Reduction of epidural fibrosis in lumbar surgery with Oxiplex adhesion barriers of carboxymethylcellulose and polyethylene oxide.

BACKGROUND CONTEXT Postsurgical epidural adhesions and fibrosis after surgery for lumbar disc herniation are a consequence of normal wound healing. The presence of fibrosis renders reoperations risky, and in some patients fibrosis may lead to nerve root tethering. PURPOSE One approach to minimizing the risk of developing epidural adhesions is to provide a barrier between the dural membrane and the healing connective tissues. The purpose of these studies was to evaluate such a barrier device. STUDY DESIGN/SETTING In vivo investigation in an animal model at a university laboratory. PATIENT SAMPLE Rabbit. OUTCOME MEASURES Gross and histomorphic evaluation. METHODS Barriers comprised of carboxymethylcellulose (CMC) and polyethylene oxide (PEO) (Oxiplex; FzioMed, Inc., San Luis Obispo, CA) were studied as devices to reduce epidural adhesion formation in rabbit laminotomy and laminectomy models. The barriers tested were either a gel alone (gel) or a gel covered with a film (gel/film combination). Two laminotomy or laminectomy sites (depending on the surgical method) were created in each rabbit at L4 and L6. One site was treated with a CMC/PEO gel, or CMC/PEO gel/film combination, and the other site served as a surgical control. Two surgical models that differed in the extent of adhesion formation at untreated injury sites and the method of injury generation were used. RESULTS Model A, which did not incorporate dural abrasion, resulted in up to 40% adhesion-free laminectomy sites in controls. Model B, which did incorporate abrasion of the dural membrane, resulted in less than 10% adhesion-free laminotomy sites in controls. Compositions of CMC/PEO gels (2.5% to 10% PEO) and films (22.5% PEO) were tested in both models. Efficacy parameters included measuring the number of sites free of epidural fibrosis and reduction in the severity of fibrosis (adhesions). Both gels and gel/film combinations consistently reduced the frequency and the extent of epidural fibrosis in both models. Gels of CMC/PEO containing a higher content of PEO (10%) and a higher molecular weight of PEO (4.4 mD) were most effective in Model B and resulted in up to 84% laminotomy sites with minimal or no epidural fibrosis, whereas controls exhibited over 90% of the sites with epidural fibrosis. Histological evaluation of the surgical sites indicated that the reduction of epidural fibrosis was accompanied by normal bone healing. In addition, these experiments demonstrated that the gel/film combination provided no additional benefit to that obtained by the gel alone. CONCLUSIONS Gels of CMC/PEO reduced epidural fibrosis and did not impair normal heal ing.