To determine whether high-dose radiation delivered to a subvolume of the prostate gland (peripheral zone) using intraoperative magnetic resonance imaging-guided brachytherapy provided comparable 5-year prostate-specific antigen (PSA) control rates to radical prostatectomy (RP) in select patients compared prospectively but in a nonrandomized setting.

Between 1997 and 2002, 322 and 196 patients with clinical Stage T1c, PSA less than 10 ng/mL, biopsy Gleason score 3 + 4 or less, and without perineural invasion underwent RP or intraoperative magnetic resonance imaging-guided brachytherapy, respectively, and had a 2-year minimal follow-up. Cox regression multivariable analysis was used to evaluate whether the initial therapy, pretreatment PSA level, biopsy Gleason score, percentage of positive biopsies, or prostate gland volume were predictors of the time to post-therapy PSA failure. PSA failure was estimated using the Kaplan-Meier method and defined using the American Society for Therapeutic Radiology Oncology consensus definition.

Only the percentage of positive prostate biopsies (P_Cox = 0.02) was a significant predictor of the time to post-treatment PSA failure. However, the distribution of this parameter between RP and brachytherapy-treated patients was not significantly different (P_chi-square = 0.25). The initial therapy did not predict for the time to post-therapy PSA failure (P_Cox = 0.18). The 5-year estimate of PSA control was 93% versus 95% (P_log-rank = 0.16) for the RP and brachytherapy patients, respectively.

Despite only partial prostatic irradiation using intraoperative magnetic resonance imaging-guided brachytherapy, similar 5-year estimates of PSA control were found for both brachytherapy and RP-managed patients.