Study on patterned photodynamic cross-linking for keratoconus

Purpose: This study examined the patterned treatment of corneal collagen cross-linking (CXL) for keratoconus to reduce the complications caused by ultraviolet (UV) irradiation. By modifying the method of UV irradiation during the cross-linking process, cross-linking with a special structure is achieved, and the cross-linking effect is analyzed and compared to that of traditional cross-linking. By constructing an animal model of keratoconus, the process and effect of corneal cross-linking can be investigated more fundamentally. These studies provide valuable references for future cross-linking precision improvement and specialization. Method: By injecting exogenous collagenase into the corneal stroma of rabbits, the balance between collagenase and collagenase inhibitor in the corneal stroma was disrupted, the collagen fiber structure of the cornea was broken to simulate the pathogenesis of keratoconus, and an animal model of keratoconus was thus constructed. Two custom cross-linking patterns were designed with reference to the cable dome structure, and these two special patterns were irradiated and cross-linked by a DMD chip. The cross-linking effect was evaluated by optical coherence tomography (OCT), corneal topography and corneal biaxial tensile tests. The experimental rabbits were divided into four groups: group A, cross-linking of the bird's nest structure; group B, cross-linking of the honeycomb structure; group C, cross-linking of the traditional spot structure; and group D, normal (without modeling and cross-linking). Result: Following collagenase treatment, the collagen fiber structure of the rabbit cornea was destroyed, the central thickness of the cornea was reduced, the mechanical properties of the cornea were weakened, and no keratitis, ulcers or haze occurred. After the three cross-linking treatments, the morphology of the cornea improved, the density of the stromal layer increased, and the mechanical properties were enhanced. For the improvement of keratoconus mechanical properties, the average relative difference (Delta) of the four outcome measures was 61.98% for bird's nest cross-linking versus keratoconus (Wilcoxon rank sum test, P = 0.024), 16.13% for honeycomb cross-linking versus keratoconus (Wilcoxon rank sum test, P = 0.025), and 21.07% for traditional spot cross-linking versus keratoconus (Wilcoxon rank sum test, P = 0.014). All these differences are statistically significant. Conclusion: All three methods of cross-linking can improve the morphology and tissue structure of keratoconus and significantly improve the biomechanical properties of the cornea. Among them, the corneal cross-linking of the bird's nest structure attains the best biomechanical properties, followed by the corneal cross-linking of the traditional spot structure and of the honeycomb structure. This suggests that similar or better cross-linking effects can be achieved by designing custom structures with less UV exposure. This provides a direction for future research on better and more accurate pattern cross-linking treatments.