Enterocutaneous Fistula in a Dog Secondary to an Intraperitoneal Gauze Foreign Body

Case Report

A 6-year-old, spayed female German shepherd crossbreed had a chronic draining tract in the left flank. The dog had no medical or surgical conditions, except that ovariohysterectomy had been performed 4 years earlier. The veterinarian that performed the ovariohysterectomy was consulted and indicated that the dog had significant postoperative hemorrhage that necessitated another surgery.

Nine months before the dog was presented, a subcutaneous, firm mass measuring 4 × 5 cm developed in the left flank area. Four months before the dog was presented, the referring veterinarian examined the mass, but no treatment was pursued. Within 2 weeks of presentation, the mass had opened and was draining. The area was biopsied and surgically explored by the referring veterinarian, and numerous fistulous tracts were noted. Biopsy results indicated marked, chronic, regional, neutrophilic, and pleocellular nodular dermatitis and panniculitis. The suture line subsequently dehisced, and a second similar surgery (including placement of a Penrose drain) was performed. Despite the drain and antibiotic therapy (orbifloxacin 68 mg q 24 hours; amoxicillin trihydrate/clavulanate potassium 375 mg q 12 hours, intermittently for the past 6 months), the area did not heal over, and a continued serosanguineous to purulent discharge was evident.^a,b

On presentation, the dog was bright, alert, and responsive. Physical examination revealed a draining tract (10 to 15 mm) of the left flank that was not painful when palpated. The tract was surrounded by a circular area (4 to 5 cm in diameter) of fibrotic, alopecic skin that was thought to be scar tissue. The abdomen was tense when palpated.

A complete blood count (CBC) was within normal limits. Results of serum biochemical analysis showed mild increases in amylase (1072 U/L, reference range 248 to 1031 U/L), aspartate transaminase (35 U/L, reference range 17 to 32 U/L), and creatine kinase (167 U/L, reference range 34 to 149 U/L). The results of a urinalysis were within normal limits.

On abdominal ultrasonography, an ill-defined mass in the left retroperitoneum cast an acoustic shadow that made visualization of the left kidney difficult. The deep tissue surrounding the mass appeared edematous, with tracts deep into the left retroperitoneal space. The right kidney appeared normal.

The dog was placed under general anesthesia, and survey abdominal radiography and fistulography were performed. Survey films were unremarkable. For the fistulography, a Foley catheter was inserted into the fistula, and the bulb was inflated with sterile saline. Water-based, iodinated contrast was injected slowly to effect during cinefluoroscopy. The fluoroscopic imaging showed contrast being delivered through the catheter into an irregular retroperitoneal pocket; a small channel appeared that led into the jejunal lumen. A diagnosis of enterocutaneous fistula was made [Figures 1 through 4].

Exploratory surgery via a midline abdominal approach revealed a large cluster of adhesions near the left retroperitoneal space. The jejunum was involved in two places: one proximal jejunum and one midjejunum. The descending colon was also adhered to the large mass surrounding the caudal mesenteric artery [Figure 5]. The caudal mesenteric artery necessitated ligation to fully resect the mass. The right kidney appeared to be enlarged, and the left kidney was small and firm. On dissection, the left ureter was found to be incorporated into the mass, and it appeared fibrotic [Figure 6]. The midjejunum was adhered to both the mass and the body wall, and it was perforated into the retroperitoneum, thus providing a source of drainage to the left flank. The adherent proximal jejunum was dissected free, and the serosal defect was sutured with 4-0 polydioxanone in a simple continuous pattern. The perforated midjejunal segment was dissected from the retroperitoneum and resected. A functional end-to-end anastomosis of the two ends of the jejunum was performed using a gastrointestinal anastomosis linear cutter stapling device and a thoracoabdominal-55 stapler. ^c The left kidney and ureter were removed because of extensive involvement in the mass, their fibrotic appearance, and suspected lack of function. The mid-descending colon was also resected because of concern of blood supply due to the necessary ligation of the caudal mesenteric artery. The colon was anastomosed with 4-0 polydioxanone in a simple continuous pattern.

The abdomen was closed routinely, and anesthetic recovery was uneventful. Postoperative management included antibiotics (cefoxitin 30 mg/kg q 8 hours for 5 days), pain medications (morphine 2 μg/kg per minute and ketamine 10 μg/kg per minute, both given via constant-rate infusion), and central venous pressure monitoring. Five days after surgery, the antibiotic was switched to amoxicillin trihydrate/ clavulanate potassium (375 mg q 12 hours for 10 days).^b Cultures were not performed.

Histopathological examination of the mass confirmed a granulating fibrosis with a large, necrocellular focus surrounding foreign material. The foreign material was found to be a gauze sponge [Figure 7].

One week later, surgery was performed to débride and close the fistula from the flank side. A 6-cm incision was made over the left paralumbar fossa and fistula site. The fistulous tracts were removed using a combination of blunt and sharp dissection, and the area was thoroughly lavaged with copious amounts of saline. After a 0.5-inch Penrose drain was placed, the incision was closed with 3-0 polydioxanone in a simple continuous pattern in the subcutaneous layer and then with 3-0 nylon in a simple interrupted pattern in the skin. Recovery from anesthesia was uneventful. The amount of drainage from the site was minimal, and the drain was removed 2 days after surgery. The dog has since recovered completely, with no evidence of ongoing disease at the 3-year follow-up.

Discussion

Gauze foreign bodies are an uncommon and unnecessary complication of surgery. To the authors’ knowledge, no published data exist on the incidence of retained surgical foreign bodies in animals. The occurrence is likely higher than believed because of the reluctance of veterinarians to report this complication.¹ Also, the incidence rate is likely to be at least similar to or higher than that reported in human medicine. The incidence of retained surgical foreign bodies in humans is estimated to be one in every 1000 to 1500 intraabdominal procedures.² The risk of retained sponges in veterinary medicine is difficult to assess. Merlo and Lamb have argued that the less rigorous operative protocols and fewer trained scrub nurses (to perform sponge counts) in veterinary medicine places animals at higher risk than humans.¹ However, they also recognize that because surgical procedures are less complex and abdominal cavities are smaller, animals are likely to be at lower risk than humans.¹ Identified risk factors for retained foreign bodies in humans include emergency surgery, unexpected change in surgical procedure, and obesity.²

When the dog in this report previously underwent ovariohysterectomy, a second emergency surgery was necessary because of significant postoperative hemorrhage. Four years later, the dog developed a chronic draining tract that was secondary to a foreign body. This tract was appropriately termed a fistula, because it was a connection between two epithelial-lined structures; but most draining tracts in small animals are actually sinus tracts, because they are a connection between mesothelial- and epithelial-lined structures.³

Draining tracts are fairly common in small animals and frequently occur in the lumbar, paralumbar, and flank areas; but they also have been reported in many other areas of the body, including the head, neck, paws, and inguinal and gluteal regions.^4–6 Most are associated with foreign bodies, but they can be a diagnostic challenge. The list of differential diagnoses is long and includes neoplasia, infection (bacterial, fungal, parasitic), osteomyelitis, bony sequestra, and many others.^4,7 In addition to a thorough history and physical examination, other diagnostic tests are often needed, including a CBC and serum biochemical analysis, urinalysis, cytology, biopsy with histopathology, macerated tissue cultures (aerobic, anaerobic, fungal, mycobacterial), radiography, sinography, ultrasonography, magnetic resonance imaging, and surgical exploration.^7,8

In humans, a surgical sponge left in the abdomen is often referred to as a “gossypiboma” (gossypium: Latin for cotton; boma: Swahili for a place of concealment).⁹ A retained surgical sponge most commonly forms an aseptic inflammatory reaction with encapsulation; it may be detected incidentally or as a space-occupying mass. This mass can be misdiagnosed on gross inspection as a neoplasm. Another possible outcome after a surgical sponge has been retained is the formation of an abscess with chronic internal or external fistula formation. Some retained surgical sponges have remained undetected for several years.^9,10

Other than the enterocutaneous fistula in the dog of this case report, only one other instance of an enterocutaneous fistula (which was secondary to pelvic trauma) has been reported in a dog.¹¹ In stark contrast, enterocutaneous fistulae in humans appear to be very common,^12,13 with most cases related to recent abdominal surgery. Other reported causes include inflammatory bowel disease, trauma, radiation therapy, neoplasia, diverticulitis, and ischemic bowel.^14,15 The prognosis in humans is guarded, with mortality rates of 5% to 20% due to sepsis, fluid and electrolyte disturbances, and malnutrition.¹²

In animals, gauze sponges left in the abdomen tend to incite an inflammatory reaction and the formation of a sterile abscess or granuloma, adhesions, or draining tracts.¹⁶ In one case series, five of eight animals that had gauze foreign bodies were presented with a draining tract.¹ Encapsulated foreign bodies can be a diagnostic challenge, because clinical signs may be nonspecific.¹⁶ Aids to diagnosis of retained surgical foreign bodies include radiography (especially when radiopaque sponges are used), sinography or fistulography with cinefluoroscopy, ultrasonography, and computed tomography. In one case series, a sensitivity of 87% and a specificity of 100% were reported for sinography in the diagnosis of foreign bodies.⁴ Numerous reports also document the use of ultrasonography for the diagnosis of retained surgical sponges. The most commonly reported sonographic findings are a hypoechoic mass with a hyperechoic center.^1,8,16 In this particular case report, fistulography with cinefluoroscopy identified the need for exploratory surgery. The authors believe this is the first report documenting an enterocutaneous fistula as a sequela of a retained surgical sponge in a dog.

Conclusion

Retained surgical foreign bodies should be considered in animals that are presented with a draining tract or mass. A thorough history, including details of all prior surgical procedures, should always be taken. Additionally, a retained foreign body should be considered for any chronic, nonspecific condition in an animal with a history of surgery. Surgeons operating under emergency conditions or on animals that are overweight should consider using radiopaque sponges, performing sponge counts, or following a policy of using only large laparotomy sponges in abdominal procedures. Although helpful, these practices should only be aids to surgical technique and should not be substitutes for diligence and attention to detail to minimize the incidence of retained foreign bodies.