Rectovaginal and Rectourethral Fistulas


Fistulas between the genitourinary tract and the digestive tract present a particular conundrum for the surgeon. Not only are they uncommon, but the technical aspects of repair in conjunction with managing etiology appropriately present a multidisciplinary challenge to even the most experienced surgeon. Not only do they present particular issues with respect to quality of life, but also complicated management of septic complications, management of radiated tissue, and the complicated dance of staged repair. In this chapter, we will discuss the aspects of diagnosis, classification, and management of rectovaginal and rectourethral fistulas (RUFs).

Rectovaginal Fistula

Rectovaginal fistulas (RVFs) are defined as epithelialized communications between the rectum and vagina. They are uncommon, comprising approximately 5% of all anorectal fistulas, but present a significant challenge for both the patient and the surgeon. Although they are rarely associated with mortality, they are associated with a significant degree of morbidity not only involving anatomic issues, but with a profound negative impact in quality of life from a social, psychological, and sexual point of view. Because of the rarity of RVFs, there is little well-described literature documenting systematic management, and the multiplicity of approaches highlights the fact that no single approach is uniformly successful.

Etiology

The causes of RVF are many, the most common being obstetric trauma. Approximately 2% of all vaginal deliveries are associated with advanced perineal injury, and 3% of these patients subsequently developed RVFs. These are typically associated with a prolonged second stage of labor resulting in ischemic necrosis of the rectovaginal septum. This is particularly common in developing countries where there is a lack of good obstetric care. RVFs are present in up to to 3.5 million women in Africa, with over 100,000 new cases each year. Other risk factors include shoulder dystocia in the fetus, high forceps deliveries, third- and fourth-degree tears, and midline episiotomy. Worldwide, an RVF develops from every 1000 vaginal births.

Crohn disease is the next most common cause. Up to 10% of female patients with Crohn disease develop RVFs. These are associated with a very high recurrence rate, and it is not uncommon for multiple procedures to be required to help these patients. A multidisciplinary approach, including medical management to ensure quiescent anorectal disease, is critical. Other autoimmune diseases such as Behçet disease may also be associated with RVF.

Anorectal sepsis, arising from cryptoglandular infection, may also be associated with RVF. These are often a result of advanced anterior cryptoglandular infection. Other gastrointestinal (GI) infections, such as diverticulitis, may be associated with fistula to the vagina; however, these are colovaginal fistulas, arising almost exclusively in patients who have undergone hysterectomy when pelvic abscesses fistulize through the apex of the vaginal cuff. Other more rare infectious etiologies include tuberculosis, lymphogranuloma venereum, human immunodeficiency virus (HIV), and cytomegalovirus (CMV).

Iatrogenic causes of RVF are not uncommon after a variety of pelvic operations. These include sequelae of leak after low colorectal anastomosis in women, neoadjuvant radiation is a significant risk factor, and the treatment of rectal cancer is responsible for 13% of RVFs. Urogynecologic procedures are also a risk factor; mesh erosion after pelvic organ prolapse repair may result in RVF. Leak after ileal J-pouch creation in the setting of ulcerative colitis or familial adenomatous polyposis (FAP) may lead to the development of a fistula. Spontaneous delayed presentation of pouch vaginal fistula is often a manifestation of previously unsuspected Crohn disease. RVF may also present as a complication of the procedure for prolapse and hemorrhoid (PPH) and the Starr procedure.

Pelvic malignancy with local invasion may also be an etiology of RVF. Often, the new development of a RVF may herald recurrent disease of cervical, rectal, vaginal, or uterine cancer, especially after local radiation for palliation. This rate was as high as 18% in patients who received external beam radiotherapy using older radiation sources in addition to interstitial brachytherapy for advanced gynecologic malignancy.

Other, less common causes include congenital fistula, mechanical necrosis of the rectovaginal septum due to fecal impaction and stercoral ulceration, long-standing pessary use, and trauma.

Clinical Manifestations

The primary presenting symptom of RVF is the new passage of fecal material or gas through the vagina. This is often misdiagnosed as fecal incontinence or attributed to vaginal infection. Chronic urinary tract infection may also be an associated symptom. In patients who develop RVF as a result of obstetric injury, antecedent fecal incontinence is as high as 50%. As a result, the presentation is often unclear, and likely a large number of these fistulas remain undiagnosed. If there are associated symptoms from the GI tract—including diarrhea, abdominal pain, or significant mucous discharge—inflammatory bowel disease may be suspected and an appropriate work-up should be undertaken. Regardless of the etiology, symptoms attributable to RVF may be ignored by the patient, and the level of suspicion in the clinical picture needs to be high.

Diagnosis

Diagnosis of RVF can often be accomplished in the office, when symptoms are overt. A careful physical exam will often yield direct or circumstantial evidence of a fistula. Digital rectal exam may reveal an indurated tender fistula tract. Exam, including anoscopy, may reveal presence of stool within the vaginal vault. Anoscopic exam, accompanied by a vaginal speculum exam, may yield visible granulation tissue at the level of the tract opening. In many cases, the fistula tract is difficult to identify in the office. It is not uncommon that, due to local inflammation or prior surgery, examination under anesthesia may be required to provide patient comfort and enhance visualization. With optimal lighting and exposure, internal openings in the anal canal or low rectum may become visible. Other operative maneuvers include filling of either the rectum or vagina with liquid while simultaneously insufflating air. The site of bubbling may demonstrate a fistula site that was previously difficult to visualize.

Adjunctive testing is sometimes required. The tampon test is often successful in demonstrating subtle, hard-to-identify fistulas. A tampon is placed in the vagina and a rectal enema of diluted methylene blue dye is carefully administered, to avoid overflow and leakage. The patient is then asked to ambulate for a period of 20 to 30 minutes. The tampon is then inspected, and evidence of blue staining along the length of the vagina indicates the presence of a fistula. If these tests do not yield a definitive diagnosis, a barium enema may be helpful. Clear lateral views and the use of balloon occlusion of the anal canal with positioning of the balloon as distal as possible are important.

If operative examination under anesthesia is undertaken and a fistula is identified, a seton should be placed. In addition, it is very important to assess for any evidence of local inflammation that might be associated with active Crohn disease as well as to manage any associated sepsis. It is important to inspect the entire rectovaginal septum and examine for any obvious sphincter defects, scarring, or stenosis. All of these are factors that significantly affect the plan of management.

As with complex cryptoglandular fistula disease, the goals of evaluation prior to definitive repair are to (1) identify the fistula, (2) manage any acute sepsis, (3) establish adequate drainage, (4) ascertain anatomy, including location and size of fistula as well as any surrounding injuries, and (5) determine etiology.

Once the fistula has been identified and any acute issues related to sepsis have been managed, further diagnostic studies will help in planning the surgical approach. Endoanal ultrasound is the diagnostic test of choice to identify associated sphincter injuries. The adjunctive use of hydrogen peroxide may also help to identify the anatomy of the fistula relative to the surrounding sphincter. Pelvic magnetic resonance imaging (MRI) and endorectal coil MRI are also helpful. Preferential use of ultrasound versus MRI is often institution- and practitioner-specific. Sphincter defects in association with RVF are very common. In a reported series of patients who sustained RVF after vaginal delivery, all demonstrated sphincter injury on endoanal ultrasound.

Pelvic floor physiology testing including anal manometry and pudendal nerve terminal motor latency testing may also be employed. Although the results of such testing rarely affect the surgical management of RVF, pudendal nerve damage may be predictive of postoperative continence issues. The association of fecal incontinence in the setting of RVF should prompt a work-up for sphincter injury and would point to sphincter repair as an appropriate treatment even if other alternatives are available.

If malignancy is identified in the setting of RVF, appropriate staging and work-up related to the malignancy itself takes priority. Consideration for treatment of the fistula is secondary to the treatment of malignancy. Management of the fistula should be considered at the time of definitive malignancy management. It is not uncommon for a RVF to develop after treatment of locally advanced anal or cervical cancer involving the rectovaginal septum. Definitive repair should be delayed until the malignancy has been managed.

Classification

Accurate classification of RVFs is not merely a descriptive exercise; it also helps to inform appropriate management. RVFs are classified in several ways: When they are classified relative to the sphincter complex, they are described as either high or low. High fistulas are defined as those above the sphincter complex, and low fistulas are those at or below the level of the pelvic floor, typically at the dentate line. Low fistulas may also be referred to as anovaginal. Fistulas following traumatic obstetric injuries are almost always the low type and associated with sphincter injury. Similarly, cryptoglandular sources of RVFs are almost always low. High fistulas generally have a more complicated etiology, either iatrogenic secondary to anastomotic leak or related to inflammatory bowel disease or malignancy.

Fistulas may also be classified as either simple or complex. Simple fistulas are located in the middle or lower portion of the rectovaginal septum, are small (<2.5 cm in diameter), and are caused by local trauma or infection. Complex fistulas are either larger than 2.5 cm, located in the upper portion of the rectovaginal septum, or secondary to neoplasia, diverticulitis, or inflammatory bowel disease ( Table 151.1 ).

TABLE 151.1
Classification of Rectovaginal Fistulas
Classification Anatomy Etiology
High Above sphincter complex Iatrogenic, irritable bowel disease (IBD), malignancy
Low (anovaginal) Below majority of sphincter Cryptoglandular, trauma, obstetric injury
Simple Middle or lower part of right ventricular septum, <2.5 cm Cryptoglandular, trauma, obstetric injury
Complex Upper sphincter, >2.5 cm Neoplasia, IBD, diverticular

Treatment

Conservative and Medical Management

Although the mainstay of treatment for the majority of RVFs is surgery, there are a few exceptions. On occasion, patients with small fistulas and minimal symptoms choose to pursue bowel management. A small number of fistulas due to obstetric trauma may close spontaneously in the immediate (within 3 months) postpartum period. Hyperbaric oxygen has been reported to be associated with successful healing in a small series of patients with RVFs from obstetric trauma.

Medical therapy, particularly immunomodulation, has played an increasing role in the management of RVFs in patients with Crohn disease. Although the initial trial using infliximab in 1999 showed short-term healing in 56% to 68% of abdominal and perianal fistulas, no RVFs were included. Subsequently the ACCENT II trial looked at 25 women with Crohn RVF. With infliximab, 60.7% healed initially, but long-term closure (at 14 weeks) was significantly lower, with a closure rate of 44.8%. Compared with other fistulas, the rate of RVF healing with infliximab is substantially less, and it has been suggested that this may be because of the relatively thin, nonmuscular, poorly vascularized rectovaginal septum. Use of MRI has shown that although fistulas have apparent healing, they may actually simply become less symptomatic with less drainage even though tracts persist on radiographic studies. Furthermore, closure of the fistula was associated with the development of an abscess in 10% of patients, presumably because the external opening heals over before the tract has healed. Poritz et al. have suggested that use of infliximab does not avoid the need for surgery in more than 70% of patients; however, such patients may be rendered relatively asymptomatic and have a reasonable quality of life before requiring surgical intervention. Infliximab and other immunomodulators are increasingly used as an adjunct to surgery. In a review of 65 women with Crohn disease and RVF, use of immunomodulators such as infliximab or adalimumab in addition to 6-mercaptopurine or azathioprine within 3 months prior to surgery was associated with successful healing ( P = .009). Biologics such as infliximab are commonly used in combination with surgery for Crohn RVF and appear to improve outcome.

Preoperative Management

Preoperative preparation is according to practitioner preference as well as the type of procedure planned. Simple anal- and vaginal-based repairs require only rectal preparation with saline or phosphate enemas the morning of surgery. If a more extensive procedure is planned or if diversion will be necessary, formal bowel preparation is recommended. Perioperative antibiotics and deep venous thrombosis (DVT) prophylaxis are administered per guidelines. Smoking, recognized as a significant predictor for adverse outcome in the repair of RVF, must stop.

Surgical Management

As with any complex anorectal procedure, conditions optimal for healing should be present prior to surgery. The goals of surgery are to preserve continence while achieving repair of the fistula.

The presence of active anorectal or anovaginal sepsis is a contraindication to definitive repair. Drainage of any abscess and establishment of definitive drainage or placement of a seton is required. We recommend delaying definitive surgery in such cases for at least 4 to 6 weeks until all sepsis and inflammation have resolved. Similarly, for treatment of RVF in the immediate postpartum period, we recommend a delay of at least 3 months to allow acute inflammation and sepsis to subside. Frequently spontaneous healing will be seen. Optimization of stool bulking is helpful, and in the setting of incontinence, sacral nerve stimulation may be an effective bridge to definitive repair. In this interval, clarification of anatomy via endoanal ultrasound or MRI can be undertaken.

The four general categories for surgical approach are transanal, transvaginal, transperineal, and transabdominal. Table 151.2 summarizes the various indications and outcomes for these approaches.

TABLE 151.2
Etiology of Rectovaginal Fistulas
Obstetric injury Episiotomy, third- and fourth-degree perineal lacerations
Inflammatory bowel disease Crohn disease
Iatrogenic Anorectal surgery (fistulotomy)
Vaginal surgery (hysterectomy, rectocele repair)
Abdominal surgery (hysterectomy, low anterior resection, J pouch, procedure for prolapse and hemorrhoids)
Infectious Cryptoglandular abscess, diverticulitis, tuberculosis
Neoplastic Anal cancer, rectal cancer, vaginal cancer, cervical cancer
Radiation-induced External beam radiation, brachytherapy

Transanal Approaches

Fistulotomy

A simple fistulotomy is the laying open of the fistula tract. This is usually approached in two stages. A seton is placed, followed by assessment of the degree of sphincter involvement. The second stage of treatment involves cutting of the remaining overlying tissue to lay open the tract. The success rate of this fistulotomy approaches 100%. However, this involves division of a varying amount of soft tissue and possibly sphincter muscle and therefore increases the risk of incontinence. In most cases, such an outcome is unacceptable; therefore simple fistulotomy alone is rarely appropriate.

Advancement Flap

The sliding-flap repair for the treatment of patients with RVF was first reported by Noble in 1902. He advocated splitting the rectovaginal septum, dissecting the lower end of the rectum from the vagina, and drawing the anterior wall down through and external to the anus. Since that time, many modifications of the sliding-flap technique have been proposed. In 1948, Laird described the use of a flap of mucosa, submucosa, and some fibers of the internal sphincter. Kodner et al. advocated the use of a flap similar to the Laird technique. Other authors have advocated the use of a flap of mucosa, submucosa, and the full thickness of the internal sphincter. Incorporating part of the internal sphincter is generally necessary to ensure that the flap is of adequate thickness. Regardless of the thickness of the flap used, the procedure is generally best for patients with simple low fistulas who have not had previous repairs. Patients with RVFs from obstetric injuries (without an associated sphincter defect) and those with Crohn RVFs without associated proctitis are good candidates for sliding-flap repair. We primarily advocate performance of the endoanal advancement flap, since this addresses the repair from the side of high pressure.

After appropriate mechanical and antibiotic bowel preparation, we typically perform this procedure with the patient in the prone jackknife position, which offers optimal exposure of the anterior rectal wall. A probe is inserted through the fistula to help identify it. A U -shaped or trapezoidal flap is outlined, with the base cephalad and twice the width of the apex. This helps to minimize the chance that the distal flap or flap edges will become ischemic. Dissection is started 1 cm distal to the fistula and consists of rectal mucosa, submucosa, and partial thickness of the underlying internal sphincter, including the fistula opening at the apex. We advocate using needle-tip cautery for precise dissection and elevation of the flap in a caudad-to-cephalad manner. The flap should be mobilized 3 to 4 cm proximal to the fistula to ensure tension-free closure. Instillation of a dilute epinephrine solution is an option to facilitate dissection and minimize blood loss ( Fig. 151.1 ). The fistula tract is curetted clear of all granulation tissue and the defect in the remaining internal sphincter is closed primarily with interrupted absorbable sutures. If the patient has a significant sphincter defect, a sphincteroplasty can be performed at this time.

FIGURE 151.1, Endorectal sliding flap. The patient is placed in the prone jackknife position and the fistula is demonstrated (inset, arrow) . (A) The flap should extend for at least 4 cm, and the base should be at least two times the width of the apex. (B) The flap should include mucosa and submucosa in addition to a portion of the internal sphincter muscle. (C) The flap is raised and dissection is performed laterally to permit a tension-free closure. (D) The internal sphincter muscle is plicated over the area of the fistula. (E) Excess flap, including the site of the internal opening of the fistula, is trimmed. (F) The flap is secured with absorbable sutures.

The apex of the flap, including the fistula tract, is excised. The flap is then sutured in place at the cardinal points with full-thickness interrupted 2-0 absorbable sutures, followed by interrupted sutures in a bisecting manner until closure is achieved. The vaginal opening is left open to facilitate drainage.

Success varies considerably (29% to 100%) due to the heterogeneity of the patient groups, variations in surgical technique, and whether initial or ultimate success rates are reported. Common causes of flap failure include ischemia of the flap and hematoma and/or the development of infection under the flap. The number of previous repairs, the presence of a concomitant sphincter defect, and the cause of the RVF all influence successful healing. If a patient has had one or two previous RVF repairs, the success rate with a sliding-flap repair decreases significantly. Therefore a sliding-flap repair should generally not be considered in a patient whose previous repairs have failed. Sphincter function should be assessed and concomitant significant sphincter defects repaired at the time of sliding-flap procedure. The success rate for patients undergoing flap repair and sphincteroplasty with or without levatoroplasty was significantly higher than the success rate for those who underwent flap repair only (80% vs. 41%; P < .02). As a result, some surgeons have advocated that anal ultrasonography and manometry be performed to detect occult sphincter defects in patients undergoing repair of RVFs; sphincter defects, however, in the majority of cases, can be determined by a thorough history and physical examination and then confirmed on manometry and ultrasound. The underlying cause of the fistula may also determine the success of a flap repair. Patients with obstetric injuries as the cause of the RVF have a better outcome than patients with inflammatory bowel disease.

Fibrin Glue

The use of fibrin glue for the repair of RVFs is limited but represents a safe approach without risk of incontinence. In this approach, the fistula tract is identified and curetted to remove granulation tissue. Fibrin glue is injected into the fistula tract from either side until it exits the secondary opening and allowed to polymerize and form a seal. The goal of fibrin glue injection is to provide a temporary mechanical obstruction of the fistula, which then will serve as a scaffold to allow autologous tissue ingrowth. Although the theory is elegant, as with complex anorectal fistulas, experience is disappointing and is almost uniformly a failure, most commonly due to extrusion of the plug due to the short length of the fistula. Modifications in the technique including closure of the internal opening and use of adhesive antibiotics have not improved the outcome.

Fibrin glue has also been used as an adjunct with other procedures such as the endorectal advancement flap. In one series, fibrin glue was combined with an endorectal advancement flap in 12 patients; the failure rate was 50%, which was not significantly different from that with use of an endorectal advancement flap alone. Despite initial enthusiasm, the technique is generally not used for RVFs.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here