Five-year biochemical recurrence-free and overall survival following high-dose-rate brachytherapy with additional external beam or radical prostatectomy in patients with clinically localized prostate cancer

Abstract

IntroductionHigh-dose-rate brachytherapy (HDR-BT) with external-beam radiation therapy and radical prostatectomy (RP) are common treatment options for clinically localized prostate cancer. The aim was to describe risk factors for biochemical recurrence (BCR) and death, as well as BCR rates and overall survival (OS) rates in both treatment groups. Patients and methodsOverall, 5,619 patients with localized prostate cancer underwent either RP (n = 5,200) or HDR-BT (n = 419) between 1999 and 2009. Median follow-up time was 72.4 months. Kaplan-Meier analyses and multivariable Cox regression analyses were performed for the overall cohort and for a propensity score–matched cohort to predict BCR and OS rates. Within the matched cohort, stratified analyses were repeated for HDR-BT alone (n = 206) and HDR-BT plus androgen deprivation therapy (ADT) (n = 213). Sensitivity analyses were performed to adjust for prostate-specific antigen rebound. ResultsThe 5-year BCR-free survival rates were 82.1% vs. 80.3% (P<0.01) for RP and HDR-BT, respectively. Corresponding 5-year OS rates were 97.1% vs. 92.4% (P<0.01). In the propensity score–matched cohort, 5-year BCR-free survival rates were 80.3% vs. 77.1%; P = 0.06 and 5-year OS rates were 95.7% vs. 92.4%; P = 0.5. In multivariable models, the overall HDR-BT exerted no significant effect on BCR, and the same results were recorded in the matched cohort. In stratified analyses, HDR-BT alone vs. RP increased BCR risk (1.45; P<0.01); conversely, HDR-BT plus ADT vs. RP decreased BCR risk (hazard ratio = 0.66; P = 0.02). ConclusionsFirst, RP offers equivalent oncological control without the need for concurrent hormone therapy and its morbidity. Second, patients who have RP avoid ADT (2%) and the need for salvage and adjuvant external-beam radiation therapy is low at 11% and 3%, respectively.