A rectovaginal fistula (RVF) is an abnormal epithelium-lined communication between the anorectum and the vagina and/or perineum,1 usually arising as a complication of an underlying disease process or injury. The most common cause is obstetric trauma,2 which accounts for approximately 90% of all RVFs.3 Other causes are Crohn’s disease,4,5 complications of pelvic surgery, pelvic irradiation, infections, and malignancy.
An RVF is a debilitating condition resulting in severe symptoms, psychological and physical impairment, and a profound impact on quality of life and activities of daily living. Several surgical treatment options for RVF have been described, ranging from use of biomaterials, such as plugs and fibrin glue, to anal and transperineal approaches, such as advancement flaps, episioproctotomy, and tissue interposition flaps, to abdominal procedures with bowel resection and anastomosis or permanent stoma formation.
The choice of treatment is patient-tailored and considers cause, involvement of the anal sphincter,6 patient’s characteristics, and expectations. Reports on the success rates of surgical treatments for RVF are extremely variable, and patients may require more than 1 procedure to achieve definitive closure.7,8 Therapeutic algorithms have been proposed9; however, most studies regarding RVF are retrospective case series with a limited number of patients. None of the recommendations of the latest clinical practice guidelines for the management of anal fistulas and RVF published by the American Society of Colon and Rectal Surgeons were based on high-quality evidence.10
The use of gracilis muscle interposition (GMI) for the treatment of RVF was first described by Gorenstein et al, in 1988. The length of the gracilis muscle and the location of its vascular pedicle make it particularly suitable for transposition into the rectovaginal septum as a perineal repair. GMI consists of harvesting the gracilis muscle from the medial aspect of the thigh, dividing the tendon to the knee, and transposing it into the perineal area while preserving the vascular pedicle. The muscle is placed between the rectum and the vagina, where it is sutured in place. GMI is rarely used as a first-line procedure, but it is recommended for the treatment of complex and recurrent RVFs.10 A systematic review published in 2014 by Takano et al11 on GMI for complex perineal fistulas reported a success rate ranging between 0% and 100%. Hotouras et al,12 in a systematic review of 17 studies comprising 106 patients who underwent GMI for the treatment of RVF and pouch-vaginal fistulas, reported a 33% to 100% success rate.
The objective of the present systematic review is to summarize the currently available data on the efficacy of GMI for RVF with the aim of providing more robust evidence to aid the decision-making process when dealing with complex, persistent, and recurrent RVFs.
MATERIALS AND METHODSA systematic literature search was conducted on March 16, 2022, and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines and the proposed guidelines for reporting Meta-analyses of Observational Studies in Epidemiology (MOOSE).13,14 Our protocol was registered in PROSPERO (CRD42022319621).
Data Sources and Search StrategyA literature search including MEDLINE, Embase, the Cochrane Library, and Web of Science was performed by the investigators. The search strategy included the text terms “fistul*” and “gracil*” in “All fields” and the related MeSH terms. No language or date restrictions were applied. The reference lists of the selected studies and of previous systematic reviews were searched manually to identify all pertinent articles.
Inclusion and Exclusion Criteria and Study SelectionRandomized controlled trials, prospective cohort studies, case–control studies, retrospective cohort studies, and case series including adult patients (aged 18 years or older) undergoing GMI for the treatment of rectovaginal fistulas were included. Studies were excluded according to the following criteria: 1) articles with less than 5 adult patients who underwent GMI for RVF; 2) articles in which the primary success rate of GMI for the treatment of the population of interest could not be retrieved from the full text or by contacting the authors; 3) reviews, guidelines, book chapters, and editorials. Conference abstracts were included if the reported data were considered sufficient as recommended by Scherer et al.15
The identified abstracts were independently screened by 2 reviewers (A.O.P. and M.M.). The full-text review of potentially eligible studies was conducted independently by 2 reviewers (A.O.P. and M.M.). In case of disagreement between the reviewers, a final decision was made with the senior authors (T.L.H. and S.R.S.).
Data Extraction and SynthesisData from the included studies were extracted using a structured REDCap form.16,17 Data included characteristics of the study (study design, setting, time period), patient demographics (age, BMI, time between diagnosis and GMI, number of previous procedures, RVF cause), perioperative data (operative time, length of hospital stay, stoma presence, early postoperative complications), and long-term outcomes (follow-up, success rate, definition of success rate, rate of stoma closure, success rate after subsequent procedures, quality of life, fecal continence, sexual function). Data extraction was performed by one author and verified by another. Data were summarized in a narrative synthesis using descriptive statistics when appropriate.
The quality of evidence for each study was rated using the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence Table.18
The risk of bias was independently assessed by 2 reviewers (A.O.P. and M.M.) using the Methodological Index for Nonrandomized Studies (MINORS) tool.19 Any divergence in the risk of bias evaluation was resolved by consensus.
Statistical AnalysisStatistical analyses were performed with MetaXL (EpiGear International Pty Ltd., Sunrise Beach, Australia, version 5.3). Pooled success rates with 95% CIs were calculated with the inverse variance method. A subgroup analysis of success rate, including only studies with at least 11 patients who underwent GMI for RVF, was also performed. In case only the median and range were available, this was converted to mean and SD using the method described by Wan et al.20
In addition, we conducted a random-effects meta-analysis with the inverse variance method, including studies that compared GMI for Crohn’s disease versus other causes.
Heterogeneity between studies was assessed with the Cochrane Q test and I2 statistics. A random-effects model was used to calculate the pooled success rate if significant heterogeneity was detected (p < 0.1 and I2 > 50%). A fixed-effects model was used if a p value was >0.1 and I2 <50%. The Doi plot and Luis Furuya-Kanamori asymmetry index were used to check for publication bias.
RESULTSAfter removal of duplicates, 662 records were screened and 68 were sought for full-text retrieval. Twenty studies were included (Fig. 1).21–40
FIGURE 1.:PRISMA flow diagram of study screening and selection. PRISMA = Preferred Reporting Items for Systematic review and Meta-Analysis.
The characteristics of the included studies are reported in Table 1.21-40 Of the 20 included studies, 19 were case series (18 retrospective, 1 prospective), whereas 1 was a prospective cohort study. Six studies (30%) included only our population of interest21–26; 11 studies (55%) also included other types of fistulas27–37 and 3 studies (15%) also included other interventions.38–40 The risk of bias evaluation is reported in Table 2 and Supplemental Figure 1 at https://links.lww.com/DCR/C146. The funnel plot and DOI plot of the included studies are shown in Supplemental Figures 2 at https://links.lww.com/DCR/C147 and 3 https://links.lww.com/DCR/C148.
TABLE 1. - Characteristics of the included studies Author, year Country Study design Study period Single center Total number of patients in the study Total number of patients included in the review Age (y) BMI (kg/m2) Cause No. of previous repairs Time between diagnosis and GMI Cannom, 201340 United States Retrospective case series 1992–2011 No, bicentric 41 20 Median 45 (range, 18–84) Obstetric injury and CD 32/41 patients (78%) had previous failed repairs (range, 1–8) Chen, 201332 China Retrospective case series May 2009–March 2012 Yes 19 11 Median 47 (range, 16–80) Median 21.9 (range, 15.6–29.1) Rectal cancer = 4CD = Crohn’s disease; GIST = GI stromal tumor; GMI = gracilis muscle interposition; GYN = gynecologic.
留言 (0)