Rabbit Model of Corneal Scarring: Objective Validation Using Vision-Independent Imaging Metrics

Abstract

Purpose : Evolving regenerative approaches for corneal stromal opacification need to be tested in valid animal models of corneal scarring. This study aimed to compare the thickness, reflectivity, light scattering and surface irregularity of experimentally induced corneal scars in rabbit eyes with post-keratitis scars in human eyes.

Methods : This was prospective study, with two arms: (i) The human arm included 14 eyes of 14 patients with healed ulcerative keratitis; and (ii) the rabbit arm included 6 eyes of 6 New Zealand white rabbits. A 3-mm central ulcer was experimentally induced in the rabbit eyes using a standardized technique, which involved manually removing 200-250 microns of the superficial stroma followed by alger-brush application. Eyes in both groups underwent slit-lamp photography, high-resolution anterior segment optical coherence tomography (AS-OCT) and Scheimpflug imaging using identical diagnostic platforms and standardized image capturing protocols at 2-months post-ulceration. Parameters studied were (i) percentage change in corneal thickness on AS-OCT; (ii) corneal epithelial: stromal reflectivity (ES ratio) on AS-OCT; (iii) corneal light scattering at the anterior and mid-stromal level using Scheimpflug densitometry; and (iv) corneal flattening using Scheimpflug tomography.

Results : The mean change in corneal thickness in the rabbit model was comparable to human eyes (28.4±13.5% vs. 19.9±13.1%, p=0.11). The mean ES ratio was 0.82 ± 0.1 and 0.75 ± 0.2 in human and rabbit eyes, respectively (p=0.26). The mean increase in corneal light scattering on densitometry was 360.3±155.6% in human eyes and 260.8±94.6% in rabbit (p=0.51). The mean corneal flattening in the affected area was 18.7±2% and 17.8±9.6% in rabbit and human eyes respectively (p=0.81).

Conclusions : The results indicate that the rabbit model of corneal scarring created in this study was objectively comparable to post-ulcerative keratitis scars seen in humans, in terms of the change in corneal thickness, reflectivity, light scattering and surface irregularity. This sets the stage for animal studies evaluating the efficacy of regenerative approaches in this rabbit model of corneal stromal scarring.

This is a 2020 ARVO Annual Meeting abstract.