While anterior vaginal prolapse is most common, loss of apical support is usually present in women with prolapse that extends beyond the hymen [1, 2]. There is growing recognition that adequate support for the vaginal apex is an essential component of a durable surgical repair for women with advanced prolapse [3, 4]. Because of the significant contribution of the apex to anterior vaginal support, the best surgical correction of the anterior and posterior walls may fail unless the apex is adequately supported [5, 6]. While recognition of apical defects is one of the biggest challenges in the preoperative evaluation of pelvic support defects, surgical correction of the apex has several good options with relatively high success rates. Apical suspension procedures can be broadly separated into those performed transvaginally and those performed abdominally. Abdominal procedures can be performed via laparotomy or using conventional laparoscopic or robotically assisted-laparoscopic techniques. Although precise estimates are not available, most studies suggest that the vaginal approach is most common with 80–90 % of procedures being performed through this route [7–11]. The individual woman’s surgical history and goals, as well as her individual risks of surgical complications, prolapse recurrence and de novo symptoms affect surgical planning and the choice of procedure for apical POP.

Sacrospinous ligament suspension

One of the most popular and widely reported transvaginal procedures for correcting apical prolapse is sacrospinous ligament suspension (SSLS). First described in 1958 [12], this procedure suspends the vaginal apex to the sacrospinous ligament, either unilaterally or bilaterally, typically using an extraperitoneal approach. Observational series and clinical trials suggest that while apical recurrence after SSLS is uncommon (2.4 to 19 %), recurrence of anterior vaginal prolapse is more problematic (6 to 28.5 %; Table 1). A meta-analysis by Morgan et al. found an overall failure rate at any site of 28.8 % (95 % CI 18.4–36.3 %), with failure of the anterior segment seen in 21.3 % (17.3–25.3 %), apical segment in 7.2 % (95 % CI 4.0–10.4 %) and posterior segment in 6.3 % (95 % CI 4.2–8.4 %). Whether the relatively high rate of anterior vaginal prolapse recurrence seen with SSLS is due to the posterior deflection of the vaginal axis, as many authors suggest [15, 17, 18, 20], or simply represents a general predilection of anterior support to fail after pelvic reconstructive surgery remains unknown [32]. Reoperation rates after SSLS range from 1.3 to 37 %, with all but two series reporting rates of less than 9 % (Table 1).

Table 1 Outcomes of sacrospinous ligament suspension (SSLS) procedures
Full size table

Information on the functional or QOL outcomes of SSLS is limited. Maher et al. demonstrated significant improvements in condition-specific and generic QOL after SSLS, similar to that after abdominal sacral colpopexy [27]. A meta-analysis of randomised and observational studies found a pooled average for failure to provide relief of prolapse symptoms after SSLS of 10.3 % (95 % CI 4.4–16.2 %) [33]. The pooled average for failure to provide patient satisfaction after SSLS in this analysis was 13 % (95 % CI 7.4–18.6 %) [33]. Although infrequent, serious complications associated with SSLS include buttock pain and sacral/pudendal neurovascular injury. In a review of 22 studies that included 1,229 SSLS procedures, 3 patients (0.2 %) had life-threatening haemorrhage from sacral or pudendal vascular injury and the overall transfusion rate was 2 % [22]. Buttock pain occurred in 3 % of subjects, the vast majority of which resolved by 6 weeks postoperatively [22].

Uterosacral ligament suspension

Uterosacral ligament suspension (USLS) was first described by Miller [34] in 1927 and was later popularised by Shull in the late 1990s [3]. The USLS suspends the vaginal apex to the proximal remnants of the uterosacral ligaments using an intraperitoneal surgical approach. This procedure restores the vagina to its normal axis, avoiding the retroflexion associated with SSLS. The current evidence supporting the use of USLS is limited primarily to uncontrolled retrospective case series and evaluation of these data confirm a mean objective success rate of 85 % (range 48–96 %) and a mean reoperation rate for prolapse of 5.8 % (range 0–12 %; Table 2). A meta-analysis performed by Margulies et al. found pooled rates of anatomical success (POPQ stage 0–1) of 81.2 % (95%CI 67.5–94.5 %) for the anterior segment, 98.3 % (95 % CI 95.7–100 %) for the apical segment and 87.4 % (95 % CI, 67.5–94.5 %) for the posterior segment [47]. Postoperative prolapse symptoms were reported in 5 of the 11 studies in this review and were relieved in 82–100 % of patients. These promising results are balanced by a ureteral kinking/injury rate of 1–11 % with this procedure [47]. A review of 700 consecutive vaginal prolapse surgeries found intraoperative ureteral kinking/injury of 5.9 % directly attributable to USLS. However, 87 % were identified at cystoscopy before the completion of the index surgery and were relieved by removing suspension sutures intraoperatively with no long-term consequences for the patient [48]. Only 3 of the 355 USLS (0.9 %) performed in this series required additional procedures to relieve or correct ureteral obstruction or injury. Margulies et al. identified 10 studies, including a total of 820 women, that reported on the perioperative complications of USLS [47]. The ureteral reimplantation rate in this series was only 0.6 %. Blood transfusions were reported in 1.3 %, cystotomy in 0.1 %, and bowel injury in 0.2 %. To date, no clinical trials comparing the USLS with sacral colpopexy or with SSLS have been published. The Pelvic Floor Disorders Network has an ongoing RCT comparing USLS with SSLS that includes 440 subjects who will be followed for 2 years; results from this trial are expected in 2013 [49].

Table 2 Outcomes of transvaginal uterosacral vault suspension procedures
Full size table

While the USLS is traditionally performed using an intraperitoneal approach, Dwyer and Fatton have described an extraperitoneal variant of the USLS [50, 51]. In their series of 123 consecutive women undergoing an extraperitoneal USLS, 93 also received anterior and/or posterior synthetic mesh. The overall anatomical success (POPQ stage 0–1) at a mean follow-up of 2 years (range 6 months to 5 years) was 85.5 %, with apical success of 95.4 % [51]. The reoperation rate for recurrent prolapse was 7 %. Ureteral injury occurred in only 1.7 %; however, the blood transfusion rate was 4.9 % and the rate of mesh exposure was 19.3 %.

Abdominal and laparoscopic USLS techniques have also been described. Lowenstein et al. reported a retrospective review of 107 women who underwent prolapse surgery that included an abdominal USLS [52]. In the 75 patients who completed the 1-year follow-up, 12 % reported recurrent or persistent prolapse symptoms and 7 % had an anatomical failure (POPQ stage 2 or greater). Complications were relatively few; however, erosion of the apical sutures (expanded PTFT, Gore-Tex) occurred in 9 % at an average time of 56 months (range 3–75 months) [52]. Rardin et al. reported a retrospective comparison of 96 patients undergoing vaginal USLS with 22 undergoing a laparoscopic USLS procedure and found no significant differences in perioperative morbidity or anatomical or subjective outcomes [53].

Mayo/McCall’s culdoplasty

Like the USLS, the Mayo/McCall’s culdoplasty uses the proximal uterosacral ligaments to suspend the vaginal apex. The major difference is that with the Mayo/McCall procedure the uterosacral ligaments are plicated in the mid-line to obliterate the posterior cul-de-sac. While commonly performed, data describing the outcomes for this procedure are limited (Table 3). Colombo and Milani retrospectively compared the outcomes of a modified McCall’s culdoplasty with those of SSLS (n = 62 in each group) [20]. Recurrence after the McCall’s culdoplasty (Baden–Walker grade ≥2) was 15 % 4 to 9 years after surgery and was not significantly different from the SSLS group. Recurrent anterior vaginal prolapse occurred less frequently in McCall’s group than in the SSLS group (6 vs 21 %, p = 0.04; OR 4.1 [95 % CI 1.3 to 14.2]) [20]. A large retrospective series of 693 women from the Mayo clinic described an 82 % satisfaction rate on subjective follow-up with few complications [54]. The rate of subsequent prolapse repair in this population was 5.2 %. A retrospective case series of 411 women undergoing Mayo culdoplasty found that a more dorsal “deep” placement of sutures through the uterosacral ligaments reduced the incidence of ureteral obstruction compared with other published series [57].

Table 3 Mayo/McCall’s culdoplasty
Full size table

Levator myorrhaphy

In 1961, Francis and Jeffcoate described their retrospective series using levator myorrhaphy in which a wide midline plication of the levator ani muscles is performed to which the vaginal cuff is fixed [58]. A large sponge pack in the rectum is used to avoid overplication and bowel dysfunction. Five of the 35 women responding to the questionnaire had transient ureteral complications, one requiring reoperation. Seventeen women were quite satisfied, while 6 were dissatisfied. Natale et al. compared high levator myorrhaphy with USLS in a randomised clinical trial of 229 women with stage 2–4 prolapse [59]. All women underwent a hysterectomy and all received placement of polypropylene mesh in the anterior vaginal segment. Anatomical success was not significantly different between groups. The mean total vaginal length was significantly shorter after levator myorrhaphy (7.9 vs 8.9 cm, p = 0.04). Urinary, bowel and sexual function did not differ between groups postoperatively. Intraoperative ureteral obstruction was less common in the levator myorrhaphy group (0 vs 7.9 %); however, all cases of ureteral obstruction in the USLS group were corrected intraoperatively with suture removal/replacement with no additional interventions required [59]. Other complications including mesh erosion were similar in the groups.

Iliococcygeus fascia fixation

There are no randomised trials that support the use of this procedure. Several case series have provided some information. Shull reported that apical support was optimal in 39 out of 42 of patients (83 %), but 8 others had apical or other defects [60]. Meeks and colleagues reported a 96 % objective cure in 110 women followed for up to 13 years [61]. In a retrospective case–control study, Maher and colleagues reported similar subjective (91 vs 94 %) and objective (53 vs 67 %) cure rates with iliococcygeus fixation (n = 50) compared with sacrospinous fixation (n = 78) [62].

Transvaginal mesh apical prolapse

Two randomised control trials evaluated transvaginal polypropylene meshes in apical prolapse. Sokol et al. reported a multicentre double-blinded RCT comparing uterosacral colpopexy and native tissue repair (n = 33) with a monofilament polypropylene mesh kit (Prolift, n = 32; Ethicon, Somerville, NJ, USA) for stage 2 or greater uterovaginal prolapse or vaginal prolapse [63].

At 1 year the conventional surgery group had no subsequent surgical interventions compared with 15.6 % in the mesh group (p = 0.017) including 3 for prolapse surgery (2 sacral colpopexy and 1 iliococcygeous fixation) and 2 interventions for mesh exposure. The objective failure rate (any stage 2 or greater prolapse) was 70 % in the conventional surgery group vs 63 % in the mesh group (p > 0.05). The subjective failure rate was also similar in the two groups, 9.1 % in native tissue repairs vs 3.8 % in the mesh group. One patient was transfused and 2 inadvertent cystotomies occurred in the mesh group with no perioperative complications reported in the native tissue group. No differences were seen between the groups utilising a wide variety of validated outcome tools. Unfortunately, owing to the ethics committee imposing a stopping criterion of 15.6 % mesh exposure rate, the study did not recruit the appropriate sample size and is underpowered to detect a significant difference between the groups if it exists.

Maher et al. recently reported results from a randomised trial comparing laparoscopic sacral colpopexy (LSC; n = 53) with a total vaginal mesh kit (Prolift; Ethicon, Somerville, NJ, USA) (n = 55) [64]. LSC was associated with longer operating times (mean difference +52 min [95 % CI 41.5–62.6]), decreased hospital stay (mean difference −0.5 days [95 % CI −0.93 to −0.10]) and quicker return to normal activities (mean difference −5.3 days [95 % CI −8.4 to −2.3]). Two years after surgery, objective success (overall POPQ Stage 0 or 1) was seen in 77 % of the LSC group compared with only 43 % of the TVM group, p < 0.001) [64]. Also, reoperations were significantly higher in the TVM group (22 %) than in the group that received LSC (5 %, p = 0.006).

As seen in Table 4 the success rate of transvaginal meshes for apical prolapse in level 3 evidence is significantly higher and ranges from 87 to 100 % for monofilament polypropylene meshes with mesh erosion rates varying from 0 to 15 %.

Table 4 Outcomes of transvaginal mesh kits used for apical repairs
Full size table

Sacral colpopexy

Since its introduction by Lane in 1962 [73], sacral colpopexy has been proven to be an effective and durable technique for correcting apical prolapse. In 2010, approximately 34,000 sacral colpopexies were performed in the USA, representing 11 % of all prolapse surgeries performed during that time period [11]. Traditionally, sacral colpopexy has been performed via a laparotomy (i.e. abdominal sacral colpopexy), but the use of laparoscopic and robotic approaches is increasing.

Abdominal sacral colpopexy

Observational studies and clinical trials suggest that abdominal sacral colpopexy (ASC) is a highly effective procedure for apical prolapse. The success rate of ASC, when defined as lack of apical prolapse, ranges from 78 to 100 % (Table 5). When success is defined as no recurrent prolapse in any segment the published success rates are 56–100 %. A systematic review of ASC performed by Nygaard et al. reported a median reoperation rate for recurrent prolapse of 4.4 % (range 0–18.2 %) and for postoperative stress incontinence of 4.9 % (range 1.2 to 30.9 %) [110]. Clinical trials demonstrate significant improvements in prolapse symptoms, urinary function and quality of life after ASC [27, 106]. There is level 1 evidence that ASC has superior anatomical outcomes compared with SSLS, but this is balanced by longer operating time, longer recovery and higher cost [111]. There are no randomised trials comparing ASC with ULS or with transvaginal mesh procedures. Given the prolonged recovery and unique complications associated with laparotomy, many surgeons reserve sacral colpopexy for patients with apical prolapse thought to be at high risk of failure from a vaginal approach, often considering such factors as age, comorbidities, history of previous prolapse surgery and vaginal length [3, 4, 10, 32, 112]. Unfortunately, there are too few published data to allow an evidence-based decision about which patient with POP will be best served by an ASC relative to other techniques.

Table 5 Abdominal sacral colpopexy outcomes
Full size table

Some surgeons have attempted to decrease mesh complications of ASC by using biological materials instead of synthetic mesh. However, the current evidence suggests that biological materials, whether allograft or xenograft, produce inferior anatomical outcomes compared with synthetic mesh, particularly polypropylene, without decreased graft-related complications. Level 1 evidence supports the superiority of polypropylene mesh to fascia lata for objective anatomical support following ASC [109, 113]. A randomised trial of 106 women undergoing ASC compared polypropylene mesh with cadaveric fascia lata and found superior anatomical outcomes in those who received polypropylene at 1 year (success 91 vs 68 %, p = 0.007) and 5 years after surgery (93 vs 62 %, p = 0.02) [109, 113]. There were no differences in graft-related complications overall between the two groups. Several retrospective case series support these data [114, 115, 116). Similarly, level 3 evidence suggests that use of xenografts such as porcine dermis and small intestinal submucosa might also have inferior anatomical success rates compared with polypropylene mesh, with similar rates of graft-related complications [117, 118].

Beyond mesh erosion, reported complications of ASC are generally consistent with those of other major open pelvic surgeries. The systematic review by Nygaard et al. reported that wound complications occurred in 4.6 % (range 0.4 to 19.8 %), haemorrhage or transfusion in 4.4 % (0.2 to 16.9 %), cystotomy in 3.1 % (0.4 to 15.8 %), ureteral injury in 1.0 % (0.8 to 1.9 %) and bowel injury in 1.6 % (0.4 to 2.5 %) [110]. One in 20 women in the CARE trial experienced significant gastrointestinal morbidity after sacral colpopexy. Of 322 women in the study, 19 had symptoms of possible ileus or small bowel obstruction; of these, 4 underwent reoperation for small bowel obstruction, 11 were readmitted for medical management, and 4 had a prolonged initial hospitalisation for gastrointestinal symptoms [119].

Abdominal sacral colpopexy vs sacrospinous ligament suspension

To date, there have been three RCTs that directly compare ASC with SSLS [17, 27, 96]. The Cochrane review on the surgical management of POP by Maher et al. summarises these studies and concludes that these trials provide level 1 evidence that there were no statistically significant differences in objective failure at any site (any pelvic organ prolapse RR 0.77, 95 % CI 0.39 to 1.53), subjective failure (RR 0.53, 95%CI 0.25 to 1.09), reoperation for POP (RR 1.46, 95 % CI 0.19 to 1.11) or patient satisfaction (RR 0.82, 95 % CI 0.32 to 2.06) [111]. However, ASC was superior to SSLS for the following outcomes: prolapse ≤ stage 2 (RR 0.29, 95 % CI 0.09 to 0.97), recurrent vault prolapse (RR 0.23, 95 % CI 0.07 to 0.77), postoperative stress urinary incontinence (RR 0.55, 95 % CI 0.32 to 0.95) and less postoperative dyspareunia (RR 0.39, 95 % CI 0.18 to 0.86). In contrast, ASC was associated with a longer operating time (weighted mean difference [WMD] 21 min, 95 % CI 12 to 30), longer time to recover (WMD 8.3 days, 95 % CI 3.9 to 12.7) and was more expensive (WMD US $1,334, 95 % CI 1,027 to 1,641) than SSLS [111].

Laparoscopic sacral colpopexy

The laparoscopic approach to sacral colpopexy has been adopted by many surgeons over the last decade as an alternative to ASC with the hopes of reproducing the high success rate of the ASC while decreasing the morbidity and delayed recovery associated with laparotomy. Multiple prospective and retrospective case series demonstrate good short to mid-term success rates, with a mean objective success rate of 90.5 % (range 60–100 %, subjective success rates of 79–98 % [120–122] and a mean reoperation rate of 5.9 % (Table 6). To date, no randomised trials have compared LSC with ASC; however, three retrospective comparisons have been published [134, 140, 141]. While results vary somewhat among studies, in general, LSC is associated with a shorter hospital stay, less blood loss, with conflicting data on operating times. Objective outcomes among the groups appear to be similar. Well-designed high-quality clinical trials are necessary to establish independently the effectiveness and safety of the LSC relative to ASC. There is level 1 evidence that LSC provides superior outcomes to total vaginal mesh procedure for women with symptomatic stage 2–4 vaginal vault prolapse, as described above. There are currently no comparative studies, randomised or not, evaluating the relative safety and efficacy of LSC and native tissue (non-mesh) vaginal POP repair.

Table 6 Outcomes of laparoscopic sacral colpopexy trials reporting more than 12 months’ review
Full size table

A recent retrospective study assessed the complication rates in 402 LSC cases [132]. This study compared patients who received concurrent laparoscopically assisted vaginal hysterectomy with those who had had previous hysterectomy. They showed no differences in intra- or perioperative complications and similar rates of mesh erosion between the two groups [132]. Overall, the complication rates for this cohort were 0.75 % for haematoma, 2.2 % for ileus or small bowel obstruction, 1.5 % for bladder injury, 0.75 % for bowel injury and 0.25 % for ureteric injury. At 1 year, the overall mesh erosion rate was 1.2 %. In contrast, Tan-Kim et al. reported on a retrospective series of 188 minimally invasive sacrocolpopexies and found a significantly higher mesh exposure rate in those who received concurrent total vaginal hysterectomy (TVH; 23 %) compared with those who were post-hysterectomy (5 %) or who received a supracervical hysterectomy (5 %) [142]. TVH was found to be an independent risk factor for mesh erosion on multivariate regression analysis in this study (OR 5.67; 95 % CI 2.88–17.10).

Despite the clinical advantages of a laparoscopic approach, the adoption of LSC has been relatively limited, probably because of the steep learning curve associated with attaining laparoscopic suturing and knot tying skills that are required to attach the mesh to the vagina and sacrum. Claerhout et al. evaluated their learning curve in the first 206 cases performed by a single surgeon [130]. Operating times declined rapidly during the first 30 procedures in this series and reached a steady state (175 min) after 90 cases. Using a cumulative sum (CUSUM) approach to evaluate operative time and failures (laparotomy, complications or anatomical failures) they found that adequate learning occurred after 60 cases [130]. Complication rates remained unchanged throughout this series. Akladios et al. found that there was a steady decrease in LSC operative time in a series of the first 48 cases performed, but that a turning point was observed after 18–24 cases [127]. Complication rates were also low throughout this series.

Robotic sacral colpopexy

Because of the relatively long learning curve required for LSC, many surgeons have turned to robotic-assisted surgery in order to offer patients a minimally invasive approach to sacrocolpopexy. Robotic surgical systems have been developed with the goal of facilitating technically difficult procedures by improving the surgeon’s vision, dexterity and ergonomics. No data have been published on the learning curve for robotic sacral colpopexy (RSC); however, expert opinion suggests that the learning curve might be shorter for RSC than with the laparoscopic approach.

The currently available data for RSC are relatively limited and consist primarily of uncontrolled case series, but meta-analysis suggests anatomical outcomes similar to those of ASC and LSC, with objective success rates reported at 60–100 % (mean 93 %), subjective success of 91–94 % and a mean mesh erosion rate of 5 % (Table 7). To date, there have been only two published studies that provide comparative data for the RSC. Geller et al. retrospectively compared 73 patients who received RSC with 105 who received ASC [125]. RSC was associated with less blood loss, longer operative time, shorter length of stay and a higher incidence of fever (4.1 vs 0 %) Anatomical outcomes of the groups 6 weeks after surgery were similar [125].

Table 7 Robotic sacral colpopexy (RSC)
Full size table

In the only randomised comparison of RSC to date, Paraiso et al. recently published a clinical trial that provides level 1 evidence that RSC results in longer operating time and increased pain and costs compared with LSC [143]. This single-centre, blinded, randomised trial compared RSC (n = 40) with LSC (n = 38) in women with stage 2–4 post-hysterectomy vaginal prolapse. Total operative time was chosen as the primary outcome for this study serving as a proxy measure for surgical efficiency. Total operative time was significantly longer in the robotic group compared with the laparoscopic group (+67-min difference; 95 % confidence interval [CI] 43–89; P < 0.001) [143]. Anaesthesia time, total time in the operating room, total sacral colpopexy time and total suturing time were all significantly longer in the robotic group. Participants in the robotic group also had significantly higher pain at rest and with activity during weeks 3 through 5 after surgery and required longer use of non-steroidal anti-inflammatory drugs (median, 20 compared with 11 days, P < 0.005). The robotic group incurred greater costs than the laparoscopic group (@mean difference + $1,936; 95 % CI $417–$3,454; P = 0.008) [143]. Both groups demonstrated significant improvement in vaginal support and functional outcomes 1 year after surgery, with no differences between groups. It is worth noting that the surgeons in this study had considerable experience of LSC.

A meta-analysis of observational studies on robotic gynaecological surgery, found that the currently available evidence shows that for most gynaecological procedures studied robotic surgery achieved a shorter hospital stay and less blood loss than open surgery [150]. However, no clinically significant improvements were noted when robotic surgery was compared with conventional laparoscopic surgery in benign gynaecological procedures [150]. The current evidence, while limited, suggests that these conclusions are also applicable for RSC (Table 8). RSC probably has a shorter learning curve than LSC and thus may be more generalisable; however, published evidence for this is currently lacking. In surgeons with advanced laparoscopic skills, RSC offers no clinical benefit compared with LSC and results in longer operating times, greater cost and greater postoperative pain.

Table 8 Compares abdominal (ASC), laparoscopic (LSC) and robotic sacral colpopexy (RSC)
Full size table

Obliterative procedures

Obliterative surgery, such as total colpocleisis (also called colpectomy/colpocleisis) or the LeFort partial colpocleisis, corrects POP by reducing the pelvic viscera back into the pelvis and closing off the vaginal canal either in part or entirely [151]. Obliterative procedures are less commonly performed in Europe, Asia and Australia compared with the USA, and are usually reserved for women who are elderly, medically compromised and no longer sexually active [152]. The purported advantages of obliterative surgery in this population are decreased operative time, decreased perioperative morbidity, and an extremely low prolapse recurrence risk. The obvious disadvantage is the elimination of the potential for vaginal intercourse. Preoperative counselling is essential when choosing between the obliterative and reconstructive options. A systematic review of colpocleisis published in 2006 noted that colpocleisis appears to be nearly 100 % effective in correcting pelvic organ prolapse; however, until recently little was known about postoperative functional or quality of life outcomes [151]. In the last few years a number of reports evaluating symptom improvement and changes in quality of life after colpocleisis have been reported [153–156]. Overall, these series have found high rates of patient satisfaction and significant functional improvement, with low rates of regret for loss of sexual function [153–155]. Barber et al. reported results from a multicentre study of obliterative surgery using a prospective cohort design with a concurrent control group of age-matched women undergoing vaginal reconstructive surgery [153]. Despite permanent alterations in sexual function, significant improvements in bladder, bowel and prolapse symptoms as well as body image were noted after surgery with no differences between those who received colpocleisis and those who underwent reconstructive surgery. Additionally, significant and clinically important improvements were noted in bodily pain, vitality, social functioning, role-emotional, and mental health summary scales of the SF-36 [153]. Similarly, in a study on a retrospective cohort of women over the age of 65 comparing women who underwent colpocleisis (n = 45) and a similar group of women who underwent reconstructive surgery with transvaginal mesh (Prolift, Ethicon Women’s Health and Urology) Murphy et al. found that improvements in condition-specific quality of life and postoperative patient satisfaction were comparable in the two treatment groups [157].

The Pelvic Floor Disorders Network has reported on a large series of women undergoing colpocleisis (n = 153) with 1-year follow-up [154]. All pelvic symptom scores and related bother significantly improved at 3 and 12 months, and 125 patients (95 %) said they were either “very satisfied” or “satisfied” with the outcome of their surgery [154]. Bothersome stress and urge incontinence were present before surgery in 54 % and 41 % of subjects respectively. Forty percent of subjects received a concurrent mid-urethral sling at the time of their colpocleisis and the rates of bothersome stress and urge incontinence 1 year after surgery were 14 % and 15 % respectively. Similarly, bothersome bowel symptoms were present in 77 % of subjects at baseline. One year after surgery, the majority of bothersome bowel symptoms resolved, particularly obstructive and incontinence symptoms, and development of new bowel symptoms was uncommon (0–14 %) [156].

While obliterative procedures are predominantly performed in elderly, frail women who often have multiple co-morbidities, the rate of serious adverse events after this procedure appears to be low. In general, major complications due to performance of surgery on the elderly (e.g. cardiac, pulmonary and cerebrovascular complications) occur at a rate of approximately 2 % [151]. Major complications due to the surgery itself (e.g. pyelonephritis, blood transfusion) occur at a rate of approximately 4 % [151]. A systematic review of published series of colpocleisis from 1966 to 2004 reported a surgical mortality rate of approximately 1 in 400 cases [151]. One complication that appears to be uniquely associated with obliterative surgery is the development of de novo rectal prolapse after surgery [158, 159]. Collins et al. in a retrospective cohort of 916 women undergoing vaginal POP surgery at one institution found that the incidence of postoperative full-thickness rectal prolapse in women who were 65 years old or more who underwent obliterative surgery was 3 out of 74 (4.1 %; 95 % CI, 1.4–11), with an estimated odds ratio of 22 (95 % CI, 2.3–196; P < 0.002) compared with women who were  65 years old or more who underwent reconstructive surgery [158].

Conclusions

  • Level 1 evidence suggests that ASC has a higher success rate than sacrospinous colpopexy, with less SUI and postoperative dyspareunia. ASC had greater morbidity including operating time, inpatient stay, slower return to activities of daily living and higher cost (grade A).

  • ASC has the lowest inpatient cost compared with LSC and RSC. LSC has lower inpatient costs than RSC (grade B).

  • In single RC's the RSC had a longer operating time than both ASC and LSC (grade B). In small trials objective outcomes appear similar although postoperative pain was greater in RSC.

  • LSC is as effective as ASC, with reduced blood loss and admission time (grade C). The data relating to operating times are conflicting

  • ASC performed with polypropylene mesh has superior outcomes to fascia lata (level I) and porcine dermis and small intestine submucosa (level 3; grade B)

  • In a single RCT, LSC had a superior objective and subjective success rate and lower reoperation rate than polypropylene transvaginal mesh for vault prolapse (grade B).

  • Level 3 evidence suggests that vaginal uterosacral ligament suspension, McCall culdoplasty, iliococcygeus fixation and colpocleisis might be relatively safe and effective interventions (grade C)