Re-evaluating the Effectiveness of Corneal Collagen Cross-Linking and Its True Biomechanical Effect in Human Eyes

Abstract

Corneal collagen cross-linking (CXL), a technique where by corneal rigidity is thought to be increased through a photo polymerization reaction that results in the subsequent induction of cross-links between collagen fibres in corneal tissue, is a treatment option for keratoconus and other corneal ectasias that, in theory, addresses the underlying pathogenesis of the disease and has thus far shown to be a promising early treatment option for these conditions. However, a striking discrepancy exists between the reported biomechanical effects of CXL in vitro and the biomechanical effects of CXL in vivo, and this has not received much attention in the literature.

Despite a documented increase in corneal stiffness in vitro reported by many investigators, reports that provide evidence of measurable and consistent biomechanical change in corneal rigidity in vivo after CXL are not comparable. The absence of documented in vivo biomechanical improvement in CXL-treated corneas is a conundrum, which merits further consideration. In order to understand this discrepancy, it has been postulated that biomechanical changes induced by CXL are too subtle to be measured by the current diagnostic tools available or have characteristics not discernible to these technologies. However, the dynamic bidirectional applanation device (ORA) and dynamic Scheimpflug analyser instruments (Corvis ST) have demonstrated the ability to quantify even subtle biomechanical differences in untreated keratoconus corneas of differing ectatic degree, and document the reduction in corneal hysteresis (CH) and corneal resistance factor (CRF) in situations where the corneal stiffness is reduced, such as after laser in situ keratomileusis and surface ablation procedures. It has also been possible to demonstrate a reduced CH and CRF in patients with diabetes, smoking, glaucoma, Fuchs’ dystrophy, and corneal oedema. It is puzzling that these diagnostic tools are able to determine subtle biomechanical changes in these situations, yet fail to measure the purported change induced by CXL on corneas with progressive keratoconus. This failure to document significant and consistent biomechanical changes in corneal rigidity could suggest that CXL does not induce a simple reversal of the particular biomechanical deficits seen in keratoconus and that CXL does not make the cornea significantly more resistant to bending forces as has been previously postulated and is widely accepted. The absence of measurable biomechanical change in keratoconic corneas after CXL could be a consequence that the biomechanical strengthening that occurs, which in contrast to the marked weakening caused by pre-existing alteration of the collagen structure, disorganization of collagen fibre intertwining, and compromised structural–mechanical homogeneity that are hallmarks of keratoconic disease and are more pronounced in corneas with progressive keratoconus, is insignificant by comparison.

The changes in the cornea induced by CXL that have been described in vivo may instead be driven by a wound healing process in response to the removal of the corneal epithelial layer and subsequent exposure to riboflavin and ultraviolet-A. This paper will present evidence that upholds this hypothesis.

Source of Support

I thank Cordelia Chan, MD, for assisting and providing critical revision of the manuscript.