The Use of Synthetic Mesh in Contaminated and Infected Abdominal Wall Repairs. A Long-term Prospective Clinical Trial

We designed a prospective study, evaluating the short and the long-term outcomes of the surgical treatment of 40 consecutive patients presenting with an infected abdominal wall, compared to a cohort of 40 patients submitted to clean ventral hernia repairs. Patients were admitted between January 2012 and February 2015, and operated at the Hospital das Clinicas of the University of São Paulo School of Medicine, in Brazil. All patients included in the study group carried an active chronic mesh infection (mesh sinus, exposed mesh or mesh related enteric fistulas) resulting from a previous hernia repair, with or without an associated recurrent ventral hernia. All the operations required in this group were classified as Class IV (dirty-infected), accordingly to the CDC Wound Classifications, as adopted by the European registry for abdominal wall hernias. The only inclusion criterion was the presence of an active chronic mesh infection. The infected abdominal wall (IM) group and the outcomes of the treatment were compared to a clean-control (CC) group of patients who underwent clean abdominal wall reconstructions, originally belonging to a prospective study of the tensiometry of the abdominal wall, and operated during the same period and in the same conditions. The exclusion criteria were: giant ventral hernias with a volume ratio higher than 25%, patients on immunosuppressive therapy or using corticosteroids, patients with hepatic cirrhosis and portal hypertension, Chron´s disease, acute postoperative mesh infection, chronic mesh infections following inguinal hernia repair and emergency operations.The data assessed included gender, age, body mass index (BMI), American Society of Anesthesiologists (ASA) score, comorbidities, smoking status, cancer history, the number of previous abdominal operations, the number of prior hernia operations, the presence of a recurrent incisional hernia and/or enteric fistula and the presentation of the mesh infection. Perioperative data included operative time, anesthesia time, associated procedures, the defect characteristics and the extension of the pre-aponeurotic dissection. Further analysis in the infected mesh (IM) group included the type and position of the infected mesh, the causes for the mesh infection and the microbiology of mesh explants. Patients were followed and operated at the Abdominal Wall and Hernia Repair Unit of the General Surgery and Trauma Discipline, and five surgeons of the team conducted the operations in both groups. Informed consents were presented to the patients in the IM group upon admission. Patients in the CC group signed informed consents to participate the study, during their post-operative follow-up visits.

Surgical procedure The operations were performed through the previous surgical incision. Patients in the IM group had the infected mesh removed completely together with all the sutures, tacks or other foreign material. The surrounding fibrotic tissue and the existing sinus tracts were resected to perform a complete toilet of the abdominal wall. The abdominal cavity was entered in most of the patients, and associated or incidental procedures were made as required. After the removal of the infected mesh, or at the end of the abdominal cavity workup, all the surgical drapes, instruments and gloves were replaced. The size of the defect and the extension of the anterior abdominal wall dissection were estimated. The reconstruction of the midline was done as anatomically as possible, by repositioning of the muscles and primary closure of the aponeurosis. We did not use component separation techniques or transverse abdominal releases in any case. In some patients, a bilateral relaxing incision along the anterior rectus sheath was required to allow the re-approximation of the muscles in the midline. In all patients, a heavyweight macroporous (pore size > 75µm) monofilament polypropylene mesh (Intracorp®, Venkuri, www.venkuri.com.br) was used in the onlay position, to reinforce the repair. The mesh was fixed with multiple absorbable Vicryl® sutures, placed over the borders of the mesh, in the midline and along the relaxing incisions. The remaining dead space between the mesh and the underlying tissue was cleared entirely with interrupted sutures. The operative field was irrigated with 0,9% saline, and the subcutaneous was drained with suction drains. The scars and the exceeding skin flaps were resected. We did not use irrigation with antibiotics solution. The subcutaneous tissue and the skin were closed with interrupted sutures. Fluids and samples of the explanted mesh were sent to cultures and microbiological analysis. The same technique of repair was used in the CC group. The size of the defect was measured, and the extension of the anterior abdominal wall dissection was estimated. A bilateral relaxing incision along the anterior rectus sheath was used routinely in this group, to allow a tensionless closure of the midline.

Outcomes parameters The primary outcomes variables were the presence of any surgical site occurrences (SSO) or surgical site infection (SSI) during the first 30 days after the operation, and the development of hernia recurrence or the recurrence of mesh infection during a 36-month follow-up period. Suspected recurrences of a hernia or infection were determined by physical examination and CT scan imaging. Non-surgical complications, other operations, and deaths were registered during the follow-up period.

Statistical Analysis The chi-square test was conducted to verify the association between categorical variables in contingency tables and the Fisher exact test was adopted in 2x2 tables whenever at least one expected frequency was less than 5. The U-test was used for verifying the association between continuous data and group with two categories, and when normal distribution was observed the Student t-test was performed. The 5% level of significance was considered for all statistical tests. Statistical computer Stata software version 10.0 (StataCorp, College Station, TX) was used for conducting all statistical analysis.