To present the anatomical basis and details of a technique for an approach to the endopelvic fascia devised to preserve urinary continence.
To present the anatomical basis and details of a technique for an approach to the endopelvic fascia devised to preserve urinary continence.
For cross-sectional macroscopic observation, seven formalin-fixed specimens of the male pelvic contents including the pelvic wall were serially sectioned at a 5-mm thickness. Semi-serial sections from eight other specimens were examined histologically. Eight fresh cadavers were dissected to mimic the various steps in a retropubic radical prostatectomy. After approaching the endopelvic fascia in an anatomically determined manner to reach the paraprostatic space, the pubic bone was removed and nerves near the rhabdosphincter dissected. To assess the clinical implication of this approach, we examined the time to achieve continence in 23 consecutive patients who had a radical retropubic prostatectomy using the new technique.
Sectional macroscopic observation depicted the fascia of the levator ani as a definite structure adherent to but not fused with the lateral pelvic fascia. The thin fascia overlying the levator ani fascia and lateral pelvic fascia represented the true endopelvic fascia. Microscopically, the lower part of the fascia of the levator ani was rich in smooth muscle, which interdigitated with the framework of the rhabdosphincter. In fresh cadavers, the levator ani muscle was removed laterally still covered by its fascia, without visualizing the muscle fibres. Small branches from the pudendal nerve entered the rhabdosphincter. The mean (sd, range) distance from the lowest point of the endopelvic fascia to the point where the sphincteric branch entered the rhabdosphincter was 5.5 (1.8, 3–8) mm. The continence rate at 1, 3, 6 and 9 months after surgery using the new technique was 44%, 83%, 96% and 100%, respectively.
Preserving the fascia of the levator ani helps to protect the levator ani muscle, rhabdosphincter and pudendal nerve branches to the rhabdosphincter. In retropubic radical prostatectomy, this anatomical approach to the endopelvic fascia should preserve or allow the earlier recovery of urinary continence.
fascia of the levator ani
the lateral pelvic fascia
In urological surgery, the fascia between the pelvic wall and pelvic organs is designated the ‘endopelvic fascia’ (EPF). Urologists must incise the EPF to open a route to the paraprostatic space. Many textbooks on urological surgery advise that a white line of condensed fascia can be identified at the bottom of the EPF, and stress the importance of incising laterally to the white line to avoid unnecessary haemorrhage from prostatovesical veins . Although Myers [2,3] studied the anatomy of pelvic fascia related to retropubic radical prostatectomy (RP), we consider that treating the EPF according to the topographic anatomy of the fascia, nerves and rhabdosphincter has not been fully elucidated. We examined the configuration of the EPF by sectional and histological observation in formalin-fixed and fresh male cadavers, including observations on dissection by a procedure similar to pelvic surgery, as described recently . In addition, we considered how best to manage the EPF to protect nerves contributing to urinary continence and to the rhabdosphincter.
For the anatomical study, macroscopic sectional or histological observations were made in 15 intact formalin-fixed male pelves. These specimens were obtained from cadavers donated to Sapporo Medical University that had been fixed by arterial injection of 10% formalin solution and stored at room temperature for at least 1 month before use. The mean (range) age at death was 77 (64–90) years. The seven specimens used for sectional macroscopic observation were removed from cadavers en bloc, including all intrapelvic organs and the pelvic wall. The specimens were sectioned serially, either frontally or axially, at a thickness of approximately 5 mm. The other eight specimens were processed for histological study as follows: after dissecting the prevesical space, the pubic bone was removed and the membranous urethra, the apex of the prostate, the proximal portion of the corpora cavernosa, and the pelvic floor muscles surrounding the membranous urethra were excised en bloc. These large tissue blocks were cut to 12–15 cm3. The specimens were dehydrated, embedded in paraffin (melting point 58°C) and cut into 4–7 µm-thick semi-serial frontal or axial sections. Haematoxylin and eosin staining or immunohistochemical staining as described previously by Murakami et al. was used. The primary antibody used in the immunohistochemical evaluation was monoclonal antihuman α-smooth muscle actin (mouse IgG2a, κ; Dako, Kyoto, Japan).
In the eight fresh cadavers, dissections were sequential to mimic the various steps in retropubic RP leading up to exposure of the apex of the prostate. The mean age at death for these specimens was 78 (71–88) years. The cadavers had been donated to Sapporo Medical University within 24 h of death. No fixatives were used, and hemicorporectomy and femoral abscission were performed. The cadavers were maintained at − 20°C, and before dissection were thawed gradually to minimize tissue damage. The dissection was carried out using an operating microscope (×2.5, Surgical Acuity, Meddleton, WI). The abdominal wall, small intestine and sigmoid colon were removed, leaving the rectum intact. After widening the prevesical space and removing the prevesical fat, the EPF was exposed. After this anatomical approach to the EPF, to reach the paraprostatic space, the deep dorsal vein complex was gathered together and the apex of the prostate exposed. The pubic bone was then removed and the nerves near the rhabdosphincter dissected.
For the clinical study, 23 patients with clinically localized (T1-2N0M0) prostate cancer had a retropubic RP using the new technique for approaching the EPF. To avoid positive margins we included in this pilot study only patients with a PSA level of <20 ng/mL and biopsy specimens that did not contain Gleason grade 5 disease. The patients had a mean (sd) age of 68.2 (4.8) years, PSA level of 11.2 (6.4) ng/mL and a mean follow-up of 13.5 (2.3) months. The urinary catheter was removed 1–3 weeks after surgery. Patients who did not use incontinence pads were defined as continent. Interviews were conducted at regular intervals until the patient reported being completely continent.
By sectional macroscopic observation we identified the levator ani (LA) muscle and the fascia of the LA (FLA) in frontal and axial sections. Frontal sections along the membranous urethra in formalin-fixed cadavers showed the FLA as a distinct structure adherent to the prostate-urethral junction (Fig. 1A). In axial sections from another formalin-fixed specimen, the FLA on the right side was not attached to the lateral pelvic fascia (LPF), and only a thin fascia connected them. We considered this thin layer to be the EPF. The FLA on the left side was attached to the LPF, and there was a space between the FLA and LA (Fig. 1B). Spaces between the FLA and the LPF, or between FLA and LA, were considered artefacts created in preserving the specimen. The FLA sometimes adhered to the LPF, but these fascial structures did not fuse, and could be separated by dissection.
In the fresh cadavers we completely removed pre- and perivesical fat in a procedure similar to pelvic surgery, and identified the EPF and the lateral surface of the rectal wall. A white condensed area represented the overlap of the EPF and FLA. We set out to incise the EPF according to this anatomical knowledge, and to refine surgical techniques for establishing access to the paraprostatic space. We identified the FLA attached to the anterolateral side of the bladder and prostate underlying the thin fibrous layer (Fig. 2A). When the thin layer representing the EPF was incised within the attachment points of the FLA (Fig. 2B), the LA was removed laterally still covered by the FLA, without visualizing the muscle fibres (Fig. 2C). The incision and removal of the LA was extended carefully in an anteromedial direction to the apex of the prostate. Displacing the apex with a specially devised notched retractor, we confirmed that the rhabdosphincter was not exposed (Fig. 2D). Removal of the pubic bone identified the FLA as the plate forming the pelvic floor, with the FLA bordering intrapelvic and infrapelvic regions (Fig. 2E). This procedure was completed in six of eight specimens; in the other two, when we tried to remove the LA laterally, the FLA was fused with the LPF near the apex of the prostate. In these specimens the prostate-urethral junction was reached after exposing the LA muscle fibres.
Small branches from the pudendal nerve reached the rhabdosphincter in all fresh cadavers (Fig. 3); these represented the sphincteric branch of the pudendal nerve. This branch was anterolateral to the rhabdosphincter. The mean (sd, range) distance from the lowest point of the FLA to the point where the nerve branch entered the rhabdosphincter was 5.5 (1.8, 3–8) mm. After exposing the LA fibres beneath the FLA, the distance between the pelvic floor and the nerve entry point decreased. The defect created in the FLA rendered this nerve vulnerable to injury. After giving rise to this branch, the pudendal nerve coursed to the penile hilum to become the dorsal nerve of the penis.
Microscopically, the FLA was thick and covered the inferomedial margin of the LA, bordering intrapelvic and infrapelvic regions. The FLA did not appear strongly adherent to the prostate, differing from the impression obtained by macroscopic observation, and appearing somewhat counter to the surgical practice of many urologists who leave the FLA on the visceral surface. The FLA overlying the lower inner corner of the LA radiated to the framework of the rhabdosphincter (Fig. 4A). Staining with anti-smooth muscle actin indicated that the lower part of FLA was rich in smooth muscle, and that this component interdigitated with the rhabdosphincter (Fig. 4B). In two of the eight specimens examined histologically, many vessels were interposed between the FLA and LPF or situated under the LPF (Fig. 4C). In these specimens, excessive bleeding would have resulted unless the prostate-urethral junction was reached by exposing the LA muscle fibres.
In the clinical study, all 23 men regained continence, attained immediately after catheter removal in three (13%), within 1 week in eight (35%), and within 1, 3, 6 and 9 months in 10 (44%), 19 (83%), 22 (96%) and all, respectively.
Textbooks on urological surgery describing retropubic RP direct urologists to incise the so-called EPF to reach the paraprostatic space. Generally, the ‘EPF’ is thought to refer to the fascia in the transitional area between the pelvic wall and pelvic viscera. Many descriptions not based on anatomical study resemble one that reads, ‘After the EPF is incised just lateral to the white line, bare levator muscle fibres are viewed, which then are displaced bluntly and laterally from the lateral surfaces of prostate and rectum’. In these accounts, the existence and implications of the FLA are not considered. When we reviewed anatomically based reports of the EPF the term was used in two different senses. One meaning used by Steiner  referred to the parietal fascia lateral to the so-called arcus tendineus fascia pelvis, distinguishing the EPF from the FLA. On the other hand, Myers  equated the EPF with the FLA. We think that the intrapelvic fascial anatomy has not been fully elucidated, and that the terminology is inconsistent.
Figures in anatomical textbooks [8,9] tend to show the FLA as a very thick membrane not attached to the LPF, while the parietal pelvic fascia is not shown. Anatomists who work with formalin-fixed cadavers and have not performed surgery use the term ‘EPF’ in the same sense as ‘FLA’, and cannot understand references to ‘incision of the EPF’. We think that the anatomical discrepancy between observations from surgical and formalin-fixed cadavers has resulted in confusion about the EPF. Steiner  and Myers , who almost always used fresh cadavers for anatomical study, recognized the existence of the FLA. The present sectional macroscopic observations in formalin-preserved cadavers sharply contrasted the thick FLA and the thin parietal pelvic fascia, as being distinctly different structures (Fig. 5). While we support Steiner's opinion, the approach to the paraprostatic space has not been described in detail. In ordinary practice, surgeons who recognize the existence of the FLA and those who do not, incise two fascial planes, the parietal pelvic fascia and the FLA, when they believe that they are incising the EPF.
We used the term ‘EPF’ to refer to the parietal and visceral pelvic fascia, similar to Steiner , but we approached them by a slightly different method (Fig. 5). In the fresh-cadaver study, we incised the thin parietal and visceral pelvic fascia just medial to the attachment point of the FLA, and removed the LA laterally while covered by the FLA, without visualizing the muscle fibres. This point sometimes was anterolateral rather than lateral to the prostate side. Conventionally, the incision could be made just lateral to this point (i.e. the arcus tendineus pelvic fascia). Thus, we advocate a new approach to the EPF based on anatomical study of both formalin-fixed and fresh cadavers. However, the new approach cannot be applied to all cases; we encountered some specimens where the FLA fused with the LPF near the apex of the prostate, or where many vessels coursed between the FLA and LPF or under the LPF. In such cases, the approach between FLA and LA must be changed to avoid excessive blood loss.
Many recent studies have discussed the neuroanatomy of the rhabdosphincter. Hollabaugh et al. reported that both the pelvic and pudendal nerves supplied intrapelvic branches that coursed bilaterally, entering the external sphincter at the 5 and 7 o’clock positions. Narayan et al. measured the distance from the prostate apex to the point of nearest pudendal branch entry into the sphincter as 3.2–12.7 mm, similar to the present results (3–8 mm). While variable, this distance sometimes was very short (Fig. 3), indicating that surgeons must be very careful manipulating in the prostatic apex and its interface with the LA so as not to injure the nerve branch to the rhabdosphincter. The present new approach to the EPF should prove very useful in this regard.
The LA is considered to be important in continence mechanisms . Murakami et al. described the rhabdosphincter as acting not only as a sphincter but also as a retractor, levator, or force transmitter with the aid of the LA. The present immunostaining finding, that smooth muscle tissue in the lower part of the FLA interdigitated with the rhabdosphincter, supports this hypothesis. In terms of innervation, Juenemann et al. suggested that the rhabdosphincter and LA formed a functional complex. Interestingly, Nelson et al. described intraoperative electrical stimulation of the neurovascular bundle causing an increase in urethral pressure, functionally identifying the intrapelvic neural pathway that innervated the male sphincter. An additional increase in urethral pressure upon pelvic wall stimulation represented technical failure, but the result nonetheless emphasises the significance of the LA for urinary continence. Akita et al. reported various communications between the branches of pelvic plexus and pudendal nerve; these communicating branches penetrated the LA. Thus preserving the LA is very important for maintaining continence.
There are conflicting reports on the risk factors for urinary incontinence after RP, e.g. patient age, preoperative continence status, previous TURP, anastomotic stricture, stage of disease and, of course, surgical technique and the experience of the surgeon. Applying this new technique in the present pilot study led to a good continence rate and a rapid return of urinary control. The present continence results are better than in other recent reports from large series [16–20], where continence rates were 88.8–99.5%. In the present study, 83% of patients were continent at 3 months, which compares favourable with a report by Eastham et al., where 75% of patients were continent at 4 months.
Currently we cannot comment on whether the new surgical procedure causes more positive surgical margins because we limited patients in this pilot study to those who were strongly predicted to have localized tumours. However, it is evident from our anatomical study that the new approach does not lead to more positive margins than the usual nerve-sparing approach .
In conclusion, this anatomical study of the EPF led us to a new surgical approach to the EPF. The area under the FLA just lateral to the prostatic apex and the sphincter should be considered important for urinary continence. In retropubic RP, this new approach should facilitate preservation or early recovery of urinary continence, as preserving the FLA leads to protection of the LA and the rhabdosphincter, as well as avoiding the pudendal nerve branch to the rhabdosphincter.