Computed tomography revealed thickened small bowel in the right lower quadrant posterior to the rectus abdominus muscle, which was diffusely enlarged within the rectus sheath, with a central area of gas and increased attenuation that communicated with the thick-walled, contrast-opacified small bowel consistent with an enterorectus fistula. As shown in Figure 1, the subcutaneous tissue was free of inflammation. The patient declined recommended treatment with bowel rest, intravenous antibiotics, and steroids, leaving the hospital against medical advice. Three months later, he was readmitted with fever, chills, and abdominal pain. Repeat computed tomography showed a persistence of a fistulous connection between the small bowel and the abdominal wall as well as a non–contrast-enhancing discrete fluid collection in the subcutaneous fat of the right anterior abdominal wall not communicating with the contrast-filled rectus cavity. On clinical examination, there was fluctuance and surrounding cellulitis. This time the patient consented to treatment that consisted of incision and drainage of the abdominal wall abscess performed in the operating room, revealing pus but no evidence of enteric contents or bowel. The abscess site developed into a low-output enterocutaneous fistula and was managed with local wound care. Once the patient was optimized for elective surgery, he underwent resection of the diseased segment of ileum with ileocolic anastomosis and has had no further recurrence in the ensuing 2 years.
Indications for operation in Crohn disease include failure of medical management, intestinal obstruction, symptomatic enteric fistulae, abscess and inflammatory mass, hemorrhage, perforation, cancer and suspected cancer, and growth retardation.1The indication in this case was a symptomatic enterorectus fistula.
A fistula is an abnormal passage or communication between 2 organs (although some insist that a fistula must, by definition, connect 2 epithelialized surfaces, a survey of medical and etymologic dictionaries, including Dorland's Medical Dictionary,2Newcastle upon Tyne Medical Dictionary,3Merriam-Webster's Collegiate Dictionary,4and Oxford English Dictionary,5does not support this criterion). Fistula formation between the small bowel and adjacent structures is a disastrous result of many different pathologic and iatrogenic processes, and it still levies substantial morbidity for patients despite an improvement in the mortality from 45% in 1945 to 15% in 1975.6The cause is postoperative in 82% of cases (with roughly half from anastomotic dehiscence and half from unrecognized enterotomies).
Predictive factors for spontaneous closure of enterocutaneous fistulae include a benign cause, short tract length (<2 cm), small orifice (<1 cm in diameter), low fistula output (<200 mL/d), lack of sepsis, local infection or foreign body, no prior radiation therapy, and adequate nutritional status. Etiologic factors are also predictive of spontaneous closure. Fistulae associated with appendicitis, diverticulitis, and postoperative complications are more likely to close spontaneously than those related to inflammatory bowel disease. Favorable anatomical locations include duodenal stump, esophagus, jejunum, pancreas, and gallbladder. Unfavorable anatomical locations that are associated with a low enteric fistulae closure rate include lateral duodenum, stomach, ligament of Treitz, and ileum.6- 8
A recent report9of 1168 cases of enterocutaneous fistulae in 1 center from January 1971 to December 2000 documented a spontaneous closure rate of 37%, inclusive of all causes. The mortality rate was 5.5%, most of which was related to septic complications. Surgical intervention was required in 56% of patients, and 98% of these patients receiving definitive operations recovered.
Enterocutaneous fistulae are by far the most common of small-bowel fistulae, but enteroenteric, gastroduodenal, gastroenteric, duodenocolic, enterobiliary, coloenteric, aortoenteric, nephroenteric, duodenobronchial, vesicoenteric, cavalenteric, and enterovaginal fistulae have all been reported and reviewed elsewhere. Review of the literature revealed no other report of an enterorectus fistula.
Correspondence:Lena M. Napolitano, MD, Department of Surgery, University of Michigan Health System, 1500 E Medical Center Dr, 1C421 UH/0033, Ann Arbor, MI 48109 (firstname.lastname@example.org).
Accepted for Publication:July 13, 2006.
Author Contributions:Study concept and design: Cunningham and Napolitano. Acquisition of data: Cunningham and Napolitano. Analysis and interpretation of data: Cunningham and Napolitano. Drafting of the manuscript: Cunningham. Critical revision of the manuscript for important intellectual content: Cunningham and Napolitano. Administrative, technical, and material support: Cunningham and Napolitano. Study supervision: Napolitano.
Financial Disclosure:None reported.
Image of the Month—Diagnosis. Arch Surg. 2007;142(7):686. doi:10.1001/archsurg.142.7.686