This single center study reviews the records of 146 consecutive intestinal transplant patients between 2004 and 2010. Both pediatric and adult patients were included. No patients were excluded. As some patients did not undergo formal closure of the abdominal wall at the time of transplant, abdominal wall closure was coded as the technique used for intended definitive closure. The abdominal wall closure for all patients was reviewed and categorized as primary fascial closure or closure with any foreign material. Mesh closures were with either ADA or other mesh material. A thorough chart review was performed to assess complications specifically related to fascia reconstruction including reoperation for perioperative dehiscence, hernia or enterocutaneous fistula at any time posttransplant.
Transplants included one of three types: isolated intestine transplant (ITx; intestine alone), modified multivisceral transplant (MMVT; stomach, duodenum, pancreas and small intestine) and multivisceral transplant (MVT; MMVT + liver). All patients received similar immunosuppression, which included induction with three doses of rabbit antithymocyte globulin, with each dose preceded by solumedrol (500, 250 and 120 mg) and a single dose of rituximab, and maintenance with tacrolimus and low dose steroids. Donor size was taken into account for all recipients, with goal donor size of 50–100% of recipient size to facilitate closure of the abdominal wall. Physical measurement of donor and recipient other than height and weight was rarely used in the size match decision.
ADA is derived from deceased donor human dermis. The dermis undergoes chemical processing to remove cellular components, leaving behind the tissue matrix. Research suggests that this tissue matrix allows revascularization of the dermis, thereby allowing blood components to migrate to the mesh. This vascularization permits implantation of this biologic mesh into an infected or contaminated field, unlike synthetic mesh. In this study, placement of thick and ultra thick ADA was per manufacturer's recommendation for preparation and implantation (AlloDerm, LifeCell, A KCI Company, Branchburg, NJ, USA). This material is commercially available in the USA, Canada, Korea, Mexico and Israel. Per a company representative, there is no current plan to expand distribution of this material because of limits on importation of human tissue. Because of the expense of this material, the mesh was often cut and shaped to fit the necessary space with the mesh sewn back together, as necessary. Typical defect size was 10–20 cm in greatest dimension and most defects could be covered by a 12 × 12 cm piece of ADA. The mesh was sewn into place with 0-prolene suture with moderate tension. It is recommended that this material be under moderate tension when placed as it loosens with time. ADA can stretch by up to 50% to cover larger and atypically configured defects. In the initial experience with ADA, we experienced rapid degradation of the mesh with direct exposure of the mesh to air (open wound) or to suction devices (closed suction or vac wound dressing). For this reason, we developed a policy of early return to the operating room to reclose any open skin wound, which resulted in exposed ADA. When the mesh remained covered with skin, we found that it retained its integrity indefinitely. Closure of the skin over the exposed mesh often included developing large vascularized skin flaps, with dermis closure using a combination of large vertical mattress sutures to reduce skin tension, as well as staples. Generally, a suction drain was placed over an area of native fascia in the subcutaneous space to prevent the accumulation of fluid and allow direct apposition of the skin to the mesh. When an open wound was reclosed, the incision site was routinely covered with a large bolster dressing of gauze, followed by an abdominal binder or compression dressing to encourage apposition of skin to mesh/fascia. In a limited number of cases, the mesh was removed after initial placement if postoperative edema resolved and primary closure was possible. These patients were considered to be primary closure for this study. Wound infections were managed by opening the wound and draining infected fluid. All open wounds were considered to be colonized and wound cultures were not routinely obtained. In patients with mesh, Aquacell Ag Hydrofiber (ConvaTec, Incorporated, Skillman, NJ, USA), a moisture retaining material impregnated with ionic silver, was placed over the mesh to treat infection and to prevent desiccation, with gauze dressing changed twice daily over the Aquacell. Intravenous antibiotics, as indicated, were administered until the wound was clean, without cellulitis and the patient was afebrile without leukocytosis. When the wound was clean and healing, the patient was returned to the operating room for delayed primary closure. Porcine derived mesh was used in a limited number of patients, but it was found to degrade and it was technically more difficult to implant because it is not pliable.
Standard statistical testing was conducted using commercially available software (IBM SPSS Statistics, Version19; IBM, Armonk, NY, USA). Categorical variables were compared using chi-square testing, though most numbers were too small for statistical analysis. Retrospective analysis of data using the transplant research database at our center has been reviewed and approved by the institutional review board of the Indiana University School of Medicine.