Rectovaginal Fistulas And Rectoceles
Sharon G. Gregorcyk, MD
Department of Surgery
University of Texas Southwestern
A rectovaginal fistula is an epithelium-lined communication between the rectum and vagina. The lower fistulas are actually communicating to the anus but are still referred to as rectovaginal fistulas. While these fistulas account for less than 5% of all anorectal fistulas, they are typically more aggravating for both the patient and the surgeon.
Most rectovaginal fistulas are acquired although congenital abnormalities do exist. The acquired fistulas will be the focus here and include etiologies such as trauma (including operative, obstetric, and traumatic injuries), infection, inflammatory bowel disease, carcinoma, and radiation.
In general, obstetric trauma is the most common cause of rectovaginal fistulas. This is not the case at some institutions likely due to their referral patterns. For example, in a series from the Mayo Clinic, only 11% of their rectovaginal fistulas were secondary to obstetric injuries while 24% were due to inflammatory bowel disease (1).
While prolonged labor resulting in pressure on the rectovaginal septum can lead to necrosis and a resultant rectovaginal fistula, more commonly the fistulas are associated with a laceration or episiotomy during delivery. If the injury is not adequately repaired or if the patient develops a postoperative infection, then a fistula may result. The incidence fortunately is low with Goldaber, et al (2) reporting a series of 23,986 vaginal deliveries in which 404 women (1.7%) suffered fourth degree lacerations and only 2 of those patients (0.5%) developed a rectovaginal fistula. Another large study showed only a 0.1% incidence of rectovaginal fistulas associated with vaginal deliveries (3).
Surgical trauma is another etiology for a rectovaginal fistula. Both anorectal and vaginal operations present a risk, as well as hysterectomies, low anterior resections, and ileoanal anastomoses. The fistula may result from a direct injury during the surgery or from an infection or anastomotic leak postoperatively.
Perianal infections such as a perianal abscess or Bartholin gland abscess can also produce a rectovaginal fistula. Other infectious etiologies include diverticulitis, tuberculosis, lymphogranuloma venereum, and even aphthous vaginal ulcers in women with HIV have been reported (4-5).
While both ulcerative colitis and Crohn's disease can cause a rectovaginal fistula, it is more commonly seen in Crohn's disease. Of 886 female patients with Crohn's disease from Radcliffe's study (6), 90 (9.8%) developed a rectovaginal fistula. DeDombal's series on ulcerative colitis showed a 3.6% incidence in women (7). Carcinoma, whether it is primary, recurrent or metastatic, can lead to a rectovaginal fistula. The most common offending cancers are rectal, cervical, vaginal, and uterine. If a patient with a rectovaginal fistula has a history of any of these cancers or other perineal cancers, a biopsy of the fistula is imperative.
Radiation of the pelvis or perineum is another cause of a rectovaginal fistula. The radiation results in tissue damage and decreased vascular supply, which can lead to a proctitis and subsequent ulceration of the rectal wall. Up to one-half of these ulcers progress to rectovaginal fistulas (8). The frequency of radiation induced rectovaginal fistulas is 0.3% to 6% (9-10). The incidence increases with higher doses of radiation. These fistulas are usually to the mid or upper vagina and will occur within 2 years of treatment. One must again biopsy the rectovaginal fistula as the radiation was most likely for one of the previously mentioned cancers and thus this fistula may be a recurrence of the cancer instead of radiation induced.
The majority of rectovaginal fistulas are symptomatic with the patient complaining of the passage of stool or flatus via the vagina. A thorough history and physical exam are necessary to help identify the etiology of the fistula, as well as assess its location, any ongoing inflammation, and whether a sphincter defect is present. All of this information is essential prior to surgical treatment.
If the fistula cannot be identified on exam, the patient may require additional studies. Endorectal ultrasound and transvaginal ultrasound have both been used for this purpose. If the fistula still cannot be identified, one may give the patient a methylene blue enema with a vaginal tampon in place looking for staining on the tampon to confirm the diagnosis. Contrast studies may be beneficial for more proximal fistulas. If there is a suspicion of inflammatory bowel disease, then endoscopy is necessary. Any fistula or lesion suspicious for cancer, of course, should be biopsied as well.
If the fistula is in the area of the sphincter muscles, one should objectively assess the patient for fecal incontinence. Relying on the history is insufficient since the patient may not be able to discern fecal incontinence from the fistula leaking. In those patients with decreased anal pressures and an anterior sphincter defect on endoanal ultrasound, consideration should be given to a sphincter repair at the time of the fistula repair.
Obviously the treatment of a rectovaginal fistula must be tailored to the individual fistula. An important consideration is whether it is a simple or complex fistula. Simple rectovaginal fistulas include small (< 2.5 cm in diameter), low or mid-level fistulas resulting from infection or trauma. Large (>2.5 cm), high fistulas are considered complex as well as any fistula caused by inflammatory bowel disease, cancer, or radiation. Also the fistula that has failed multiple prior repairs is classified as complex.
Once a surgical plan has been determined, a decision on timing must be made. It is essential that one waits for the resolution of any infection or inflammation of the surrounding tissue before attempting a repair. Typically a waiting period of 3 to 6 months is recommended. Some small fistulas will actually close during this time. Consideration should also be given to whether the patient is planning on having more children. If the fistula is significantly symptomatic, one does not need to wait until the patient has had all her children. However she should be counseled about a recurrence of the fistula and c-section may be advised.
Regardless of what repair is chosen, the principles are similar. The patient receives a full bowel prep to reduce the fecal load and bacterial content. The epithelial lined fistula tract must be removed. And finally, well-vascularized tissue must be interposed between the rectum and the vagina.
Surgical Options for Simple Fistulas
There are four major approaches reported for repairing a simple rectovaginal fistula: 1) transvaginal, 2) conversion of the fistula into a complete perineal laceration with subsequent repair, 3) transperineal, and 4) transanal.
The transvaginal approach is routinely used by gynecologists but does not abide by the principle that the repair should be on the high-pressure side, namely the rectum, and thus is not favored by colorectal surgeons. The technique of converting the fistula to a fourth degree laceration and then repairing it does have the benefit of repairing any associated anterior sphincter defect. However, in those patients without a sphincter defect one risks incontinence for the patient. The transperineal approach, as proposed by Goligher, (11) involves separating the anus and rectum from the vagina and rotating the two walls in opposite directions to separate the suture lines. Modifications of this also include interposing other tissue between the two walls. This technique is technically difficult and risks damage to the rectum and vagina.
The transanal endorectal advancement flap is the most popular technique of repairing simple rectovaginal fistulas by colorectal surgeons. The rectal flap consists of mucosa, submucosa and circular muscle. A broad base is essential for adequate blood supply. The flap is mobilized proximally for a minimum of 4 cm to help prevent tension on the suture line. The distal extent of the flap, which contains part of the fistula is excised. The remaining fistula tract is excised or curetted free of granulation tissue. The rectal opening is then closed by reapproximating the muscle. The flap is then advanced down and sutured in place over the prior fistula opening. The vaginal side is left open for drainage. Care must be taken not to create an ectropion with the advancement of the flap. If the fistula, therefore, is distal to the dentate line, another technique such as a houseflap anoplasty should be employed.
The rate of success of the endorectal advancement flap is reported at anywhere from 29 to 100% with current reports supporting about a 75% success rate (12-17). Part of this variability is likely due to differences in technique. As previously mentioned, if the patient has incontinence or sphincter dysfunction and a sphincter defect is present, an overlapping sphincteroplasty may be performed in addition to the advancement flap. This additional tissue placed between the two structures can increase the success rates.
Surgical Options for Complex Fistulas
Complex fistulas as previously described require a different approach than simple fistulas. To follow are a variety of problems and various potential solutions. With all of these complex fistulas, consideration should be given to diverting the patient while the repair heals.
Due to the proximal location of these fistulas, a transabdominal approach is recommended. The fistula can simply be divided with repair of each wall and interposition of omentum. However, more commonly, the involved segment of rectum is resected and an anastomosis is performed caudad to the vaginal repair thereby separating the suture lines. Omentum can then also be interposed between the two structures.
Inflammatory Bowel Disease
Obviously medical therapy is important for the rectovaginal fistula that occurs secondary to Crohn's disease. Local repair may be attempted if the inflammation can be successfully treated. The most popular local repair in this setting is the endorectal advancement flap, which has a success rate of about 60-70% (18-19) in Crohn's patients. A vaginal approach with Crohn's disease is worth considering since the vagina is the non-diseased side. Unfortunately, regardless of the local approach, recurrences of the fistula as well as severity of the patient's disease results in a need for proctectomy about 50% of the time (6, 20).
Rectovaginal fistulas can occur with ulcerative colitis as well. Control of the disease medically may control the symptoms, but rarely does the fistula completely resolve (21). A local repair is not typically advised in the patient with ulcerative colitis. Instead, a total proctocolectomy is recommended since it will not only treat the rectovaginal fistula but will also be curative of the ulcerative colitis. An ileal pouch anastomosis may still be performed with special attention to the anastomosis keeping it separate from the fistula repair.
Radiation induced rectovaginal fistulas pose an additional hurdle in their repair. That hurdle is the damage to surrounding tissues caused by the radiation thus necessitating the introduction of normal tissue to the area. By performing a low anterior resection or coloanal procedure, not only is normal tissue brought in for the anastomosis, but also the diseased bowel, along with its other possible complications, is removed. The procedure can be very difficult and incontinence is possible with very low anastomosis.
Bricker and Johnson (22) developed a different solution to the problem by performing an onlay patch of well-vascularized sigmoid colon. The rectosigmoid is mobilized and the rectovaginal fistula exposed and debrided. The rectosigmoid is divided and the distal end is rotated down upon itself and anastomosed to the opening of the rectum. The proximal end is brought up as an end colostomy. Later, once the repair has healed and a contrast study confirms no leaks or fistulas, the colostomy is taken down and anastomosed to the loop of rectosigmoid. Bricker reported success in 19 out of 20 patients. The procedure's biggest drawback is that the radiated bowel is left in place.
Recurrent Rectovaginal Fistulas
Multiple factors may lead to a recurrence of a rectovaginal fistula. These factors include post- operative infection, extensive scarring with poor blood supply, inflammation pre-operatively, and technically inadequate repairs. Various factors should be taken into account in choosing an appropriate surgery such as the etiology of the recurrence and the prior repair technique. In some cases, the same technique may be employed. Lowry et al (23) demonstrated an 85% success rate utilizing an endorectal advancement flap for the second repair of a fistula. However, when attempted for the 3rd repair the success rate was only 55%.
In addition to some of the previously mentioned techniques for complex rectovaginal fistulas, other procedures may be utilized for the recurrent fistula including the use of other well- vascularized tissue as an interpositional graft. Tissues used include the gracilis, omentum, gluteus maximus, sartorius and rectus abdominus. A more popular interpositional graft is the Martius graft. This technique involves mobilization of the bulbocavernous muscle with its labial fat pad and tunneling it to be interposed between the rectum and vagina. It carries a success rate of up to 85% (24-27).
Fibrin glue has gained popularity in the treatment of complex anorectal fistulas. Venkatesh and Ramanujam (28) evaluated their results of 20 women with complex fistulas of whom eight had rectovaginal fistulas. The overall success rate for the 20 patients was 60%. Of those with rectovaginal fistulas, 6 out of 8 healed without evidence of recurrence.
A variety of options exist for treating rectovaginal fistulas. The approach must be tailored to the individual fistula taking into account its size, location and etiology. The fistulas resulting from radiation or inflammatory bowel disease can be particularly challenging. Still a cure is possible if a good repair with well-vascularized tissue is performed.
A rectocele is an outpouching of the anterior rectal wall and posterior vaginal wall into the lumen of the vagina. Rectoceles may occur at 3 levels: high, mid, and low. The high rectoceles are usually due to a stretching or disruption of the upper third of the vaginal wall and uterosacral ligaments. Mid level rectoceles are most common and are associated with loss of pelvic floor support. The low-level rectoceles can be caused by obstetric trauma.
Rectoceles are common, occurring in up to 81% of all females (29, 30). Of all patients with rectoceles only 23-70% have complaints of difficulty with defecation (31, 32). Thus, there are many patients with asymptomatic rectoceles. Even in those patients with complaints of constipation, the presence of a rectocele does not equal to it being the etiology of the constipation. In fact, it has been demonstrated that only 10-20% of rectoceles are clinically significant as the cause of symptoms (33, 34).
Rectoceles less than 2 cm are usually accepted as a normal finding where as those greater than 2 cm may be symptomatic (35, 36). Possible symptoms include incomplete rectal emptying, a sensation of rectal pressure, and a bulge in the vagina. The rectocele can be diagnosed on exam or by defecography.
The first attempts at treatment are medical. The patient is placed on a high fiber diet with increased water intake. Fiber supplements are routinely used. Surgery is recommended in very select cases after failure of medical management.
Criteria for surgical treatment varies from surgeon to surgeon. The repair of an asymptomatic rectocele is not advocated. The patient should have symptoms that relate to the rectocele such as the need of vaginal digitation to effect evacuation. Mere complaints of constipation are insufficient to warrant a repair as the constipation may be due to other etiologies. In fact, some of these etiologies, such as paradoxical puborectalis contraction may have led to the patient's rectocele.
The common approaches for a rectocele repair are transvaginal, transrectal, and transperineal. All 3 techniques involve the rebuilding of tissue between the rectum and vagina. Rectocele repairs are commonly performed by gynecologists who prefer a transvaginal approach. This operation is often performed concomitantly with a cystocele repair and in that case, a transvaginal repair would be preferred.
The success rates for rectocele repair can be high if good patient selection is utilized. With strict criteria applied, 80-95% of patients report good to excellent results regardless of approach (35, 37-40). Complications vary from 3-38% (37-40) and include infection, fecal incontinence, dyspareunia and bleeding.
- Lescher TC, Pratt JH. Vaginal repair of the simple rectovaginal fistula. Surg Gynecol Obstet 1967; 124:1317-21.
- Goldabar KG, Wendel PJ, McIntire DD, Wendel GD. Post-partum perineal morbidity after fourth degree perineal repair. Am J Obstet Gynecol 1993; 168:489-93.
- Venkatesh KS, Ramanyam PS, Larson DM, Haywood MA. Anorectal complications of vaginal delivery. Dis Colon Rectum 1989; 32:1039-41.
- Greenwald JC, Hoexter B. Repair of rectovaginal fistulas. Surg Gynelcol Obstet 1978; 146:443-5.
- Schuman P, Christensen C. Sobel JD. Aphthous vaginal ulceration in two women with acquired immunodeficiency syndrome. Am J Obstet Gynecol 1996; 174:1660-3.
- Radcliffe AG, Ritchie JK, Jawley PR et al. Anovaginal and rectovaginal fistulas in Crohn's disease. Dis of Colon Rectum 1988; 31:94-9.
- DeDombal FT, Watts JM, Watkinson G, Goligher JC. Incidence and management of anorectal abscess, fistula and fissure in patients with ulcerative colitis. Dis Colon Rectum 1966; 9:201-6.
- Gordon P. Rectovaginal fistula. In: Principles and Practice of Surgery for the Colon, Rectum and Anus. Quality Med. Pub Inc; 361-81, 1992.
- Cooke SA, DeMoor NG. The surgical treatment of the radiation damaged rectum. Br J Surg 1981; 68:488-92.
- Allen-Mersh TG, Wilson ET, Hope-Stone HF, Mann CV. The management of late radiation- induced rectal injury after treatment of carcinoma of the uterus. Surg Gynecol Obstet 1987; 164:521-4.
- Goligher JC. Rectovaginal fistula. In: Goligher JC, ed. YSurgery of the Anus, Rectum and Colon. (ed).4. London; Spottiswoode Ballantyne Ltd.;191-193, 1980.
- MacRae HM, McLeod RS, Cohen Z, Stern H, Reznik R. Treatment of rectovaginal fistulas that has failed previous repair attempts. YDis Colon Rectum 1995; 38:921-5.
- Watson SJ, Phillips RKS. Non-inflammatory rectovaginal fistula. Br J Surg 1995; 82:1641-3.
- Khanduja KS, Yamashita HJ, Wise WE Jr., Aguilar PS, Hartmann RF. Delayed repair of obstetric injuries of the anorectum and vagina. A stratified surgical approach. YDis Colon Rectum 1994: 37:344-9.
- Baig MK, Zhao RH, Yuen CH, Nogueras JJ, Singh JJ, Weiss EG, Wexner SD. Simple rectovaginal fistulas. Int J Colorectal Dis 2000; 15:323-27.
- Lowry AC, Goldberg SM. Management of simple rectovaginal fistula. In: Cameron JL, ed. Current Surgical Therapy, 4, 1991.
- Hoexter B, Labow SB, Moseson MD. Transanal rectovaginal fistula repair. Dis Colon Rectum 1985; 28:572-5.
- Radcliff AG, Ritchie JK, Hawley PR, Lennard-Jones JE, Northover JMA. Anovaginal and rectovaginal fistulas in Crohn's disease. Dis Colon Rectum 1988; 31:94-9.
- Sher ME, Bauer JJ, Gelernt I. Surgical repair of rectovaginal fistulas in patients with Crohn's disease: transvaginal approach. Dis Colon Rectum 1991; 34:641-8.
- Scott NA, Nair A, Hughes LE. Anovaginal and rectovaginal fistula in patients with Crohn's disease. Br J Surg 1992; 79:1379-80.
- Froines EJ, Palmer DL. Surgical therapy for rectovaginal fistulas in ulcerative colitis. Dis Colon Rectum 1991; 34:925-30.
- Bricker EM, Johnston WD. Repair of postirradiation rectovaginal fistula and stricture. Surg Gynecol Obstet 1979; 148:499-06.
- Lowry AC, Thorson AG, Rothenburger DA, et al. Repair of simple rectovaginal fistulas: Influence of previous repairs. Dis Colon Rectum 1988; 31:676-678.
- Hibbard LT. Surgical management of rectovagial fistulas and complete perineal tears. Am J Obstet Gynecol 130:139-41.
- Zacharin KF. Grafting as a principle in the surgical management of vesicovaginal and rectovaginal fistulae. Aust N Z J Obstet Gynecol 1980; 20:10-17.
- White AJ, Buchsbaum HJ, Blythe JG, Lifshets S. Use of the bulbocavernous muscle (Martius technique) for repair of radiation-induced rectovaginal fistulas. Obstet Gynecol 1982; 60:114-118.
- Boronow RC. Repair of the radiation-induced rectovaginal fistulas with or without interposition of the bulbocavernous muscle. Eur J Surg Oncol 1988; 14:171-7.
- Venkatesh KS, Ramanujam P. Fibrin glue application in the treatment of recurrent anorectal fistulas. Dis Colon Rectum 1999; 42 (9):1136-44.
- Mellgren A, Anz‚n B, Nilsson B-Y, et al. Results of rectocele repair, a prospective study. Dis Colon Rectum 1995; 38:764-8.
- Sarles JC, Arnaud A, Sielezneff I, Olivier S. Endo-rectal repair of rectocele. Int J Colorectal Dis 1989; 4:167-71.
- Halligan S, Bartram CI. Is Barium trapping in rectoceles significant? Dis Colon Rectum 1995; 38:764-8.
- Van Dam JH, Ginai AZ, Gosselink MJ, et al. Role of defecography in predicting clinical outcome of rectocele repair.Dis Colon Rectum 1997; 40:201-7.
- Block IR. Transrectal repair of rectocele using obliterative sutures. Dis Colon Rectum 1986; 29:707-11.
- Ting K-H, Mangel E, Eilbl-Eibesfeldt B. Is the volume retained after defecaton a valuable parameter at defecography? Dis Colon Rectum 1992; 35:762-7.
- Melligren A, Anzen B, Nilsson B-Y, Johansson C, et al. Results of rectocele repair. Dis Colon Rectum 1995; 38:7-13.
- Murthy VK, Orkin BA, Smith LE, Glassman LM. Excellent outcome using selective criteria for rectocele repair. Dis Colon Rectum 1996; 39:374-8.
- Arnold MW, Stewart WRC, Aguiler PS. Rectocele repair: four years' experience. Dis Colon Rectum 1990; 33:684-7.
- Sarles JC, Ninou S, Arnaud A. Rectoceles: diagnosis and treatment. Chirurgie 1999; 117:618-23.
- Van Dam JH, Huisman WM, Hop WCJ, Schouten WR. Fecal continence after rectocele repair: a prospective study. Int J Colorectal Dis 2000; 15:54-57.
- Cohen SM, Wexner SD, Binderow R, et al. Prospective, randomized endoscopic-blinded trial comparing pre-colonoscopy bowel cleansing methods. Dis Colon Rectum 1994; 37:689-96.