Ten-year disease progression and mortality rates in men who experience biochemical recurrence versus persistence after radical prostatectomy and undergo salvage radiation therapy: A post-hoc analysis of RTOG 9601 trial data

Abstract

Purpose: To compare local/metastatic disease progression and overall mortality rates in men with node-negative prostate cancer at radical prostatectomy (RP) that experience biochemical recurrence vs. persistence postoperatively and undergo salvage radiation therapy (sRT). Materials and methods: Data on 760 men who participated in the RTOG 9601 trial were extracted using the NCTN data archive platform. Patients were stratified into biochemical recurrence (nadir-PSA ≤0.4 ng/ml) or persistence (nadir-PSA >0.4 ng/ml) groups, based on the cut-off reported in the original trial. Inverse probability of treatment weighting (IPTW) methodology was utilized to minimize the baseline differences among groups. Competing-risk and Kaplan-Meier analyses estimated the impact of prostate-specific antigen (PSA) persistence vs. recurrence on local and metastatic disease progression and overall-mortality in the IPTW-adjusted model; a 2-sided P < 0.05 was considered significant. Results: All patients received sRT, and about 50% of the patients in either group received concomitant antiandrogen therapy (P = 0.951). The median follow-up was 12 years. After IPTW, the 2 groups were well-matched with standardized mean differences ∼10%. In the IPTW-adjusted cohort, the 10-year local and metastatic disease occurrence rates were 3.2% vs. 1.4% (Gray's P = 0.0001) and 28.6% vs. 10.1% (Gray's P < 0.0001) in patients with persistent vs. recurrent PSA, respectively. Similarly, the 10-year overall-mortality rates were 24.9% vs. 11.9% (Log-rank P = 0.029), respectively. Conclusions: Patients with biochemical persistence after RP are approximately 2.5 times more likely to experience local/metastatic failure and death, compared to patients with biochemical recurrence after RP, despite equivalent sRT with/without antiandrogen therapy use. These data may facilitate patient counseling and shared treatment selection.