Introduction

Although robot-assisted radical prostatectomy (RARP) has recently become the gold standard treatment for localized prostate cancer, postoperative urinary incontinence is still a complication that decreases the quality of life (QOL) of patients1. Preoperative risk factors for urinary incontinence after prostatectomy include an advanced age, higher BMI, larger prostate volume, severe preoperative lower urinary tract symptoms, and a history of prostatectomy for benign prostatic hyperplasia2. Preoperative anatomical variables measured on magnetic resonance imaging (MRI), such as the length of the membranous urethra (MU) and periurethral supporting structures, have also been associated with urinary incontinence after RARP3. However, the mechanisms underlying the development of urinary incontinence have not yet been examined in detail.

The levator ani muscle is an important anatomical and functional structure for maintaining urinary continence, and its preservation improves urinary continence4,5,6,7. It comprises the puborectalis, pubococcygeus, and iliococcygeus muscles. The puboperinialis muscle (PPM) is the medial part of the puborectalis, arising from the posterior surface of the pubis on each side and passing posteriorly beside the urethra8. The rectourethralis muscle (RUM) is located at the interface between PPM and the rectum. Right and left PPM muscle edges are located close together in an area between the urethra and rectum, immediately lateral to RUM. Since these structures may form a strong sling and influence the stabilization of the MU, we considered the preoperative status of and surgical damage to these structure to potentially be important factors contributing to urinary incontinence after RARP. In the present study, we conducted an anatomical analysis of the structures around the urethra using MRI before and after RARP, and discussed the mechanisms underlying postoperative urinary incontinence.

Methods

Patients

We conducted a prospective observational cohort study. Patients who underwent RARP in our hospital between 2014 and 2019 were included in the present study. Patients with organ-confined prostate cancer were considered to be candidates for RARP and were included in the study. Inclusion criteria were (1) clinically localized prostate cancer, (2) age < 75 years, and (3) an Eastern Cooperative Oncology Group performance status of 0. During this period, 533 patients who met the inclusion criteria underwent RARP. The Ethics Committee of Fukushima Medical University Hospital approved this study (approval number: 2614) and all patients provided written informed consent. This study was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.

Surgical technique

RARP was performed using a four-arm da Vinci Si™ surgical system (Intuitive Surgical Inc., Sunnyvale, CA, USA) by the transperitoneal posterior approach reported by Guillonneau et al. in conventional laparoscopic procedures9. This approach is initiated with transverse peritoneotomy between the bladder and rectum, followed by retrovesical dissection. After dissection of the bladder neck, the prostatic vascular pedicles are ligated and the prostate is removed. Preservation of the neurovascular bundles is recommended to patients with low or favorable intermediate risk prostate cancer. Posterior reconstruction was performed on all cases in the present study. Vesicourethral anastomosis was conducted using the Van Velthoven procedure. RARP was performed or supervised by a single surgeon (Y.K.).

One-hour pad test to evaluate urinary incontinence

Urinary incontinence was evaluated by the one-hour pad test recommended by the International Continence Society10. We considered it important to perform the assessment of urinary incontinence objectively because the anatomical analysis was objectively conducted by MRI. Patients performed the one-hour pad test 1, 3, 6, 9, and 12 months after RARP to evaluate the amount of urinary leakage. Pad test information was collected by urology nurses.

MRI examinations and imaging techniques

MRI was performed before and 9 months after RARP in all cases because the effects of postoperative inflammation in muscles needed to be excluded. Static and dynamic MRI were performed using a 1.5 T MR scanner (Magnetom Trio; Siemens Healthcare, Erlangen, Germany). Dynamic MRI consisted of a T2-weighted, single slice True Fast Imaging with Steady-state Procession (TrueFISP) sequence in the sagittal, coronal, and transverse planes. The cross-sectional area of PPM and left–right (transverse) diameter of RUM (RUM diameter) were measured on T2-weighted fast spin echo (FSE) sequences. The left and right cross-sectional areas of PPM at the height of RUM on the dorsal side of the urethra were calculated using coronal sections, and the sum of the left and right cross-sectional areas was defined as the cross-sectional area of PPM (Fig. 1A). The tissue between the urethra and rectum was considered to be RUM in the transverse plane, and the left–right diameter was measured (Fig. 1B).

Figure 1
figure 1

Anatomical analysis of the structure around the urethra using MRI before and after RARP. (A) The left and right cross-sectional areas of PPM (dotted circles) at the height of RUM on the dorsal side of the urethra were calculated using coronal sections. (B) The tissue between the urethra and rectum was considered to be RUM (double arrow) in the transverse plane, and the left–right diameter was measured. PPM puboperinealis muscle, RUM rectourethral muscle, U urethra, R rectum.

Full size image

Patients also underwent cine MRI to evaluate the antero-posterior MU diameter with a full bladder 9 months after RARP (Fig. 2). Cine MRI in static and dynamic phases was conducted during the Valsalva maneuver and T2-weighed sagittal sections were obtained. Cine acquisition was performed over a 30-s time period during the static and dynamic phases. A urologist provided patients with guidance in the examination room during MRI scanning. Changes in the MU diameter from the static (Fig. 2A) to dynamic phases (Fig. 2B) during the Valsalva maneuver by cine MRI were calculated.

Figure 2
figure 2

Antero-posterior diameter of the membranous urethra (MU diameter) in static (A) and dynamic phases (B) during the Valsalva maneuver by cine MRI.

Full size image

Statistical analysis

All data were analyzed using SPSS statistical package version 24.0 (SPSS Inc., Chicago, IL, USA). The significance of differences was considered when p < 0.05. Parametric comparisons between before and after RARP were performed using the paired t-test. Simple and multiple regression models were used to examine the relationship between two or more variables. Variables with p < 0.05 in the simple regression model were included in the multiple regression model.

Results

Finally, 323 patients (mean age, 67.2 ± 6.0 years; age range, 48–76 years) were enrolled in the present study (Fig. 3). The baseline and perioperative characteristics of patients are shown in Table 1. The average amounts of urinary leakage in the one-hour pad test obtained by weighing the pads were 5.4, 46,1, 22.1, 18.5, 14.3, and 13.0 ml before and 1, 3, 6, 9, and 12 months after RARP, respectively. When urinary continence was defined as total urine leakage of < 2 g in the one-hour pad test10, continence rates were 85.1, 44.6, 75.5, 75.5, 76.5, and 78.0% before and 1, 3, 6, 9, and 12 months after RARP, respectively (Table 2).

Figure 3
figure 3

Flowchart of the study.

Full size image
Table 1 Patient characteristics.
Full size table
Table 2 Amount of urinary leakage according to the 1-h pad test and the rate of urinary continence before and 1, 3, 6, 9, and 12 months after RARP.
Full size table

We compared the cross-sectional area of PPM and RUM diameter using MRI before and after RARP (Table 3). The preoperative cross-sectional area of PPM (2.21 ± 0.69 cm2) was significantly reduced by 19% after RARP (1.79 ± 0.60 cm2; p < 0.01). No significant change was observed in the RUM diameter between before and after RARP (1.33 ± 0.28 and 1.36 ± 0.29 cm, respectively). There was no significant correlation between the change in the cross-sectional area of PPM and preoperative anti-androgen administration in the simple regression model (p = 0.49). There was also no significant correlation between the change in the cross-sectional area of PPM and neurovascular bundle (NVB) preservation in the simple regression model (p = 0.17). In addition, this correlation was not found regardless of unilateral or bilateral neurovascular bundle (NVB) preservation (p = 0.26 and 0.13, respectively).

Table 3 Comparison of the PPM cross-sectional area and left–right diameter of RUM before and after RARP.
Full size table

We performed a correlation analysis between the amount of urinary leakage according to the one-hour pad test after RARP and BMI, neurovascular bundle (NVB) preservation during RARP or prostate volume, which are considered to be factors which may affect urinary continence11. In the simple regression model, the amount of urinary leakage 1 or 3 months after RARP positively correlated with BMI (p < 0.01 and < 0.05, respectively). The amount of urinary leakage 1, 6, or 9 months after RARP negatively correlated with NVB preservation during RARP (p < 0.05, respectively). The amount of urinary leakage 6 months after RARP positively correlated with prostate volume (p < 0.05; Table 4). The multiple regression model revealed a correlation between the amount of urinary leakage 1 months after RARP and BMI or NVB preservation (p < 0.01 and p = 0.05, respectively; Table 5).

Table 4 Comparison between the amount of urinary leakage by the 1-h pad test and the PPM cross-sectional area or left–right diameter of RUM before and after RARP: simple regression analysis.
Full size table
Table 5 Comparison between the amount of urinary leakage by the 1-h pad test and the PPM cross-sectional area or left–right diameter of RUM before and after RARP: multiple regression analysis.
Full size table

We performed a correlation analysis between the amount of urinary leakage according to the one-hour pad test after RARP and MRI parameters, such as the cross-sectional area of PPM or RUM diameter (Table 4). In the simple regression model, the amount of urinary leakage 1, 3, 6, 9, or 12 months after RARP negatively correlated with the postoperative PPM cross-sectional area (p < 0.05, < 0.001, < 0.001, < 0.001, and < 0.001, respectively). The amount of urinary leakage 1, 3, 6, 9, or 12 months after RARP positively correlated with the change in the cross-sectional area of PPM by RARP (p < 0.001) and the postoperative RUM diameter (p < 0.001, < 0.001, < 0.001, < 0.01, and < 0.001, respectively). Furthermore, positive correlations were observed between the amount of urinary leakage 3, 6, 9, or 12 months after RARP and the preoperative RUM diameter (p < 0.01, < 0.01, < 0.05, and < 0.001, respectively) as well as between the amount of urinary leakage 1 month after RARP and the change in the RUM diameter by RARP (p < 0.05).

The multiple regression model revealed a negative correlation between the amount of urinary leakage 3, 9, or 12 months after RARP and the postoperative cross-sectional area of PPM (p < 0.01; Table 5). The amount of urinary leakage 1, 3, 6, 9, or 12 months after RARP positively correlated with the change in the cross-sectional area of PPM by RARP (p < 0.01, < 0.001, < 0.001, < 0.001, and < 0.001, respectively). These results indicate that a decrease in the cross-sectional area of PPM by RARP induced urinary incontinence after RARP. Representative distribution chart to show the correlation between the amount of urine leakage 12 months after RARP and the change in the cross-sectional area of PPM is shown in Fig. 4. A positive correlation was noted between the amount of urinary leakage 6 or 12 months after RARP and the preoperative RUM diameter (p < 0.05). Furthermore, the amount of urinary leakage 1 month after RARP positively correlated with the postoperative RUM diameter (p < 0.01). These results indicate that a large RUM diameter, which reflects a large distance between the left and right PPM, is a risk factor for the development of urinary incontinence after RARP.

Figure 4
figure 4

Distribution chart to show the correlation between the change in the cross-sectional area of PPM (before and 12 months after RARP) and amount of urine leakage. Change in PPM before and 12 months after RARP: preoperative PPM − postoperative PPM.

Full size image

A correlation analysis of the amount of urinary leakage according to the one-hour pad test after RARP and the change in the MU diameter from the static to dynamic phases during the Valsalva maneuver was also performed (Table 6). The simple regression model revealed a negative correlation between the amount of urinary leakage 1, 3, 6, 9, or 12 months after RARP and the change in the MU diameter (p < 0.001). The multiple regression model also showed that the amount of urinary leakage 1, 3, 6, 9, or 12 months after RARP negatively correlated with the change in the MU diameter (p < 0.001). Although the amount of urinary leakage 1, 3, 6, 9 or 12 months after RARP negatively correlated with preoperative membranous urethral length (p < 0.05, p < 0.05, p < 0.05, p < 0.05 and p < 0.01, respectively) in the simple model, this correlation did not find in the multiple regression model (p = 0.78, p = 0.27, p = 0.33, p = 0.07 and p = 0.30, respectively). It did not correlate with postoperative membranous urethral length in the simple regression model, (p = 0.11, 0.13, 0.33, 0.30 and 0.30, respectively).

Table 6 Correlation between the amount of urinary leakage by the 1-h pad test and pre- or post-operative membranous urethral length or the change in MU diameter on cine MRI parameters after RARP.
Full size table

We also examined the relationship between the change in the MU diameter from the static to dynamic phases during the Valsalva maneuver and the cross-sectional area of PPM or the RUM diameter (Table 7). The simple regression model showed that the change in the MU diameter from the static to dynamic phases negatively correlated with the change in the cross-sectional area of PPM (p < 0.01) and the preoperative (p < 0.05) or postoperative RUM diameter (p < 0.05). The multiple regression model revealed a negative correlation between the change in the MU diameter and the change in the cross-sectional area of PPM (p < 0.05).

Table 7 Correlation between the change in MU diameter on cine MRI at 9 months after RARP and the PPM cross-sectional area and left–right diameter of RUM before and after RARP.
Full size table

Discussion

The present study investigated the relationship between the anatomical pelvic structure around the urethra by MRI and urinary incontinence after RARP, and suggested that a decrease in the postoperative cross-sectional area of PPM after RARP and a large RUM diameter affected the postoperative urethral support system and consequently induced urinary incontinence.

The levator ani muscle plays a significant role in maintaining urinary continence and its preservation improves urinary continence4,5,6,7. Retzius-sparing RARP, which preserves the levator ani and puboprostatic ligaments, has a higher continence rate than standard RARP12,13,14,15. PPM, which has components of the levator ani muscle, is responsible for the quick stop phenomenon of urination in males, and, thus, the weakening of PPM by transection, traction injury, or denervation may affect urinary continence after radical prostatectomy16. In other words, preserving the integrity and innervation of PPM is crucial for the maintenance of urinary continence16. The right and left PPM course posteriorly from the pubis to flank the urethra, and converge in the midline behind the urethra, anterior to the rectum16. RUM occupies a space between the right and left PPM17. The left and right PPM edges are consistently located in close proximity in an area immediately lateral to RUM17, suggesting that PPM and RUM function as a urethral sling to maintain urinary continence17,18 (Fig. 5A). This urethral support mechanism by the connection between PPM and RUM is important for urinary continence in women19,20. In women, urethral hypermobility and a low urethral closure pressure due to damage to the urethral support mechanism causes stress urinary incontinence21. Mid-urethral sling surgery has become the gold standard treatment for stress urinary incontinence in women22, and the effectiveness of sling surgery depends on adequate postoperative urethral mobility and the urethral closure pressure23. The loss of the prostate may cause instability in the urethral support mechanism in men after RARP, as in women11. Therefore, we considered the urethral support mechanism by the connection between PPM and RUM to be important for urinary continence in men after RARP, and that the low ability or weak strength of inherent PPM and RUM and surgical damage to PPM and RUM after RARP may induce urinary incontinence after RARP.

Figure 5
figure 5

Anatomical structure around the urethra before and after RARP and the mechanism underlying postoperative urinary incontinence. (A) PPM and RUM function as a urethral sling to maintain urinary continence. (B) Surgical interventions during RARP damage and may lead to the atrophy of PPM and the loss of its function as a urethral sling. (C) The longer preoperative distance between left and right PPM indicate looser function as a urethral sling, and these patients may originally have had weak urethral support. PPM puboperinealis muscle, RUM rectourethral muscle, U urethra, PB; pubic body, PPL puboprostatic ligament, DVC dorsal vein complex.

Full size image

We herein focused on morphological changes in PPM and RUM before and after RARP. Sohn et al. examined the anatomy of men preoperatively with MRI, and found that men with a thicker puborectalis muscle at the anorectal angle regained continence earlier after prostatectomy24. We demonstrated that the cross-sectional area of PPM decreased by 19% after RARP. Moreover, the decrease in the cross-sectional area of PPM after RARP correlated with postoperative urinary incontinence and induced inadequate urethral compression. These results suggest that surgical interventions during RARP may damage PPM itself and nerve and/or blood supply around PPM, leading to the atrophy of PPM, the loss of its function as a urethral sling (Fig. 5B), and, as a result, urinary incontinence after RARP. Therefore, to avoid postoperative urinary incontinence by loss of PPM muscle volume, it may be important to minimize damage to nerves and blood supply by refraining from using electrocautery around PPM and minimize mechanical damage to PPM itself from surgical manipulation during RARP. However, NVB preservation may not be able to avoid loss of PPM muscle volume.

RUM also may play a significant role as a urethral support system by its connection with left and right PPM to maintain urinary continence. Individual variations have been reported in the development of RUM17,18. Soga et al. showed that RUM influenced MU stabilization with the rhabdosphincter, and also that it was one of the important structures forming a urethral sphincteric complex18. In the present study, pre- and postoperative RUM diameters positively correlated with urinary incontinence; however, the change in the MU diameter from the static to dynamic phases during the Valsalva maneuver did not correlate with the pre- or postoperative RUM diameter. Therefore, although the longer pre- and postoperative distances between left and right PPM were not directly proven to indicate weaker pre- and postoperative function as a urethral sling in the present study, these patients may originally have had weak urethral support from before RARP, which may easily induce postoperative urinary incontinence (Fig. 5C). Therefore, the preoperative measurement of RUM by MRI may predict urinary incontinence after RARP.

Preoperative pelvic floor muscle training (PFMT) enhances the postoperative function of pelvic muscles, including the levator ani muscle, reduces urinary incontinence after prostatectomy, and improves QOL outcomes related to urinary continence25,26,27. PFMT with a strength training protocol to hypertrophy the levator ani muscle may yield higher continence rates25. PFMT recuperates strength and lost tone in the sling mechanism after prostatectomy, and potentially enhances the pelvic floor support of vesicourethral anastomosis and the bladder base by strengthening PPM16. Additionally, the development of surgical procedures to reinforce the urethral sling system may promote urinary continence after RARP. Vladimir et al. reported that the advanced reconstruction of the vesicourethral support technique using the levator ani muscle attenuated urinary incontinence after RARP28. Previous studies demonstrated the effectiveness of urethral sling or bladder neck suspension techniques during RARP for urinary continence29,30,31. We showed that bladder neck sling suspension during RARP promoted the early return of urinary continence29. Cestari et al. found that a proper sling procedure during RARP allowed for the restoration of the sphincteric apparatus capability to its presurgical status; however, its capability decreased after prostate removal and subsequent ureterovesical anastomosis30. The efficacy of the male sling to treat post-prostatectomy incontinence has been reported32. Our anatomical data on the relationship between pelvic muscles and urinary incontinence after RARP appear to support the effectiveness of PFMT, surgical procedures, such as the urethral sling or bladder neck suspension techniques, and postoperative male sling procedures to maintain urinary continence after RARP.

There are several limitations that need to be addressed. First, this was a single-center analysis. Furthermore, postoperative MRI was only performed 9 months after RARP in all cases because the effects of postoperative inflammation in muscles need to be excluded. Second, we conducted pre- and postoperative anatomical analyses of the structure around the urethra using MRI, but not a functional analysis, such as the urethral pressure profile. Third, Valsalva effort is variable and subjective. The degree of dynamic movement is influenced by this effort. Therefore, further detailed studies are warranted.

Conclusions

PPM and RUM play significant roles in the supportive mechanism to maintain urinary continence as a urethral support. Surgical interventions during RARP lead to the atrophy of PPM, the loss of its function as a urethral sling, and, ultimately, urinary incontinence after RARP. The preoperative measurement of PPM by MRI may predict urinary incontinence after RARP. The further development of surgical procedures to prevent or minimize damage to and reinforce this urethral support system may contribute to the mitigation of urinary incontinence after RARP.