Biological meshes for abdominal hernia: Lack of evidence-based recommendations for clinical use

Background: In the clinical literature on abdominal hernia repair, no sound criteria have been established to support the use of biological meshes as opposed to synthetic ones. Furthermore, the information on biological meshes is quite scarce, and so their place in therapy has not yet been defined. Methods: The treatment of primary and incisional ventral hernia was the target intervention evaluated in our analysis. Our study consisted of the following phases: a) Identification of the biologic meshes available on the market; b) Literature search focused on efficacy and safety of these meshes; c) Analysis of the findings derived from the literature search. The information collected this way was reviewed narratively, and presented according to standard meta-analysis. The main end-points of our analysis included infection of surgical wound at 1 month and recurrence at 12 months. Results: Our clinical literature comprised 11 trials that evaluated 5 biological meshes: Permacol (706 patients), Strattice (324 patients), Surgisis (44 patients), Tutomesh (38 patients) and Xenmatrix (22 patients). These studies generally showed a poor methodological quality. Surgical wound infection showed a wide between-study variability (95% CI: from 12.0% to 22.9%). Also the 12-month relapse rate demonstrated a wide 95% CI (from 5.0% to 19.9%). A significantly lower rate of recurrence at 12 months was found for Permacol compared with Strattice (rate difference: -14.2%; 95% CI: -22.1% to -6.2%). Discussion: Our analysis provided an overview of 5 biological meshes currently available on the market. The different types of meshes showed a marked statistical variability in the clinical outcomes. Hence, nearly all comparisons between different meshes in the two clinical end-points did not reach statistical significance. One exception was represented by the finding that cross-linked meshes had a significantly lower recurrence rate at 12 months than non-cross-linked meshes.