Dosimetric Predictors of Sexual Function Decline Following Low-Dose-Rate Brachytherapy for Prostate Cancer (Pca)

Conventional dosimetric standards for implant quality do not adequately predict patient-reported quality of life (QOL) measures following permanent prostate seed brachytherapy. We used a dynamic learning multiscale knowledge database developed to explore relationships between dose and patient-reported toxicity. Patient-reported QOL outcomes were prospectively collected for 366 patients with stage cT1–T2 PCa who received permanent low-dose-rate (LDR) prostate brachytherapy over a 10-year interval. Patients received Pd-103 seed implantation under real-time ultrasound guidance and completed a QOL instrument (IIEF-5) prior to treatment and at each follow-up. Multivariate risk models were used to evaluate dose and toxicity relationships. In addition, the database learning environment was used to examine granular dose/volume relationships for correlation with clinical endpoints. Median follow-up was 36 months from implant date (SD 28.5). Of the entire cohort, 23.7% (n = 87) received combined brachytherapy and external beam radiation therapy (EBRT); 14.7% (n = 54) had received prior androgen suppression or 5 alpha-reductase therapy. Sexual QOL data at minimum of 2 years was available in 141 men. Within this group, mean decrement per the IIEF-5 instrument was 3.76 points (SD 10.3), with 17.0% of men reporting a decrement of 10 points or greater (39% of men reporting a decline of 5 or more). There was no significant association with prior androgen suppression (P = .20) or combined EBRT use (P = .69). Using the database informatics platform, dose-volume histogram curves were queried and dichotomized according to stable sexual function versus significant decline (>10) assessed 2 years postimplant. A logistic regression model was used to generate probability surface maps exploring the effect of dose on outcomes as a function of normalized volume. Structures analyzed included the neurovascular bundles, the cavernosal nerves (delineated as a 10-mm cylindrical area caudal to the apex), prostate, urethra, and penile bulb. Logistic regression revealed a significant correlation between dose to the cavernosal nerve and total SHIM score throughout the region of the dose-volume histogram curve encompassing 10%, 20%, and 30% of the contoured cavernosal nerve volume (P < .05). No significant dose/volume correlation with sexual function emerged for any other structures analyzed. Granular dose/volume and toxicity correlation using the dynamic learning multiscale knowledge database learning environment exposed a statistically significant relationship between volumetric dose to the region encompassing the cavernosal nerves and worsened sexual outcomes in patients undergoing LDR brachytherapy. Minimization of dose to the cavernosal nerves may help to preserve sexual function after permanent prostate seed brachytherapy.