Keratoconus is a progressive eye disease in which the cornea thins and starts to become conical in shape. Although the etiology of this eye disease is not fully known, it significantly reduces the mechanical strength of the tissue.1 Corneal cross-linking (CXL) is a relatively new treatment that is currently used to halt the progression of this eye disease.2,3 This therapeutic intervention uses the photosensitizer riboflavin solution and ultraviolet-A light (UV-A) to enhance mechanical properties of the cornea by inducing cross-links in its extracellular matrix. There has been great progress in characterizing the effect of this treatment option on corneal hydrodynamic behavior, collagen fibril diameter, keratocytes, and endothelial cells, among others.4–8 Nevertheless, its exact molecular mechanisms are not fully understood. A complete understanding of the working principle of CXL is crucial for proposing new modified protocols for this treatment option.9–14
The mechanical properties of the cornea are mainly dependent on its extracellular matrix (stroma), which makes up approximately 90% of its thickness and includes the majority of collagen and proteoglycan content of the tissue.15–17 Inside the stroma, collagen fibrils are organized into 1- to 2-μm thick sheet-like lamellae, which show a depth-dependent organization (ie, the anterior lamellae are interwoven while the posterior ones are arranged parallel to the surface). In addition to the inhomogeneous microstructure of the corneal stroma, it has been shown that the riboflavin solution uptake is limited to the anterior stroma.18 Thus, it has been hypothesized that the CXL therapy should have an inhomogeneous stiffening effect over full-corneal thickness. This hypothesis has been tested before and proven true by characterizing the stiffening effect of CXL in different depths of the stroma. To the best of our knowledge, all of these previous studies have been done by cross-linking full-thickness corneas.8,19–22
The commonly used Dresden protocol has been developed such that it only affects the anterior portion of the cornea.2 If damage to the endothelium due to the UV light is circumvented, it can be hypothesized that cross-linking deeper layers of the cornea will further strengthen the tissue. This is a desirable option, especially for patients with keratectasia after LASIK or researchers working on various accelerated CXL protocols.23–25 Nevertheless, possible stiffening effects of CXL on posterior corneal layers are not fully known. In the current study, anterior and posterior flaps excised from porcine corneas were cross-linked separately with riboflavin and UV-A. Then mechanical tests were conducted to determine the amount of their stiffening because of CXL treatment. Furthermore, full-thickness porcine corneas were cross-linked from the endothelial side (although this approach is not clinically applicable) to better characterize possible effects of CXL on posterior corneal lamellae. In addition to providing new data on CXL, this study is a step forward to better understand the mechanisms involved in CXL and could assist researchers who are interested in modifying CXL protocols.
Materials and Methods
Tissue and Specimen Preparation
This study used 25 porcine eye globes, which were brought to the laboratory from a local slaughterhouse within 6 hours postmortem. All samples were used within 24 hours of the procurement. Corneoscleral rings were dissected from the eye globes and corneal thickness was measured by a digital pachymeter (DGH Technology Inc., Exton, PA). All samples were air dried (or put in saline solution if necessary) until their thickness was 800 μm to have specimens with similar initial thickness. Epithelium and endothelium were removed from 15 corneoscleral rings and full-thickness corneal discs were obtained using a 10-mm biopsy punch (Acuderm, Inc., Fort Lauderdale, FL). The remaining 10 corneoscleral rings were used to prepare anterior and posterior strips.
First, epithelium debridement was performed using the blunt edge of a scalpel. Then, they were mounted in a Descemet stripping automated endothelial keratoplasty (DSAEK) system (Med-Logics, Inc., Athens TX) to excise 10-mm anterior corneal flaps (Figure 1). A hypoosmolar solution was used to apply pressure underneath the samples in the DSAEK system. Thus, to minimize possible corneal swelling during flap creation, the endothelium was removed from the remaining posterior corneoscleral rings after anterior flaps were dissected and the samples were removed from the DSAEK system. After excising the anterior flaps, a 10-mm biopsy punch was used to obtain posterior flaps from posterior corneoscleral rings. A digital pachymeter was used measure the flap thickness at the center of each specimen. The average thickness of excised anterior and poster flaps immediately after dissection was 360 and 480 μm, respectively. A custom double-bladed cutting device26,27 was used to prepare 5-mm wide nasal-temporal strips from anterior and posterior flaps, as well as from full-thickness corneal buttons.
Schematic plot showing anterior, posterior, and full-thickness strips. (A) A Descemet stripping automated endothelial keratoplasty system was used to excise an anterior flap from the porcine cornea. (B) Anterior and (C) posterior strips were cross-linked from the epithelium side and the cut mid-stromal surface, respectively. (D) Full-thickness corneal strips were cross-linked from either the epithelial or endothelial side.
Preparation of Treated Groups
All strips were soaked in photosensitizer solution composed of 10 mg of riboflavin-5-phosphate in 10 mL of 10% dextran T500 until their thickness reached equilibrium. To this end, the thickness of strips was measured every 10 minutes using the pachymeter. An equilibrium state was defined as occurring when the difference between two successive measurements was less than 5%. The soaking time in riboflavin solution was approximately 1.5 to 2 hours to ensure sufficient riboflavin penetration in the samples.
A custom cross-linking device was used to cross-link the strips (Table 1).28,29 To this end, strips were placed on a plastic hemisphere and were subjected to a UV-A irradiance of 3 mW/cm2 for 30 minutes. The UV-A light (370 nm) source was at a distance of approximately 2 cm from the samples and drops of photosensitizer solution were continuously applied to the cornea during the treatment period. Five strips excised from the full-thickness cornea were cross-linked from the epithelial side (Figure 1D) and another five full-thickness samples were cross-linked from the endothelial side (Figure 1E). The anterior strips were cross-linked from the epithelial side (Figure 1B) and the posterior strips were cross-linked from the incised mid-stromal surface (Figure 1C). Five specimens were used for each group.
Preparation of Control Groups
Control samples were subjected to the exact same treatment as treated samples except that the UV-A light was turned off during the treatment procedure. Five control samples were used for each of the anterior, posterior, and full-thickness groups.
The strips were immediately mounted in a Dynamic Mechanical Analysis machine (TA Instruments, New Castle, DE) and 20 mN tare load was applied to remove any slack. The displacement rate was 2 mm/min and the samples were stretched to 10% strain. The experiments took less than 1 minute to complete so no bathing solution was used; the thickness was measured before and after the experiments and no significant dehydration was observed (ie, hydration variation during tests did not occur30,31). The stress-strain was plotted to compare the behavior of different groups. One-way analysis of variance with a significance level of .05 was used to statistically compare the experimental data.
Figure 2 shows the stress-strain response of anterior and posterior flaps. Comparing the behavior of anterior and posterior flaps from the control group shows that anterior corneal flaps had a stiffer tensile response compared to that of the posterior ones (P < .05). The CXL treatment increased the tensile properties only for the anterior flaps (P < .05), but had an insignificant effect on the biomechanical properties of the posterior flaps. Figure 3 reports the maximum tensile stress and tangent modulus of flaps from treated and control groups. For anterior groups, both the stress and tangent modulus increased significantly after CXL therapy (P < .05). Figure 4 compares the mechanical response of full-thickness corneal samples cross-linked from the epithelium (CXL-Top) or endothelium (CXL-Bottom) side. The CXL procedure significantly improved the mechanical properties of full-thickness porcine corneas only when these samples were treated from the epithelium side.
Tensile stress-strain behavior of anterior and posterior flaps excised from porcine cornea. Anterior flaps showed a stiffer response than posterior flaps. Furthermore, CXL enhanced the biomechanical properties of only the anterior flaps, but had no significant effect on the tensile behavior of the posterior flaps. The abbreviations CXL and CTR denote treated (corneal cross-linked) and control groups, respectively. Note that the curves representing posterior-CTR and posterior-CXL groups overlap.
Maximum tensile stress and tangent modulus of strips excised from the posterior and anterior region. A significant amount of stiffening was observed in anterior flaps (P < .05) but posterior flaps were not stiffened by CXL (P > .90). Furthermore, anterior flaps showed a significantly stronger tensile properties than the posterior flaps (P < .05). The abbreviations CXL and CTR denote treated (corneal cross-linked) and control groups, respectively.
Tensile stress-strain behavior of full thickness porcine corneas that were cross-linked from the epithelial (top) or endothelial (bottom) side (see Figures 1D–1E). CXL improved the tensile properties only when it was performed from the epithelium (top) side. The abbreviations CXL and CTR denote treated (corneal cross-linked) and control groups, respectively.
Our research shows that CXL significantly improved the mechanical properties of anterior flaps but had little effect on the tensile properties of posterior flaps. It was also found that anterior flaps had much stiffer tensile properties compared to posterior flaps (Figure 2), which is in agreement with previous studies. Randleman et al.32 found that the anterior stroma had significantly higher cohesive tensile strength than the posterior stroma. Scarcelli et al.33 used Brillouin optical microscopy to show that the anterior portion of the stroma has the highest elastic modulus in the cornea. Indentation techniques were also used to show that the Young's modulus of the anterior stroma was significantly larger than that of the posterior stroma.34,35 Furthermore, Kohlhaas et al.19 reached the same conclusion by running uniaxial tensile experiments on posterior and anterior flaps. These previous studies captured the depth-dependent corneal mechanical properties; nevertheless, actual values vary from one study to another because of different species, experimental protocols, and techniques that have been used.
The cornea shows depth-dependent tensile properties because of its depth-dependent microstructure. The corneal stroma is composed of collagen fibrils embedded in a proteoglycan matrix. The collagen fibrils are organized into sheet-like lamellae, which are stacked parallel to the surface of the cornea. The arrangement of the collagen lamellae changes through the thickness and anterior lamellae interweave markedly more than in posterior ones.15,17 This inhomogeneous architecture of the corneal stroma affects its biomechanics such that anterior layers show much higher elastic modulus than the posterior portion of the stroma.19,32–35
Previous studies have exclusively focused on characterizing the stiffening effect of CXL when full-thickness corneas were used, which replicates what is usually done in clinics. Moreover, the Dresden protocol has been designed such that it primarily affects the anterior 300 μm of the cornea to avoid UV light damage to endothelial cells. Thus, it is natural to expect the stiffening effect of CXL to also be depth dependent. Various studies in the literature confirm this conclusion. For example, Brillouin microscopy of CXL samples showed that the anterior portion of the stroma accounted for most of the mechanical stiffening.20 Moreover, mechanical tests such as uniaxial tension and indention testing on samples excised from anterior and posterior cross-linked stroma showed CXL caused a significant increase in the anterior stroma stiffness but an insignificant change in the posterior stroma stiffness.19,34 Indirect methods have also been used to reach the same conclusion.8,22 For instance, it was found that keratocyte apoptosis was primarily located in the anterior stroma when the usual surface irradiance of 3 mW/cm2 was used.8 These previous studies discussed the depth-dependent CXL primarily in terms of the absorption behavior of the riboflavin-treated cornea for UV-A. Kohlhaas et al.19 showed that approximately 70% of UV-A irradiation was absorbed within the anterior part of the cornea. Furthermore, Søndergaard et al.18 determined the riboflavin distribution in the corneal stroma and concluded that riboflavin uptake is limited to the anterior layers independent of the concentration and application time of the riboflavin solution.
The primary and novel finding of the current study was that, even when the posterior stroma was soaked in riboflavin and cross-linked in isolation, the tensile properties of the posterior flap were not improved by CXL (Figure 2). Furthermore, CXL significantly improved the mechanical properties of full-thickness porcine corneas only when they were treated from the anterior side (Figure 4). Zhang et al.36 showed that CXL creates cross-links between collagen molecules themselves and core proteins of the proteoglycans. Nevertheless, strong cross-links between collagen and proteoglycan core proteins have not been observed. Furthermore, Hayes et al.4 suggested that cross-links should mainly occur at the surface of collagen fibrils and in the proteoglycan network surrounding them. Thus, although future studies are required to fully understand why CXL did not stiffen posterior flaps and full-thickness corneas cross-linked from the endothelium side, we propose that the distinctive differences between the collagen lamella organization and proteoglycan distribution in the anterior and posterior stroma are the reason.
CXL of posterior lamellae is not common. Nevertheless, it might be considered as an option for strengthening donor tissues in vitro.37 Furthermore, previous studies tried to combine CXL with other refractive surgical procedures. For example, LASIK is a refractive laser surgery in which the surgeon severs the anterior portion of the cornea to alter central corneal curvature. This procedure could reduce corneal biomechanical stability because it alters the microstructure of the tissue by cutting and ablation of collagen lamellae. One possible complication of LASIK is corneal ectasia.38 It has been suggested that CXL treatment could increase the stiffness of the remaining cornea.23,24 The results presented here suggest that extra care should be taken in such approaches because CXL seems to have a limited effect on posterior layers. Future studies could possibly introduce the thickness of residual stromal bed as the inclusion criteria for such therapeutic options. In other words, it is envisioned that a thin LASIK flap might leave some anterior stroma that could be cross-linked, whereas a thick flap, associated with a higher laser ablation, most likely would not have any biomechanical advantage of CXL. Furthermore, many studies on different aspects of the CXL treatment have been done using normal cornea, primarily from animals and to a lesser degree from human donors. It is important to realize that the effect of this treatment procedure may be different on diseased tissue. It has been shown that keratoconus affects the corneal collagen microstructure and the proteoglycan content.39,40 The main difference between posterior and anterior layers is their ultrastructure and composition. In our study, there was an insignificant effect of CXL on the posterior flaps. Thus, we propose that modifications of the Dresden protocol, especially in diseased states such as keratoconus and corneas with ectasia after LASIK, should not generally be implemented in clinics without careful laboratory studies. In other words, surgeons should be careful in practicing such ideas on their patients without rigorous in vitro studies on keratoconic corneas.
The current study provided enough evidence to conclude that CXL had an insignificant effect on biomechanical properties of posterior layers of the porcine cornea. Thus, attempts to cross-link the posterior cornea using current technology will not affect its tensile strength. Because these layers were directly cross-linked, this peculiar observation should not be due to insufficient riboflavin uptake. This peculiar observation should not be due to the insufficient riboflavin uptake because these layers were directly cross-linked. Instead, we believe that it is because of the clear differences in the microstructure of the collagen lamellae and the specifications of collagen fibrils and proteoglycans in posterior and anterior regions. We are currently investigating this hypothesis and will present our findings in future publications.
- Andreassen TT, Simonsen AH, Oxlund H. Biomechanical properties of keratoconus and normal corneas. Exp Eye Res. 1980;31:435–441. doi:10.1016/S0014-4835(80)80027-3 [CrossRef]
- Wollensak G, Spoerl E, Seiler T. Stress-strain measurements of human and porcine corneas after riboflavin–ultraviolet-A-induced cross-linking. J Cataract Refract Surg. 2003;29:1780–1785. doi:10.1016/S0886-3350(03)00407-3 [CrossRef]
- Spoerl E, Huhle M, Seiler T. Induction of cross-links in corneal tissue. Exp Eye Res. 1998;66:97–103. doi:10.1006/exer.1997.0410 [CrossRef]
- Hayes S, Kamma-Lorger CS, Boote C, et al. The effect of riboflavin/UVA collagen cross-linking therapy on the structure and hydrodynamic behaviour of the ungulate and rabbit corneal stroma. PLoS One. 2013;8:e52860. doi:10.1371/journal.pone.0052860 [CrossRef]
- Bottós KM, Dreyfuss JL, Regatieri CV, et al. Immunofluorescence confocal microscopy of porcine corneas following collagen cross-linking treatment with riboflavin and ultraviolet A. J Refract Surg. 2008;24:S715–S719.
- Spoerl E, Mrochen M, Sliney D, Trokel S, Seiler T. Safety of UVA-riboflavin cross-linking of the cornea. Cornea. 2007;26:385–389. doi:10.1097/ICO.0b013e3180334f78 [CrossRef]
- Wollensak G, Aurich H, Pham D-T, Wirbelauer C. Hydration behavior of porcine cornea crosslinked with riboflavin and ultraviolet A. J Cataract Refract Surg. 2007;33:516–521. doi:10.1016/j.jcrs.2006.11.015 [CrossRef]
- Wollensak G, Spoerl E, Wilsch M, Seiler T. Keratocyte apoptosis after corneal collagen cross-linking using riboflavin/UVA treatment. Cornea. 2004;23:43–49. doi:10.1097/00003226-200401000-00008 [CrossRef]
- Hashemi H, Miraftab M, Seyedian MA, et al. Long-term results of an accelerated corneal cross-linking protocol (18 mW/cm2) for the treatment of progressive keratoconus. Am J Ophthalmol. 2015;160:1164–1170. doi:10.1016/j.ajo.2015.08.027 [CrossRef]
- Vinciguerra P, Rechichi M, Rosetta P, et al. High fluence iontophoretic corneal collagen cross-linking: in vivo OCT imaging of riboflavin penetration. J Refract Surg. 2013;29:376–377. doi:10.3928/1081597X-20130509-01 [CrossRef]
- Raiskup F, Pinelli R, Spoerl E. Riboflavin osmolar modification for transepithelial corneal cross-linking. Curr Eye Res. 2012;37:234–238. doi:10.3109/02713683.2011.637656 [CrossRef]
- Hafezi F, Mrochen M, Iseli HP, Seiler T. Collagen crosslinking with ultraviolet-A and hypoosmolar riboflavin solution in thin corneas. J Cataract Refract Surg. 2009;35:621–624. doi:10.1016/j.jcrs.2008.10.060 [CrossRef]
- Hatami-Marbini H, Jayaram SM. Effect of UVA/riboflavin collagen crosslinking on biomechanics of artificially swollen corneas. Invest Ophthalmol Vis Sci. 2018;59:764–770. doi:10.1167/iovs.17-22814 [CrossRef]
- Mazzotta C, Traversi C, Caragiuli S, Rechichi M. Pulsed vs continuous light accelerated corneal collagen crosslinking: in vivo qualitative investigation by confocal microscopy and corneal OCT. Eye (Lond). 2014;28:1179–1183. doi:10.1038/eye.2014.163 [CrossRef]
- Maurice DM. The cornea and sclera. In: Davson H, ed. The Eye. Orlando: Academic Press; 1984:1–158.
- Maurice DM. The structure and transparency of the cornea. J Physiol. 1957;136:263–286.261. doi:10.1113/jphysiol.1957.sp005758 [CrossRef]
- Meek KM. The cornea and sclera. In: Fratzl P, ed. Collagen: Structure and Mechanics. Boston: Springer US; 2008:359–396. doi:10.1007/978-0-387-73906-9_13 [CrossRef]
- Søndergaard AP, Hjortdal J, Breitenbach T, Ivarsen A. Corneal distribution of riboflavin prior to collagen cross-linking. Curr Eye Res. 2010;35:116–121. doi:10.3109/02713680903431847 [CrossRef]
- Kohlhaas M, Spoerl E, Schilde T, Unger G, Wittig C, Pillunat LE. Biomechanical evidence of the distribution of cross-links in corneas treated with riboflavin and ultraviolet A light. J Cataract Refract Surg. 2006;32:279–283. doi:10.1016/j.jcrs.2005.12.092 [CrossRef]
- Scarcelli G, Kling S, Quijano E, Pineda R, Marcos S, Yun SH. Brillouin microscopy of collagen crosslinking: noncontact depth-dependent analysis of corneal elastic modulus. Invest Ophthalmol Vis Sci. 2013;54:1418–1425. doi:10.1167/iovs.12-11387 [CrossRef]
- Schumacher S, Mrochen M, Wernli J, Bueeler M, Seiler T. Optimization model for UV-riboflavin corneal cross-linking. Invest Ophthalmol Vis Sci. 2012;53:762–769. doi:10.1167/iovs.11-8059 [CrossRef]
- Seiler T, Hafezi F. Corneal cross-linking-induced stromal demarcation line. Cornea. 2006;25:1057–1059. doi:10.1097/01.ico.0000225720.38748.58 [CrossRef]
- Richoz O, Mavrakanas N, Pajic B, Hafezi F. Corneal collagen cross-linking for ectasia after LASIK and photorefractive keratectomy: long-term results. Ophthalmology. 2013;120:1354–1359. doi:10.1016/j.ophtha.2012.12.027 [CrossRef]
- Wu Y, Tian L, Wang L-Q, Huang Y-F. Efficacy and safety of LASIK combined with accelerated corneal collagen cross-linking for myopia: six-month study. BioMed Research International. 2016;2016:5083069. doi:10.1155/2016/5083069 [CrossRef]
- Medeiros CS, Giacomin NT, Bueno RL, Ghanem RC, Moraes HV, Santhiago MR. Accelerated corneal collagen crosslinking: technique, efficacy, safety, and applications. J Cataract Refract Surg. 2016;42:1826–1835. doi:10.1016/j.jcrs.2016.11.028 [CrossRef]
- Hatami-Marbini H, Rahimi A. Effects of bathing solution on tensile properties of the cornea. Exp Eye Res. 2014;120:103–108. doi:10.1016/j.exer.2013.11.017 [CrossRef]
- Hatami-Marbini H, Rahimi A. The relation between hydration and mechanical behavior of bovine cornea in tension. J Mech Behav Biomed Mat. 2014;36:90–97. doi:10.1016/j.jmbbm.2014.03.011 [CrossRef]
- Hatami-Marbini H, Rahimi A. Interrelation of hydration, collagen cross-linking treatment, and biomechanical properties of the cornea. Curr Eye Res. 2016;41:616–622.
- Hatami-Marbini H, Rahimi A. Collagen cross-linking treatment effects on corneal dynamic biomechanical properties. Exp Eye Res. 2015;135:88–92. doi:10.1016/j.exer.2015.04.005 [CrossRef]
- Hatami-Marbini H. Hydration dependent viscoelastic tensile behavior of cornea. Ann Biomed Eng. 2014;42:1740–1748. doi:10.1007/s10439-014-0996-6 [CrossRef]
- Hatami-Marbini H, Etebu E. Hydration dependent biomechanical properties of the corneal stroma. Exp Eye Res. 2013;116:47–54. doi:10.1016/j.exer.2013.07.016 [CrossRef]
- Randleman JB, Grossniklaus HE, Dawson DG, McCarey BE, Edelhauser HF. Depth-dependent cohesive tensile strength in human donor corneas: implications for refractive surgery. J Refract Surg. 2008;24:S85–S89.
- Scarcelli G, Pineda R, Yun SH. Brillouin optical microscopy for corneal biomechanics. Invest Ophthalmol Vis Sci. 2012;53:185–190. doi:10.1167/iovs.11-8281 [CrossRef]
- Dias JM, Ziebarth NM. Anterior and posterior corneal stroma elasticity assessed using nanoindentation. Exp Eye Res. 2013;115:41–46. doi:10.1016/j.exer.2013.06.004 [CrossRef]
- Winkler M, Chai D, Kriling S, et al. Nonlinear optical macroscopic assessment of 3-D corneal collagen organization and axial biomechanics. Invest Ophthalmol Vis Sci. 2011;52:8818–8827. doi:10.1167/iovs.11-8070 [CrossRef]
- Zhang Y, Conrad AH, Conrad GW. Effects of ultraviolet-A and riboflavin on the interaction of collagen and proteoglycans during corneal cross-linking. The Journal of Biological Chemistry. 2011;286:13011–13022. doi:10.1074/jbc.M110.169813 [CrossRef]
- Arafat SN, Robert M-C, Shukla AN, Dohlman CH, Chodosh J, Ciolino JB. UV crosslinking of donor corneas confers resistance to keratolysis. Cornea. 2014;33:955–959. doi:10.1097/ICO.0000000000000185 [CrossRef]
- Randleman JB, Russell B, Ward MA, Thompson KP, Stulting RD. Risk factors and prognosis for corneal ectasia after LASIK. Ophthalmology. 2003;110:267–275. doi:10.1016/S0161-6420(02)01727-X [CrossRef]
- Meek KM, Tuft SJ, Huang Y, et al. Changes in collagen orientation and distribution in keratoconus corneas. Invest Ophthalmol Vis Sci. 2005;46:1948–1956. doi:10.1167/iovs.04-1253 [CrossRef]
- Akhtar S, Bron AJ, Salvi SM, Hawksworth NR, Tuft SJ, Meek KM. Ultrastructural analysis of collagen fibrils and proteoglycans in keratoconus. Acta Ophthalmol. 2008;86:764–772. doi:10.1111/j.1755-3768.2007.01142.x [CrossRef]
|Treatment target||Porcine corneas|
|Fluence (total) (J/cm2)||5.4|
|Soak time and interval (minutes)||30|
|Treatment time (minutes)||30|
|Light source||Custom made|
|Irradiation mode (interval)||Continuous|
|Abbreviation in the manuscript||CXL|