A reliable animal model of corneal stromal opacity: Development and validation using in vivo imaging.

Purpose: To validate an animal model of corneal stromal opacity by using objective vision-independent in vivo imaging metrics.

Methods: This was a prospective study, with two arms: (i) observational human arm which included 14 patients with healed unilateral ulcerative keratitis; and (ii) experimental rabbit arm, which included 6 New Zealand white rabbits. A 3-mm central wound was created in the left eye of the rabbits by manually removing 200-250 μm of the superficial stroma, followed by rotating-burr application. Both groups underwent photography, high-resolution anterior segment optical coherence tomography, and Scheimpflug imaging using similar diagnostic platforms and standardized image capturing protocols. Parameters studied were relative change in (i) corneal thickness; (ii) corneal epithelial: stromal (E:S) reflectivity ratio; (iii) corneal stromal light scattering using densitometry; and (iv) central corneal keratometry.

Results: In the experimental arm, there was a significant decrease in corneal thickness (273 ± 51.3 vs. 407.3 ± 10.3 μm, p = 0.0038), E:S reflectivity ratio (0.71 ± 0.09 vs. 0.99 ± 0.06, p = 0.0018), and keratometry (40.4 ± 2.3 vs. 45.8 ± 0.9D, p = 0.0033) and increase in densitometry (54.2 ± 11.65 vs.18.7 ± 3.8 GSU, p = 0.0001) from baseline, which stabilized at 4 to 8-weeks post-wounding (p > 0.3632). At 8-weeks, the relative change from baseline in corneal thickness (28.4 ± 13.5% vs.22.4 ± 13%, p = 0.368), E:S reflectivity ratio (28.1 ± 11.5% vs. 30.6 ± 8.9%, p = 0.603), corneal densitometry (204.17 ± 97.3% vs. 304.9 ± 113.6%, p = 0.1113), and central corneal keratometry (13.6 ± 6.9% vs. 18.9 ± 7.4%, p = 0.1738) in rabbits was similar to human corneal scars.

Conclusion: The animal model of corneal opacification was objectively comparable to human post-keratitis scars and can be valuable for in vivo evaluation of emerging therapies for corneal opacities.