Preoperative MRI-Measured Membranous Urethral Length as a Predictor for Urinary Continence after Prostate Cancer Surgery

Abstract

This thesis addresses prediction of post-prostatectomy urinary continence, which is of major importance to patients and clinicians. Urinary incontinence significantly impacts quality of life, and some patients may choose non-surgical treatments if their risk of post-surgery incontinence is unacceptably high. The membranous urethra is a small segment of the urethra surrounded by pelvic floor musculature. In theory, a longer membranous urethra on pre-operative imaging may signify a more robust sphincter complex and a lower probability of involuntary urinary leakage after surgery. In practice, data from several studies suggest MRI-measured membranous urethral length (MUL) is one of few predictors of post-prostatectomy continence. However, it is currently not clear how to use MUL when counseling patients since proposed prediction models have never been externally validated. To address this issue, we systematically reviewed the existing literature, and performed a meta-analysis, allowing us to conclude, with a moderate degree of confidence, that longer MUL is predictive of urinary continence after prostatectomy. We also observed that MUL measurement techniques vary between studies, posing a significant risk for miscalibration when applied to clinical practice. We then externally validated a published continence prediction model, by Jeong et al. We observed that the model has good calibration and discrimination when continence is defined as 0-1 incontinence pad/24h but does not perform well in predicting complete urinary continence (no incontinence pads needed). In conclusion, we now know the benefits and limits of a MUL prediction model that can be applied to our future patients.