Postprostatectomy incontinence (PPI) is a major complication of prostatectomy. Although pelvic floor muscle training (PFMT) is a good treatment of PPI, there is no evidence of how the muscle movement affects recovery of continence. MRI (magnetic resonance imaging) is a good tool to evaluate the risk factors of PPI such as membranous urethral length (MUL). In the present study, we evaluated some dynamic factors of prostatectomy patients using cine MRI to identify the risk factors of PPI and reveal the contribution of pelvic floor muscle to recovery of continence.

A total of 128 male prostate cancer patients were enrolled in the present study. The all prostatectomy was robot assisted laparoscopic surgery. Cine MRI was performed preoperatively and at 6 months after surgery. Continence was defined as pad free or safety pad. We defined bladder neck elevation length during PFMT as “bladder elevation length (BEL)”. Other static or dynamic parameters were also measured. Patients who got recovery of continence within a month were divided into Continence-group (n=48), and others were into Incontinence-group (n=80). We analyzed the parameters between two groups using Mann-Whitney U test or chi-square test, and between pre-post operations using Wilcoxon signed rank test. Preoperative parameters were also analyzed using multivariable analysis. Kaplan-Meier analysis with log-rank testing was also applied for preoperative BEL.

In the two groups, there was no significant difference in any parameters of patient characteristics, perioperative data, pathological results and with or without nerve-sparing technique. In Continence-group, preoperative BEL was significantly longer than Incontinence-group (10.4 vs 8.2 mm; P<0.001). Postoperative BEL of Continence-group also tended to be longer (9.9 vs 8.9 mm; P=0.057). Only in postoperative state, posterior urethrovesical angle (PUVA) of Continence-group was significantly smaller than Incontinence-group at both resting and voiding phase (130 vs 135 º; P=0.005, 138 vs 143 º; P=0.026). Postoperative membranous urethral length (MUL) of Continence-group was significantly longer (14.5 vs 12.4 mm; P<0.001). Multivariate analysis showed that preoperative BEL significantly contributed to getting recovery of continence (HR=0.96, P=0.016). Patients with longer preoperative BEL (>8.8mm) got recovery of continence significantly faster than patients with shorter (<8.8mm) (log-rank test; P=0.038).

Preoperative BEL could be a great predictor for early recovery of continence after radical prostatectomy. This parameter is acquired from dynamic and highly accurate data of cine MRI. Moreover, cine MRI could detect not only anatomical features, but also actual function of pelvic floor muscle. We and the patient can consider his own risk of PPI using cine MRI. Moreover, long BEL might indicate better PFMT, and PFMT biofeedback using cine MRI has a potential to achieve early recovery of continence.
Some reports demonstrated that smaller PUVA made recovery of continence earlier, and the results of the present study supported the theory. Surgical effort to preserve MUL would also result in preventing PUVA opened.

A novel dynamic parameter of BEL was strongly related to recovery of early continence after radical prostatectomy. This is the first study to show that cine MRI is useful for digitizing the level of PMFT skill and predicting the risk of PPI.