Introduction

The identification of patients less likely to recover erectile function (EF) after surgery is crucial for counseling and for the early administration of proerectile treatments. We aimed at developing a model to predict EF recovery in prostate cancer (PCa) patients treated with robot-assisted radical prostatectomy (RARP).

Methods

We included 833 PCa patients treated with RARP between 2006 and 2016. Postoperative UC recovery was defined as being pad-free over a 24-hour period. Early continence was defined as UC within 60 days from surgery. Postoperative EF was defined as an Erectile Function-Erectile Function (IIEF-EF) domain score ≥22. Kaplan-Meier and Cox regression analyses assessed the impact of early UC on EF recovery. Covariates were age, preoperative IIEF-EF, Charlson comorbidity index (CCI), nerve-sparing, adjuvant radiotherapy (aRT), and early UC recovery. Predictors of EF were used to develop a novel risk score based on the cumulative number of risk factors. Kaplan-Meier analyses assessed the impact of the risk score on EF rates. A decision-curve analysis (DCA) assessed the net benefit associated with the use of our model.

Results

Median preoperative IIEF-EF was 25. Overall, 746 (90%) patients were treated with nerve-sparing surgery and 54 (6.3%) patients received aRT. Median follow-up was 36 months. The proportion of patients who experienced early UC recovery was 337 (40.5%). At 3-year follow-up, the UC and EF recovery rates were 85.9% and 45.8%. The 3-year EF rates were higher in patients who were continent within 2 months compared to those incontinent at this time point (53.8 vs. 40.4%; P<0.001). In multivariable analyses, age <65 years, a preoperative IIEF-EF ≥22, the receipt of nerve-sparing surgery, no aRT, and early UC recovery were associated with an increased probability of EF recovery (all P≤0.01). Based on these factors, a risk score predicting EF was calculated. When patients were stratified according to the risk score (≤2 vs. 3 vs. ≥4), the 3-year were 10.9 vs. 32.0 vs. 60.6%, respectively (P<0.001). At the DCA, clinical risk prediction improved for men with a probability of EF recovery between 15 and 60%.

Conclusions

Early UC recovery is associated with the probability of subsequent EF recovery. A risk score based on pre- and postoperative characteristics to predict EF recovery should be used for patient counseling in the early postoperative setting and for the identification of candidates for more aggressive proerectile therapies.

Funding

none