Identifying the predictors of acute urinary retention following magnetic-resonance-guided prostate brachytherapy

Purpose: Larger prostate gland volumes have been associated with long-term urinary morbidity in prostate interstitial radiation therapy utilizing ultr asound image guidance technique. This study was performed to identify the c linical and technical predictors of acute urinary retention following magne tic-resonance (MR)-guided prostate interstitial brachytherapy. Methods and Materials: Fifty patients underwent MR-guided prostate brachyth erapy between December 1997 and March 1999. Patient selection was limited t o men with stage T1cNXM0 disease, PSA of less than 10 ng/mL, biopsy Gleason score not more than 3 + 4, and endorectal coil MR stage T2 disease. Dosime try plans were developed in the operating room and (125)Iodine sources were implanted using MR real-time guidance. The peripheral zone (PZ) of the pro state gland was defined as the clinical target volume (CTV) and the minimum prescribed dose to the CTV was 137 Gy. The volumes of the PZ, transition z one (TZ), and total prostate gland volume were also determined by MR. Indiv idual source strength ranged from 0.35 to 0.54 mu Gym(2)/h (NIST 99, median 0.46 mu Gym(2)/h) and the total implanted activity ranged from 17.0 to 43. 1 mCi (median, 28.1 mCi) using 43-120 seeds (median, 79). The seeds were pl aced using MR-compatible biopsy needles (14-28, median, 19). Results: The ability of clinical (MR defined prostate, PZ, and TZ volumes) and technical (number of catheters, number of seeds implanted, and total ac tivity) factors to predict AUR for 50 men undergoing MR-guided prostate int erstitial brachytherapy were evaluated using univariable and logistic regre ssion multivariable analyses. Six men (12%) experienced AUR within 24 h aft er removal of the Foley catheter subsequent to prostate brachytherapy. The total number of seeds (p = 0.05), MR determined prostate volume (p < 0.01), and the MR-determined TZ volume (p < 0.01) were significant predictors of AUR on univariable analysis. Utilizing a multivariable logistic regression analysis, the TZ volume was the only significant predictor of AUR (p < 0.01 ). The prostate volume is highly correlated to the TZ volume (Spearman corr elation coefficient of 0.91) and was thus significant in the univariable an alysis; however, the prostate volume did not add prognostic value in multiv ariable analysis. Conclusion: Benign prostatic hyperplasia (BPH) resulting in an enlarged TZ volume, is the most important predictor of AUR following MR-guided prostate interstitial radiation therapy. Although AUR was significant (60%) in men with moderate BPH (TZ volume greater than or equal to 50 cc), it was also s elf-limiting. (C) 2000 Elsevier Science Inc.