An investigational time-course study using an in vivo ovine laminectomy model for the neurohistopathological evaluation of hemostatic agents.

Abstract

Background: Preclinical in vivo analysis of hemostatic agents is a fundamental requirement in the assessment of implant safety and efficacy prior to utilization in the clinical operative setting. The purpose of this study served to investigate the hemostatic, peri-operative outcomes and histopathologic responses following epidural application of a novel hemostatic agent in an ovine lumbar laminectomy model. Despite routine utilization of hemostatic agents, the potential for inflammation and compression of neural structures, fibrosis, and neurotoxicity remains a clinical concern. Epidural fibrosis resulting from hemorrhage creates chronic pain and is often refractory to treatment. Experimental endpoints, including intraoperative hemostasis, perioperative neurologic assessment, clinical pathology, and neurohistopathology, were used to evaluate the translational efficacy of this model.

Methods: Nine healthy crossbred Suffolk sheep (skeletally mature females, 1.5-2 years old, approximately 150 pounds) underwent posterior surgical lumbar laminectomies at the L3 and L5 levels followed by epidural application of a novel hydrophobically modified chitosan-based hydrogel hemostatic agent (test article). The Surface Bleeding Severity Scale (SBSS) score was recorded immediately after completion of the laminectomy at baseline, and then at 3-, 6-, and 10-minute intervals following application of the hemostatic agent. Postoperative survival analysis, specifically focused on animal recovery and neurological status was performed. Clinical pathology including comprehensive vet screens (CVS), complete blood count (CBC) with differential, and cerebrospinal fluid (CSF) assays were completed. Animals were humanely sacrificed at 12 days (n=6), 30 days (n=1), 60 days (n=1), and 90 days (n=1). Necropsy was performed at the time of sacrifice and the operative levels were axially sectioned for histopathologic evaluation.

Results: There were no neurologic, vascular, or infectious complications and animals exhibited normal recovery for this 2-level laminectomy surgical model. Complete intraoperative hemostasis was achieved in 16 out of 18 operative levels by the 3-minute interval and all 18 laminectomy levels demonstrated complete hemostasis by the 6-minute interval. Clinical pathology results were unremarkable. Histological characterization exhibited normal inflammatory and healing responses of the tissues at all postmortem time points, without evidence of abnormal spinal cord changes or infection. Residual test article was observed to decrease between the 30- and 60-day intervals, with no residual material in the epidural space observed at 90 days.

Conclusions: This pilot study demonstrates the translational utility of the ovine model in evaluating the safety and efficacy of a novel hemostatic agent in the lumbar spine. Based on these early findings in intraoperative hemostasis, postoperative survival, and histologic analysis, additional studies of hemostatic agents using the ovine model are encouraged.