Recombinant Human Albumin PEG Hydrogel with Dual Functions: Sealing CSF Leak and Mitigating Peridural Fibrosis in a Large Animal Durotomy

Abstract

BACKGROUND CONTEXT: Biodegradable hydrogels have potential for sealing CSF leak or inhibiting postoperative peridural fibrosis, but few have been developed and investigated to show dual functions or efficacy. We evaluated the effectiveness of 2 formulations of a novel, self-polymerizing, absorbable, biocompatible hydrogel in prevention of CSF leak and reduction of peridural adhesions in a canine model. Functionalized PEG crosslinked to recombinant human albumin [rHA] produces a hydrogel of high crosslink density, matrix strength with increased surface adherence. METHODS: L2 and L5 laminectomies with 1 cm midline durotomies were created and sutured, except for the caudal 2 mm to create CSF leak. 24 animals were randomly assigned to 1 of 4 groups (each n56): control (no sealant), hydrogels DS1/2 & DS1/3 [Neomend] and DuraSeal [Covidien]. Hydrogels (equivalent volumes) were sprayed onto the durotomy area and formed a flexible, adherent seal, confirmed by Valsalva to 20 cm H2O O 1 min. 3 animals from each group were sacrificed at day 28 and 3 at day 120. Tissue was harvested and histological sections prepared to evaluate dural healing, residual gel, and inflammatory and tissue responses. Peridural scar formation was quantified. Spinal cord, spinal nerves, ganglia were reviewed for pathological changes. RESULTS: All animals recovered without treatment related ill effects or problems. CSF leaks in hydrogel animals were sealed intraoperatively and remained leak free. Controls had CSF subcutaneous accumulation. 28 days: hydrogels showed low severity foreign body and bioresorbing response primarily comprised of epithelioid and vacuolated macrophages with fewer lymphocytes and no dural/subdural inflammation. Fibrous dural remodeling, integrity-healing in DS1/2, DS1/3 was advanced vs. controls (p#0.002) and DuraSeal (p#0.022). No intradural adhesions were observed in DS1/2; DuraSeal and DS1/3 exhibited low levels of intradural dura-pia adhesions (p>0.013 vs. DS1/2). Residual sealant varied from large (DuraSeal), to intermediate (DS1/2) and low (DS1/3). Quantitative variations can be explained by the bioresorption rate (DS1/3 fastest) and material swelling (DuraSeal greatest).120 days: hydrogels were completely resorbed; scattered vacuolated macrophages were visualized. Complete and comparable healing quality: minimal dural fibrous thickening and less extensive peridural fibrosis in hydrogel groups vs. controls (p#0.025). DS1/3 showed the least peridural fibrosis, although tissue planes where the gel had been were identified in hydrogel groups. Peridural connective tissue appeared looser and more tenuous vs. controls where denser connective tissue comprised from muscle layers to dura. DS1/2 and DS1/3 exhibited better intradural adhesion prevention vs. DuraSeal (p#0.035). Ossification was greatest in DS1/3. CONCLUSIONS: DS1/2 and DS 1/3 hydrogels were effective in sealing CSF leak, and mitigated peridural fibrosis without affecting dura or laminectomy healing. At 28 days, DS1/2 and DS1/3 showed advantages in adhesion prevention and dural healing compared to DuraSeal; at 120 days the appearance of hydrogel groups was more similar. Hydrogels showed excellent biocompatibility characteristics. Differences in tissue reactions are likely related to material compositions. The DS rHA hydrogels were effective in preventing CSF leak, reducing fibrosis, with histological optimal appearance of healing. Such hydrogels may have significant potential and provide advantage in human spine surgery.