The prevalence of rectocele in women ranges from 12.9 to 18.6 % and the average annual incidence is estimated to be 5.7 cases per 100 women years [1, 2]. Approximately 225,000 operations are performed every year in the United States for pelvic organ prolapse and repair of posterior vaginal wall is performed in between 40 and 85 % [3–5].

Terms used to describe the supportive tissue utilised for posterior wall prolapse repair date back to a publication by Francis and Jeffcoate in 1961 [6]. The concept of site-specific repair stems from observations by Richardson, who felt that discrete defects in what he termed rectovaginal fascia should be addressed by a directed repair of the defect, thus producing a more anatomical repair in his opinion [7]. In recent years there has been a clearer understanding of the anatomy of the posterior vaginal wall and its support; however, surgical studies in this area continue to utilise terms that are not anatomically based. The goal of this review is to provide a brief overview regarding current understanding of the anatomy of the support of the posterior vaginal wall, discuss various anatomical defects that can clearly contribute to posterior vaginal wall prolapse, and finally update the previous ICI report by citing the various studies that have reported outcomes for the surgical correction of posterior vaginal wall prolapse.

Anatomy of the posterior vaginal wall

Historically, pelvic organ support, as it relates to the anterior and posterior vaginal wall compartments, has been described in relation to supportive tissues termed endopelvic fascia. In the anterior segment this has been called pubocervicovesical fascia and in the posterior segment this has been called rectovaginal fascia. Histological studies have noted that what has previously been termed fascia is actually vaginal muscularis in both the anterior and posterior segments [8]. With respect to the rectovaginal fascia it is now well appreciated that there is no such layer between the posterior vaginal wall and the anterior wall of the rectum. At the level of the mid vagina histological assessment of the posterior vaginal wall from the lumen of the vagina to the lumen of the rectum notes the following layers: the vaginal epithelium, the lamina propria of the vagina, the fibromuscular wall of the vagina (smooth muscle cells, elastin and type II collagen), the adventitia, the outer muscular wall of the rectum, the inner muscular wall of the rectum, the lamina propria of the rectum and the rectal mucosa.

DeLancey performed cross-sections on nulliparous and multiparous cadavers and noted that the support of the posterior compartment was maintained by a complex interaction of connective tissue and levator ani muscle [9]. He went on to note that the support of the posterior vaginal wall was best divided into three separate and distinct levels of support. Level III support, being the most distal portion of the vagina, is provided by the perineal membrane and the rectovaginal septum. This level of support has strong attachments to the levator ani complex and is thus less susceptible to pelvic pressure transmission that may cause prolapse. Level II, or the mid-vagina, is supported by its attachments of the vaginal muscularis laterally to the fascia of the levator ani muscles. Level I support, or the upper vagina, is supported by the cardinal–uterosacral ligament complex.

This same group of investigators recently discussed posterior vaginal wall anatomy in a review article and likened it to an open container. The front wall of the container would be formed by the posterior vaginal wall while the bottom of the container is made up of the perineal body and anal sphincters. The levator ani muscles form the lateral sides of the container and the levator plate where the muscles decussate behind the rectum to create the iliococcygeal raphe form the back wall of the container. The uppermost portion of the container would be that created by the attachment of the posterior vaginal wall to the uterosacral ligament, which extends below the peritoneum. All of these boundaries are subject to defects that can give rise to different structural failures [10].

Anatomical defects that may contribute to prolapse of the posterior vaginal wall

The patient who presents with prolapse of the posterior vaginal wall either in isolation or in conjunction with prolapse of other segments of the pelvic floor could potentially have a posterior enterocele, a rectocele, or a sigmoidocele [11]. These three conditions can occur in isolation or in conjunction with each other and will commonly be accompanied by a perineal defect and/or a widened genital hiatus. While it is beyond the scope of this article to discuss clinical and radiographic mechanisms that can be used to differentiate these various defects, suffice it to say that they can all result in descent of the posterior vaginal wall to various degrees.

The published literature continues to classify posterior vaginal repairs into what has been termed a “traditional technique”, which implies that the repair has been supplemented with a levator ani muscle plication in the midline, or a “site-specific technique”, which implies that discreet defects in the rectovaginal fascia are identified and repaired and no levator plication is performed. To date we are unaware of any studies that have addressed how often a posterior enterocele and or sigmoidocele coexist with a rectocele and how the presence of these defects impacts on ultimate surgical outcomes. Based on our current understanding of the anatomy of the posterior vaginal wall and perineum it is clear that the defect specific repairs involve plication of the fibromuscular layer of the posterior vaginal wall and based on the initial level of dissection this tissue may be found on the anterior wall of the rectum or may have to be mobilised off the vaginal epithelium to allow an appropriate tension-free plication. In patients with advanced prolapse and a widened genital hiatus the only way to address the gaping vagina is to routinely perform a distal levatorplasty. In the authors’ opinion, future surgical studies assessing outcomes of prolapse repair involving the posterior vaginal wall should take into consideration these points, and realise that these procedures are not mutually exclusive and that a combination of the techniques, especially in cases of advanced prolapse, is commonly required. Other types of repairs that have been reported include transanal repairs, transperineal mesh (biological or synthetic) augmented repairs, and abdominal sacral colpopexy, in which the mesh attachment is extended down to the distal portion of the posterior vaginal wall and/or perineum. The authors have also observed that aggressive reattachment of the uppermost portion of the full thickness of the posterior vaginal wall (level III support) to the uterosacral ligament provides significant support to the posterior vaginal wall in patients with high rectoceles or rectoceles in conjunction with a posterior enterocele (Fig. 1).

Fig. 1
figure 1

Demonstrates that the rectocele (a) is corrected with suspension of the posterior vaginal wall to the uterosacral ligaments

Full size image

Midline plication or traditional posterior colporrhaphy

The mean reported anatomical success rate with this type of repair is 83 % (range 76–96 %) with a mean postoperative dyspareunia rate of 18 % (range 5–45 %) and 26 % using vaginal digitation to defecate. (Table 1)

Table 1 Midline plication or traditional posterior colporrhaphy
Full size table

Site-specific defect repair

This technique is similar to traditional post-repair in terms of dissection. The aim of the repair is for the surgeon to identify and individually correct breaks in the rectovaginal septum. Traditional levatorplasty is avoided. The mean anatomical success rate is 83 % (range 56–100 %) with 18 % postoperatively needing vaginal digitation to defecate and 18 % experiencing postoperative dyspareunia (Table 2).

Table 2 Site-specific posterior vaginal repair
Full size table

Abramov et al. retrospectively compared the midline fascial plication and discrete site-specific repair for rectoceles [25]. They noted a significantly higher recurrence rate of rectoceles following the discrete site-specific repair (32 % vs 13 % following midline fascial plication (P = 0.015). The correction of the rectovaginal fascia defect that allows entrapment of faeces on straining in significant rectoceles may be too large to be repaired using the discrete approach [22] and appears to be corrected by the more robust midline fascial plication.

In a randomised control trial Paraiso et al. compared three techniques for rectocele repair in a prospective randomised trial [26]. Patients were randomised to receive either a traditional repair (n = 37), a site-specific repair (n = 37) or a site-specific repair augmented with porcine small intestine mucosa (n = 32). Preoperatively, all patients had stage II or greater posterior vaginal wall prolapse. The objective anatomical failure rate was highest in the graft-augmented group (12 out of 26) at 1 year, which was statistically significantly worse than results in the site-specific group (6 out of 27) and traditional repair (4 out of 28). There was no significant difference in subjective symptoms (worsening prolapse or colorectal symptoms) or dyspareunia among the three groups.

Recently, Sung et al. reported on a double blind multicentre randomised control trial comparing native tissue repair (70) or native tissue porcine subintestine submucosal (SIS) graft (67) for symptomatic grade 2 rectocele [27]. The native tissue repair involved either a midline plication or a site-specific repair, with the majority undergoing site-specific repair. In the graft group this repair was augmented with porcine SIS overlay. At 1 year there was no difference between the groups in objective and subjective success rates or in resolution of defecatory symptoms. Postoperative dyspareunia rates were not significantly different at 7 % in the native tissue group and 12.5 % in the graft group.

In a single trial Vijaya [28] reported at 6 months’ follow-up that superior support of the posterior vaginal wall was attained after the fascial plication compared with levator ani repair. Block randomisation was performed with 26 women randomised to each operation. Allocation concealment, power analysis and status of reviewers were not reported. Anatomical outcomes were reported via Pop-Q point AP and a variety of quality of life assessments were performed without reporting of the data.

Transanal repair of rectocele

Three trials have evaluated transanal vs transvaginal repairs of rectoceles [29–31]. Each trial had slightly different inclusion criteria. Kahn and Stanton included women who had symptoms of prolapse or impaired rectal evacuation with incomplete emptying on isotope defaecography and normal compliance on anorectal manometry [29]. Nieminen et al.’s included women with symptomatic rectoceles not responding to conservative therapy. Importantly, women with compromised anal sphincter function and other symptomatic genital prolapse were excluded. In both trials the vaginal repair was performed by gynaecologists and the transanal repair by colorectal surgeons. In Kahn and Stanton’s trial the posterior vaginal wall repair was performed using levator plication and in Nieminen’s trial the rectovaginal fascia was plicated. Farid et al.’s inclusion criteria required women to have a rectocele larger than 2 cm on defaecography with symptoms including digitation, incomplete evacuation, excessive straining and dyspareunia. Women with a compromised anal sphincter complex or recurrent prolapse, rectal prolapse, intussusception, or anismus were excluded. The surgery was performed within the surgery department and blinded examiners utilised defaecography, anal manometry and a modified obstructed defecation syndrome patient questionnaire to report outcomes.

Based on these three trials we can conclude that the results for transvaginal repair of rectocele are superior to those of transanal repair of rectocele, in terms of subjective and objective outcomes. In women with rectocele alone recurrent rectocele occurred in 2 out of 39 women in the vaginal group and 7 out of 48 following the transanal repair, a difference that did not reach statistical significance. Postoperative enterocele was, however, significantly less common following vaginal surgery compared with the transanal group.

Farid et al. [31] reported on outcomes of three types of rectocele repair comparing transperineal repair using levatorplasty with transanal repair and noted conclusions similar to the two previously discussed trials. The rectocele on defaecography was significantly smaller in the transperineal group (with or without levatorplasty) compared with the transanal repair. Also, functional outcome based on a modified obstruction defecation syndrome patient questionnaire was better after transperineal repair than after transanal repair.

Puigdollers et al. reported results from a prospective cohort of women with rectocele and constipation who underwent surgery via either an endorectal or a transperineal approach based on surgeon preference [32]. At the end of 1 year an overall subjective improvement in constipation was reported in 43 % (P < 0.001) and the need to splint decreased in 52 % (P = 0.001).

Thornton et al. [33] reported in a single non-randomised study outcomes for a cohort of women with symptomatic rectocele who were treated laparoscopically (n = 40) vs transanally (n = 40). Level 2B evidence from this study supports the superiority of the transanal approach for symptom relief (55 % vs 28 %, P < 0.02), but lower postoperative dyspareunia rates (22 % vs 36 %) using the laparoscopic approach.

Van Dam et al. [34] performed a combined transvaginal and transanal repair in 89 women who were evaluated at a follow-up of 52 months. The anatomical success rate was 71 % (defined as no persistent or recurrent rectocele on defaecography at 6 months). However, de novo dyspareunia was reported in 41 % of women and there was a deterioration in faecal maintenance in 7 patients.

Graft augmented rectocele

Sand et al. [16] compared posterior repair with and without mesh and noted that rectocele recurrence appeared equal with and without polyglactin (vicryl) mesh augmentation (7 out of 67 vs 6 out of 65). Neither Paraiso et al. or Sung et al. noted any benefit to augmenting a native tissue repair with a porcine subintestine submucosal graft overlay [26, 27]. Mesh exposure was not reported in these trials. Altman et al. reported on a prospective evaluation of insertion of a 7 × 4cm porcine dermis graft at 3 years and found a 40 % recurrence rate on examination ,and while there was a significant decrease in rectal emptying difficulties compared with preoperatively less than 50 % reported cure of rectal emptying issues [35]. There was no change in the rate of anal incontinence or dyspareunia post-intervention.

Modified sacrocolpopexy

The abdominal route has been employed in the correction of posterior vaginal wall prolapse when a co-existing apical defect requires surgery. The technique is a modification of sacrocolpopexy with extension of the posterior mesh down to the distal posterior vaginal wall and/or the perineal body. The procedure has been reported completely abdominally or as a combined abdominal and vaginal approach. Table 3 summarises a series of studies that have reported on extended posterior fixation of sacrocolpopexy mesh.

Table 3 Abdominal repair (posterior extension of colpopexy mesh)
Full size table

Summary

  • Transvaginal repair of posterior vaginal wall prolapse continues to be reported as a traditional repair with levatorplasty, midline fascial plication without levatorplasty or site-specific repair. Level 1 and 2 evidence suggest that midline plication posterior repair without levatorplasty has superior objective outcomes compared with site-specific posterior repair (grade B).

  • Higher dyspareunia rate reported when levatorplasty is employed (grade C).

  • Transvaginal approach is superior to the transanal approach for repair of posterior wall prolapse (grade A).

  • To date no studies have shown any benefit of mesh overlay or augmentation of a suture repair for posterior vaginal wall prolapse (grade B).

  • While modified abdominal sacrocolpopexy results have been reported, data on how these results would compare with traditional transvaginal repair of posterior vaginal wall prolapse are lacking.