ASAIO Journal 2000

Management of a High-Output Postoperative Enterocutaneous Fistula with a Vacuum Sealing Method and Continuous Enteral Nutrition
SUNG H. HYON,* JORGE A. MARTINEZ-GARBINO, MARIO L. BENATI,* MARCELO E. LOPEZ-AVELLANEDA,* NICOLAS A. BROZZI,* AND PABLO F. ARGIBAY*

A postoperative enterocutaneous stula is one of the most complex medical problems. Its treatment may become longlasting, wearisome, and its outcome often is disappointing. Here, we describe the use of a novel device to treat a 67year-old patient with a postoperative, high-output enterocutaneous stula. A semipermeable barrier was created over the stula by vacuum packing a synthetic, hydrophobic polymer covered with a self-adherent surgical sheet. To set up the system, we constructed a vacuum chamber equipped with precision instruments that supplied subatmospheric pressures between 350 and 450 mm Hg. The intestinal content was, thus, kept inside the lumen, restoring bowel transit and physiology. The stula output was immediately reduced from a median of 800 ml/day (range, 400 1,600 ml/day), to a median of 10 ml/day (range, 0 250 ml/day), which was readily collected by the apparatus. Oral feeding was reinitiated while both parenteral nutrition and octreotide were withdrawn. No septic complications occurred, and the peristular skin stayed protected from irritating intestinal efuents. Both the stula orice and the wound defect fully healed after 50 days of treatment. We believe this method may serve as a useful tool to treat selected cases of high-output enterocutaneous stulas without the need for octreotide or parenteral nutrition. ASAIO Journal 2000; 46:511514.

scopic sealing with brin, or coverage of the bowel defect by extra-peritoneal surgical maneuvering of local tissue, have been reported with promising results.5 8 We investigated the use of a nonsurgical approach, adapted from previously described methods,9 11 in which a semipermeable barrier is created over the stula orice by applying high levels of subatmospheric pressure in a vacuum tight system. The purpose of this communication is to describe the design and function of the system and to report on its performance in the treatment of a patient with a high output (500 ml/day), jejunoileal enterocutaneous stula. Materials and Methods Two distinct components make up the system: a vacuum chamber with its precision instruments attached and a vacuum-tight seal xed over the stula on the patient. Noncollapsible light-weight tubing connects both components.

The Vacuum Chamber

As depicted in Figure 1, vacuum is rst accumulated in a custom made, stainless steel rigid chamber with capacity for 40 L. The variation of subatmospheric pressure inside the chamber is continuously displayed by a precision vacuometer (Weksler Glass Thermometer Corp., Boca Raton, FL) attached to one of its sides. Upper and lower limits of vacuum may be set by the operator with a vacuostat (Danfoss, Nordborg, Denmark), which is connected to an electronic solenoid valve (Jefferson S.A., Buenos Aires, Argentina). The valve, interposed between the chamber and the vacuum source, opens when the lower limit is reached (increasing vacuity inside the chamber) and closes once the upper limit attained. Two disposable uid collection asks (Medi-Vac, Baxter HealthCare Corp., Valencia, CA.), serially connected at the patient outlet, are intended to prevent entrance of liquid into the chamber. Either a central vacuum line or an electric vacuum pump (Westinghouse/Lammert Pump, Sargent Welch Scientic Co., Addison, IL) may serve as the source of subatmospheric pressure. When a pump is used, its activation is controlled by the vacuostat, which simultaneously opens the solenoid valve and switches the engine on, or closes the valve and switches the engine off.

cation of surgical procedures on the gastrointestinal tract. Other known causes are radiation therapy, inammatory bowel disease, trauma, and cancer.1 Conventional, nonsurgical therapy is the rst choice and consists of total parenteral nutrition (TPN), administration of somatostatin analogues (octreotide), bowel rest, prevention of sepsis, and local wound care. The rate of spontaneous closure with the above-mentioned strategy ranges from 60 to 78% of affected patients.2 4 Those who do not heal may require one or more re-operations, but precise timing and patient selection seems to be of paramount importance. Other less conventional strategies, such as uoroscopically guided percutaneous catheterization of the stula tract, endo-

E nterocutaneous stulas most commonly arise as a compli-

From the *Department of Surgery and the Unit of Experimental Medicine, Hospital Italiano de Buenos Aires, Argentina. Submitted for consideration June 1999; accepted for publication in revised form December 1999. Reprint requests: Pablo F. Argibay, MD PhD, Department of Surgery, Hospital Italiano de Buenos Aires, Gascon 450, Buenos Aires 1181, Argentina.

Patient Components
To set up the vacuum-tight seal on the patient (Figure 2), the stula orice is covered by completely lling the depth of the 511

512

HYON ET AL.

Figure 1. Custom made stainless-steel vacuum chamber with precision instruments attached to it. The vacuostat controls the opening and closure of the solenoid valve. Vacuum source may be a high-wall suction or a vacuum pump.

abdominal wall defect with a synthetic, hydrophobic polymer. A karaya- or pectin-based skin barrier paste is used next to create a dam following the contour of the abdominal wall defect. An oversized piece of self-adhesive operative drape sheet (Steri-Drape, 3M Health Care, St. Paul, MN), is nally applied covering the entire surface of the polymer and the paste dam, extending over the surrounding abdominal skin. A rigid plastic cannula passing beneath the drape sheet is used to connect the hydrophobic polymer with the tubing from the vacuum chamber. Once the whole system is activated, high suction (350 to 450 mm Hg) exerted under the drape sheet causes the loosely arranged polymer to tightly compress against the wound bed, creating a mesh-like structure that functions as a semipermeable barrier over the stula. The hydrophobic nature of the polymer prevents this barrier from becoming oversaturated

with liquid, thus prolonging its capacity to maintain vacuity in the system. The overall effect is an almost total occlusion of the stula. Any efuent, including neighboring abscess formations, is readily sucked and directed through the mesh to the collection asks. In this setting, bowel content is kept inside the lumen, restoring normal enteral physiology and transit, provided no distal obstruction is present.

Case study
MB, a 67-year-old overweight man was referred to us for treatment of an enterocutaneous stula. Three weeks earlier the patient had undergone an emergency laparoscopic cholecystectomy with conversion to open abdominal surgery because of technical difculty. Eight hours later, the patient was re-explored for intestinal uid coming through the surgical

Figure 2. High-output enterocutaneous stula (a), over which a synthetic polymer has been loosely arranged and air-sealed with a self-adhesive operative drape (b). As soon as suction is applied into the system, the polymer compresses against the wound bed, creating an occlusive barrier over the stula orice (c).

VACUUM SEALING FOR HIGH-OUTPUT ENTERAL FISTULA

513

wound. Upon identication of a perforated small bowel, segmental resection and primary end-to-end anastomosis was practiced. One week postoperatively, the patient developed wound dehiscence and two jejunoileal stulas, one located at the upper and another one at the lower aspects of the wound. At our facility, the patient presented with two enterocutaneous stulas draining an average of 900 ml/day (range, 400 1,600 ml/day). His body weight had reduced by approximately 17%, and his renal function was deteriorating (plasma creatinin, 2.74 mg/dl). Initially, the patient was put on TPN, bowel rest, octreotide (300 g/day subcutaneously), and a gauze dressing that was changed several times a day. Because no favorable response was observed, treatment with the vacuum packing method was initiated on day 15 after stula onset. Both the upper and lower stulas were occluded with the polymer and covered with the drape sheet for vacuum packing. The vacuostat was adjusted to uctuate between 350 and 450 mm Hg. Results Fistula output, which had reached its highest point (1,600 ml/day), fell to 250 ml at 24 h, 0 ml at 48 h, and 50 ml at 72 h. Median output for the 50-day period that composed the treatment was 10 ml/day (range, 0 250 ml/day). Oral feeding was initiated, completing calorie requirement with a nasojejunal catheter, whereas TPN and octreotide administrations were both stopped on day 10 of treatment. On two occasions during the course of therapy, the rst time on day 13 and the second time on day 28, vacuum packing was interrupted to observe spontaneous stula response. Efuent volume immediately increased to 950 ml and 750 ml, respectively, leading us to reinstall the system. Fistula output progressively decreased thereafter, coming down to 0 ml on day 38. Vacuum packing was continued for 12 more days with equal results, suggesting complete stula closure. Treatment was stopped, and the patient was discharged home 5 days later with the stula healed. On follow-up, he continues to be well 1 year later. Discussion Although most enterocutaneous stulas heal with bowel rest, TPN, octreotide, and local wound care in a relatively short period of time, unresponsive cases may become a complex medical problem. A mortality rate of 34% has been reported in these patients,12 sepsis being responsible for as many as 80% of all deaths.13 Factors associated with failure to achieve spontaneous closure include output 500 ml/day, eversion of the mucosa, a short stula tract, an abdominal wall defect, jejunoileal location, residual cancer, distal obstruction, severe sepsis, Crohns disease at the stula site, and patient malnutrition.14 TPN is widely used, but it requires a central venous line and patient hospitalization with increased costs and morbidity. Octreotide administration seems to reduce stula output, but its real inuence on clinical outcome is still controversial.15,16 Fistulas not responding to conservative therapy are usually operated on, but laparotomy with resection of the bowel segment containing the stula followed by end-to-end anastomosis is not always possible due to general, as well as local, patient condition. Less invasive surgical techniques designed

to cover the bowel defect with abdominal muscles or skin grafts have been used with promising results.7,8 Recently, radioscopically guided percutaneous catheterization of the bowel lumen through the stula tract showed success rates of 90% and 65% for high-output and low-output stulas, after 39 days and 45 days of treatment, respectively.5 Other investigators have used a commercially available brin tissue adhesive to endoscopically seal gastroduodenal stulas with excellent results.6 Local wound care using closed systems with suction or drain sumps have been described previously.9 11,1719 We investigated the utility of a method based on the creation of a semipermeable, still occlusive barrier over the stula orice by means of applying high levels of subatmospheric pressure into a vacuum-tight system. Our patient, who had been on conventional therapy for 2 weeks without responding, decreased his stula output from 1,600 ml/day to 250 ml at 24 h, and to 0 ml at 48 h. Efuent volume was maintained very low thereafter (median, 10 ml/day; range, 0 250 ml/day), until day 38 of treatment, when it fell to 0 ml. Vacuum packing was prolonged for 12 more days to secure denitive stula closure. From the observation that the system kept the intestinal content inside its lumen, enteral feeding was initiated according to patient tolerance. Both TPN and octreotide were withdrawn on day 10 of therapy without any negative impact on patient well being. After a short term of calorie supplementation through a nasojejunal catheter, this was also withdrawn and the patient continued with oral intake only. Each time the stula was vacuum sealed, tightness was maintained for varying periods of time, the longest for 4 days. This method eliminated the need for complex and repeated gauze dressing changes while the peristular skin stayed dry and protected. After 50 days of treatment and 5 more days of in-hospital observation, the patient was discharged home with the stula completely healed, having recovered 8 of the 18 kg of lost body weight. Laboratory values at discharge included creatinin, 1.3 mg/dl; cholesterol, 145 mg/dl; total protein, 7.28 g/dl; and albumin, 3.95 g/dl. We here describe the successful closure of two lateral stulas that presented with eversion of the mucosa but without a complete diversion of the bowel ends. One other patient, who is currently under treatment for a jejunoileal stula with complete separation of the bowel ends, has shown a progressive reduction in the exposed mucosa over a week of treatment with the vacuum packing system. However, it remains to be elucidated at this point in time whether our method alone will sufce to completely close the stula. We conclude that selected cases of high-output enterocutaneous stulas may be treated by the vacuum sealing method, which allows for full enteral feeding and suspension of TPN and octreotide administration. References
1. Berry SM, Fischer JE: Classication and pathophysiology of enterocutaneous stulas. Surg Clin North Am 76: 1009 1018, 1996. 2. Sitges-Serra A, Jaurrieta E, Sitges-Creus A: Management of postoperative enterocutaneous stulas: The roles of parenteral nutrition and surgery. Br J Surg 69: 147150, 1982. 3. Nubiola P, Badia JM, Martinez-Rodenas F, Gil MJ, Segura M,

514

HYON ET AL. 12. Levy E, Frileux P, Cugnenc PH, Honiger J, Ollivier JM, Parc R: High-output external stulae of the small bowel: Management with continuous enteral nutrition. Br J Surg 76: 676 679, 1989. 13. Reber HA, Roberts C, Way LW, Dunphy JE: Management of external gastrointestinal stulas. Ann Surg 188: 460 467, 1978. 14. Campos ACL, Meguid MM, Coelho JCU: Factors inuencing outcome in patients with gastrointestinal stula. Surg Clin North Am 76: 11911198, 1996. 15. Sitges-Serra A, Guirao X, Pereira JA, Nubiola P: Treatment of gastrointestinal stulas with Sandostatin. Digestion 54(Suppl 1): 38 40, 1993. 16. Sancho JJ, Di Costanzo J, Nubiola P, Larrad A, Beguiristain A, Roqueta F, Franch G, Oliva A, Gubern JM, Sitges-Serra A: Randomized double-blind placebo-controlled trial of early octreotide in patients with postoperative enterocutaneous stula. Br J Surg 82: 638 641, 1995. 17. Lange MP, Thebo LM, Tiede SM, McCarthy B, Dahn MS, Jacobs LA: Management of multiple enterocutaneous stulas. Heart Lung 18: 386 390, 1989. 18. Meehan PA, Mayz EJ: Nursing management of an open abdominal wound. Crit Care Nurse 8: 29 34, 1988. 19. Wooding Scott M, Montgomery B, Coleman D: No wound is too big for resourceful nurses. RN 51: 2225, 1988.

4. 5. 6. 7. 8.

9. 10. 11.

Sancho J, Sitges-Serra A: Treatment of 27 postoperative enterocutaneous stulas with the long half-life somatostatin analogue SMS 201995. Ann Surg 210: 56 58, 1989. Paran H, Neufeld D, Kaplan O, Klausner J, Freund U: Octreotide for treatment of postoperative alimentary tract stulas. World J Surg 19: 430 434, 1995. DHarcour JB, Boverie JH, Dondelinger RF: Percutaneous management of enterocutaneous stulas. AJR Am J Roentgenol 167: 3338, 1996. Eleftheriadis E, Tzartinoglou E, Kotzampassi K, Aletras H: Early endoscopic brin sealing of high-output postoperative enterocutaneous stulas. Acta Chir Scand 156: 625 628, 1990. Kearney R, Payne W, Rosemurgy A: Extra-abdominal closure of enterocutaneous stula. Am Surg 63: 406 409, 1997. Sarfeh IJ, Jakowatz JG: Surgical treatment of enteric bud stulas in contaminated wounds: A riskless extraperitoneal method using split-thickness skin grafts. Arch Surg 127: 10271037, 1992. Cunha-Madeiros A, Rosado-Soares CE: Treatment of enterocutaneous stulas by high-pressure suction with a normal diet. Am J Surg 159: 411 413, 1990. Fernandez ER, Cornalo AO, Gonzalez D, Viella V: Nuevo enfoque en el tratamiento de las stulas enterocutaneas postquirurgicas. Rev Argent Chir 62: 117127, 1992. Dearlove JL: Skin care management of gastrointestinal stulas. Surg Clin North Am 76: 10951109, 1996.