Access to the PDF text
Service d'aide à la décision clinique

Free Article !

Journal Français d'Ophtalmologie
Volume 37, n° 5
pages 371-376 (mai 2014)
Doi : 10.1016/j.jfo.2013.11.012
Received : 31 July 2013 ;  accepted : 22 November 2013
Comparison of transepithelial corneal collagen crosslinking with epithelium-off crosslinking in progressive keratoconus
Comparaison entre le cross-linking du collagène cornéen transépithélial et le cross-linking conventionnel dans le kératocône évolutif

I. Kocak a, A. Aydin a, b, , F. Kaya a, H. Koc c
a Service d’ophtalmologie, hôpital de Nisa, Istanbul, Turkey 
b Clinique universitaire d’ophtalmologie, faculté de médecine, université d’Istanbul Medipol, Istanbul, Turkey 
c Service d’ophtalmologie, hôpital de Kent, Kutahya, Turkey 

Corresponding author. Özel Nisa Hastanesi, Göz Hastalıkları Servisi, Fatih Caddesi, Yenibosna, Turkey.

To evaluate the safety and efficacy of transepithelial corneal collagen crosslinking (TE-CXL) as compared to epithelium-off crosslinking (epi-off CXL) in progressive keratoconus.


Records of keratoconus patients treated with TE-CXL or epi-off CXL were reviewed retrospectively. Patients were included if they had at least 12months follow-up. Pre- and postoperative measurements of visual acuity, refractive errors, keratometry, corneal topography and pachymetry were assessed and compared.


There was no statistically significant difference between two groups at baseline in terms of demographic, refractive and corneal parameters. Mean maximum cone apex curvature (apical K) increased from 51.62±5. Eighty-four diopters (D) to 53.70±5.49 D in the TE-CXL group (n =17), and decreased from 52.02±4.07 D to 51.22±3.51 in the epi-off CXL group (n =19) at the end of the follow-up period. The difference between two groups was statistically significant (P =0.0002). An increase of1D in apical K was observed in two of 19 eyes (11%) in the epi-off CXL group, and 11 of 17 eyes (65%) in TE-CXL group at the last follow-up visit, compared to baseline (P <0.0001). Fourteen patients in the epi-off CXL group exhibited corneal edema that resolved without haze with topical corticosteroid treatment by 4months. No postoperative corneal edema was observed in TE-CXL group.


Although it is safe and well tolerated, TE-CXL does not effectively halt the progression of keratoconus. Epi-off CXL appears to be effective in stopping progression and even improves corneal parameters.

The full text of this article is available in PDF format.

Évaluer l’efficacité et l’innocuité du cross-linking du collagène cornéen transépithélial (TE-CXL) par rapport au cross-linking conventionnel (épi-off CXL) dans le traitement du kératocône évolutif.

Patients et méthodes

Les patients atteints de kératocône traités par TE-CXL ou épi-off CXL ont été revus rétrospectivement. Les patients ont été inclus s’ils avaient au moins 12 mois de suivi. Les mesures pré- et postopératoires de l’acuité visuelle, des erreurs de réfraction, de la kératométrie, de la topographie cornéenne et de la pachymétrie ont été évaluées et comparées.


Il n’y avait pas de différence statistiquement significative entre les paramètres préopératoires des deux groupes concernant les donnés démographiques, les données réfractives et topographiques cornéennes. La valeur moyenne de la courbure maximale du sommet du cône (apical K) a augmenté de 51,62±5,84 dioptres (D) à 53,70±5,49 D dans le groupe de TE-CXL (n =17) et diminué de 52,02±4,07 D à 51,22±3,51 D dans le groupe d’épi-off CXL (n =19) à la fin de la période de suivi. La différence entre les deux groupes était statistiquement significative (p =0,0002). Une augmentation supérieure à 1 dioptrie d’apical K a été observée dans 11 % (2/19) du groupe d’épi-off CXL et 65 % (11/17) du groupe de TE-CXL, lors de la dernière visite de suivi, par rapport aux mesures préopératoires (p <0,0001). Quatorze patients dans le groupe d’épi-off CXL ont présenté un œdème cornéen toujours résolutif avant 4 mois avec une corticothérapie topique. Il n’a pas été observé d’œdème cornéen postopératoire dans le groupe de TE-CXL.


Malgré une sécurité et une tolérance démontrées, le TE-CXL ne semble pas arrêter efficacement la progression du kératocône. Le cross-linking conventionnel est plus efficace dans la stabilisation du kératocône en améliorant même ses paramètres topographiques cornéens.

The full text of this article is available in PDF format.

Keywords : Keratoconus, Corneal collagen crosslinking, Transepithelial

Mots clés : Kératocône, Cross-linking du collagène cornéen, Transépithélial


Keratoconus is a non-inflammatory, progressive, bilateral disorder of the eye in which the cornea assumes a conical shape secondary to biomechanical changes [1]. This ectatic corneal disorder has an incidence of approximately 1/2000. Its onset is usually at puberty and progresses until the fourth decade of life, after which it typically stabilizes [2]. Progression of keratoconus was a challenging issue to handle, until the definition of corneal collagen crosslinking (CXL) which is a novel surgical procedure increasing the rigidity of the corneal tissue. It has been reported that CXL is effectively stopping the progression of corneal ectatic diseases including keratoconus by creating extra crosslinks in the corneal stroma [3, 4, 5, 6].

CXL promotes the photopolymerization of collagen fibers induced by ultraviolet A (UVA) light in the presence of riboflavin as a photosensitizing substance. The effect of CXL is based on the augmentation of the number of intrafibrillar and interfibrillar covalent bonds. Since riboflavin cannot easily penetrate intact corneal epithelium, removal of epithelium is necessary for the classic CXL procedure (epi-off CXL). However, the removal of epithelium can cause severe postoperative pain and temporary visual blurring. To avoid these problems, some authors designed transepithelial CXL (TE-CXL) based on the use of a special riboflavin solution which can penetrate the intact epithelium [7]. In this study we aimed to evaluate and compare the efficacy and safety of epi-off CXL and TE-CXL procedures.


Subjects enrolled in the study were consecutive keratoconus patients who underwent TE- or epi-off CXL at the Nisa Hospital, Istanbul, Turkey, from January 2009 to July 2012. The study was conducted in accordance with the Declaration of Helsinki and approved by the institutional ethics committee. Informed consent was signed by all patients. We retrospectively reviewed the data of patients with a diagnosis of keratoconus who underwent TE- or epi-off CXL. Inclusion criteria were presence of keratoconus, clear cornea at slit-lamp examination, central corneal thickness greater than 400μm, evidence of progression documented by an increase of at least 1.00 diopter (D) in the maximum cone apex curvature (Apical K) or an increase of at least 1.00 D in central corneal astigmatism over the previous 6 months. Exclusion criteria included apical scarring, active keratitis, any other anterior or posterior segment disorder, and concomitant autoimmune disease. Contact lens wear was discontinued one month before the initial examination in all patients.

All patients had a full ophthalmologic examination, including the objective refraction, uncorrected and best spectacle corrected Snellen visual acuity (UCVA and BSCVA), slit lamp examination, posterior segment examination, intraocular pressure measurement, keratometry, pachymetry and corneal topography (Orbscan II, Bausch & Lomb, Rochester, NY, USA), at the preoperative and postoperative visits. Postoperative follow-up visits were scheduled at the 1st, 4th,7th, 30th day and the 3rd, 6th, 12th month after surgery, and then yearly. Patients with at least 12months follow-up were included in the study.

Surgical procedure

Epi-off CXL was performed under topical anesthesia. Proparacaine 0,5% was administered every 5minutes beginning 20minutes before surgery. Miosis was induced with pilocarpine 2% before surgery. Standard preoperative preparation with 5% povidone iodine was performed. The eyelids and eyelashes were covered with a sterile drape. After mechanical debridement of corneal epithelium over the central 9–10mm, 0,1% Riboflavin in 20% dextran solution (Ricrolin, Sooft, Montegiorgio, Italy) was instilled topically every 2minutes for 30minutes. The cornea was exposed to UVA light of 366–374nm at an irradiance of 3,0 mW/cm2 for 30minutes (Vega CBM-X-Linker, CSO, Florence, Italy). Meanwhile riboflavin instillation was continued every 2minutes. At the end of the procedure, antibiotic and corticosteroid drops were administered and a therapeutic soft contact lens was placed on the cornea. Contact lens was removed following epithelial healing. Antibiotic drops were continued for 1 week. Corticosteroid drops were gradually tapered after 1 week, and stopped at postoperative first month follow-up visit.

Same surgical procedures were performed for TE-CXL, but the corneal epithelium was not removed. Besides a special solution of riboflavin 0,1% in 15% dextran, containing riboflavin penetration enhancers such as sodium EDTA, tromethamine, and bihydrate sodium phosphate (Ricrolin TE, Sooft, Montegiorgio, Italy) was used. Outcome measures at baseline and postoperative follow-up visits were reviewed from patient charts and prepared for statistical analysis.

Statistical analysis

Intra-group comparison of multiple related samples was performed with Friedman ANOVA test. Comparison of related two samples was performed with paired two-sample t -test in case of normal distribution, and with Wilcoxon matched pairs test in the absence of normal distribution. Inter-group comparisons were performed with t -test in case of normal distribution, and Mann-Whitney U test was used if there is not normal distribution. Two tailed distribution outcomes were accepted for P values. All analyses were performed with Statplus software (Analysoft, USA). P values<0.05 were considered to be statistically significant.


Thirty-six eyes of 36 patients were included in the study. Characteristics of patients before surgery are shown in Table 1. Seventeen eyes were recruited in TE-CXL group, and 19 in epi-off CXL. There was no statistically significant difference between two groups at baseline in terms of demographic, refractive and corneal parameters (Table 1).

Refractive and corneal parameters of TE-CXL group at the baseline and during postoperative follow-up are presented in Table 2. Statistically significant deterioration was observed in mean keratometric value (K), maximum keratometric value (Kmax), maximum cone apex curvature (apical K) and corneal central thickness (CCT) measurements (P =0.0023, 0.0025, 0.0001 and 0.0042, respectively). There was no statistically significant change in BSCVA, spherical equivalent of refractive error, and minimum keratometric value (Kmin) measurements (P =0.5, 0.16 and 0.06, respectively). A slight improvement was observed in central corneal astigmatism and irregularity index of central 3mm (P =0.014 and 0.0091).

In epi-off CXL group, statistically significant improvement was observed in mean K, Kmax, apical K, and central corneal astigmatism (P <0.0001, P =0.004, 0.001, 0.003, respectively). There was no a statistically significant change in UCVA, BSCVA, spherical equivalent, Kmin, and irregularity index of central 3mm as shown in Table 3 (P =0.07, 0.14, 0.4, 0.1, 0.07, respectively).

Postoperative changes of corneal and refractive parameters were calculated using the difference between the measurements of last follow-up visit and baseline. The comparison of postoperative changes between two groups is presented in Table 4. There was no statistically significant difference between postoperative changes of two groups, concerning BSCVA, spherical equivalent, irregularity index of central 3mm, and CCT. However, epi-off CXL group showed statistically significantly better outcome in postoperative changes of corneal parameters such as mean K, apical K, central corneal astigmatism (P =0.01, 0.0002, and 0.03, respectively).

An increase of1D in apical K was observed in two of 19 eyes (11%) in epi-off CXL group, and 11 of 17 eyes (65%) in TE-CXL group at the last follow-up visit, compared to baseline (P <0.0001). In epi-off CXL group, apical K decreased in 13 eyes, and remained unchanged in four eyes. In TE-CXL group, apical K decreased in four eyes, and remained unchanged in two eyes. No postoperative complication was observed in TE-CXL group. Fourteen patients of epi-off CXL group showed corneal edema that resolved without haze with topical corticosteroid treatment in 4months.


An intact corneal epithelium absorbs 30–33% of UVA irradiation, and with additional effect of riboflavin in the epithelium, this blockage rate reaches to approximately 85%. In addition to UVA blockage, an intact corneal epithelium avoids the riboflavin infiltration to corneal stroma, because of its hydrophilic macromolecular structure. It is reported that, in case of epithelial removal, stromal dose of riboflavin reaches a level of a hundred times of that when epithelium is intact [8, 9]. Due to these facts, corneal epithelium debridement is an indispensable part of the standard CXL procedure [4]. However, the removal of epithelium can cause some complications such as severe postoperative pain, temporary visual blurring, epithelial healing problems, haze, viral reactivation, even corneal melting. To prevent these serious problems, some researchers have performed a number of studies on a special riboflavin solution, which can penetrate the intact epithelium. In these studies, it has been reported that some substances such as benzalkonium chloride, EDTA, and trometamol may enhance the epithelial penetrance of hydrophilic macromolecules [10, 11, 12, 13, 14, 15, 16]. EDTA has an effect of breaking intercellular bonds, thus it facilitates epithelial permeability. Another material, topical anesthetic tetracaine 1%, is reported to be used for loosening epithelial tight junctions in order to facilitate epithelial permeability [17]. Based on these findings, transepithelial CXL has been defined for the treatment of progressive keratoconus without debridement of corneal epithelium [16, 17, 18, 19].

In the present study, we analyzed retrospectively the results of TE-CXL in 17 eyes and epi-off CXL in 19 eyes. In TE-CXL group, Mean K, Kmax, apical K, and CCT showed deterioration at the end of the follow-up time (mean±SD: 22,12±7,61months). However, in epi-off group, statistically significant improvement was observed in mean K, Kmax, apical K, and central corneal astigmatism (mean follow-up±SD: 17,47±6,79months). Mean CCT of epi-off CXL cases decreased at 3rd month of CXL, but then increased gradually and reached to baseline level finally (Table 3). This can be due to false-thinner measurement of Orbscan II in edematous corneas as we had a high rate of long lasting corneal edema in epi-off CXL cases. Orbscan pachymetric measurement is reported to give thinner results in case of an obscuring factor such as haze or edema [20]. In our study, ectatic progression continued after CXL in 11% of epi-off CXL cases, and 65% of TE-CXL cases.

There are controversial reports about the efficacy of TE-CXL in the literature. In a recent study, Koppen et al. [16] analyzed the efficacy of TE-CXL in 53 eyes of 38 patients with progressive keratoconus. They reported that TE-CXL was less effective than standard CXL in stabilizing progressive disease. Buzzonetti et al. [21] evaluated the results of TE-CXL in young patients under 18years, and they concluded that this technique did not effectively halt keratoconus progression in children. However, Filipello et al. [18] reported that TE-CXL appeared to stop keratoconus progression with a significant improvement in visual and topographic parameters in a patient group of 20 keratoconus cases. Similarly, it has been reported that the improvements in topographical parameters provided by TE-CXL were comparable to those obtained after standard CXL in pediatric keratoconus, in the study of Magli et al. [22]. In another study, Spadea et al. [23] reported that TE-CXL seemed moderately effective in keratoconic eyes with ultrathin corneas (331 to 389μm).

In a confocal microscopy study, Touboul et al. [24] have compared the early corneal healing process following conventional, transepithelial, and accelerated CXL protocols. They have reported that the anterior corneal stroma showed significant changes following conventional and accelerated CXL in early postoperative period. However, the cornea appeared to be unaltered following TE-CXL in confocal microscopy. This finding may explain the reason for the inefficiency of TE-CXL that we observed in our study. Recently, a new riboflavin impregnation method has been described to increase the efficiency of TE-CXL. Bikbova et al. [25] have used iontophoresis to enhance the transepithelial penetration of riboflavin in 22 eyes of 19 patients with progressive keratoconus. They have reported that TE-CXL by iontophoresis appeared to halt progression with a statistically significant improvement in topographic parameters.

In conclusion, although TE-CXL is a safe and comfortable procedure, there are conflicting opinions about its efficacy. Based on our results, TE-CXL does not effectively halt the progression of keratoconus. However, this conclusion should be interpreted with caution due to the small number of patients, limited follow-up period, and potential biases related to the retrospective nature of our study.

Disclosure of interest

The authors declare that they have no conflict of interest concerning this article.

 Presented at the 47th Annual Congress of the Turkish Society of Ophthalmology, November 2013, Antalya, Turkey.


Feder R., Kshettry P. Non-inflammatory ectatic disorders Cornea Philadelphia: Elsevier Mosby (2005).  966
Rabinowitz Y.S. Keratoconus Surv Ophthalmol 1998 ;  42 : 297-319 [cross-ref]
Spörl E., Huhle M., Kasper M., Seiler T. Erhöhung der Festigkeit der Hornhaut durch Vernetzung[nl]Artificial stiffening of the cornea by induction of intrastromal cross-links Ophthalmologe 1997 ;  94 : 902-906
Spoerl E., Huhle M., Seiler T. Induction of cross-links in corneal tissue Exp Eye Res 1998 ;  66 : 97-103 [cross-ref]
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 [cross-ref]
Wollensak G., Spoerl E., Seiler T. Riboflavin/ultraviolet-A-induced collagen crosslinking for the treatment of keratoconus Am J Ophthalmol 2003 ;  135 : 620-627 [inter-ref]
Pinelli R. Corneal collagen crosslinking: is it necessary to remove epithelium? J Intraocular Implant Refract Soc India 2008 ;  4 : 28-34
Baiocchi S., Mazzotta C., Cerretani D., Caporossi T., Caporossi A. Corneal crosslinking: riboflavin concentration in corneal stroma exposed with or without epithelium J Cataract Refract Surg 2009 ;  35 : 893-899 [cross-ref]
Rathore M.S., Majumdar D.K. Effect of formulation factors on in vitro transcorneal permeation of gatifloxacin from aqueous drops AAPS PharmSciTech 2006 ;  7 : 57
Nakamura T., Yamada M., Teshima M., Nakashima M., To H., Ishikawa N., and al. Electrophysiological characterization of tight junctional pathway of rabbit cornea treated with ophthalmic ingredients Biol Pharm Bull 2007 ;  30 : 2360-2364 [cross-ref]
Uematsu M., Kumagami T., Kusano M., Yamada K., Mishima K., Fujimura K., and al. Acute corneal epithelial change after instillation of benzalkonium chloride evaluated using a newly developed in vivo corneal transepithelial electric resistance measurement method Ophthalmic Res 2007 ;  39 : 308-314 [cross-ref]
Chang S.W., Chi R.F., Wu C.C., Su M.J. Benzalkonium chloride and gentamicin cause a leak in corneal epithelial cell membrane Exp Eye Res 2000 ;  71 : 3-10 [cross-ref]
Majumdar S., Hippalgaonkar K., Repka M.A. Effect of chitosan, benzalkonium chloride, and EDTA on permeation of acyclovir across isolated rabbit cornea Int J Pharm 2008 ;  348 : 175-178 [cross-ref]
Pauly A., Meloni M., Brignole-Baudouin F., Warnet J.M., Baudouin C. Multiple endpoint analysis of the 3D-reconstituted corneal epithelium after treatment with benzalkonium chloride: early detection of toxic damage Invest Ophthalmol Vis Sci 2009 ;  50 : 1644-1652
Kikuchi T., Suzuki M., Kusai A., Iseki K., Sasaki H., Nakashima K. Mechanism of permeability-enhancing effect of EDTA and boric acid on the corneal penetration of 4-[1-hydroxy-1-methylethyl]-2-propyl-1-[4-[2-[tetrazole-5-yl]phenyl] phenyl] methylimidazole-5-carboxylic acid monohydrate (CS-088) Int J Pharm 2005 ;  299 : 107-114 [cross-ref]
Koppen C., Wouters K., Mathysen D., Rozema J., Tassignon M.J. Refractive and topographic results of benzalkonium chloride-assisted transepithelial crosslinking J Cataract Refract Surg 2012 ;  38 : 1000-1005 [cross-ref]
Hayes S., O’Brart D.P., Lamdin L.S., Doutch J., Samaras K., Marshall J. Effect of complete epithelial debridement before riboflavin-ultraviolet-A corneal collagen crosslinking therapy J Cataract Refract Surg 2008 ;  34 : 657-661 [cross-ref]
Filippello M., Stagni E., O’Brart D. Transepithelial corneal collagen crosslinking: bilateral study J Cataract Refract Surg 2012 ;  38 : 283-291 [cross-ref]
Leccisotti A., Islam T. Transepithelial corneal collagen crosslinking in keratoconus J Refract Surg 2010 ;  26 : 942-948 [cross-ref]
Altan-Yaycioglu R., Pelit A., Akova Y.A. Comparison of ultrasonic pachymetry with Orbscan in corneal haze Graefes Arch Clin Exp Ophthalmol 2007 ;  245 : 1759-1763 [cross-ref]
Buzzonetti L., Petrocelli G. Transepithelial corneal cross-linking in pediatric patients: early results J Refract Surg 2012 ;  28 : 763-767 [cross-ref]
Magli A., Forte R., Tortori A., Capasso L., Marsico G., Piozzi E. Epithelium-off corneal collagen cross-linking versus transepithelial cross-linking for pediatric keratoconus Cornea 2013 ;  32 : 597-601 [cross-ref]
Spadea L., Mencucci R. Transepithelial corneal collagen cross-linking in ultrathin keratoconic corneas Clin Ophthalmol 2012 ;  6 : 1785-1792 [cross-ref]
Touboul D., Efron N., Smadja D., Praud D., Malet F., Colin J. Corneal confocal microscopy following conventional, transepithelial, and accelerated corneal collagen cross-linking procedures for keratoconus J Refract Surg 2012 ;  28 : 769-776 [cross-ref]
Bikbova G., Bikbov M. Transepithelial corneal collagen cross-linking by iontophoresis of riboflavin Acta Ophthalmol 2014 ;  92 : e30-e34

© 2014  Elsevier Masson SAS. All Rights Reserved.
EM-CONSULTE.COM is registrered at the CNIL, déclaration n° 1286925.
As per the Law relating to information storage and personal integrity, you have the right to oppose (art 26 of that law), access (art 34 of that law) and rectify (art 36 of that law) your personal data. You may thus request that your data, should it be inaccurate, incomplete, unclear, outdated, not be used or stored, be corrected, clarified, updated or deleted.
Personal information regarding our website's visitors, including their identity, is confidential.
The owners of this website hereby guarantee to respect the legal confidentiality conditions, applicable in France, and not to disclose this data to third parties.
Article Outline