Enterocutaneous Fistula

Content below is from a textbook reference chapter (Sabiston Ch50 - Diverticular Disease & Miscellaneous Problems), not primary literature. See also Crohn Disease’s “Penetrating disease” subsection for the Crohn-specific fistula discussion — this page covers the general small bowel fistula (mostly postoperative/iatrogenic) presentation.

Epidemiology and causes

Despite improvements in surgical nutrition and critical care, mortality from enterocutaneous fistulas remains high — 10% in recent reports. Improvements in outcome focus on prevention and, when fistulas occur, prompt recognition and intervention. Multidisciplinary care is critical for improving fistula outcomes. Enterocutaneous fistulas are most commonly iatrogenic, with 75–85% occurring during surgical intervention (e.g., anastomotic leakage, injury of the bowel or blood supply, erosion by suction catheters, laceration of the bowel by wire mesh or retention sutures). The remaining 15–25% of fistula occurrences are associated with predisposing conditions such as Crohn disease, malignant disease, radiation enteritis, diverticulitis, intraabdominal sepsis, or trauma.

Recently, the popularization of damage control laparotomy and staged management of the open abdomen has led to a more virulent form of small bowel fistula referred to as an enteroatmospheric fistula. These patients typically present with an open segment of intestine exposed through a large fascial defect, without a surrounding epidermal margin.

Enterocutaneous fistulas are classified according to their location and volume of daily output (Table 50.12 in the source lists favorable vs. unfavorable factors for nonoperative closure — favorable: surgical etiology, appendicitis/diverticulitis cause, transferrin >200 mg/dL, no bowel obstruction/discontinuity/infection/inflammation, length >2 cm, end fistula, output <200 mL/24h, no sepsis with balanced electrolytes, early tertiary-center referral; unfavorable: ileal/jejunal/ nonsurgical etiology, IBD/cancer/radiation cause, transferrin <200 mg/dL, distal obstruction/bowel in discontinuity/adjacent infection or inflammation, length <2 cm or lateral/multiple fistulas, output

500 mL/24h, sepsis with electrolyte disturbances, delayed tertiary-center referral). These factors dictate treatment and morbidity/mortality rates — proximal fistulas are associated with higher output, greater fluid/electrolyte loss, and greater loss of digestive capacity; distal fistulas tend to have lower output, making them easier to manage and more likely to close spontaneously. High-output fistulas are those that discharge ≥500 mL per 24 hours.

Multiple factors prevent spontaneous fistula closure: retained foreign body, radiation enteritis, inflammatory bowel disease or infection, epithelialization of the fistula tract, neoplasm, and distal obstruction.

Clinical manifestations

Recognition of enterocutaneous fistulas is usually not difficult. The typical presentation is a febrile postoperative patient with an erythematous wound. When a few skin sutures are removed, a purulent or bloody discharge is noted; leakage of enteric contents then occurs, sometimes immediately but often within 1–2 days. The diagnosis rarely eludes the surgeon for long. Small bowel fistulas can also present with generalized peritonitis, though this is less common.

Treatment

Once a fistula is identified, management should focus on prompt IV fluid resuscitation and consideration of potential factors that could prevent spontaneous closure. Successful management requires a coordinated staged approach in three phases:

1. Stabilization

Historically, malnutrition and fluid losses were the leading causes of death in patients with small bowel fistula; with better nutritional support and critical care, sepsis has become the most common cause of death. Prompt fluid resuscitation and electrolyte replacement should occur on recognition of a fistula. Sepsis control is critical, and early CT scanning may be invaluable in identifying undrained abscesses, complete distal obstructions, or generalized intraabdominal sepsis with peritonitis (Box 50.7 in the source lists a full treatment strategy: sepsis control — radiologic abscess drainage, relaparotomy on demand [minimally invasive if possible], consideration of other infectious foci; optimization of nutritional status; wound care; anatomy characterization; timing of surgery — clinically stable, psychologically willing, albumin >25 g/L, convalescence >6 weeks; surgical strategy — one-stage procedure, careful adhesiolysis, wedge excision of intestinal resection, limiting anastomoses, covering sutures with healthy tissue, avoiding compromised areas). Once sepsis is controlled and the patient is resuscitated, effluent control with skin protection and adequate nutrition are necessary. Fistula output is best controlled by intubation of the fistulous tract with a drain. Protection of the skin around the fistulous opening is important to prevent excoriation and skin destruction — most easily accomplished with a stomahesive product plus zinc oxide/aluminum paste/karaya powder applications; a suction catheter can be brought out through the end of the stomahesive bag, cut to just fit the fistulous opening, allowing collection and accurate measurement of output. TPN has been an important advance in managing high-output enterocutaneous fistulas and significantly decreases malnutrition incidence — particularly valuable in stabilization for immediate nutritional repletion while the fistula is being delineated. If the patient can meet calorie goals without TPN, especially when a high-output fistula is not present, enteral feeding is preferable and recommended.

2. Staging and supportive care

Once sepsis is controlled and nutritional therapy instituted, the fistula must be adequately staged — combined fluoroscopic contrast studies, fistulography if necessary, and CT, along with the patient’s clinical behavior, characterize the anatomy/underlying pathology. Some advocate conservative management for up to 3 months to allow spontaneous closure; others have shown that of small intestinal fistulas that closed, >90% did so within 1 month, <10% closed after 2 months, and none closed spontaneously after 3 months. In one large retrospective study, the majority of enterocutaneous fistulas closed spontaneously (54%), while 18% needed definitive surgery at a later time; operative mortality was 9.8% while recurrence rate was 8%. Uncomplicated proximal fistulas have higher spontaneous closure rates, with duodenal fistulas closing within 2–4 weeks. A reasonable management plan is to follow a 6-week period of convalescence, at which point surgical management should be considered if closure has not been obtained — though if the preoperative albumin level is above 25 g/L, spontaneous closure is unlikely and should not prompt immediate reexploration at 8 weeks. In general, a period of 3–6 months is beneficial to allow the profound inflammatory response associated with intraabdominal sepsis to subside completely and for adhesion formation to stabilize — this provides a better opportunity for safe/successful operative intervention. As with enteroatmospheric small bowel fistulas, it may take several months to stabilize the complex abdominal wound associated with the fistula.

Several adjuncts have been proposed to assist spontaneous fistula closure and abdominal wound management, though none are supported by vigorous level I data. Bowel rest with TPN therapy may improve fistula closure rates and time to closure in high-output fistulas. Low-output fistulas can be successfully managed with enteral therapy while avoiding known parenteral nutrition complications. Dysmotility agents (loperamide, codeine) can assist attempts at enteral therapy. Newer techniques such as fistuloclysis (the distal limb of a proximal fistula is intubated and enteral therapy delivered to the distal bowel) have proved effective. Several randomized trials evaluated octreotide’s role in fistula management — octreotide decreases fistula output (useful in high-output fistulas) but has not convincingly improved spontaneous closure rates. Vacuum devices are valuable for enteroatmospheric fistulas to help contract the open abdominal wound around the associated fistula. Care should be taken to avoid direct contact with visceral contents as this can cause new fistulas. Skin grafting up to the fistula has also been used with an open abdomen, with graft success rates up to 80% in some series. Patients who could not be discharged before definitive repair also have higher mortality risk.

3. Definitive management

If the fistula persists despite adequately addressing the patient’s nutritional, fluid, and wound needs, reoperative intervention will ultimately be necessary for some patients. Surgery is most easily accomplished by entering the previous abdominal wound, with great care to avoid further damage to adherent bowel. The preferred operation is fistula tract excision and segmental resection of the involved segment of intestine and reanastomosis. Simple closure of the fistula after removal of the fistula tract almost always results in fistula recurrence. If an unexpected abscess is encountered or if the bowel wall is rigid and distended over a long distance (making primary anastomosis unsafe), exteriorization of both ends of the intestine should be accomplished. Various bypass procedures have also been described as part of a staged approach, in which exclusion of the segment containing the fistula is accomplished in the first reoperation, with another operation required later for resection of the involved segment and fistula tract — though this is certainly not the preferred surgical management except in extreme circumstances. Basic surgical considerations include attempting a one-stage procedure, careful adhesiolysis, addressing compromised tissues with wedge excision or intestinal resection, covering sutures with viable tissues, and avoiding friable areas not directly involved with the fistula.

Summary

Enterocutaneous fistulas occur most commonly as a result of a previous operative procedure. Once identified, a three-phase approach of stabilization, staging/supportive care, and (in some cases) definitive surgical intervention is necessary. Most fistulas heal spontaneously within 6 weeks; if closure is not achieved after 6 weeks, surgery is indicated.

Open items / gaps

• No primary literature yet in the wiki independently appraising the octreotide RCTs, fistuloclysis technique, or the large retrospective spontaneous-closure-rate study cited above — currently only the textbook’s narrative summary.