The Pivotal Role of Negative Pressure Wound Therapy in the Management of Enteroatmospheric Fistula: A Year-Long "Obstacle Marathon"

© 2024 HMP Global. All Rights Reserved.

Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of Wounds or HMP Global, their employees, and affiliates.

EAF is an abnormal communication between the gastrointestinal tract and the atmosphere. It is not an actual fistula, however, because it lacks a fistulous tract.¹ EAF can be caused by anastomotic leak, temporary abdominal closure, adhesions between edematous intestinal loops and the abdominal wall, surgical site infection, burst abdomen, and intestinal ischemia.¹ EAF remains among the most significant postoperative complications faced by the general surgeon after abdominal surgery.² When it affects the small intestine, EAF represents an enormous problem for the patient, not only physically, but also mentally and socially.³ The reported incidence of EAF is 7% to 12% in patients undergoing open abdomen management in damage control surgery,⁴ 2% to 25% in patients with abdominal trauma, 20% to 25% in patients with abdominal sepsis, and 50% in patients with pancreatic necrosis.⁵

The primary goals of managing EAF are to control and divert intestinal contents outside the abdominal cavity, protect surrounding tissues from retraction, and promote wound healing. Achieving these goals is not easy. Patients with EAF have high morbidity rates due to fluid and electrolyte loss, acid-base imbalance, increased catabolism, vitamin and trace element deficiencies, and wound complications.⁶ The mortality rate associated with EAF is 40%.⁵

Several techniques have been proposed for managing EAF. NPWT, which was initially used for the management of chronic leg ulcers, diabetic foot ulcers, and pressure injuries, is being used for more and more indications. The use of NPWT in the open abdomen has increased exponentially.³ The use of bladder catheters, nipples, endoscopic stents, vascular grafts, and fistula funnel, among other options, in the management of EAF has been described in the literature.

A male patient weighing 72 kg and with a BMI of 26.45 underwent Hartmann reversal surgery 7 months after undergoing Hartmann surgery for sigmoid diverticulitis complicated by perforation. On POD 5 after the Hartmann reversal surgery, the patient had an anastomotic leak with ischemia of the descending colon and the transverse colon. This was managed with resection of the ischemic colon and the creation of a terminal ileostomy on the last ileal loop on the right side. The patient was admitted to intensive care unit for 9 days after ischemic colon resection surgery.

Upon the patient's return to the surgery department, complete dehiscence of the laparotomy wound was managed with NPWT with instillation and dwell time. Polyvinyl alcohol dressing (V.A.C. Whitefoam Dressing; Solventum) was used in the deep portion of the wound (in contact with muscle fascia and loops), and the medium-sized V.A.C. Veraflo Dressing (Solventum; hereafter "NPWT dressing") applied on top was appropriately shaped to remain below the edges of the aspirated wound for 14 days.⁷ A sponge bridge was also placed from the wound to the site of the previous colostomy on the left side (Figure 1).

On POD 23 after the second operation, the appearance of the first small orifice (<5 mm) of EAF (output 100 mL/day) was noted, treated with stitches in PDS 3/0 in frozen abdomen characterized by loops exposed and attached to the diastasated muscle bands. Total parenteral nutrition was started.

The second orifice of the EAF appeared on POD 28. It was managed with placement of an intraluminal Foley catheter with subcutaneous tunneling in an attempt to convert the aforementioned EAF to an enterocutaneous fistula (Figure 2).

This attempt failed, so the skin margins were synthesized, and the nipple was positioned and isolated at the base using stoma paste (Stomahesive Paste; ConvaTec) (Figure 3). The nipple was sectioned at the top and passed through the Natura Two-Piece Stomahesive Skin Barrier (ConvaTec; hereafter "2-piece skin barrier") without the application of NPWT.

The seal was not optimal. The loss of enteric fluid around the nipple caused skin maceration. Thus, it was decided to apply a Fistula Funnel (Fistula Solution; hereafter "fistula funnel") in addition to the NPWT system and a stoma collection system in an attempt to block the EAF (Figure 4). Dressing change was done under deep sedation every 3 days. The output was 1200 mL/day. The authors of the current case report added somatostatin to therapy to reduce the amount of intestinal secretions.

The third orifice appeared on POD 45. Initially, an Isolator Strip -3M+KCI was applied (Figure 5), with poor results. Therefore, 2 baby bottle type nipples were applied⁸ (Figure 6). The nipples were held in place with polyvinyl alcohol dressings and medium-sized NPWT dressing (Figure 7) and were isolated with stoma paste, after which they were connected to the 100-mm 2-piece skin barrier for people with obesity (Figure 8). At the end of this procedure, the patient weighed 55 kg, with a BMI of 20.2. The EAF output was 1600 mL/day. Physical therapy was begun on POD 45.

On POD 70, using a radiopaque guide wire, it was discovered that the 3 fistulous orifices were arranged on the same section of the jejunal loop. Thus, a 20-mm × 100-mm enteral stent (Niti-S Enteral Colonic Covered Stent; Taewoong Medical Co Ltd) was positioned and connected to the fistula funnel for application of negative pressure therapy and the ileostomy collection system (Figure 9). The EAF output was 1200 mL per day. On POD 70 the treating physicians combined parenteral nutrition with oral nutrition after nutritional counseling in a short bowel diseases reference center.

On POD 134, the displaced enteral stent was removed and NPWT stopped. The most cranial fistulous orifice then was isolated in a subxiphoid position by positioning and placing an inverted nipple in the intestinal lumen, the nipple was subsequently connected to a convex ostomy system (Figure 10) (Natura Two-Piece Durahesive Skin Barrier with Convex-It Technology [ConvaTec]), treating it as if it were a stoma. The patient weighed 48 kg, with a BMI of 17.63.

On POD 171, bleeding from the stoma fistulous orifice was noted, and the patient underwent blood transfusion. He weighed 42 kg, with a BMI of 15.43. On POD 176, purulent material was noted near the stoma fistulous orifice (Figure 11).

On POD 184, to address persistent purulent secretion, the patient underwent surgery to mobilize the loop (with estimated fistulous orifice) by moving it approximately 8 cm caudally, almost into the umbilical region, while also synthesizing the muscle bands above and below the fistulous orifice (Figure 12). This adjustment guaranteed improved wound cleansing with complete blocking of the stoma orifice⁹ and favoring closure of the abdomen. The patient weighed 38 kg, with a BMI of 13.96. On POD 184 the nipple was applied and reversed into the intestinal lumen and advanced silver wound dressings were applied to stimulate closure by secondary intention.

On POD 201, parietal bleeding was managed conservatively with hemostatic gauze and blood transfusion. NPWT was suspended. On POD 208, the inverted nipple was repositioned in the intestinal lumen and NPWT was started.

On POD 253, the patient underwent resection of the fistulated loop, extensive vitreolysis of the entire small intestine, and mechanical jejunojejunal laterolateral anastomosis to reestablish the canalization towards the previous terminal ileostomy on the right side.

Management of the open abdomen began with positioning of an AbThera SensaT.R.A.C. device (Solventum; hereafter "temporary abdominal closure device") (Figure 13), which was replaced every 3 days; this device was used to monitor the anastomosis and step-by-step fascial closure in the following 12 days (Figure 14). Gradual fascial closure was associated with step-by-step synthesis of the skin and application of traction systems on pieces of drainage tube on the still dehiscent portion of the wound to reduce fascial retraction.

On POD 265, after complete closure of the fascial plane was verified (Figure 15), treatment began using NPWT with polyvinyl alcohol dressings applied to the synthesized muscle fascia (Figure 16).

On POD 279, the patient was discharged with NPWT to be replaced every 3 days on an outpatient basis. The length and width of the skin dehiscence were measured, the edges were protected with connection gauze, and the size of the polyvinyl alcohol dressing was gradually reduced (Figure 17). Terminal ileostomy performed on the right side was found to function well.

The patient received continued physiotherapy and walking support. He weighed 42.5 kg, with a BMI of 15.61.

On POD 295, NPWT was stopped. Outpatient care continued with silver gauze dressing until connective tissue formed and nearly complete skin closure was noted (Figure 18).

Complete closure of the skin was noted on POD 358. At that time, the patient weighed 51 kg, with a BMI of 18.73.

In the current case, it was difficult to achieve the objective of blocking enteric fluid to promote muscle fascia and periwound skin healing. At times, the perilesional skin was so burned by enteric secretions that application of the adhesive sheets used with negative pressure therapy was not possible (Figure 19). The initial difficulty was the number of orifices, which quickly increased to 3 and were difficult to isolate. It was impossible to confirm based on magnetic resonance enterography or transit time of diatrizoate whether the orifices affected a single loop or multiple jejunal loops. Negative pressure therapy in conjunction with use of a fistula funnel, nipple, or enteral stent played a decisive role in treatment because applying negative pressure on the appropriately shaped dressing helped keep the device in place. Use of NPWT avoided excessive retraction of the muscular fascial planes of the abdomen, stimulated granulation, protected the skin from chemical irritation due to secretion of enteric fluid, and allowed the application of a stoma collection system. Relatively prolonged use of NPWT to manage the fistulous orifice was aimed at resolving the frozen abdomen, which did not allow debridement of the loops and resection of the loop with EAF.

One hundred eighty-four days after the anastomotic leak due to ischemia, the appearance of pus allowed partial access to the abdominal cavity, at which time partial mobilization of the fistulated jejunal loop was done. However, the loop could not be resected because the proximal and distal heads did not come together; thus, the treating physicians decided to exploit this mobilization to move the stoma orifice toward the navel, because the application of stoma bags near the xiphoid was always challenging. Two hundred fifty-three days after the anastomotic leak, it was possible to access the abdominal cavity. At that time, with the apparent reduction in the edema of the loops of the small intestine, complete vitreolysis with resection of the fistulated loop was performed, which is the definitive treatment for this severe complication. The decision not to close the abdomen and to manage it as an open abdomen with a temporary abdominal closure device was made with the intent to monitor the tightness of the jejunojejunal anastomosis, knowing that this anastomosis was being placed in an abdomen from which purulent material occasionally emerged.

At each temporary abdominal closure device dressing change, a cranial and caudal section of the muscle fascia was synthesized. However, it was impossible to synthesize the skin, which was subjected to traction and closure by secondary intention on an outpatient basis.

The current case report has the typical limitations of such reports. There are so few cases that no surgeon can gain the level of experience desired to manage this condition. In the case of EAF, the clear goal is to block intestinal secretions and perform subsequent resection of the affected intestinal tract. However, the methods to achieve this objective have not been codified. Each clinical case is managed differently depending on the extent of the fistula, the location on the intestinal tract, the number of orifices, and the nutritional status of the patient. The ideal device does not always exist. In some moments it is necessary to modify existing ones or combine different devices (ie, "invent something").

EAF is a dramatic event for the patient and a daily challenge for the surgeon. There is no ideal treatment approach that is valid for all cases. Spontaneous closure of an EAF is unlikely but feasible in the case of a single, deep lesion with limited output and when intestinal continuity is preserved.¹⁰ In 1964, Chapman et al¹¹ described 4 fundamental principles of managing intestinal fistulas: intravenous correction of fluid loss, drainage of abscesses, control of fistula output, and skin protection; Chapman reported that malnutrition was the leading cause of death.

Enteral nutrition must always be attempted because it preserves the mucosal barrier and its immunological function and prevents malnutrition; absolute contraindications to enteral nutrition are intestinal discontinuity and reduced intestinal length (<75 cm).¹² In the current case, surgical resection of the fistulous tract was somewhat delayed (occurring on POD 253) compared with the median of 187 days postoperatively indicated by Wainstein et al.¹⁰ The delay in the current study was due to the failure of the first resection attempt in which pus appeared and the loop was mobilized (POD 184). The episodes of spontaneous bleeding in the current case were attributed to the phases of liver failure the patient experienced.

Authors: Raffaele Porfidia, MD¹; Simona Grimaldi, MD²; Maria Giovanna Ciolli, MD¹; Pietro Picarella, MD¹; and Sergio Grimaldi, MD¹

Affiliations: ¹Casa Di Cura Villa Dei Fiori, Naples, Italy; ²University of Milan–La Statale, Milan, Italy

Disclosure: The authors disclose no financial or other conflicts of interest.

Correspondence: Raffaele Porfidia, MD; Casa Di Cura Villa Dei Fiori, General and Oncological Surgery, Corso Italia 157, Acerra, Naples 80011 Italy; dottoreporfidia@gmail.com

Manuscript Accepted: July 1, 2024