Biological characterisation of morphological alterations of the cornea following collagen cross-linking (CXL) treatment

Abstract

Introduction: Keratoconus (KC) is a disease included in a broader range of conditions called corneal ectasia which leads to spontaneous, non-inflammatory, self-limiting ectasia of the cornea that results in localized corneal thinning with protrusion of the thinned cornea. In recent years, corneal crosslinking (CXL) treatment with ultraviolet light (UVA) and riboflavin has been introduced as a minimally invasive treatment option to arrest the progression of keratectasia and it has revolutionized the treatment of KC. Several animal models, including rabbits, have been used to study the morphological alterations of the cornea following CXL treatment. Few studies have investigated the corneal morphological alterations following CXL treatment in porcine and rabbit corneas due to their structural similarity to the human cornea. Currently, no studies that have compared the normal corneal collagen fibrillar (CF) diameter, interfibrillar distance and interlamellar distance between human and these animal models. In addition, none have compared the morphological alterations of collagen fibrils (CFs) in both porcine and rabbit corneas with human, to identify an animal model that reacts in a similar manner to the human cornea, following CXL treatment.

Aims: In the present study, we have compared the normal morphology of CF’s in porcine and rabbits to find an animal model comparable to human. These models (rabbit and porcine) were also investigated to find the optimal animal model that mimics reactions similar to human, to CXL treatment. Finally, we have investigated the effects of crosslinked corneas maintained in vitro for two weeks. We have also explored the possibility of extending the time a crosslinked cornea that can be kept in vitro beyond 2 weeks.

Study 1: Materials and methods: The normal morphology of porcine (n=5), rabbit (n=5) and human (n=5) corneas were analysed using light and Transmission Electron microscopy (TEM). Results & conclusion: Histologically, all five layers (epithelium, Bowman's layer, stroma, Descemet membrane and endothelium) of the cornea were visible in all the three species. The CF diameter, interfibrillar distance and interlamellar distance were compared between porcine, rabbit and human corneas for any statistically significant difference. It was evident that the CF diameter and interfibrillar distance of porcine and rabbit corneas were significantly different (p<0.001) from the human corneal values, but not from each other. While neither of animal models were structurally identical to the human cornea, they are both relatively similar to being used as models to study the biomechanical effects of external insults/treatments to be extrapolated to the human cornea.

Study 2: Materials & methods: Porcine (n=10), rabbit (n=10) and human (n=10) corneas were treated with CXL treatment. The central corneal thickness (CCT) was measured before and after CXL treatment. Each treated and control corneas were bisected, and one half was used for histological examination; the other half imaged by Transmission Electron Microscope (TEM) to investigate the morphological changes of corneal CF’s. These parameters were compared to identify the optimal animal model that mimics reactions of human following CXL treatment. Results & conclusion: The CXL treatment leads to an increase in CF diameter, a decrease in the interfibrillar distance, and a decrease in interlamellar distance. The effect of the CXL treatment was greatest in the anterior region of the corneal stroma, followed by the intermediate region, followed by the posterior region. All the species showed statistically significant keratocyte apoptosis and a significant reduction in the CCT following CXL treatment. When the magnitude of the change post-CXL treatment is taken into consideration, it was evident that porcine corneas were more suitable as an animal model to study reactions of CXL treatment.

Study 3: Materials & methods: The effects/changes to the morphology of the CF’s in crosslinked corneas in vitro for two weeks were investigated, by maintaining 15 human corneas in a culture medium. Results & conclusion: The results revealed that the effect of crosslinking (increase in collagen fibre diameter, decrease in interfibrillar and interlamellar distance) does not significantly change during the 2 week post-treatment. In the next part of the in vitro study, porcine (n=12) and human (n=3) corneas were investigated for the possibility of extending the viability of the crosslinked cornea in vitro beyond 2 weeks. However, we were unable to maintain the crosslinked corneas for more than two weeks, and we postulate that epithelial damage due to de-epithelialization before crosslinking treatment has an adverse effect on maintaining a healthy endothelium in vitro, hindering the de-epithelialized cornea from being maintained in a culture medium for a longer period of time.