Enterovesical Fistula Treatment & Management

Updated: Dec 24, 2020
  • Author: Joseph Basler, MD, PhD; Chief Editor: Bradley Fields Schwartz, DO, FACS  more...
  • Print

Medical Therapy

Nonsurgical treatment of colovesical fistulae may be a viable option in patients who cannot tolerate general anesthesia or in selected patients who can be maintained on prolonged antibacterial therapy for symptomatic relief.

Colovesical fistulae in patients with diverticulitis who are deemed to be a surgical risk have been managed conservatively. In highly select patients, nonoperative therapy has been reported as a viable treatment option. Six patients observed for 3-14 years encountered little inconvenience and were without significant complications while on intermittent antibacterial therapy alone. [55] In another study, six patients who declined surgical intervention were monitored and were found to exhibit no significant changes in renal function, and urosepticemia was not documented. [56]

If the fistula closes spontaneously, which occurs in as many as 50% of patients with diverticulitis, requirements for resection depend on the nature of the underlying colonic disease. Some patients tolerate a colovesical fistula so well that surgery is deferred indefinitely. However, although some small studies have suggested conservative management as a reasonable option, no randomized controlled trials have supported conservative management, and careful selection with close follow-up is stressed.

For enterovesical fistulae due to Crohn disease, medical therapy is the first choice. [57] Zhang et al reported that 13 of 37 patients with Crohn disease achieved long-term remission of enterovesical fistulae over a mean of 4.7 years through treatment with antibiotics, azathioprine, steroids, and/or infliximab. Significant risk factors for surgery included sigmoid-originated fistulae and concurrent Crohn disease complications such as small bowel obstruction, abscess formation, enterocutaneous fistula, enteroenteric fistula, and persistent ureteral obstruction or urinary tract infection. [58]

In a study analyzing the outcomes of 97 patients with enterovesical fistulae due to Crohn disease, the use of anti-TNF agents was associated with an increased rate of remission without need for surgery (hazard ratio 0.23, 95% confidence interval 0.12–0.44; p < 0.001). [59]  A review of Crohn disease–related internal fistulae, including 16 enterovesical fistulae, treated with anti-TNF agents (infliximab or adalimumab) reported a cumulative surgery rate of 47.2%, and a fistula closure rate of 27.0% at 5 years from the induction of anti-TNF therapy. [60]

Patients with advanced carcinoma may be treated with catheter drainage of the bladder alone or supravesical percutaneous diversion.


Surgical Therapy

Open surgery

Colovesical fistulae can almost always be treated with resection of the involved segment of colon and primary reanastomosis. Fistulae due to inflammation are generally managed with resection of the primarily affected diseased segment of intestine, with repair of the bladder only when large visible defects are present. The bladder usually heals uneventfully with temporary urethral catheter drainage. Suprapubic tube diversion is an option but is not necessary. [61]

Historically, staged procedures were used to treat colovesical fistula. Staged repairs may be more judicious in patients with large intervening pelvic abscesses or in those with advanced malignancy or radiation changes. Most cases do not involve abscesses. If an abscess is present, spontaneous drainage through the fistula into the bladder may alleviate the immediate need for drainage if the bladder is emptying under low pressure. Further operations may be delayed pending culture results and after adequate antibiotic therapy has reduced the inflammation. A one-stage operation is recommended for patients in good general health who have a well-organized fistula and no systemic infection. [62]

A diverting colostomy, with or without urinary diversion, may be used as a long-term solution for palliation or severe radiation damage in cases of advanced cancer.

Endoscopic treatment

A review of the literature reveals one reported case of a colovesical fistula treated with transurethral resection with no evidence of recurrence in more than 2 years of follow-up. [63] With the development and advancements of hemostatic sealants, endoscopic injection of these materials is possible as a minimally invasive treatment. One concern would be the presence of foreign material in direct contact with the urine possibly acting as a nidus for stone formation. Few clinical trails have studied the application of these sealants, and this author does not recommend their use from an endoscopic approach.

Laparoscopic treatment

Several reports suggest that laparoscopic resection and reanastomosis of the offending bowel segment is possible as a minimally invasive treatment. [64, 65, 66, 67, 68] However, an abdominal incision is still required for removal of the affected intestinal segment intact for pathological assessment to rule out cancer.



Preoperative Details

The usual preoperative medical evaluation and staging (in the case of suspected or diagnosed cancer) should be performed. In addition, a preoperative mechanical and antibiotic bowel preparation is performed. At this author's institution, this includes an oral lavage with polyethylene glycol & electrolytes (GoLYTELY or its equivalent) and oral neomycin and erythromycin base. A second-generation cephalosporin is generally administered intravenously for antibiotic prophylaxis.

Other variations of this bowel preparation, such as colonic irrigation with a povidone-iodine solution, have also been used successfully. Surgeon preference dictates which is used. The goal is to clear as much fecal content and as many bacteria as possible before resection to allow uncomplicated healing after successful surgery.


Intraoperative Details

The colon is mobilized proximal and distal to the fistula. Pinching the colon off the bladder with blunt dissection may be possible, but separating the two organs usually requires a careful and tedious sharp dissection. [53]

Inflammatory fistulae

Diverticulitis is generally managed with blunt dissection of the colon from the bladder, resection of the colon, and primary anastomosis. Often, when the colon is freed from the bladder, the bladder does not contain an actual opening. Many of these fistulous tracts are tiny, and, if the opening into the bladder is not apparent, it can be demonstrated by distending the bladder via a catheter with fluid that contains methylene blue. A large visible opening can be closed in two layers with interrupted absorbable sutures. Smaller lesions can be left alone. [4]

Fibrin sealant closure of a contaminated fistula has been described, with no evidence of fistula recurrence at 4 years. [69] The diseased bowel is resected, and a primary anastomosis is usually created. If suitable omentum is available, it may be interposed with tacking sutures between the bladder and bowel. Extensive inflammatory involvement of the bladder wall, once thought to require partial bladder resection, does not necessarily require removal of any part of the bladder. Excision of involved bladder tissue is necessary only for carcinoma. [4]

Carcinoma-related fistulae

To avoid tumor spillage, a circumscribing incision around the tumor mass and through the bladder wall is made. Frozen sections of the margins are sent for histologic analysis. Further resection is undertaken as indicated, and, if frozen section analysis results eventually return as negative, a multilayered closure and omental interposition are performed. This may help reduce postoperative complications and the risk of recurrence.

Radiation-induced fistulae

Surgery to manage radiation-induced fistulae can be difficult. In severe cases, the colorectal and adjacent organs are matted together with no natural planes, making mobilization and resection hazardous. In this situation, diverting proximal colostomy or ileostomy is advisable. In milder cases in which resection can be safely performed, a descending anal anastomosis, with or without a colonic J pouch, can be performed. [51]

The urinary system can be left intact with catheter drainage, although healing in this situation is slow and may require longer periods of catheterization. Typically, surgical separation of the genitourinary and gastrointestinal systems is required, and staged operations are more commonly performed because of the poor quality of tissues. When healing is not expected, a transverse colon conduit is often successful at restoring quality of life. Ileal and sigmoid conduits are less favorable because they have often been in the field of radiation. [24]


Most colovesical fistulae enter the bladder well away from the trigone. When fistulae enter the bladder close to the trigone, avoid periureteral dissection to prevent devascularization. If identification is difficult, ureters can be stented intraoperatively or observed either endoscopically or through the vesicostomy after intravenous injection of indigo carmine or methylene blue.


Surgical management of the bladder varies. The technique of bladder repair (ie, excision versus oversewing) is not critical, and small defects do not require any particular repair. As long as adequate bladder drainage is provided, variations in bladder management are unlikely to affect the patient outcome. When available, omentum should be applied to the serosal surface . This may be particularly beneficial in the setting of acute traumatic injury to both the bladder and rectum, to aid in healing and may prevent future fistula formation. [27] To date, no studies have demonstrated that the choice of absorbable suture, the number of layers of closure, or the type of postoperative bladder drainage significantly affects outcomes.

Dziki et al conducted a retrospective study of 59 patients with enterovesical fistula from benign causes who underwent an open surgical single- stage procedure with bowel anastomosis, without protective stoma. In 18 patients (31%) with urinary bladder leakage observed, the bladder wall defect was closed with a single No. 0 absorbable polyglactin 910 (Vicryl) suture. No bladder leakage occurred postoperatively in any of the 59 patients, suggesting that closure of the bladder defect is not necessary in cases where a leak is not demonstrated from the bladder intraoperatively. [70]


Postoperative Details

A nasogastric tube can be left in place or the patient can continue on nothing by mouth (NPO) status until bowel function returns, depending on surgeon preference. The use of rectal stimulatory suppositories (for high nonrectal fistulae) may hasten the return of bowel function. Concomitant treatment with parenteral or low-residue enteral feeding may be appropriate. Treatment with steroids is continued in patients with Crohn disease, but slower healing of the bladder should be anticipated. Bladder drainage is continued, taking care to ensure low-pressure unobstructed urine flow.



After repair of fistulae caused by benign disease, the urinary catheter is left in place for 5-7 days or longer depending on the level of inflammation and size of the repair. The patient remains on appropriate antibiotics (ie, based on preoperative culture findings and sensitivity). At the next observation, a repeat urine culture with sensitivity is obtained. The author’s preference is to perform a gravity cystography with postdrainage films to confirm healing before catheter removal. Antibiotics are continued for 24-48 hours after catheter removal until the culture results are documented as negative.

Thereafter, the primary enteric process is treated as indicated, and the patient is periodically observed with urinalysis and cultures as indicated. Patients are usually aware of the symptoms of recurrence and should be encouraged to return early if they experience any indication of infection, pneumaturia, or fecaluria.

If cancer resection is performed, observational colonoscopy and CT scanning are obtained as indicated based on tumor histology findings and stage. Periodic cystoscopy may also be indicated because of the possibility of local recurrence in the detrusor muscle. Cystoscopy is especially important if the margin status of the tumor is questionable.

Certainly, any hematuria in the postoperative period should be carefully evaluated with upper tract imaging and cystoscopy.



In a 1988 study, Woods et al reported a 3.5% operative mortality rate and a complication rate of 27%. [43] Fistula recurrences have been reported in 4%-5% of patients. Most other studies have not reported such high operative mortality rates, except in the cases of severely ill patients with other significant medical problems.

Short-term complications include the usual potential problems after general surgery (eg, fever, atelectasis, slow return of bowel function, catheter-related UTI, deep vein thrombosis [DVT], wound breakdown and infection). These complications are largely preventable with incentive spirometry, early ambulation, thromboembolic hose or anticoagulation in susceptible patients, and appropriate wound-closure techniques.

Long-term complications include persistent bladder leak (usually observed after radiotherapy for carcinoma), recurrence of a fistula (also more likely after radiotherapy), pelvic/abdominal abscess (from a leaking anastomosis), cutaneous fistulization (also from a leaking anastomosis), and bowel obstruction (from adhesions or recurrent diverticulitis).

Consider recurrent cancer in the abdomen or previously involved bladder wall when patients return with signs of bowel obstruction, new hematuria, or irritative voiding. Repeat CT scanning, serum carcinoembryonic antigen (CEA) measurement, urine culture and cytology, and cystoscopy are indicated in these settings.


Outcome and Prognosis

In a retrospective record review of 76 patients diagnosed with enterovesical fistula over a 12-year period, the complication rate in those treated with single-stage repair was not statistically different from that in patients who underwent multistage repair. [71]

In general, the overall outcome and prognosis are excellent in patients with non–radiation-induced or cancer-induced fistulae. Such patients usually respond well to resection of the diseased colon and have no significant urinary sequelae.

The prognosis in patients with colon carcinoma and fistulization is less favorable because the involvement of the bladder usually heralds a more aggressive tumor that often is metastatic at the time of detection.

Radiation-induced fistulae are more likely to recur, but the long-term patient prognosis may be better if the malignancy for which the radiation was administered has been controlled.

Reported outcomes of laparoscopic surgery for enterovesical fistulas has been limited to small series and observational studies, with conversion rates that range from 0%-36%. Larger studies are needed to determine the saftey and efficacy of this approach. [72]


Future and Controversies

Future treatment of typical enterovesical fistulae may focus on development and refinement of minimally invasive surgical techniques, such as laparoscopic and robotic, to shorten recovery time and to potentially decrease hospital stay. The benefit and complications of preoperative ureteral stenting for intraoperative identification of the ureter may become a particular area of interest for clinical study. As of the latest review of this article, there are few clinical data to demonstrate the benefit of ureteral stents in preventing ureteral injury.

New modalities in neoadjuvant chemotherapy may allow further bladder preservation strategies. Trends in radiation oncology that permit minimization of collateral organ damage (eg, conformal external beam radiotherapy) and the use of tumor-specific radiosensitizing agents may be highly useful in preventing radiation-induced fistulae.

Improved surgical techniques, including laparoscopic procedures that greatly enhance visualization of the operative field, hold promise for fewer fistula-related complications of gynecologic and urologic procedures.

Advancements in molecular biology and the development of medications to manipulate inflammatory mediators may eventually produce more specifically targeted therapies to decrease the risk of fistula formation in inflammatory conditions, particularly Crohn disease.

Given the low incidence of colovesical fistula, the development of and recruitment for prospective randomized studies is difficult.