Long-term cancer control outcomes in patients with clinically high-risk prostate cancer treated with robot-assisted radical prostatectomy: results from a multi-institutional study of 1100 patients.

BACKGROUND: Long-term cancer control outcomes in clinically high-risk prostate cancer (PCa) patients treated with robot-assisted radical prostatectomy (RARP) remain unknown.
OBJECTIVE: To report on long-term biochemical recurrence (BCR)-free survival, clinical recurrence (CR)-free survival, and salvage therapy rates in these patients. Given the heterogeneity of high-risk patients, a second objective was to stratify them according to their BCR risk (using preoperative parameters), in an effort to counsel them better preoperatively regarding their cancer control outcomes. DESIGN, SETTING, AND PARTICIPANTS: We evaluated 1100 D'Amico high-risk PCa patients who underwent RARP between 2002 and 2013 at three tertiary care centers. OUTCOME MEASURES AND STATISTICAL ANALYSES: Outcomes consisted of BCR-free survival, CR-free survival, and salvage therapy rates. Regression tree analysis stratified patients into novel risk groups based on preoperative characteristics and corresponding BCR risk. Kaplan-Meier curves estimated BCR-free survival, CR-free survival, and salvage therapy rates in the entire cohort and after stratification according to the novel risk groups (RGs). RESULTS AND LIMITATIONS: Median age and prostate-specific antigen (PSA) were 63 yr and 6.5 ng/ml, respectively. Biopsy Gleason score (GS) was ≥8 in 57.7%. Mean follow-up was 53 mo (median: 49 mo). At 10 yr, BCR-free survival, CR-free survival, and salvage therapy rates were 50%, 87%, and 37%, respectively. Regression tree analysis stratified patients into five novel RGs): RG1, very low risk (GS ≤6); RG2, low risk (PSA ≤10 ng/ml; GS: 7); RG3, intermediate risk (PSA ≤10 ng/ml; GS ≥8); RG4, high risk (PSA >10 ng/ml; GS: 7); RG5, very high risk (PSA >10 ng/ml; GS ≥8). In these RGs, the 10-yr BCR-free survival rates were 86%, 70%, 36%, 31%, and 26% (p<0.001), respectively; the 10-yr CR-free survival rates were 99%, 96%, 85%, 67%, and 55% (p<0.001), respectively; and the 10-yr salvage therapy rates were 9.8%, 16%, 42%, 47%, and 64% (p<0.001), respectively.
CONCLUSIONS: Most patients with clinically high-risk PCa treated with RARP alone remain CR free at long term. Nonetheless, almost 37% of the patients at 10 yr require salvage therapy. Our novel tool allows accurate stratification of these heterogeneous patients according to their BCR, CR, and salvage therapy risks. This may help inform patients preoperatively about their cancer control outcomes postoperatively. PATIENT
SUMMARY: Robot-assisted radical prostatectomy confers lasting long-term oncologic control in most high-risk prostate cancer patients. Our novel risk grouping might serve as a useful tool for setting expectations and counseling patients regarding their cancer control outcomes.