Corneal collagen cross-linking (CXL) has become a widely accepted alternative for the treatment of progressive corneal ectasia in the last decade. The classic Dresden protocol involves corneal impregnation with the chromophore riboflavin (RF) for 30 minutes, followed by 30 minutes of photosensitization by ultraviolet A (UVA) irradiation, resulting in significant corneal stiffening.1,2 In an attempt to reduce total treatment time and patient discomfort, several adaptations to the Dresden protocol have been suggested.3,4 High-fluency cross-linking protocols reduce treatment time by increasing radiation intensity according to the Bunsen-Roscoe law, limiting cumulative delivered energy bellow 5.4 J/cm2 to prevent endothelial cell damage.2,5 However, controversy exists about the efficacy of such protocols, as the effect is variable and the treatment depth is reported to be lower than that of the classic Dresden protocol.6–10
Recently, our group described a novel method for corneal stiffening using the bacteriochlorophyll derivative WST11 in a dextran solution [WST11 with dextran (WST-D)] and near-infrared light (NIR), using parameters similar to the Dresden protocol.11 Reduction of the total treatment time could benefit patients and reduce treatment costs. Therefore, in this study, we examined the efficacy of shortened protocols for WST-D/NIR corneal stiffening. We compared 30, 5, and 1 minute of NIR irradiation at a constant irradiance of 10 mW/cm2, in in vivo and ex vivo New Zealand White (NZW) rabbit models.
One hundred four corneas of NZW rabbits were included in this study and treated ex vivo (n = 34) or in vivo (n = 18). Treated corneas of eyes in both groups were deepithelialized mechanically, impregnated with WST11-dextran T500 (WST-D) for 20 minutes as described by Marcovich et al11 and irradiated with NIR light at 10 mW/cm2 using a diode laser at 755 nm for either 30, 5, or 1 minute (Fig. 1). Corneoscleral strips were cut from all eyes and used for biomechanical testing.
Ex Vivo Rabbit Model
Paired whole globes of 3-month-old NZW rabbits were obtained from a local abattoir within hours after enucleation and transported on ice in a humid container until further processing. One eye of each pair was randomly chosen, and allocated to one of the 3 study groups. Full mechanical deepithelialization of corneas in the treatment group, including the limbus, was manually performed and microscopically confirmed before impregnation with WST-D. Contralateral corneas, serving as controls, were left untouched until processing into strips and were not deepithelialized, similar to the methodology used in the in vivo model.
In Vivo Rabbit Model
NZW rabbits were housed with ad libitum access to food and water at the Core Animal Facility of the Weizmann Institute of Science (Rehovot, Israel). All experimental procedures were approved by the Institutional Animal Care and Use Committee, in adherence to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.12 At the age of 3 months, rabbits were anesthetized by intramuscular injection of 5 mg/kg xylazine (Vitamed, Binyamina, Israel) and 35 mg/kg ketamine (Rhone Merieux, Lyon, France), topped off with additional ketamine if longer anesthesia was needed. Corneas of the left eyes, allocated to the treatment group, underwent mechanical deepithelialization up to ∼2 mm of the limbus under microscopic confirmation. Right eyes, serving as paired controls, were left untouched except for eyelid closure during the entire procedure, in accordance with ARVO regulation.12 After treatment, ophthalmic ointment containing dexamethasone 0.1%, neomycin and polymyxin B (Maxitrol; Alcon, Puurs, Belgium) was applied to the treated eyes once daily, until full epithelial closure was confirmed by fluorescent dye and slit-lamp examination. Four weeks after treatment, the rabbits were killed using intravenous injection of pentobarbital sodium (CTS Chemical Industries Ltd, Kiryat Malakhi, Israel).
WST-D was prepared from WST11 (Steba Laboratories Ltd, Rehovot, Israel) at a concentration of 2.5 mg/mL in double-distilled water containing 20% 500-kDa dextran (Leuconostoc spp. Mr 450,000–650,000; Sigma-Aldrich, St. Louis) and 0.9% NaCl, with pH adjusted to 7.2–7.3 with sodium hydrogen carbonate (Merck, Darmstadt, Germany), if needed. Corneas in the treatment group were impregnated for 20 minutes using a cup filled with approximately 1 mL of WST-D placed on top of the deepithelialized cornea. After 20 minutes, remaining WST-D was removed, and the corneas were slightly rinsed with a few milliliters of Dulbecco's phosphate-buffered saline (Biological Industries, Beit Haemek, Israel).
A NIR diode laser with tunable output of up to 1 W at 755 nm (CeramOptec, Bonn, Germany) was used to irradiate the impregnated corneas. Irradiance of 10 mW/cm2 at the corneal apex was confirmed using a power meter (NOVEL; Ophir Optronics Ltd, Jerusalem, Israel). Directly after impregnation, corneas were placed under the NIR laser for 30, 5, or 1 minute. In the 30-minute group, a drop of Dulbecco's phosphate-buffered saline (ex vivo) or Tears Naturale Free (Alcon, Fort Worth, TX) (in vivo) was applied every 5 minutes during irradiation, to maintain corneal hydration.
Biomechanical Testing of Corneal Stiffness
Central corneal thickness (CCT) was measured using an ultrasound pachymeter (Humphrey ultrasonic pachymeter, Humphrey Instruments, San Leandro, CA), before and after treatment. For strip extensiometry, 4 ± 0.2 mm wide strips were cut from all corneas in the superior–inferior direction, with 2 to 3 mm of sclera on both ends. By placing the corneoscleral buttons on paraffin hemispheres matching the corneal curvature, and using a double-bladed cutter, the strips were cut without stretching the tissue. Samples were stored separately, for less than 3 hours, in humid Eppendorf containers on ice before extensiometry testing. All strips were placed horizontally between the clamps of a microcomputer-controlled extensiometer with a 200-N load cell (Minimat; Rheometric Scientific GmbH, Bensheim, Germany) at a gauge length of 6 mm, set at a deformation speed of 5 mm/min. The clamps were tightened with a controlled maximal force of 9 cN·m using a calibrated screwdriver (Torqueleader, Surrey, United Kingdom). Stress is the applied force normalized by the cross-sectional area before testing, and strain is the displacement normalized by the gauge length in percentage. The obtained stress–strain curves were considered to consist of 3 different regions as is common to biological tissue (Fig. 2).13 It was hypothesized that region 1 reflects elongation of the sample without stretching followed by uncrimping of collagen (toe region), region 2 represents load bearing of collagen fibers (linear region), and region 3 indicates breakage of interfibrillar bonds due to increased shear stress between collagen fibers, before complete rupture of the sample at the end of region 3 (failure region). Using a self-written MATLAB script and graphical user interface (MATLAB R2015b, MathWorks, Inc, Natick, MA), the borders of a linear section within region 2 were manually chosen 3 times, in a masked and random-sample order. A linear fit was applied, the slope of which represents the tangent elastic modulus, often called the Young modulus in the ophthalmic literature. The average of 3 values obtained from each stress–strain curve was used for further statistical analysis.
Sample Size and Statistical Analysis
The sample size was chosen based on the previously published literature on WST-D/NIR corneal stiffening, indicating a mean difference between control and treated corneas of 13.9 and 20.3 MPa for an ex vivo and in vivo rabbit model, respectively, with a standard deviation of 8.3 MPa.11 This yielded a sample size of 8 and 4 corneas for the ex vivo and in vivo models, respectively, with 90% power and 0.05 probability of a type I error. The sample size was increased by 50% to account for a possibly decreased treatment effect after shorter irradiation times.
Statistical analysis was performed by the Bioinformatics and Biological Computing department at the Weizmann Institute of Science, using Statistica, version 12 (StatSoft, Inc, Tulsa, OK). Results of both ex vivo and in vivo biomechanical experiments and corneal pachymetry were analyzed using repeated-measures analysis of variance (ANOVA), with treatment (treatment vs. control) as the within-subject effect, and irradiation time as the between-subject effect. The mean tangent elastic moduli and pachymetry data of control and treated corneas in each of the 3 irradiation time groups were compared using dependent Student t tests. The level of statistical significance was set at 0.05 for all analyses.
Ex Vivo Biomechanics
One pair in the 5-minute irradiation group was excluded from further analysis because of a software error during measurement of the control eye. The tangent elastic moduli for the ex vivo irradiation time groups are shown in Figure 3 and Table 1. The mean elastic moduli after WST-D/NIR treatment were significantly higher in all 3 groups. The mean difference between treated corneas and controls measured 8.46 [95% confidence interval (CI, 6.61–10.31), n = 11], 10.25 [95% CI (7.58–12.91), n = 11], and 7.95 [95% CI (4.40–11.51), n = 11], in the 30-, 5-, and 1-minute groups, respectively (P < 0.001 for all comparisons).
Factorial ANOVA showed a significant treatment effect (F1,30 = 152.26, P < 0.001). The interaction between treatment and irradiation time was not significant (F2,30 = 0.93, P = 0.4).
In Vivo Biomechanics
Of the 18 pairs treated in vivo, 3 (1 pair from each irradiation time group) were excluded from biomechanical testing, leaving 5 pairs in each group to be analyzed. Pairwise exclusion was due to problems with 4 different corneas—difficulties reading central corneal pachymetry (1 control cornea in the 5-minute group) or failure to cut a 4-mm-wide parallel strip (paired control and treated corneas in the 1-minute group and 1 treated cornea in the 30-minute group). Figure 3 and Table 2 show the tangent elastic moduli for the in vivo groups. One month after WST-D/NIR treatment, the mean elastic moduli increased significantly in the 30- and 5-minute groups, with mean differences between treated and control corneas of 6.63 MPa [95% CI (2.9–10.3), n = 5, P < 0.01], and 5.74 MPa [95% CI (4.6–6.8), n = 5, P < 0.001], respectively. In the 1-minute group, the mean elastic modulus of the treated corneas was not significantly higher than that of the paired controls, with a mean difference of 0.85 MPa [95% CI (−4.85 to 6.55), n = 5, P = 0.7].
Factorial ANOVA showed a significant treatment effect (F1,12 = 28.33, P < 0.001). The interaction between treatment effect and irradiation time was significant (F2,12 = 4.71, P = 0.031).
No significant differences in CCT were seen between control and treated eyes before treatment ex vivo and in vivo and before biomechanical measurement in vivo (Table 3). Factorial ANOVA showed no significant difference in CCT between the irradiation time groups in vivo at baseline (F1,2 = 0.161, P = 0.854), whereas ex vivo, there was a significant difference (F1,2 = 86.23, P < 0.001).
This study shows that WST-D/NIR treatment provides significant corneal stiffening, even with a reduced irradiation time of only 1 minute ex vivo and 5 minutes in vivo, without increasing irradiance. A comparable and significant stiffening effect of WST-D/NIR treatment was observed in all but the 1-minute in vivo group (Fig. 3). Shortening the irradiation time is not only beneficial in terms of patient comfort but also makes WST-D/NIR corneal stiffening more suitable for children and less cooperative patients.
When the stiffening effect is expressed percentually, as is common in the literature, a large difference will be noted between the ex vivo and in vivo models (mean increase ex vivo: 108.6%, 214.4%, 131.4%; in vivo: 70.5%, 50.6%, 7.5%, after 30, 5, or 1 minute of irradiation, respectively). We believe that the in vivo results are to be considered as true values, as treatment and measurement were performed under stable and physiological circumstances. Ex vivo eyes show a trend in stiffness after treatment, which is in line with the in vivo effect. The ex vivo treatment effect is most likely skewed because of variability that exists between abattoir rabbits, which are not bred and kept for research purposes. This emphasizes the importance of conducting in vivo experiments. Besides the differences between ex vivo and in vivo results, we observed a decrease in the elastic modulus after WST-D/NIR treatment in 1 cornea in the 1-minute irradiation in vivo group, compared with its paired control. After reviewing our records carefully, we do not have an explanation for this unexpected finding, which may be related to the inherent variability associated with strip extensiometry.
In agreement with the main objective of this study, namely to establish the efficacy of shortened WST-D/NIR treatment protocols, and to keep animal numbers to a minimum, in conjunction with ethical guidelines, we did not include an RF/UVA control group in this study. The NZW rabbit and extensiometry models are accepted in corneal CXL research, allowing indirect comparison with the literature. Wollensak et al evaluated corneal stiffening in NZW rabbits after RF/CXL and reported a 79% increase in the elastic modulus at both 1 day and 3 months after treatment. However, the treatment effect at 3 months was compared with a control group measured immediately after treatment.14 Such a comparison neglects age-related stiffening in the treatment group during follow-up, resulting in overestimation of the treatment effect. Taking into consideration variation in strip extensiometry readings and age-related corneal stiffening in the control group of our study, we conclude that the stiffening after WST-D/NIR treatment (30 and 5 minutes of NIR irradiation) is comparable to RF/UVA CXL after 30 minutes of UVA irradiation.
NIR light causes no intrinsic damage to cells by itself and is, within ranges, considered safe for patients and the environment by the International Commission of Non-Ionizing Radiation Protection (ICNIRP).15 NIR can however cause thermal damage to ocular structures, in particular to the retina, at high irradiance and/or exposure time. The ICNIRP considers NIR irradiance at 10 mW/cm2, as used in this study, to be safe to the cornea and lens for a duration longer than 1000 seconds.15 For shorter exposure times, the thermal effect decreases exponentially, and irradiance safety limits follow the formula 1.8·t −3/4 W/cm2 (t in seconds).16 For 5 and 1 minute of irradiation, this would mean a safety threshold of 25 and 83 mW/cm2, respectively. The irradiance applied in our current protocols can thus be considered safe, without the need for absorbance of light before reaching the endothelium, as is the case with toxic UVA light.
In this study, NIR irradiation time was shortened, whereas irradiance remained constant at 10 mW/cm2, resulting in a decrease in irradiation dose from 18 to 3 J/cm2 and 0.6 J/cm2 in the 30-, 5-, and 1-minute groups, respectively. Our results show a nonlinear positive correlation between irradiation time and corneal stiffening with a nonproportional increase in stiffening, reaching a plateau phase as the irradiation time is increased, suggesting saturation of the effect (ie, cross-linking bond formation). To our knowledge, RF/UVA CXL studies focusing on reducing irradiation time have only done so while proportionally increasing irradiance, to maintain the same total delivered energy of 5.4 J/cm2. This is suggested to be the maximal delivered energy that does not induce endothelial toxicity.2 These so-called “high-fluency” protocols show a similar nonlinear positive correlation between the irradiation time and biomechanical effect.17,18 Hammer et al concluded that oxygen depletion is likely the limiting factor in the rate of cross-link formation with such protocols, and they stated that a higher irradiance accompanying a shorter irradiation time depletes oxygen faster, resulting in a decreased stiffening effect.17 The current study shows that 5-minute irradiation at 10 mW/cm2, instead of an expected 60 mW/cm2 according to RF/UVA high-fluency protocol logic, results in an elastic modulus of approximately 15 MPa, similar to the stiffening effect observed after 30 minutes of irradiance at 10 mW/cm2. Future studies are needed to determine whether shortened WST-D/NIR protocols with linearly increased irradiance can provide a greater stiffening effect compared with the shortened constant fluency protocol with 5 minutes of NIR irradiation presented in this study.
Clinically, WST-D/NIR treatment could provide a safe alternative to RF/UVA CXL, particularly for patients with a CCT below 400 μm, for whom RF/UVA is considered unsafe. The safe nature of NIR light and the ability to prevent WST from reaching the endothelium by controlling the dextran concentration in the WST-D solution ensure safe treatment regardless of corneal thickness. Theoretically, WST-D/NIR could even allow targeted local treatment of selected (thin) regions in the cornea, which may prove sufficient or even preferable to arrest corneal ectasia.19 In a previous study of our group, endothelial viability after WST-D/NIR treatment was shown by the TUNEL assay.11 Also, edema was not observed in any of the WST-D/NIR-treated corneas, at any time point, and the CCT before biomechanical testing was similar for treated and control eyes, indicating a functional corneal endothelium. Because of the safe nature of NIR light, WST-D/NIR treatment potentially has a superior safety profile for both patients and care providers compared with RF/UVA CXL.
Although a larger sample size may have resulted in a different overall effect in the 1-minute irradiation in vivo group, sample sizes were kept small in the in vivo groups in adherence to the ARVO statement for the Use of Animals in Ophthalmic and Vision Research.12 Nonetheless, this study established a proof of concept for the efficacy of shortened WST-D/NIR protocols for corneal stiffening.
In summary, in this study, we have shown that WST-D/NIR corneal stiffening can be effectively achieved using a significantly shortened protocol, both ex vivo and in vivo. Photoexcitation using safe NIR light overcomes the major limitations of RF/UVA CXL, related to the potential toxicity of UVA light. Importantly, corneas thinner than 400 μm, as often is the case with progressive corneal ectasia, may benefit from WST-D/NIR treatment. A significantly shortened treatment protocol, as presented in this study, may reduce patient discomfort.
The authors thank Dr Ron Rotkopf of the Bioinformatics and Biological Computing Department at the Weizmann Institute of Science for his assistance and consultation concerning statistical analyses for this study.
1. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol. 2003;135:620–627.
2. Wollensak G, Spoerl E, Wilsch M, et al. Endothelial cell damage after riboflavin-ultraviolet-A treatment in the rabbit. J Cataract Refract Surg. 2003;29:1786–1790.
3. Krueger RR, Herekar S, Spoerl E. First proposed efficacy study of high versus standard irradiance and fractionated riboflavin/ultraviolet A cross-linking with equivalent energy exposure. Eye Contact Lens. 2014;40:353–357.
4. Lytle G. Advances in the technology of corneal cross-linking for keratoconus. Eye Contact Lens. 2014;40:358–364.
5. Bunsen RW, Roscoe HE. Photochemical Researches–Part V. On the measurement of the chemical action of direct and diffuse sunlight. Proc R Soc Lond. 1862;12:306–312.
6. Touboul D, Efron N, Smadja D, et al. Corneal confocal microscopy following conventional, transepithelial, and accelerated corneal collagen cross-linking procedures for keratoconus. J Refract Surg. 2012;28:769–776.
7. Ozgurhan EB, Sezgin Akcay BI, Yildirim Y, et al. Evaluation of corneal stromal demarcation line after two different protocols of accelerated corneal collagen cross-linking procedures using anterior segment optical coherence tomography and confocal microscopy. J Ophthalmol. 2014;2014:981893.
8. Cingü AK, Sogutlu-Sari E, Cınar Y, et al. Transient corneal endothelial changes following accelerated collagen cross-linking for the treatment of progressive keratoconus. Cutan Ocul Toxicol. 2014;33:127–131.
9. Kanellopoulos AJ. Long term results of a prospective randomized bilateral eye comparison trial of higher fluence, shorter duration ultraviolet A radiation, and riboflavin collagen cross linking for progressive keratoconus. Clin Ophthalmol. 2012;6:97–101.
10. Cınar Y, Cingü AK, Turkcu FM, et al. Accelerated corneal collagen cross-linking for progressive keratoconus. Cutan Ocul Toxicol. 2014;33:168–171.
11. Marcovich AL, Brandis A, Daphna O, et al. Stiffening of rabbit corneas by the bacteriochlorophyll derivative WST11 using near infrared light. Invest Ophthalmol Vis Sci. 2012;53:6378–6388.
12. Association for Research in Vision and Ophthalmology. Statement for the use of animals in ophthalmic and visual research. 2016. Available at: http://http://www.arvo.org
/about_arvo/policies/statement_for_the_use_of_animals_in_ophthalmic_and_visual_research/. Accessed February 18, 2017.
13. Fratzl P, Misof K, Zizak I, et al. Fibrillar structure and mechanical properties of collagen. J Struct Biol. 1998;122:119–122.
14. Wollensak G, Iomdina E. Long-term biomechanical properties of rabbit cornea after photodynamic collagen crosslinking. Acta Ophthalmol. 2009;87:48–51.
15. ICNIRP (International Commission on Non-ionizing Radiation Protection). Guidelines on limits of exposure to broad-band incoherent optical radiation (0.38 to 3 microM). International Commission on Non-Ionizing Radiation Protection. Health Phys. 1997;73:539–554.
16. Kourkoumelis N, Tzaphlidou M. Eye safety related to near infrared radiation exposure to biometric devices. ScientificWorldJournal. 2011;11:520–528.
17. Hammer A, Richoz O, Arba Mosquera S, et al. Corneal biomechanical properties at different corneal cross-linking (CXL) irradiances. Invest Ophthalmol Vis Sci. 2014;55:2881–2884.
18. Wernli J, Schumacher S, Spoerl E, et al. The efficacy of corneal cross-linking shows a sudden decrease with very high intensity UV light and short treatment time. Invest Ophthalmol Vis Sci. 2013;54:1176–1180.
19. Scarcelli G, Besner S, Pineda R, et al. Biomechanical characterization of keratoconus corneas ex vivo with Brillouin microscopy. Invest Ophthalmol Vis Sci. 2014;55:4490–4495.
Keywords:Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
cross-linking; photosensitizing agents; WST-D; near-infrared light; keratoconus