Development of a novel in vivo canine corneal fibrosis model and therapeutic potential of Suberanilohydroxamic Acid (SAHA) for corneal scarring in vivo

Abstract

Purpose: A reproducible in vivo large animal model of corneal scarring is required for bench-to-bedside translational research. We hypothesized that a canine corneal alkali burn in vivo would result in significant focal fibrosis without damage to limbal stem cells and topical application of SAHA would decrease fibrosis.

Methods: Seven research 9-12 months Beagle dogs were used. Corneal opacity was obtained from a 6 mm axial corneal alkali burn using 1 N NaOH at 15 vs 30 seconds for initial dog and at 20 seconds for remaining dogs. Baseline optical coherence tomography (OCT) was performed to determine the degree of corneal opacity. Six dogs were randomly assigned into 2 groups (3 dogs/group): A) vehicle (DMSO, 2 μL/mL); B) anti-fibrotic treatment (50 μM SAHA). Efficacy of SAHA as a corneal anti-fibrotic agent and ocular health was graded daily according to modified McDonald-Shadduck (mMS) scoring system. At each study aims’ endpoint, OCT was repeated; corneal histopathology, special stains, and immunohistochemistry (IHC) were performed. Normal corneas of dogs euthanized for reasons unrelated to the study were used as negative controls for histopathology and IHC.

Results: No significant difference in mMS scores between control and SAHA-treated groups was detected (p=0.264). Abnormalities noted on histopathology (epithelial hyperplasia, stromal fibrosis, edema and neovascularization) did not differ significantly between negative controls and treatment groups. Smooth muscle actin staining was significantly greater in group B than negative controls (p=0.032). OCT measurements detected significantly thicker corneal epithelium in group B compared to baseline images (p=0.007). Total corneal thickness of both treatment groups was significantly greater than baseline images (A, p<0.001; B, p<0.001); total corneal thickness of group B was significantly greater than group A (p=0.004).

Conclusions: The canine corneal alkali burn is well tolerated and performed without damaging corneal limbal stem cells. Histopathology and OCT findings demonstrated reliable and reproducible induction of fibrosis in the cornea. No significant difference in fibrotic changes was detected after treatment with SAHA in this small study size (i.e. pilot study). Additional studies are warranted.