Background
Corneal melting with perforation is a severe, vision-threatening complication of corneal disorders such as corneal ulceration, chemical burn and autoimmune keratitis. In the acute stage, the urgent approach is to limit inflammation by directing against the cause as well as to optimize epithelial healing [
1]. Furthermore, surgical procedures, including tissue adhesive [
1,
2], amniotic membrane transplantation [
1,
3], conjunctival flaps [
1,
3], pericardial membrane graft [
4], are done to temporarily maintain the integrity of the globe. In general, this involves a multistage surgery [
1]. The final tectonic keratoplasty is performed to restore visual function. The surgical methods include full-thickness penetrating keratoplasty, lamellar corneal patch graft, and deep anterior lamellar keratoplasty [
1,
5‐
8], depending on the size and location of perforation. In these procedures, it is essential to have graft material for repair of the corneal defect readily available.
Shortage of graft material for repair, particularly in developing countries, has prompted the development of bioengineered tissue alternatives. Tissue engineering relies on the use of three-dimensional porous scaffolds to provide appropriate microenvironment to induce regeneration of injured tissues and organs. The porous collagen-glycosaminoglycan copolymer matrix (CG matrix) is composed of type I collagen and chondroitin 6-sulfate. The collagen of the corneal stroma is largely type I collagen. Previous study demonstrated that CG matrix in eyes could modulate the healing procedure of conjunctival wound, reducing scarring contraction and promoting the formation of a near-normal subconjunctival stroma [
9]. The CG matrix also serves as a three-dimensional scaffold for cell migration and proliferation on surgical bleb defect to maintain the size of bleb and repair the leakage [
10,
11]. CG matrix supplies good biocompatibility on the ocular surface. Therefore, CG matrix has a potential to be used as an alternative graft material for repair of corneal thinning by suppling thicker extracelluar matrix in the wound bed. However, considering that diameter, spacing, and spatial orientation of the collage fibrils in the corneal stroma is essential for corneal transparency [
12], we tested the healing effect of CG matrix on corneal thinning in a rabbit model. This study provides preliminary results for further advanced studies.
Discussion
Corneal thinning by trauma, surgery, infection, or inflammation triggers a series of corneal wound healing processes and corneal matrix remodeling, including stromal keratocyte apoptosis, epithelial migration, myofibroblast proliferation, and fibrosis [
17,
27,
28]. However, these responses might lead to angiogenesis and compromise the restoration of corneal transparency [
18,
19]. In addition, if incomplete wound healing persists, it could cause no improvement of corneal thickness and even perforation [
20]. Currently, the ultimate management to correct stromal scarring and corneal thinning is corneal transplantation. Full or partial-thickness corneal grafts are effective means of restoring transparency and thickness, but this procedure relies on fresh donated cadaveric cornea. Due to shortage of donor cornea, it is imperative to find viable alternatives to corneal tissue.
Type I collagen accounts for about 85% of the fibrillar collagen in human corneal stroma. In the form of heterotypic fibrils with type V collagen, type I collagen are crucial for corneal transparency [
21]. The porous collagen-glycosaminoglycan copolymer matrix (type I collagen and chondroitin 6-sulfate) was designed as temporary scaffolds with stiffness that maintains corneal structure in corneal thinning, while the surrounding corneal stroma tissue regenerates and replaces the original scaffold over time. In this study, CG matrices were implanted in the corneal stroma of 24 rabbits after deep lamellar keratectomy and observed over a 90-day period thereafter. On day 3, the grafted cornea showed infiltration of acute inflammatory cells around CG matrix. On day 7, new vessels began to appear at the periphery of the corneal wound. The CG matrix degraded completely between days 28 and 60. On day 60, CG matrix disappeared and central corneal scar formation with peripheral vessels ingrowth were found. And on day 90, the cornea of the grafted eyes cleared (Table
1). From the histopathologic findings, the intensity of α-SMA staining increased progressively from 7th to the 14th day then decreased gradually overtime but remaining positive at day 60. Therefore, CG matrix induced stronger inflammatory reaction and delayed wound healing process up to the 3rd month as evidenced by persistence of α-SMA stained cells [
22‐
26].
Inflammation is a fundamental process in corneal wound healing. The infiltrating inflammatory cells and cytokines activate keratocytes differentiation into fibroblasts and myofibroblasts [
22‐
26]. They migrate and accumulate in the provisional matrix of the wound site and secrete and deposit collagen [
27,
28]. Severe inflammation might overwhelm the antiangiogenic mechanism of cornea and might give rise to a secondary ingrowth of blood vessels from the limbus into the central cornea [
18,
29,
30]. Prolonged myofibroblasts activation and ongoing deposition of repair matrix would cause corneal scarring and opacification [
31]. These two physiologic conditions may complicate corneal wound healing resulting to poor vision. With tissue engineered CG matrix providing the 3-dimensional porous scaffold and appropriate microenvironment to promote more fibroblast and myofibroblast repopulation, the healing processes are modified and optimized [
9‐
11].
During the degradation of CG matrix, the patterns of cell migration and proliferation changes and the coexisting inflammation activates more fibroblasts and myofibroblasts. These keep depositing multiple elements of extracellular matrix to increase the corneal thickness [
32]. By remodeling the healing stroma, and replacing the disorganized repair matrix with regular corneal extracellular matrix, a better transparency can be achieved [
33,
34]. In addition, the number and the diameter of new vessels slowly decreases as the remodeling progresses [
35].
In our study, the complete re-epithelization of the cornea wound on ungrafted eyes was achieved on day 14 but stroma remodeling was finished on day 28. In the grafted eyes, these results were observed on day 60 and day 90, respectively. These demonstrated that the wound healing proceeded from a proliferation phase to a stromal remodeling stage. Utsunomiya et al. in 2014 also observed the wound healing process after corneal stromal thinning shifted from an acute wound healing phase to a remodeling phase by anterior segment optical coherence tomography (OCT) [
36]. The time to complete corneal wound healing in grafted eyes was influenced by the process and timing of CG matrix degradation. As a result, the intensity of corneal inflammation induced by CG matrix combined with the duration of CG degradation influenced the degree of stromal thickening, scarring and neovascularization.
CG matrix for repair of corneal thinning must be strong and implantable to maintain the corneal shape and curvature at the early stage. Additionally, the material should be able to maintain the 3-dimensional structure to support cell adhesion, migration and turnover into host extracellular matrix over time. In this study, we used soft contact lens and lateral tarsorrhaphy to anchor the CG matrix to corneal wound without further invasive procedures. This procedure minimizes disturbance of epithelium and maintains corneal shape without suturing, thereby avoids stimulation of an aggressive wound healing response. Though the main purpose of contact lens and tarsorrhaphy was to keep the CG graft in situ, they both also exerted certain pressure and compressed the CG matrix making it less prone to deformation. The CG matrix in our study was constructed by blending type I collagen with chondroitin 6-sulfate, with glutaraldehyde crosslinking. Compression test indicated that the matrix, upon compression by external force, becomes more rigid and becomes more difficult to deform [
10]. In previous study, 2 mm thickness of collagen/C-6-S copolymer was used to repair conjunctival defect and the implantation reduced contraction and promoted the formation of a nearly normal subconjunctival stroma. The matrix was almost completely degraded on day 28 [
9‐
11]. Therefore, 2 mm thickness collagen matrix with compression molding was used to resist the external forces by tarsorrhaphy and contact lens as well as enzymatic degradation. The majority of degradation of the CG matrix occurred in the first 4 postoperative weeks. Previous studies reported degradation rates of different collagen scaffolds to be 4 to 5 weeks, similar to what we observed in our study [
37‐
39]. Based on the histopathologic findings of this study (Fig.
3), the 3-dimensional scaffold was maintained to a certain degree without collapsing, and this allows for cell migration and proliferation to happen. At the early stage, CG matrix maintains the good crosslinking structure to maintain the original porous structure without collapsing and result in the predictable randomized collagen deposition pattern. At the later stage, degrading CG matrix became less rigid and was flattened in a more parallel way by the pressure of tarsorrhaphy on it and resulted in better corneal clarity in the grafted group than expected.
In this study, CG matrix significantly increased the thickness of the healed cornea compared with ungrafted ones on day 90 (
P < 0.001). From histological findings, the increased thickness resulted from additional lamella of new collagen deposition. Moreover, pentacam successfully measured the true central corneal thickness of the grafted eyes due to the improved clarity of the cornea. These results demonstrated that the new collagen ingrowth can undergo further remodeling to increase the corneal transparency. Although the corneas of grafted eyes were relatively hazy compared with the ungrafted ones, CG matrix was successful as a 3-dimensional temporary scaffold for corneal regeneration to regain the corneal thickness [
39]. Corneal transparency greatly depends on the organization of the type I collagen fibrils, especially their diameter and regular lamellae organization [
21]. While the slow degradation of CG matrix and inflammation induced by CG matrix lead to thicker cornea, both also resulted in corneal haziness. Therefore, finding the balance between corneal thickening and corneal haziness induced by CG matrix must be overcome. The limitation of this work is that corneal healing response from injury differs based on the nature of the insults such as chemical burn [
40].
This study showed the healing effect of CG matrix on corneal thinning by injury. In the future, CG matrix may be designed with a different degradation rate so as to optimize stromal regeneration, or with a different porous size, diameter and arrangement to better regulate the assembly of cornea fibrils and the organization of the extracellular matrix to maintain corneal transparency [
41,
42].