This systematic review of uterosacral ligament suspension provides a metaanalysis of anatomic outcomes and a summary of subjective outcomes. A successful anatomic outcome was considered present when women had “optimal” or “satisfactory” (pelvic organ prolapse quantification system stage 0 or 1) outcomes. In the anterior, apical, and posterior compartments, the pooled rates for a successful outcome were 81.2% (95% confidence interval [CI], 67.5–94.5%), 98.3% (95% CI, 95.7–100%), and 87.4% (95% CI, 67.5–94.5%). In the anterior compartment, women with preoperative stage 2 prolapse were more likely than those with preoperative stage 3 prolapse to have a successful anatomic outcome (92.4% vs 66.8%; P = .06). Outcomes, with respect to subjective symptoms, were reassuring; however, it was not possible to pool data because of methodologic differences between studies.
Pelvic floor dysfunction leads to surgery for 11% of women in their lifetime. Approximately 29% of these women will require a second surgery, and the annual cost of surgery for prolapse alone is >$1 billion. Surgical correction of vaginal vault prolapse has been described through abdominal, laparoscopic, and transvaginal approaches, with >40 different procedures described in the literature. Today, 2 of the most commonly reported transvaginal, native tissue, apical prolapse repairs are the sacrospinous ligament suspension and the uterosacral ligament suspension. Our group recently has reviewed outcomes after sacrospinous ligament suspension, and we sought to review the published anatomic outcomes after transvaginal uterosacral ligament suspension.
The principals of uterosacral ligament suspension have been described by several surgeons in the past 80 years. In 1927, Miller used “lifting sutures” of chromic catgut to reattach the vaginal cuff to “the peritoneum and underlying fascial and muscular structures at the base of the sacro-uterine ligament, approximately 1-1/2 inches below the promontory of the sacrum.” The McCall’s posterior culdeplasty, described in 1957, used nonabsorbable sutures to obliterate “the redundant cul-de-sac of Douglas by a series of continuous sutures so as to suspend it by the uterosacral ligaments.” In 2000, Shull et al described a “high” uterosacral ligament suspension in which 3 nonabsorbable sutures are “placed in the ligament on either side… to secure the superior aspect of the transverse portion of pubocervical and rectovaginal fascia” to the vaginal cuff. Outcomes for this type of vault suspension have been reported in the recent literature.
Our objective was to perform a systematic review and metaanalysis of anatomic outcomes after uterosacral ligament suspension, focusing on the transvaginal approach using either a McCall’s culdeplasty or a high uterosacral ligament suspension as described by Shull et al.
Data sources
Search strategy
A Medline search with the terms uterosacral or culdeplasty was performed for articles that were indexed between 1966 and May 2007, with the assistance of a medical research librarian at the University of Michigan Taubman Medical Library. These terms were used because they identified a larger number of citations than more specific terms such as high uterosacral ligament suspension or modified McCall’s culdeplasty , thereby increasing the breadth of our initial search. We then cross-referenced the citations that were identified with the terms uterosacral or culdeplasty with the MeSH heading treatment outcome to verify that all articles that described outcomes were included in our search results. We also searched the references listed in the articles that we reviewed to identify additional studies not picked up on our initial search strategy. The identification of these additional studies by reference review brought to our attention the alternate spelling of culdoplasty (as opposed to culdeplasty ). Repeating the Medline search strategy mentioned earlier with the alternate spelling, 21 additional abstracts were identified; however, none of them met the criteria for full text article review.
Abstract and full text review
The abstracts were reviewed by 1 of the authors (R.U.M.) who was blinded to the authors. Studies were excluded if they did not (1) pertain to pelvic organ prolapse, (2) report treatment outcomes, or (3) report original research (eg, review articles).
The full text articles were reviewed in a blinded fashion by 2 of the authors (R.U.M. and D.M.M.) with a standardized, data abstraction form. Studies were included if (1) preoperative prolapse severity and postoperative vaginal support were described with either the pelvic organ prolapse quantification system (POP-Q) or Baden-Walker Halfway system ; (2) a transvaginal high uterosacral ligament suspension, McCall’s culdeplasty, or modification of the McCall’s culdeplasty was described; (3) mean or median follow-up time was at least 12 months.
Studies were excluded if an abdominal or laparoscopic surgical approach was used or the shortening and reattachment of the uterosacral ligaments with or without the plication or posterior cul de sac were not described.
For each study, data were collected regarding the sample size, mean or median number of months followed, number lost to follow-up evaluation, the system used to describe vaginal support, mean or median preoperative prolapse stage/grade in each vaginal compartment, the number of women with postoperative stage/grade in each compartment of vaginal support, and patient characteristics. When the mean (SD) of a POP-Q point such as Ba, C, or Bp was provided, it was used to determine the mean preoperative prolapse stage/grade. When the distribution of support across stages or grades of prolapse was reported, the median severity preoperative prolapse stage/grade was identified.
In those studies that met the inclusion criteria, data were abstracted regarding subjective outcomes. Symptoms of prolapse, urinary incontinence, bowel dysfunction (constipation, splinting of the perineum or vagina to evacuate bowel movements, and fecal incontinence), and sexual function were collected and compared. The measures that were used to assess subjective outcomes ranged from symptoms elicited by authors/surgeons to specific items on validated questionnaires to summary scores of validated questionnaires. There was no generally accepted standard for any of the subjective assessments.
Statistical methods
The primary outcomes were the respective numbers of women in the anterior, apical, and posterior compartments by POP-Q stage. Women with “optimal” (POP-Q stage 0) and “satisfactory” (POP-Q stage 1) postoperative support, as described in the National Institutes of Health guidelines for researchers, were considered to have a successful anatomic outcome. Rates for women with stages 2 and 3 postoperative support were also calculated. Because 3 of the 4 studies that used the Baden-Walker vaginal profile dichotomized anatomic outcomes, we were able to calculate only a pooled rate for less than grade 2 support. A random effects model was used to determine pooled estimates for vaginal support with the POP-Q staging system and the Baden Walker profile. Cochran’s Q test was used to assess for heterogeneity (α ≤ .10) among studies. When heterogeneity was observed, an analysis was undertaken to explore whether preoperative severity of prolapse might account for the differences among studies. The rates of postoperative vaginal support were compared for studies that reported mean or median preoperative stage 2 and 3 prolapse with the use of the Fleiss Z-test for differences in proportions, incorporating the standard error of the pooled rates. Alpha was set at .05 (2-tailed).
It was not feasible to pool the results with respect to subjective outcomes. This was due to differences among study cohorts in preoperative symptom status, between procedures that could have an effect on symptoms (eg, antiincontinence procedures on urinary incontinence, posterior colporrhaphy on bowel dysfunction), and between outcomes that were assessed.
Results
The MEDLINE search identified 341 citations. All abstracts that were reviewed in our search were in English. There were 23 citations that were published in full text in other languages (5 French, 5 German, 5 Chinese, 3 Japanese, 2 Polish, 2 Portuguese, 1 Hebrew), but none of these studies met the criteria for full-text review. The criteria for full-text review were met by 25 studies. A search of bibliographies identified an additional 7 articles and led to a full-text review of 32 articles. The results of the literature review are summarized in the Figure .
The review process identified the 11 studies that are listed in Table 1 . There were 930 patients who underwent either a high uterosacral ligament suspension (82.2%) or a McCall’s culdeplasty (17.8%). Ten of the 11 studies reported a mean age in the seventh decade. The median length of follow up was 25 months, with a range of 12-84 months. The mean percentage of patients lost to follow up was 7.2%. Five studies reported no patients lost to follow up; 2 studies had <10% lost to follow up; and 4 studies had 15-22% lost to follow up. An overlap in patient population was identified in 2 articles. The study published by Silva et al was the long-term follow up of the series published by Karram et al (ie, same hospital, overlapping dates of surgery, senior author of the second study was first author of the first study). Both studies were included in calculation of the anatomic success rates. This was possible because the postoperative anatomic outcome was described with the use of different systems (Baden-Walker in the first and POP-Q in the second). Patients figured in the calculation of the pooled rates of support once for the Baden-Walker grading system and once for POP-Q staging data. However, when complications were considered, the 110 patients from the article by Silva et al were eliminated from the analysis of complication rates.
| Study | Design | n | Lost to follow-up, n | Mean age, y | Median parity, n | Procedure | Follow-up period, mo | Vaginal support classification | Preoperative stage/grade, n | Concurrent procedures | Postoperative ≥stage/grade 2, n/N | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Any site | Anterior | Apex | Posterior | Any site | Anterior | Apex | Posterior | ||||||||||
| Barber et al | Prospective non-randomized | 46 | 7 | 65 | 3 | HUSLS/repair of endopelvic fascia defects with suspension of the vaginal cuff to the proximal uterosacral ligaments | 15.5 a (4–41) | POP-Q | NR | 3 | 2 | 2 | Hyst, 28 (61%); AR, 27 (59%); PR, 20 (44%); UI, 35 (76%) | NR | 7/39 (18%) | 2/39 (5%) | 8/39 (21%) |
| Amundsen et al | Retrospective | 33 | 0 | 63 | NR | HUSLS/suspension of the vaginal cuff to the proximal uterosacral ligaments with endopelvic fascial defects repair | 28 b (6–43) | POP-Q | NR | 3 | 1 | 2 | Hyst, NR; AR, NR; PR, NR; UI, 33 (100%) | NR | 0/33 | 2/33 (6%) | 4/33 (12%) |
| Montella and Morrill | Retrospective | 51 | 8 | 61 | 3 | McCall’s/apical prolapse with good posterior fornix support | 12 (exactly) | POP-Q | NR | NR | 3 | NR | Hyst, 43 (100%); AR, 34 (79%); PR, 30 (70%); UI, 10 (23%) | NR | NR | 1 (2%) | 3/43 (7%) |
| Antovska and Dimitrov | Prospective non-randomized | 32 | 0 | 49 | 2 | Modification of McCall’s or HUSLS | 25 b (9–42) | POP-Q | NR | 2 | 1 | 2 | Hyst, 0; AR, NR; PR, NR; UI, NR | NR | 4/32 (13%) | 0/32 | 3/32 (9%) |
| Wheeler et al | Retrospective | 36 | 8 | 64 | 3 | HUSLS and cystocele repair, augmented with porcine dermis graft | 17 a (6–43) | POP-Q | NR | 3 | 1 c (mean C, –2.8) | 2 | Hyst, NR; AR, 36 (100%) d ; PR, 34 (94%); UI, 32 (89%) | NR | 14/28 (50%) | NR | NR |
| Wheeler et al | Retrospective | 35 | 3 | 64 | 3 | Modified HUSLS/correction of apical prolapse | 23 b | POP-Q | NR | 3 | 1 c (mean C, –2.5) | 2 | Hyst, NR; AR, 35 (100%); PR, 35 (100%); UI, 32 (91%) | NR | NR | 0/32 | NR |
| Silva et al | Retrospective | 110 | 38 | 64 | 3 | HUSLS | 61 b (42–90) | Baden-Walker (preop); POP-Q (postop) | NR | 2 | 3 | 2 | Hyst, 27 (38%); AR, 42 (58%); PR, 63 (88%); UI, 23 (32%) | 12/72 (16%) | 5/72 (7%) | 2/72 (3%) | 10/72 (14%) |
| Colombo and Milani | Retrospective case-control (SSLF group excluded) | 62 | 0 | 58 | 2 | McCall’s/“advanced uterovaginal prolapse” | 84 b (48–108) | Baden-Walker | NR | 3 | 3 | 3 | Hyst, 62 (100%); AR, 56 (90%); PR, 57 (92%); UI, 27 (43%) | 9/62 (15%) | 4/62 (6%) | 3/62 (5%) | 3/62 (5%) |
| Shull et al | Prospective nonrandomized | 302 | 4 | 66 | NR | HUSLS | 14 b | Baden-Walker | NR | 3 | 2 | 1 | Hyst, NR; AR, NR; PR, NR; UI, NR | NR | 10/198 (5%) | 2/198 (1%) | 4/198 (2%) |
| Karram et al | Retrospective | 202 | 34 | 60 | 3 | HUSLS | 22 b | Baden-Walker | ≥2 | NR | NR | NR | Hyst, 90 (54%); AR, 101 (60%); PR, 132 (79%); UI, 59 (35%) | 21/168 (13%) | NR | 2/168 (1%) | NR |
| Koyama et al | Prospective nonrandomized (Inmon group excluded) | 21 | 0 | 65 | 2 | McCall’s/uterine prolapse without enterocele | 26 b | Baden-Walker | NR | 2 b (18/21 ≥grade 2) | 2 b (18/21 ≥grade 2) | 1 b (6/21 ≥grade 2) | Hyst, 21 (100%); AR, 17 (81%); PR, 12 (57%); UI, e 0 (0%) | NR | 4/21 (19%) | 1/21 (5%) | 1/21 (5%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
