Predicting Erectile Dysfunction after Highly Conformal, Hypofractionated Radiotherapy to the Prostate

Abstract

Background: Erectile dysfunction (ED) is common after prostate cancer treatment. It has been studied for conventional radiotherapy, but associations in the hypofractionated radiotherapy context are less clear. This study aimed to determine which factors are predicted for worsening ED after highly conformal, modestly hypofractionated radiotherapy to the prostate. Methods: Two hundred and twelve patients treated with 6000 cGy in twenty fractions across four centers were included in this study. Demographic, clinical, and dosimetry factors were then evaluated for post-treatment declines in erectile function using logistic regression and an explainable machine learning-based neural network. Results: 212 patients with a median follow-up of 3.6 years were evaluated. A total of 104 (49%) patients received androgen deprivation therapy. Prior to treatment, 52 (25%) patients were on ED medication. Mean doses to the penile bulb, penile crus, and penile shaft were 2490 (IQR: 1529–3656) cGy, 2095 (1306–3036) cGy, and 444 (313–650) cGy, respectively. Fifty-nine (28%) patients had a worsening of ED after treatment. On multivariable analysis, only the mean dose to the penile shaft [OR >345 vs. ≤345: 4.47 (1.43–13.99); p = 0.010] and pretreatment use of ED medication [OR yes vs. no: 12.5 (5.7–27.5; p < 0.001)] predicted for worsening ED. The neural network confirmed that the penile shaft mean dose and pre-treatment ED medication use are the most important factors in predicting ED. Conclusions: Pre-treatment ED and penile shaft dosimetry are important predictors for ED after hypofractionated radiotherapy for prostate cancer.