Standard Biocompatibility Studies Do Not Predict All Effects of PVA/CMC Anti-Adhesive Gel in vivo

PVA/CMC (polyvinyl alcohol/carboxymethyl cellulose) hydrogel fulfills various physiochemical properties required for an adhesion barrier and has shown good anti-adhesion properties in previous in vivo studies. In this investigation, we assessed the in vitro and in vivo biocompatibility of PVA/CMC gel and compared this to the functionality and promotion of wound healing for two surgical indications.

Standardized ISO10993 in vitro and in vivo biocompatibility studies, comprising cytotoxicity, genotoxicity, acute systemic toxicity, delayed contact and maximization sensitization test, intracutaneous reactivity and local muscle implantation, were performed on PVA/CMC gel. In the functional studies, PVA/CMC gel was applied - on the one hand - to a rabbit abdominal wall model enforced with a polypropylene mesh for testing the anti-adhesion properties and - on the other hand - to an end- to-end anastomosis model that was selected for surveying potential influences of different dosages of PVA/CMC gel on anastomotic wound healing.

The ISO10993 methods indicated generally good biocompatibility properties, such as the absence of cytotoxic and mutagenic effects as well as no signs of systemic toxicity and sensitization potentials. No irritation effects were observed after the intracutaneous injection of lipophilic PVA/CMC sesame oil extract. However, the injection of hydrophilic PVA/CMC physiologic saline extract induced slight irritation. Following rabbit muscle implantation of the PVA membrane for 2, 4, 12, 26 and 52 weeks, a slight irritant effect was observed at 12 weeks due to the peak of phagocytosis. In the functionality tests, PVA/CMC gel showed good anti-adhesive effects in the abdominal wall model enforced with the mesh, with significantly lower and less tense adhesions compared to the untreated control. However, moderate signs of inflammation, especially in the spleen were observed after the intra-abdominal implantation of 3.3 ml PVA/CMC gel per kg body weight. In the end-to-end anastomosis model, PVA/CMC gel had no influence on wound healing. For dosages of 1-6 ml gel per treatment, no signs of intestinal leaks were detected, and tensile strength was equal to that of the untreated control, but again more moderate signs of inflammation in the spleen were observed at a dosage >3 ml.

Comparing the standardized ISO10993 methods, anti-adhesive PVA/CMC gel displays good biocompatibility. However, those methods do not seem to be sensitive enough because the rabbit abdominal wall and the end-to-end anastomosis models display more effects with respect to the dosage and routes of the intra-abdominal resorption of PVA/CMC gel - with the recommended <2 ml PVA/CMC gel per kg body weight as a secure dosage.