GI Anomalies: Colovesical Fistula

During hospitalization for aortic valve replacement, a 45-year-old man with rheumatoid arthritis and an extensive history of cardiac disorders was found to have a urinary tract infection (UTI) and pneumaturia. He had had recurring UTIs and pneumaturia for the past 2 months. Review of systems was otherwise negative.

The patient had diverticulitis of 6 months’ duration that resolved 3 months before his current hospitalization. He also had hypertension and hyperlipidemia. He had had an appendectomy and mediastinal lymph node biopsy for benign lymphadenopathy. His family history was significant for diabetes mellitus and sarcoidosis. He denied use of alcohol or drugs; he was a former smoker but had quit in his youth.

Temperature was 37ºC (98.6ºF); blood pressure, 107/40 mm Hg; heart rate, 87 beats per minute; respiration rate, 18 breaths per minute. Weight was 216 lb (he had lost 50 lb in the past 6 months); height, 6.2 ft. Oxygen saturation was 96% on room air. The patient appeared mildly cushingoid.

A grade 2/6 diastolic murmur was most audible along the right second intercostal space; this finding was consistent with aortic insufficiency. The abdomen was soft without hepatosplenomegaly or palpable masses; bowel sounds were active and normal. The patient’s history and physical findings strongly suggested colovesical fistula.

A CT cystogram was negative for fistula but revealed gas in the bladder, which could have been introduced during contrast dye injection (A). Cystoscopy revealed a small lateral bladder wall tumor but no fistula. Evaluation of the tumor was deferred. Before discharge, the patient underwent abdominal and pelvic CT scans that were also negative for fistula but positive for air in the bladder (B).

Two weeks later, the patient was readmitted for continued pneumaturia, UTI, and possible sepsis. After a water-soluble contrast enema revealed only diverticulosis (C), an exploratory laparotomy was performed. During the procedure, a colovesical fistula was finally identified and corrected.

Any process that promotes inflammation and development of scar tissue may create an environment conducive for fistula formation. The most common risk factor associated with colovesical fistula is diverticulitis; it was reported in 75% of cases in one study, although the frequency in other studies ranges from 21% to 90%.¹ Other risk factors are malignancy, postoperative irradiation, Crohn disease, previous abdominal surgery, and trauma.¹

Pneumaturia and fecaluria are 2 hallmark findings.² Other characteristic findings are suprapubic pain, urinary frequency, dysuria, vesical tenesmus, and a history of recurrent UTIs and urine passed through the rectum.

Abdominal CT is currently the best available diagnostic modality^1,3; however, colonic surveillance, barium enemas, cystoscopy, and cystography are also widely used. Frequently, a constellation of CT findings—namely, air in the bladder, thickened bowel and/or thickened bladder wall, and colonic diverticula—points to the diagnosis.1 Actual visualization of the fistula tract is rare. Newer modalities, such as 3-dimensional CT scanning, may provide greater direct visualization.⁴

One study examined the efficacy of poppy seed ingestion to reveal enteric fistulas. After oral ingestion of poppy seeds, their recovery in subsequent urine samples provided indirect but compelling evidence of the presence of a colovesical fistula. The poppy seed test has a 100% sensitivity; the cost is 100 times lower than the average diagnostic modality.⁵

Treatment typically consists of resection of the fistula and closure of the affected organs.