lymph nodes (46-50 Gy) with ADT (6 m-2 years). Median follow-up was 50 months (12 m-142 m). Biochemical failure was defined according to Phoenix definition. Imaging studies were used to identify local, regional or metastatic failure. Four different VHR/HR groupings were formed using the 2014 and revised 2015 NCCN guidelines. Differences were examined using Kaplan Meier (KM) estimates with log rank test and uni- and multivariate Cox regression analysis (MVA).
Results: Failure occurred in 30/203 patients (15%). Median time to failure was 30 m (4 m-76 m). KM estimate of 4 year biochemical disease free survival (b-DFS) for the entire cohort was 87% (95%CI: 82-92%). Four year KM survival estimates for b-DFS, PCSS and OS were comparable for each NCCN subgroup. On univariate analysis, the NCCN subgroups were not predictive of b-DFS at 4 years, however, DMFS was worse for both VHR subgroups (p = .03and .01) respectively. Cox univariate analysis was also significant for: PSA ≥40 ng/ml p = 0.001; clinical stages T2c p = .004, T3b p = .02 and > 4 cores with Gleason score 8-10 p < .03. On MVA, only PSA ≥ 40 ng/ml was predictive for b-DFS or MFS at 4 years (HR: 3.75 and 3.25, p < 0.005).
Conclusion: Patients with HR and VHR disease treated with CRT and ADT had good outcomes. Stratification into HR and VHR sub-groups provided no predictive value. Only PSA ≥40 ng/ml predicted poor outcomes on MVA. Distant failure was dominant and local recurrence rare, suggesting that improved systemic treatment rather than intensification of local therapy is needed. Patients with high-risk prostate cancer are most often treated with conformal dose escalated radiation therapy with androgen deprivation. Stratification into high versus very high-risk subgroups using 2014 or revised 2015 National Comprehensive Cancer Network (NCCN) criteria did not impact treatment outcomes. Only Prostate Serum Antigen (PSA) ≥40 ng/ml was predictive of poor prognosis. Distant failure was dominant and local recurrence uncommon which challenges the notion that intensification of local therapy will further improve outcomes in patients with high-risk disease.