A Novel Technique To Skin Graft Abdominal Wall Wounds Surrounding Enterocutaneous Fistulas

Abstract

Background:

Enterocutaneous fistulas often are associated with large abdominal wall wounds. Successful skin grafting of these sites is difficult, as the bed is constantly bathed by enteric contents. A method to graft these sites successfully would provide an important advance in their treatment.

Methods:

The medical records of patients undergoing skin grafting of a site around an enterocutaneous fistula were reviewed. The amount of fistula output at the time of grafting was recorded. The method of grafting, as well as the means of protecting the graft from enteric exposure, were noted. Skin grafts were evaluated for the extent of “take.”

Results:

Seven patients met the inclusion criteria. After 1–2 weeks, the graft take was 90% in three patients, 80% in two patients, and 50% in two patients. After 1 month, there was complete epithelialization in 85% of patients, and the remaining patient had most of the site epithelialized. This healing allowed placement of an ostomy appliance in all patients. The fistula output was >400 mL per day in 70% of the patients. Multiple techniques were used to divert enteric flow away from the graft, but the most common was placement of a negative pressure dressing that concomitantly secured the graft and allowed enteric diversion.

Conclusion:

The presence of a high-output enterocutaneous fistula does not preclude successful skin grafting. Such grafting can accelerate wound healing as well as improve skin and site hygiene by allowing the placement of an ostomy device.

The formation of an enterocutaneous fistula (ECF) can be one of the more difficult complications seen by surgeons. The management of these patients is often a long and arduous process. There are several causes of fistulas, including prior abdominal surgery, abdominal trauma, and inflammatory bowel disease [1]. Many of these patients have risk factors such as multiple operations, malnutrition, or sepsis, which contribute to the difficulty in treatment and increase the risk of morbidity and death [2]. In patients with a wound surrounding the fistula, complications can be even more prevalent. In these situations, enteric contents can spill onto tissue, causing disruption of normal tissue planes and inflammation that can prolong and complicate healing [3].

In the absence of a fistula, large soft-tissue wounds can be grafted with skin. The standard technique for grafting cannot be used in patients with an enterocutaneous fistula, however, as the spillage of enteric contents prevents success. As such, these wounds generally must heal by secondary intention. This process can take several months and is a large contributor to morbidity in this patient population. These wounds may require regular debridement in order to maintain healthy tissue, and this process can contribute to larger wounds and long healing times [4]. Additionally, there is a substantial physiologic burden that can lead to impaired immune function and malnutrition.

Despite the advances modern surgery has made, the overall morbidity and mortality rate of patients with fistulas remains high. Most of the current studies have shown a slight decrease in the mortality rate for patients with ECFs to around 10–20% with little or no change in morbidity [5,6]. The fact that patient outcomes have not improved substantially reflects the difficulty in treatment posed by an ECF. The general treatment goals consist of early diagnosis and drainage of the fistula, control of infection and sepsis, and nutritional support [7]. Current treatment is aimed at promoting spontaneous closure; if this is not successful, operative intervention with fistula resection and anastomosis is completed [8]. Delayed surgery is associated with fewer deaths [9]. The most common treatment of wounds surrounding an ECF is to allow the wound to heal by secondary intention, using local care to protect the site [10]. More recently, vacuum-assisted closure (VAC) has been used to decrease healing time and promote granulation [11]. Attempts at diverting the enteric contents from the rest of the wound bed have been described, with various degrees of success.

Ideally, these sites would be grafted with simultaneous diversion of the enteric contents to allow graft “take” and to minimize the adverse affects of an open site. Herein, we describe our successful technique for grafting abdominal wounds surrounding an ECF.

Patients and Methods

The medical records of patients admitted to a university hospital from 2002–2007 were reviewed. Patients were identified by querying operating room records for patients with both an enterocutaneous fistula and a skin graft procedure. We then searched for additional cases by evaluating all skin grafts completed by surgeons in the Division of Acute Care Surgery. These patients were included in the study if they also had an ECF. Each chart was reviewed for patient demographics, co-morbidities, hospital length of stay, treatment method, skin graft take at 1–2 weeks, and postoperative complications. Outpatient followup visits were reviewed to determine the final outcome. Additionally, specific attention was placed on the operative techniques used for the placement of the skin graft, along with the technique for isolation of the fistula output. Permission to complete this study was obtained from the Institutional Review Board.

Results

Seven patients underwent skin grafting of abdominal wounds that abutted an ECF during the study period (Table 1). They consisted of four women and three men with an average age of 60.8 years (range 34–79 years). The hospital stay ranged from 3–152 days, with an average of 52 days. None of the patients died. All of the fistulas formed postoperatively and were associated with multiple prior abdominal operations. Fistula origin was small bowel in three of the patients, colon in two patients, and unknown in two patients. Output was high (>500 mL/day) in five patients, low in one patient, and unrecorded in the seventh. Skin grafting was attempted an average of 172 days (range 15–307 days) after the appearance of the ECF. The average size of the skin graft was 310 cm².

Table 1.

Characteristics of Study Patients

	Age (years)	Co-morbidities	Fistula output (mL/day)	Fistula origin	Graft size (cm²)	Time frame ^a	1–2 week take percentage
1	77	Hypertension, chronic obstructive pulmonary disease, coronary artery disease	Unknown	Small bowel	312	87	70–80
2	66	Obesity	600–1,400	Colon	500	30	50–60
3	57	Obesity, hypertension, hypothyroidism, schizophrenia	500–1,500	Unknown	150	55	50
4	34	None	500–1,000	Unknown	280	16	90
5	62	Hypertension, gastroesophageal reflux, type 2 diabetes mellitus, atrial fibrillation	600–900	Colon	132	285	95
6	79	Chronic obstructive pulmonary disease, diabetes mellitus, anemia, obesity	“High output”(actual amount not specified)	Small bowel	480	282	99
7	51	Ovarian cancer, hypothyroidism	“Low output” (actual amount not specified)	Small bowel	50	152	80

Time in days from appearance of fistula to skin grafting.

Prior to skin grafting, all patients were on complete bowel rest and receiving total parenteral nutrition (TPN). The caloric needs were determined by a clinical dietician, and the TPN was formulated by clinical pharmacists. In six patients, prealbumin concentrations were checked three days–three weeks preoperatively and were 7–27 mg/dL. There was no correlation between the prealbumin concentration and graft take. We generally do not rely on laboratory measures of nutritional status. Rather, we use the appearance of healthy granulation tissue as the marker of readiness to accept skin grafting. No patient received octreotide as an adjunct to fistula management.

Sites were prepared for grafting using a VAC device (KCI, San Antonio, TX). Grafting took place when healthy granulation tissue was seen. We did not culture the wounds prior to grafting. In all patients, the skin was harvested in the standard fashion with a dermatome. The donor site was either the left or the right thigh for every patient. All the grafts were meshed in the standard fashion to either 1:1.5 or 1:2.

The most common technique—a modified wound VAC—was used on five patients (Fig. 1). The graft was either stapled or sutured at the periphery. A hole was then cut in the graft over the area of the fistula, and a Vaseline^®-impregnated dressing (Adaptic^TM, Johnson & Johnson, New Brunswick, NJ) was placed over the graft. In two patients, a silver sponge was used, and the standard black sponge was used in the other three patients. A defect was cut in the portion of the sponge directly overlying the fistula. Ostomy paste was then distributed along the interior of the sponge defect, with care to create a thin coat, because if too thick a coat was placed, the paste tended to spread out and occlude the fistula output, thus contaminating the graft. Transparent adhesive was placed over the sponge, and suction was attached to the VAC. The suction was set to a minimum of 75 mm Hg and a maximum of 125 mm Hg. This value was not standardized, and the reasons for the different pressure settings could not be determined from the medical records. The overall effect was a sponge that held pressure against the graft and diverted the fistula output. The VAC output was not recorded reliably in the charts. The fistula effluent was diverted into a separate container that was not attached to the VAC. This output did not change immediately after the procedure. An ostomy appliance could then be placed on top of the VAC for easy collection of the enteric contents.

FIG. 1.

Surgical technique. (A) Wound with healthy bed of granulation tissue surrounding enterocutaneous fistula. (B) Placement of skin graft with hole to allow drainage of fistula contents. (C) Placement of vacuum-assisted sponge, the key step in the technique. A Vaseline^®-impregnated dressing (Adaptic, Johnson & Johnson, New Brunswick, NJ) is placed over the skin graft with a defect cut out directly over the fistula (not pictured). A defect is cut in the sponge overlying the fistula. A thin layer of ostomy paste is spread along interior of sponge along surface of defect. Paste is added to occlude holes in sponge to prevent negative pressure from the VAC from suctioning enteric contents into the sponge, which would contaminate graft. If paste is applied in too thick a layer, it will spread and occlude fistula opening when negative pressure is applied. This will lead to graft contamination. (D) Application of adhesive dressing. Transparent dressing should overlap ostomy paste completely down to base of wound. If there is any ostomy paste not covered by the dressing, additional pieces of transparent adhesive should be added until full paste coverage is achieved. Successful placement of VAC device is indicated by sponge that holds continuous suction without diversion of any enteric contents into sponge.

One patient failed the above technique because of graft contamination discovered on postoperative day one. The ostomy paste had not been applied properly, which allowed leakage of contents onto the graft. The VAC was removed, and a modified system was used. Kerlix Kendall™ (Covidien, Norwalk, CT) gauze was placed over a new Vaseline^® dressing, and an adhesive dressing was placed over the Kerlix. Two small cannulas were inserted into the Kerlix, and physiological saline was infused into the Kerlix at a low rate through one of the cannulas and removed by the other, which was attached to suction. This continuous irrigation system was changed daily.

Two other methods were used, one on each of the remaining patients. One patient underwent grafting with diversion of fecal flow using a catheter. A Foley catheter was inserted into the fistula, and the balloon was inflated to prevent leakage onto the graft. A collection bag was attached to the other end of the catheter to collect the output. The technique described above was then used to place a VAC on top of the graft. The final patient had the fistula closed superficially with sutures, and the graft was laid down on top of the entire wound with no defect created. A VAC was then placed on top of the graft, and suction was applied. Two days later, the fistula reopened spontaneously, and wet-to-dry dressings were applied to prevent graft contamination.

Graft take measured between one and two weeks postoperatively ranged from 50–99%. By one month, all patients showed complete epithelialization, and all of the patients were able to be outfitted with an ostomy device directly on the graft. The two patients with outputs exceeding 1,000 mL per day had graft takes of 50%–60% at one week. The remaining patients had graft takes between 70% and 99% at one week.

Discussion

Enterocutaneous fistulas still represent a major cause of morbidity and death in patients undergoing high-risk abdominal operations. Once these ECFs occur, management becomes complex and often is associated with a prolonged hospital course [12]. Enteric contents spilled onto the skin are toxic to tissues and can result in the creation of an abdominal wound or increase the size of an existing wound secondary to necrosis and debridement. Currently, the most common method of treatment is labor intensive, with regular dressing changes.

In theory, the most effective technique for treating these wounds would be to use skin grafting to allow rapid healing. When an open wound is present, an ostomy appliance cannot be fitted around the fistula, and the constant bathing of the wound with intestinal effluent inhibits healing.

Our technique of simultaneous skin grafting and placement of an ostomy device was effective in the treatment of abdominal wall defects with an associated ECF. Complete wound coverage was seen in most patients by two weeks and in all patients by one month. The use of a VAC in our technique has a two-fold purpose. The first is to increase the likelihood of skin graft survival. Seromas, infection, and shearing forces are the main causes of graft failure [13]. By using a VAC, the graft is secured to the underlying granulation tissue by negative pressure, preventing both the accumulation of fluid beneath the graft and any shearing caused by graft movement. Also, studies have shown consistently that the use of negative pressure improves graft survival [14]. The other effect of our technique is to divert the enteric contents away from the graft site. By using ostomy paste to occlude the sponge around the defect, complete diversion of enteric contents is accomplished, and graft survival is maximized. Our patients with early contamination had reduced early graft take. This result underscores the importance of complete enteric diversion away from the graft during healing. Even with initial contamination, the early recognition and changing of the VAC resulted in half of the graft being salvaged and no need for repeat grafting.

Several other methods for skin grafting around fistulas have been reported in the literature. One technique is placement of dressings and flavin wool to absorb discharge and prevent contamination [15]. Those authors reported 100% graft take, and once the graft was stable, an ostomy device was applied to divert the fistula contents. This technique requires frequent dressing changes and may not contain high-output fistulas completely, allowing graft contamination. The need to place dressings between the graft and the fistula also decreases the amount of open wound that can be grafted. By using the VAC, we are able to graft closer to the fistula. Also, by having the ostomy device applied right away, we are able to divert enteric flow regardless of output volume.

Another technique involves either closing the fistula locally with sutures or placing a catheter to suction the fistula [16]. With this technique, graft take was excellent in the initial series. However, other studies have shown that local closure of fistulas is ineffective. Additionally, this method requires substantially more wound care than our technique.

Goverman et al. used a method referred to as the “fistula-VAC” to divert the enteric contents from abdominal wall defects [17]. Our technique adds to this method by placing ostomy paste on the inside of the sponge adjacent to the fistula. We believe this step adds protection from enteric contamination. The earlier investigators reported graft take in all three patients and eventual ostomy placement directly onto the graft. They noted the importance of complete diversion of enteric contents for graft survival, which is similar to our experience.

There are limitations to our study worth noting. The series is small, and a larger number of patients would have been desirable to demonstrate the consistency and durability of the technique. Also, because we had multiple surgeons performing the procedure, the technique may not have been exactly the same for every patient. At the same time, we have demonstrated that by following the basic principles of the procedure, similar results can be obtained in different individuals. This may show that an individualized approach is most appropriate. Moreover, the study is retrospective, so we are dependent on reports for information on graft take. However, the endpoint of complete healing that was seen in every patient is an objective measure not subject to estimation.

In conclusion, we reviewed data from patients with ECFs and open abdominal wounds treated with skin grafting and a modified VAC. All grafts showed complete healing by one month, and every patient was able to be fitted with an ostomy appliance. Our data support our belief that skin grafting of abdominal wall defects can be done in the presence of a high-output fistula using a novel method of diverting the enteric contents into an ostomy appliance attached to a wound VAC.

Footnotes

Author Disclosure Statement

No conflicting financial interests exist.

Presented at the Thirtieth Annual Meeting of the Surgical Infection Society, Las Vegas, Nevada, April 18–20, 2010.

References

Berry

, Fischer

. Classification and pathophysiology of enterocutaneous fistulas. Surg Clin North Am, 1996; 76:1009–1118.

Foster

3rd , Lefor

. General management of gastrointestinal fistulas: Recognition, stabilization, and correction of fluid and electrolyte imbalances. Surg Clin North Am, 1996; 76:1019–1033.

Schecter

, Ivatury

, Rotondo

et al. Open abdomen after trauma and abdominal sepsis: A strategy for management. J Am Coll Surg, 2006; 203:390–396.

Attinger

, Janis

, Steinberg

et al. Clinical approach to wounds: Debridement and wound bed preparation including the use of dressings and wound-healing adjuvants. Plast Reconstr Surg, 2006; 117:72S–109S.

Hollington

, Mawdsley

, Lim

et al. An 11-year experience of enterocutaneous fistula. Br J Surg, 2004; 91:1646–1651.

Evenson

, Fischer

. Current management of enterocutaneous fistula. J Gastrointest Surg, 2006; 10:455–464.

Schecter

, Hirshberg

, Chang

et al. Enteric fistulas: Principles of management. J Am Coll Surg, 2009; 209:484–491.

Draus

Jr , Huss

, Harty

et al.

Enterocutaneous fistula: Are treatments improving?

Surgery, 2006; 140:570–576.

Haffejee

. Surgical management of high output enterocutaneous fistulae: A 24-year experience. Curr Opin Clin Nutr Metab Care, 2004; 7:309–316.

10.

Becker

, Willms

, Schwab

. Small bowel fistulas and the open abdomen. Scand J Surg, 2007; 96:263–271.

11.

Wainstein

, Fernandez

, Gonzalez

et al. Treatment of high-output enterocutaneous fistulas with a vacuum-compaction device: A ten-year experience. World J Surg, 2008; 32:430–435.

12.

Dearlove

. Skin care management of gastrointestinal fistulas. Surg Clin North Am, 1996; 76:1095–1109.

13.

Ratner

. Skin grafting: From here to there. Dermatol Clin, 1998; 16:75–90.

14.

Scherer

, Shiver

, Chang

et al. The vacuum assisted closure device: A method of securing skin grafts and improving graft survival. Arch Surg, 2002; 137:930–933.

15.

Cheong

, Ong

, Lim

et al. Successful split-thickness skin grafting in a contaminated wound with an enterocutaneous fistula. Plast Reconstr Surg, 2005; 115:1221–1222.

16.

Dumanian

, Llull

, Ramasastry

et al. Postoperative abdominal wall defects with enterocutaneous fistulae. Am J Surg, 1996; 172:332–334.

17.

Goverman

, Yelon

, Platz

et al. The “fistula VAC,” a technique for management of enterocutaneous fistulae arising within the open abdomen: Report of 5 cases. J Trauma, 2006; 60:428–431.