A prediction model relating the extent of intraoperative fascia preservation to erectile dysfunction after nerve-sparing robot-assisted radical prostatectomy.

Abstract

Robot-assisted radical prostatectomy (RARP) is performed in patients with prostate cancer. Unfortunately, 10-46% of patients may still suffer from limited erectile function (EF) after RARP. This study aimed to develop a prediction model based on the extent of fascia preservation (FP) and postoperative EF after RARP. A previously developed FP score quantizing the extent and regions of nerve-preservation was determined in a cohort of 1241 patients who underwent RARP. The predictive value of the FP score for post-prostatectomy EF (following the international index erectile function (IIEF) score, EF domain) was analyzed. To increase the predictive value of the scoring system, the FP regions were related to postoperative EF, nerve distribution and co-morbidity factors. Finally, a prediction model for EF was developed based on the studied cohort. When corrected for the preoperative IIEF-EF, the FP score was shown to be a significant denominator for IIEF (p = 2.5 × 10^- 15) with an R² of 35%. Variable selection performed using the Akaike information criterion led to a final prediction model for postoperative IIEF after nerve-preservation based on the FP score. Furthermore, patient's age, preoperative IIEF score, CCIS and use of clips for nerve sparing were significantly associated with postoperative IIEF-EF. More anterior fascia preservation was correlated with better EF outcome and age was a strong independent predictor of EF outcome. In older men, the relative benefit of more extensive fascia preservation was at least similar to younger men, despite a lower baseline IIEF-EF score. Quantitative nerve-sparing FP scoring could be related to the postoperative IIEF-EF and integrated into a multivariate prediction model, which includes with age, use of surgical clips, the Charlson Comorbidity Index Score (CCIS), and preoperative IIEF-EF. When further validated the prediction model could provide patients and care-givers a qualitative estimation of EF outcome after RARP.