Rectovaginal Fistula

Fifth chapter of Part II ("Anorectal Disease") — the female-anatomy analog of Rectourethral Fistula (ASCRS Ch14), which explicitly flagged this chapter as the related-but- separate next topic. An abnormal connection between the anal canal or rectum and the vagina, most often from obstetric injury or Crohn's disease, but also from infection, surgical complications, radiation, or malignancy.

Classification

By tract level: low (tract between the distal anal canal — dentate line or below — and the vaginal fourchette), high (connecting the upper vagina, at the cervix level, with the rectum), or middle (somewhere between). "Anovaginal fistula" and "low rectovaginal fistula" are often used interchangeably.

By complexity: simple (low, small-diameter <2.5cm communication, from obstetric injury or infection) vs. complex (higher communication, larger opening, or from radiation, cancer, or pelvic surgical complications).

Symptoms

Flatus and/or stool per vagina, pain, dyspareunia, local skin/mucosal irritation — severity ranges from minimal to severely debilitating.

Etiology (Table 15.1 in the source chapter)

Obstetric injury — reported as the most common cause of RVF in many series, though the true contribution is hard to pin down (in single-center US series: 23–24% of RVF from obstetric injury; in a Norwegian population-based study of 182 women with enterogenital fistula, only 13% were classified as RVF specifically from obstetric injury). Results from pressure necrosis of the rectovaginal septum during prolonged labor, laceration of the sphincter complex during delivery, or episiotomy. Injuries identified at delivery should be repaired immediately under optimal conditions by a dedicated, experienced surgical team.
Crohn's disease — a 2019 systematic review/population-based analysis found 1% RVF prevalence among females with Crohn's; a Netherlands population-based cohort (728 females with CD) found a 3% cumulative 10-year RVF diagnosis probability. RVF incidence has decreased in the biologic era: in the IBD-South Limburg cohort, the cumulative 5-year rectovaginal rate declined from 5.7% (1991–2005) to 1.7% (2006–2011). Anal fistula may be Crohn's initial manifestation, or occur years before/after proximal luminal disease diagnosis — fistula incidence increases with both disease duration and more distal luminal disease extent, highest with proctitis (see Crohn Disease for that page's broader perianal Crohn's content). As with other RVF etiologies, controlling associated infection/sepsis is the necessary first management step.
Infliximab has proven fistulizing-perianal-Crohn's efficacy. A subset analysis of ACCENT-2 (RCT of infliximab vs. placebo): among 25 women with RVF induced with infliximab, 16 (64%) responded with ≥50% tract closure; responders randomized to maintenance infliximab vs. placebo showed 45% vs. 43% closure at week 52 (no long-term healing benefit), but a meaningfully longer short-term duration of closure with infliximab (45 vs. 25 weeks). A 2016 systematic review (16 studies, 137 Crohn's-RVF patients) found 63% had some infliximab response, 38% a complete response, but lacked long-term outcome data. Bottom line per the chapter: infliximab is "worth a try" for Crohn's RVF and helps some patients, but is far from a panacea — and aside from surgery, infliximab alone may be sufficient for healing in some cases.
Early stem cell data: an uncontrolled series of 10 Crohn's-RVF patients treated with rectal injection of allogeneic expanded adipose-derived stem cells showed a 60% healing rate — promising but early. Stronger data since (ASCRS Ch47): a phase 3 RCT of 212 CD patients with complex fistulas (not RVF-specific) found adipose-derived stem cell injection achieved 56.3% vs. 38.6% fistula closure vs. placebo (p=0.010); a separate, smaller RCT of stem cell injection specifically for RVF found 51% vs. 35% remission at 24 weeks. See Crohn Disease's "Perianal (anorectal) Crohn's disease" section for a flagged discrepancy — a BSG-sourced bullet elsewhere on that page cites a phase III trial of the same drug class (Cx601/darvadstrocel) that failed its primary endpoint, plausibly a different (negative confirmatory/US) trial of the same drug rather than a contradiction, but not independently verified.
Surgical treatment spans the same complexity spectrum as non-Crohn's RVF (draining seton through proctectomy with permanent ostomy), including ERAF, gracilis or bulbocavernosus (Martius) flap, and rectal resection with coloanal anastomosis. Surgical success is generally lower for Crohn's than non-Crohn's RVF (e.g. one Cleveland Clinic Florida series: 78% per-procedure success for non-Crohn's vs. 44% for Crohn's RVF) — though as with non-Crohn's RVF, most patients are eventually healed after multiple attempts. Recent anti-TNF/steroid/ immunomodulator use did not negatively impact surgical RVF-healing success in one study — a reassuring finding for surgical timing decisions. Temporary diverting ostomy's benefit in Crohn's RVF repair "remains controversial" (no RCT exists); the Cleveland Clinic approach favors temporary diversion for redo repairs, technically challenging repairs, and suboptimal tissue conditions. Risk factors for permanent fecal diversion in perianal Crohn's include complex fistula, anal stenosis (see Anal Stenosis), history of rectal resection, fecal incontinence, and prior temporary fecal diversion.
Cryptoglandular (abscess/Bartholin's cyst abscess) — 2–40% of surgically-treated RVF series in retrospective studies; initial treatment focuses on sepsis eradication (see Anorectal Abscess for that page's fuller cryptoglandular-sepsis-control content), with subsequent surgery timing individualized by symptoms, sphincter integrity, and soft-tissue condition.
Anastomotic and other surgical complications — fistulization of a colorectal anastomosis to the vagina occurs in up to 10% of women undergoing low anterior resection (see Rectal Cancer and Colorectal Anastomotic Complications for LAR-specific complication content). Fecal diversion is generally the recommended first step to resolve acute inflammation/ infection; diversion alone healed 6/16 (37%) colorectal-anastomotic-vaginal fistulas within 6 months in one series. Persistent fistulas need neo-colorectal anastomosis, ERAF, or gluteal-fold flap interposition. Rarer iatrogenic causes: stapled hemorrhoidopexy (see Hemorrhoids) or stapled transanal rectal resection (STARR) — see Rectal Prolapse Repair's existing note on STARR's rectovaginal-fistula risk, which this chapter corroborates as a recognized, if rare, cause. Anterior rectal-lesion transanal excision also carries this risk (see Local Excision of Rectal Neoplasms).
Radiation injury — RVF after radiation for cervical cancer or other pelvic malignancies develops anywhere from 5 months to 20 years post-treatment (median interval 20 months in one review). Many radiation-RVF patients are maintained long-term on fecal diversion alone; other series report high resolution rates using a bulbocavernosus flap or rectal resection with coloanal anastomosis.

Evaluation

Thorough history (duration of symptoms, vaginal discharge nature/volume, day/night bowel movement frequency/consistency, ability to defer defecation/flatus, fecal incontinence, pad use/changes per day as a drainage-volume proxy, urinary symptoms — pneumaturia/fecaluria suggesting a coexistent enterovaginal fistula, past medical/surgical/childbirth history). Awake clinic exam is an important first step: anoperineal inspection (perineal body integrity, episiotomy/laceration-repair scars), digital exam of the anorectum/vagina (sphincter assessment, fistula location/size), vaginoscopy and anoproctoscopy (fistula location/size, stricture, proctitis, other mucosal abnormalities). If awake exam is insufficient, exam under sedation/general anesthesia should be considered.

Tampon test: a tampon or gauze pad is inserted into the vagina, left in place for several hours, then inspected for fecal staining — an occasionally useful confirmatory test. A modified version uses a clean gauze/lap pad in the vagina plus dilute Betadine solution instilled into the rectum via a 60cc syringe, then inspecting the vaginal pad for Betadine staining.

Diagnostic imaging is used selectively: MRI is the preferred modality to define fistula trajectory/tract number and detect undrained fluid collections or local pathology; CT is inferior to MRI for pelvic soft-tissue contrast; fluoroscopic studies (rectal contrast enema) help evaluate a colorectal anastomosis for patency/fistula; vaginography has shown high RVF-detection sensitivity; endoluminal ultrasound may have a role for anal sphincter examination specifically but has limited readability/surrounding-anatomy view. Endoscopic evaluation (rectum/colon) may be considered when Crohn's disease or radiation injury is the suspected etiology. Urology/gynecology consultation may be needed for suspected coexistent rectourethral (see Rectourethral Fistula) or enterovesical fistula, or when gynecologic intervention is needed. If an ostomy is being considered, preoperative site marking/education is advised.

Surgical techniques

No single surgical technique has been found superior; healing rates range 50–85% depending on case series/patient characteristics — reflecting genuinely comparable, technique-agnostic efficacy more than a clear winner. Choice depends on symptom severity, patient health, fistula etiology, sphincter integrity, rectal condition (proctitis, stricture), vaginal epithelium pliability/length, and history of prior failed repair attempts. Multiple repair attempts are often required: in one 125-patient retrospective analysis (184 total procedures), only 57% were healed after the initial repair, rising to 87% after multiple operations; a more recent 79-patient review found 72% ultimately healed, requiring a mean of 3.6±2.4 (range 1–10) surgical procedures.

General approach selection: for patients with anal sphincter injury, deficient perineal tissue, or a gaping introitus, a perineal approach is preferred. If the perineum/sphincters are intact, the fistula can be repaired through the rectum or vagina, avoiding perineal-incision trauma. Colorectal surgeons traditionally favor a transanal approach (comfort/experience, plus the philosophy that focusing repair on the high-pressure rectal side aids healing), though transvaginal approaches are worth considering specifically when avoiding anal sphincter stretch from retractors is advantageous.

Preoperative optimization: smoking cessation, weight loss, diabetes control, improved bowel consistency/frequency; mechanical bowel prep preference varies (many prefer an enema-only clean-out plus single-dose perioperative antibiotics). Postoperative dyspareunia/bowel-function-change risk should be discussed preoperatively. Postop complications include infection, bleeding, perineal sepsis; perineal skin breakdown is common and sometimes left to heal by secondary intention. Patients are restricted from sexual activity, tampon use, excessive leg stretches, and strenuous activity until the 6-week postsurgical visit.

Perineal approach

Episioproctotomy — the chapter authors' preferred approach for women with sphincter or perineal injuries. Opens the fistula tract, creating a defect similar to a fourth-degree perineal laceration; fistula tract edges excised, rectovaginal septum mobilized until healthy pliable tissue is reached, meticulous hemostasis, layered closure (rectal mucosa running 3-0 absorbable suture; levators/rectovaginal septum approximated avoiding deep bites/tension that risk levator spasm/dyspareunia; sphincter muscles overlapped with mattress sutures; vaginal mucosa closed running locked 3-0 absorbable; transverse perineal muscles repaired; perineal skin loosely approximated). Reported success 64–100%.
Transverse perineal repair — transverse perineal-body incision, dissection proximal to the fistula tract, vaginal/rectal wall mobilized and fistula edges excised, two-layer rectal wall closure, midline rectovaginal septum/levator approximation, vaginal epithelium closure.

Transrectal approaches

Endorectal advancement flap (ERAF) — well described in the literature, by far the most common approach in the colorectal literature. Near-180-degree curvilinear incision distal to the fistula opening; mucosa/submucosa/rectal-wall flap dissected off the rectovaginal septum; mobilization proximal to the fistula to avoid tension; fistula tract excised, opening closed with absorbable suture; flap trimmed and sutured to the distal cut end. Vaginal/perineal openings left to heal by secondary intention. Overall reported success rates: 43–93% (wide range reflecting patient/technique heterogeneity).
Rectal sleeve advancement — reserved for select situations where the distal rectum is diseased but the proximal rectum is normal (Crohn's, radiation, prior surgery). Dentate-line mucosectomy, cephalad dissection to healthy tissue, 90–100% rectal mobilization, healthy proximal rectum brought down tension-free and sutured to the neodentate line. Typically considered in patients whose only other option is total proctocolectomy or permanent fecal diversion — i.e., a rectum-preservation option of last resort short of proctectomy.

Vaginal approach

Least commonly reported in the colorectal literature. Advantages: avoids extensive mobilization of diseased rectum and the sphincter-stretch/tissue trauma associated with rectal-flap anal retractors. Favored in patients with active Crohn's rectal disease; endovaginal flaps have shown outcomes similar to endorectal flaps in comparative studies. Technique: posterior vaginal flap raised over the fistula, rectal/vaginal fistula orifices identified and repaired with absorbable suture, levator ani approximated in the midline, vaginal flap advanced over the repair. Alternative "flapless" technique (coring out the fistula tract, freshening the edges): 67% healing in a 15-patient series.

Tissue transposition repairs

For more complicated repairs (prior failed attempts, Crohn's, radiation-induced fistula), tissue transposition with muscle or pedicled adipose increases bulk and obliterates dead space.

Martius flap — harvested from the labia (fat + bulbocavernosus muscle), preserving the posterolateral vascular pedicle (branch off the internal pudendal artery), rotated/interposed between the closed fistula edges. Success rates 60–100%.
Gracilis muscle interposition — offers greater bulk of healthy vascularized tissue; harvested from the thigh, tunneled toward the perineum, positioned between rectum and vagina. Overall success 53–92%.

Bioprosthetic products

Several biosynthetic products developed for fistula-tract closure (biological matrix theorized to promote inflammatory response/scar formation while minimizing dissection/trauma) — but success rates are low, 20–35%, and initial enthusiasm has diminished.

Abdominal approaches

Warranted for RVF originating in the middle third of the rectum or upper vagina, or with severely damaged tissue after surgery/radiation. Healthy-tissue anterior dissection between rectum and vagina, fistula tract excision, and omentum interposition have satisfactory reported success. In most cases, resection of the diseased rectum with a lower-level anastomosis is necessary — two main techniques: immediate anastomosis (when distal rectal tissue is normal) or delayed coloanal anastomosis (Turnbull-Cutait), reserved for other fistulous connections or an internal opening close to the suture line — the proximal bowel is prolapsed out of the anus for 5–7 days, then amputated and fixed to the neodentate line at a second operation. Temporary fecal diversion is highly recommended in the setting of abdominal procedures and radiation. Permanent colostomy should be considered for patients with severe radiation injury.

Conclusion

RVF are uncommon but can pose a very challenging problem; the colorectal surgeon is likely to encounter the most complicated cases. A thorough understanding of the disease process and surgical options is imperative to give patients the best opportunity for fistula healing — often requiring multiple attempts regardless of technique chosen.

Rectourethral Fistula — the male-anatomy analog; shares this page's ERAF technique and general multidisciplinary/staged-repair philosophy.
Crohn Disease — RVF as a Crohn's manifestation, with distinct epidemiology (declining incidence in the biologic era), infliximab efficacy data (ACCENT-2 subset analysis), and generally lower surgical success than non-Crohn's RVF.
Colorectal Anatomy & Embryology — Denonvilliers' fascia/rectovaginal septum anatomy underlying this page's etiology and repair-approach content.
Rectal Cancer, Colorectal Anastomotic Complications — anastomotic-vaginal fistula as a recognized LAR complication (up to 10% of women).
Hemorrhoids, Rectal Prolapse Repair, Local Excision of Rectal Neoplasms — each already flagged rectovaginal fistula as a rare complication of stapled hemorrhoidopexy, STARR, or anterior rectal-lesion excision respectively; this chapter corroborates and contextualizes those mentions within the fuller RVF etiology/treatment picture.
Anorectal Abscess, Anal Stenosis — cryptoglandular sepsis-control principles and permanent- diversion risk factors shared with this page's Crohn's-RVF content.

Open items / gaps

No primary-literature source yet for the ACCENT-2 infliximab subset analysis or the allogeneic stem cell series — both known only via this textbook chapter's synthesis; the ACCENT-2 data in particular is load-bearing for this page's Crohn's-RVF medical-therapy content and would be a good candidate for independent ingest.
ASCRS Ch47 added a much stronger stem-cell-therapy data point (a 212-patient phase 3 RCT, 56.3% vs. 38.6% closure) and a separate RVF-specific stem-cell RCT (51% vs. 35% remission at 24 weeks) — see "Medical management" above. Flagged a discrepancy against Crohn Disease's BSG-sourced content citing a failed phase III trial of the same drug class (Cx601/darvadstrocel) — not independently resolved; see that page's "Perianal (anorectal) Crohn's disease" section for the full flag. Neither trial has been independently primary-sourced.
Surgical technique success-rate ranges throughout are wide and drawn from heterogeneous case series — the chapter itself notes no technique has been shown superior, which likely reflects real patient-population heterogeneity as much as technique equivalence.

Colorectal LLM Wiki