31. Urinary Tract Fistula
A fistula is defined as an extra-anatomic communication between two or more epithelial or mesothelial lined body cavities or the skin surface. Fistula can occur as a result of congenital anomalies, malignancy, inflammation or infection, tissue trauma, or iatrogenic causes, such as surgical injury or radiation. There have been reports of fistula formation since ancient times, involving connections from the urinary tract to a myriad of bodily cavities and organs. Organ systems immediately adjacent to the urinary tract are the most commonly affected, specifically the reproductive and gastrointestinal systems. Presenting signs and symptoms of urinary fistula are dependent on the termination point of the fistula, the fistula size, concomitant infection or inflammatory processes, and associated malignancy or other medical conditions.
The principles of general fistula management are applicable to all urinary tract fistulas and should be addressed prior to any planned intervention. Issues of nutrition, infection, and malignancy can significantly alter risk factors for initial fistula formation, the approach to repair, and the risk of recurrence following a given intervention. As many urinary fistulas in the industrialized world are iatrogenic, prevention of fistula development is paramount. Intraoperative and early postoperative identification of urinary tract injury allows for immediate management and minimizes the possibility of a fistula.
Principles of treatment and surgical repair of a urinary tract fistula
Elimination of infection
Evaluation for malignancy
Adequate exposure of the fistula tract
Debridement of devitalized or ischemic tissue
Careful dissection to maintain separation of involved organ cavities and hemostasis
Removal of foreign bodies or synthetics
Repair with well-vascularized healthy tissue flaps
Multiple layer closure with non-overlapping tension-free suture lines
Removal of distal obstruction
Maintain adequate urinary tract drainage
Awareness of medicolegal implications
Urologic fistulae are nearly always unexpected occurrences with potentially life-altering implications. The diagnosis is often accompanied by significant distress and frustration. Patients should be approached in a forthright manner, with great care taken to validate their concerns and to present a multimodal treatment strategy that addresses these concerns. Treating physicians must also be mindful of the potential medicolegal implications of the diagnosis, taking great care to set appropriate expectations and documenting meticulously.
Vesicovaginal fistula (VVF) are the most common acquired fistula of the urinary tract . It is defined as a communication between the bladder and vagina, resulting in continuous urinary leakage. Descriptions of vesicovaginal fistulas have been well documented since ancient times, although early attempts at repair met with little success. In 1852, Sims published his method for the surgical treatment of VVF using a transvaginal approach, followed by Trendelenburg in 1888, who successfully performed a transabdominal VVF repair [2, 3].
Etiology and Risk Factors
The etiology and prevalence of VVF differ in various parts of the world. In the industrialized world, the most common cause of VVF is iatrogenic injury during gynecologic, urologic, or other pelvic surgery, accounting for greater than 75% of cases [2, 4, 5]. Hysterectomy is the most common procedure associated with lower urinary tract injury, with most of the remainder a result of general surgical pelvic procedures, urogynecological procedures such as anterior colporrhaphy, cystocele repair, or incontinence surgery, or other urologic procedures . In a study of 207 VVF repairs by Eilber et al., the cause was reported as 83% from abdominal hysterectomy, 8% from vaginal hysterectomy, 4% from radiation, and miscellaneous in 5% .
A review of 25,998 obstetric and gynecologic procedures performed in a Turkish center over 3 years found that bladder injuries were reported in −0.49% of gynecologic operation and 0.18% of obstetric operations .
The overall rate for iatrogenic bladder injury at the time of hysterectomy is between 0.5% and 1.0%, while the incidence of fistula is approximately 0.1–0.2% [9, 10]. The primary risk factor for the development of VVF following hysterectomy appears to be intraoperative injury. Iatrogenic cystotomy, tissue necrosis from cauterization injury, or suture placement through both the bladder and vaginal wall can predispose to postoperative fistula formation. Tissue ischemia and necrosis lead to fibrosis and inflammation between the bladder and vagina, eventually allowing formation of an epithelialized tract. This most commonly occurs at the apex of the vagina at the level of the vaginal cuff . Preoperative risk factors include prior cesarean section or uterine surgery, endometriosis, infection, diabetes, arteriosclerosis, pelvic inflammatory disease, and prior pelvic radiation . Additionally, abdominal hysterectomy is three times more likely to result in bladder injury compared to vaginal hysterectomy.
In the industrialized world, radiation is also a significant cause of complicated urinary tract fistula. The incidence of radiation-induced fistula is dependent on the type, dose, and location of radiation, as well as the specific malignancy undergoing treatment. Urinary fistula rates of 1.6% have been reported following radiation treatment for cervical carcinoma . VVF from radiation may occur as long as several decades following treatment . Biopsy of the fistula tract in such cases should be strongly considered prior to any definitive therapy to exclude recurrence of the primary malignancy. Malignancy-induced VVFs can occur with locally advanced cervical, vaginal, and endometrial carcinomas and account for approximately 3% of fistulas . The management of malignant fistulas may be very different from the benign type combining extirpative surgery with subsequent reconstruction and/or complete urinary diversion.
In the developing, non-industrialized world, VVF most commonly results from complications of childbirth. The incidence of obstetric fistula in developing countries is approximately 0.3–0.4% of deliveries, or between 1 and 4 per 1000 vaginal deliveries [16, 17]. In a study surveying 14,070 reproductive age women in Ethiopia, 1% experienced obstetric fistula in their lifetime. Women who gave birth ten or more times were far more likely to develop a fistula than those with four or fewer childbirths . Routine prenatal and perinatal obstetrical care is limited, as is access to general healthcare. Additionally, pelvic size may be small due to poor nutritional status and/or an early age of marriage and conception . Prolonged obstructed labor due to cephalopelvic disproportion can cause pressure necrosis of the anterior vaginal wall, bladder, bladder neck, and proximal urethra. The “obstructed labor injury complex” which occurs in such individuals includes variable degrees of urethral loss, stress incontinence, renal failure, vesicovaginal fistula, rectovaginal fistula, rectal atresia, anal sphincter incompetence, vaginal stenosis, osteitis pubis, and foot drop . Obstetric fistulas tend to be larger than iatrogenic gynecological VVF, with necrosis of large parts of the anterior or posterior vaginal wall and/or urethra, distally near the true pelvis and pubis. Repair can be exceedingly complicated due to the large areas of necrosis and poor adjacent tissue quality due ischemia and/or inflammation as well as trophic skin changes due to large volume urine loss.
Evaluation and Diagnosis
The most common presentation for vesicovaginal fistula is persistent, continuous urinary drainage from the vagina. The amount of drainage is variable and may be directly related to the size of the fistula tract. Pain is uncommon but can be present in cases with extensive skin irritation or prior radiation. VVF should be distinguished from urinary incontinence due to other causes including stress, urge, and overflow incontinence, as well as ureterovaginal or urethrovaginal fistula.
Commonly utilized procedures during patient examination for the evaluation of stress urinary incontinence, vesicovaginal fistula, and urethrovaginal fistula
Intravesical indigo carmine/methylene blue
Visualize leak with cough through the urethral meatus = stress incontinence
Intravaginal Pad Test
Intravesical indigo carmine/methylene blue/Uribel
Distal pad blue = stress incontinence or urethrovaginal fistula
Double Dye Test
Intravesical indigo carmine/methylene blue
And oral phenazopyridine
Proximal pad blue = VVF
Distal pad blue = stress incontinence or urethrovaginal fistula
Middle/proximal pad blue = VVF
Upper pad orange = Uretero-vaginal fistula
Cystoscopy An endoscopic evaluation should be performed in all patients with suspected VVF. Immature fistulas are often surrounded by bullous edema and do not have a distinct ostium. Mature fistulas are variably sized with smooth, distinct margins. In many cases, especially from iatrogenic VVF, the fistula site will be located on the posterior bladder wall, at or just above the intertrigonal ridge, frequently with multiple pits present, making it difficult to localize the specific tract. In cases where identification of the fistula is difficult, cystoscopic passage of a guide wire via the fistula tract can confirm the exact location of the fistula within the bladder and the vagina simultaneously.
Imaging Evaluation of VVF should include both bladder and upper tract imaging. A voiding cysto-urethrogram (VCUG) may objectively determine the presence and location of the fistula tract. With bladder filling, the contrast will opacify the vagina, usually best seen in a lateral image projection. Voiding images are occasionally necessary to visualize small VVF, as the increase in intravesical pressure will facilitate fistula drainage. A complete VCUG in the evaluation of VVF includes filling, voiding, and drainage films in multiple projections (A-P, lateral, and oblique). CT or a specialized CT cystogram may be utilized for the evaluation of VVF in certain centers .
Ureteral injury or ureterovaginal fistulas can be present in up to 12% of postsurgical VVF; therefore, upper-tract evaluation is obtained routinely . This can be accomplished easily and successfully with intravenous urography, CT urography, or MR urography. Retrograde pyelograms may be utilized if the distal ureter is not well visualized, and a concomitant ureterovaginal fistula is suspected but has not been demonstrated [1, 12]. Delayed visualization of contrast within the vagina on CT urogram or direct contrast extravasation into the fistula tract on CT cystogram provides alternate means of evaluation, with the added ability to detect additional intra-abdominal pathology .
The goal of treatment of VVF should be the timely and complete cessation of urinary leakage with minimal effect on normal urinary and genital function.
Conservative measures can be considered for small fistulas, typically less than 2–3 mm in diameter. This consists of continuous bladder drainage with an indwelling catheter along with anticholinergic medications to manage symptoms. In properly selected patients, fistula closure can occur after 2–3 weeks . Pooled data suggest a 13% spontaneous closure rate for fistulas managed with initial catheterization . Ongoing drainage of urine from the vagina after placement of a foley catheter indicates a persistent fistula tract and other methods should be considered for treatment as appropriate. Prolonged foley catheter drainage in a persistent fistula will cause considerable patient discomfort and is doomed to fail.
Cystoscopic electrocoagulation of the epithelialized fistula tract in conjunction with bladder catheterization may provide some additional benefit. Stovsky et al. reported successful ablation in 11/15 patients with fistula tract diameters less than 3.5 mm . Fibrin sealant has also been utilized with some success to plug the fistula tract, presumably until tissue ingrowth occurs . Again, if persistent leakage is noted with indwelling urethral catheterization following electrocoagulation, other methods should be considered in the short term.
The timing of intervention for VVF is a contentious issue. Classic teaching advocates delaying repair several months after diagnosis to allow for stabilization of inflamed or necrotic tissue and recovery from the inciting event. More recently, however immediate intervention has become the preferred approach, at least in uncomplicated iatrogenic fistula. Early repair can minimize patient discomfort and anguish without compromising surgical repair [12, 31, 32, 33–35]. In complex cases, however, such as those involving continued infection, obstetric etiology or radiation, a waiting period of anywhere from 1 to 12 months may be necessary to allow demarcation of inflamed or devascularized tissues [20, 36, 37]. During any such waiting period, special attention should be paid to skin protection (incontinence pads, barrier creams) and to nutritional status.
Once the decision to pursue definitive repair has been made, thoughtful surgical planning is essential to maximize chances of success. The first attempt at VVF repair is typically offers the best opportunity for success, free from some of the scarring, anatomical distortion and revascularization that often complicate salvage procedures.
VVF can be repaired via a transvaginal or transabdominal approach. There is no “correct” approach, and each option has advantages and disadvantages. Whereas vaginal repairs are typically outpatient procedures that can be done immediately and regardless of surgical history, abdominal repairs introduce the potential complications involved in abdominal surgery, often in patients who have recently had a complicated abdominal surgery. Transabdominal repairs are associated with greater blood loss and longer hospital stay . Transvaginal surgery may be challenging in treating fistulas located high at the vaginal cuff in a deep vagina, or in patients with a narrowed vagina due to radiation, and in patients who are unable to be placed in a high-lithotomy position. In such patients, an abdominal or minimally invasive approach (laparoscopic/robotic) can be considered. An abdominal approach is necessary in patients requiring a concomitant ureteral reimplantation or in the case of complex fistulas involving adjacent organs. Regardless of the approach taken, ureteral catheter placement should be considered in the case of any fistula located close to the ureteral orifices to avoid inadvertent injury to these structures.
Surgical management of vesicovaginal fistula; comparison of transabdominal and transvaginal approaches to repair
Delayed (3–6 months)
Reimplant possible if indicated
Reimplant not possible
No change in vaginal depth
Risk of vaginal shortening
Labial, peritoneal, gluteal, gracilis
Large fistula, high fistula in narrow vault, radiation, failed vaginal approach, other procedures (augment)
Low fistulas, failed transabdominal repair
It should be noted that in certain extreme cases where repair is not feasible or not possible, urinary diversion should be considered. This is most common in situations involving active pelvic malignancies, multiple failed repairs, severe radiation damage or other cases of extreme tissue loss [42–44].
The classic transabdominal VVF repair was described by O’Conor in 1980 . Transabdominal approaches for fistula repair include supravesical or transvesical approaches, and laparoscopic/robotic techniques. The O’Conor transabdominal VVF repair has been well described . The patient is positioned in a low lithotomy position, with access to the vagina and abdomen. Ureteral catheters may be placed and are recommended if the fistula is near the ureteral orifices or the trigone. A lower midline incision is performed and the bladder is mobilized. The bladder is then bivalved vertically to the level of the fistula, and dissection is continued distally to open the vesicovaginal space, 2–3 cm distal to the fistula site. Following mobilization of the vaginal wall from the bladder wall distal to the fistula tract, the fistula tract is excised, and the vaginal wall is closed with running synthetic absorbable suture (SAS). The bladder is closed in multiple layers with running SAS. An additional layer of tissue can be placed between the suture lines utilizing an omental interposition flap or peritoneal flap. It is important to secure the interpositional flap distally beyond the fistula.
A later adaption, known as the transvesical approach, mitigated some of the morbidity associated with the complete bivalving of the bladder. Instead, the bladder is opened via an anterior wall midline cystotomy. The VVF tract is visualized on the posterior wall, were it is then circumscribed and excised. Following mobilization of the vesicovaginal space surrounding the fistula site, the vaginal and vesical tissues are closed separately. A flap of adjacent bladder tissue may be advanced to avoid overlapping suture lines as described by Gil-Vernet .
Interest in minimizing the morbidity of transabdominal VVF repair led to the advent of minimally invasive approaches. First reported in 1994 by Nezhat et al., laparoscopic VVF repair has been described in several case series in both transvesical [47–49] and extravesical  approaches, with various modifications, with or without the utilization of omental or peritoneal flaps. A literature review found success rates across multiple case series to be 93.5%, comparable to open repair, with a complication rate of 2.3% . Advantages of minimally invasive surgery, such as improved visualization, decreased blood loss, shorter length of stay and decreased convalescence are well established. One particular advantage of minimally invasive transvesical techniques is the use of a limited posterior cystotomy, typically less extensive than anterior cystotomy in typical open cases and, as such, is likely less morbid than the formal bivalving of the bladder required in the classical O’Conor procedure . Potential drawbacks of laparoscopic VVF repair include the longer operative time, potential injury to intraabdominal structures (as compared to the transvaginal approach) and the advanced laparoscopic skill necessary to successfully complete the procedure.
The advent of robotic surgery helped to overcome some of the perceived drawbacks of pure laparoscopy, most notably a steep learning curve pertaining to dissection and suturing. By the mid-2000s series of robotic-assisted laparoscopic VVF repair began to appear in the literature. The largest series to date by Bora et al. described robotic assisted laparoscopic repair of 30 VVF. Average fistula size was 10.3 mm and 11 of the 30 fistulas were characterized as “complex” (prior failure, prior radiation, obstetric cause). Their technique involved cutting down posteriorly directly on to the fistula, guided by manipulation of a traversing open-ended catheter. Eighteen patients underwent interposition flaps (epiploic, omental, peritoneal). They reported two recurrences, with an overall success rate of 93.3%. No complications were reported . A recent review performed by the same group of smaller published cases and case series of robotic VVF repair corroborated the impressive success rate . However, it must be cautioned, that such an approach should not be attempted by those who are unskilled in robotic techniques.