Share this article on:

Transepithelial Cross-Linking in Keratoconus Patients: Confocal Analysis

Filippello, Massimo; Stagni, Edoardo; Buccoliero, Donato; Bonfiglio, Vincenza; Avitabile, Teresio

doi: 10.1097/OPX.0b013e318269c8e5
Clinical Report

Purpose. To analyze the modifications of the corneal layers after transepithelial (TE) corneal cross-linking (CXL).

Methods. Twenty eyes of 20 patients with progressive keratoconus were treated with TE-CXL (without epithelium debridement), and their corneas analyzed at different times, up to 18 months, by confocal microscopy using a Nidek Confoscan 4.

Results. The epithelial layer, after an initial disorganization due to the TE technique, returned to a normal state after 2 weeks. Keratocytes and nerve fibers in the stroma decreased in number (approximately 25%), to return to a normal density at about 6 months after the procedure. A demarcation dense line indicative of effective cross-linking could be observed after 6 months at a depth of around 60 μm below the Bowman membrane. The endothelium remained unaffected at all examination times.

Conclusions. The TE-CXL procedure appears to be safe and effective at least up to 18 months. The absence of epithelium debridement (necessary in the classical technique) makes the procedure more comfortable for the patient, and allows the treatment of thinner corneas (between 360 and 400 μm) and problematic patients.



Clinica Di Stefano-Velona, Catania, Italy (MF, ES), and Clinica Oculistica II, Azienda Universitario-Ospedaliera “Policlinico-Vittorio Emanuele,” Catania, Italy (DB, VB, TA).

Received February 11, 2012; accepted June 6, 2012.

Massimo Filippello Clinica Di StefanoVelona Via S. Euplio 16, 95124 Catania, Italy e-mail:

The advent of corneal cross-linking (CXL) drastically improved the expectations and perspectives of patients afflicted by ectatic corneal conditions (keratoconus, pellucid marginal degeneration, and ectasia further to refractive surgery).

It has been widely demonstrated that CXL treatment increases corneal rigidity, thus helping to slow down or, sometimes, even stop the evolution of corneal ectasia.14 If left untreated, corneal thinning will result in a decreased visual acuity in the early ectatic phases and in alterations of corneal anatomy in the later phases, which can only be solved by cornea transplantation.5,6

Traditional CXL (“epi-off”) is done in three steps: denudation of the Bowman layer by debridement of corneal epithelium, to allow penetration of riboflavin; permeation of the corneal stroma by riboflavin; and irradiation of the permeated stroma by a ultraviolet (UV)-A laser source.7 Owing to the necessity of epithelium removal, traditional CXL can only be performed on corneas with a thickness >400 μm. To overcome this limitation and to improve patients' comfort during the procedure a transepithelial cross-linking (TE-CXL, “epi-on”) technique has been developed, based on the weakening of corneal epithelium tight junctions by a Tris–ethylenediaminetetraacetic acid (EDTA) solution, in which riboflavin is solubilized (Ricrolin-TE®, Sooft Italia). In this way, the small riboflavin molecule can penetrate into the corneal stroma through the paracellular way, without the need for epithelium removal.8 Indeed, it has already been shown that TE-CXL improves patient compliance during and after the procedure and may slow down their keratoconus progression.8

In this study, we have used confocal microscopy to analyze the morphological modifications of the corneal layers after the TE-CXL procedure.

Back to Top | Article Outline


Confocal observations were performed on a sample of 20 eyes belonging to 20 (14 male and 6 female) patients, aged between 14 and 42 years (Table 1), affected by progressing keratoconus (Krumeich stages II and III)9 and treated with TE-CXL. In all patients, the fellow eye was not treated.

The procedure involves the use of a silicone ring reservoir, 3 mm high and 12 mm wide, with a 0.3 mm flange with the double function of protecting the corneal limbus during the UV-A irradiation phase and stabilizing the ring in its position by the pressure exerted on it by the eyelids, so that a blepharostat is not required. The ring can be filled with 3 drops (15 μl) of TE riboflavin solution (Ricrolin-TE), which is left standing for 30 min over the cornea so that the penetration enhancer allows riboflavin permeation of the corneal stroma.8 After removal of the excess riboflavin from the corneal ring, the following irradiation phase uses the same UV device used for the traditional CXL (Vega, CSO, Italy),7 lasting 30 min, with the power set at 3 mW/cm2. During the pulsed irradiation phase,7 a drop of TE riboflavin is instilled every 3 to 5 min.8

Confocal microscopy analysis with Confoscan 4 (Nidek Technologies) was then performed in all patients after the instillation of 1 drop of Viscotears gel (carbopol 980, 0.2%; Medivis) and 1 drop of benoxinate (oxybuprocaine, 0.4%; Sandoz) in the conjunctival fornix (to prevent blinking) and on the tip head of the confocal instrument.

The confocal microscopy analysis was performed about 30 min after treatment, and was repeated 15 days and 1, 3, 6, 12, and 18 months after the procedure. All corneal layers were carefully examined at each recorded time point. Only relevant changes were reported at the time point they have been observed.

The examination was done in all patients using an “auto mode” setting, acquiring 350 images per eye and exploiting an accommodative target in a central position as well as with a shift upward, after flattening the instrument on the cornea. This procedure allowed analyzing the two lowest millimeters beneath the apex of the cone. Because the epithelial layer is often thinned out near the corneal apex, this area should be the preferred part for the passage of TE riboflavin and the most affected by morphological changes.

The endothelium was also inspected for morphological signs of suffering, and an endothelial cell count was performed.

This study followed the tenets of the Declaration of Helsinki. It has been performed with informed consent and following all the guidelines for experimental investigations required by the institutional review board and ethics committee of the institutions to which the authors are affiliated.

Back to Top | Article Outline


Thirty minutes after TE-CXL treatment, the epithelial sheet appeared thinner than before treatment, with hyper-reflective and exfoliating cells (Fig. 1A), whereas basal cells showed irregular edges (Fig. 1B). This picture is likely the consequence of the loss of the most superficial cell layers due to TE enhancer effects and to the thermal impact of the procedure.

Two weeks after treatment, owing to the high turnover of this tissue, the corneal epithelial sheet was again regular in density and morphology in all its cell layers, and the hyper-reflecting cells disappeared (Fig. 2).

Images of the subepithelial nervous plexus showed no particular anomalies soon after the procedure (30 min), although in most cases, it is difficult at this stage to visualize nerve fibers owing to the hyper-reflectivity of the reacting anterior corneal stroma (Fig. 3).

One month after the procedure, an approximate 25% reduction could be observed in the number of fibers of the subepithelial nervous plexus (Fig. 4A), most likely as a result of the short-term toxicity of the UV-A irradiation. The number and morphology of fibers returned to a normal state at about 6 months after the procedure (Fig. 4B).

In four patients, a curvilinear evolution of the subepithelial nervous plexus was evident 1 year after treatment, probably caused by the regeneration of nervous fibers in areas in which the cornea was irregularly irradiated during the TE-CXL procedure (Fig. 5).

Confocal analysis of the corneal stroma at 3 months after treatment showed an approximate 25% reduction in the number of keratocytes in the anterior stroma, likely due to induction of apoptosis by UV-A irradiation (Fig. 6A). Repopulation of keratocytes gradually occurred, being already evident after sixth months (Fig. 6B).

It is worthy of notice that the repopulation of keratocytes in the anterior stroma, at a depth of approximately 60 μm below Bowman layer, is associated, in almost all patients treated with TE-CXL, with a constant hyper-reflectivity of keratocyte nuclei (activated keratocytes), a possible reaction to an effective cross-linking procedure (Fig. 7).

Exploring deeper into the corneal stroma, at about 95 to 110 μm, it was possible to observe hyper-reflecting bands arranged approximately parallel to each other (also known as demarcation line), like “bridges” connecting the keratocytes of the anterior stroma (Fig. 8, arrow).

In all examinations performed up to the 12th month after the procedure, the endothelial layer was always intact, with endothelial cells showing the same number and morphology as before treatment (2388 ± 25 and 2354 ± 32 cells/mm2, respectively, before and 6 months after TE-CXL).

Back to Top | Article Outline


CXL is becoming the elective procedure to treat keratoconus, to stop its progression, and to improve the visual acuity of affected patients.10,11 The traditional technique requires debridement of the corneal epithelium to let the riboflavin solution permeate the corneal stroma, and then UV-A irradiation to activate riboflavin and trigger the biochemical reactions leading to the rearrangement of the stroma and the arrest of its degeneration.7 Follow-up of these patients has shown that the obtained results are stable up to 5 years.11 However, removal of the corneal epithelium is a highly distressing event, and delayed re-epithelization may occur, further complicating the healing process.12 Therefore, TE techniques that do not require epithelium debridement have been developed.13 These techniques rely on the presence of chemical agents in the riboflavin formulation that loosen the epithelial barrier, thus allowing penetration of riboflavin into the stroma in presence of the epithelium. Benzalkonium chloride used in one of the formulations13 achieves this effect through its toxic properties on epithelial cells.14,15 EDTA in trometamol (Tris buffer) is the epithelial enhancer used in our formulation.8 EDTA is a well-known chelator of divalent calcium and magnesium ions, which are important to keep cell to cell junctions.16,17 Thus, when the calcium and magnesium ion concentrations are lowered by EDTA, the cells loosen their contacts, and the epithelial sheet becomes permeable to riboflavin penetration8 in a non-toxic manner. The efficacy of this technique up to 18 months of observation has been demonstrated in a recent publication.8

In this pilot study, we used confocal microscopy to document the morphologic changes in the corneas of patients who underwent TE-CXL. It is known that the limitations of corneal confocal analysis are related to its poor repeatability, and include the fact that different fields are usually analyzed at different time points. However, in this study, we performed our analytical procedure using the Nidek Confoscan 4, which offers a good detailed view of corneal structures and a good degree of repeatability, owing to its specialized software that allows taking pictures of the same areas at the different time points.

The obtained results show that after an initial disarrangement of its superficial layers (30 min after UV-A treatment), the corneal epithelium returned to its original state already at 15 days after the procedure (Figs. 1 and 2). Such disarrangement could be the result of either the toxic effect of irradiation and riboflavin or the activity of the permeation enhancer, added to the formulation precisely for this purpose. Using confocal analysis, benzalkonium chloride which has also been used as TE enhancer,13 has shown dramatic disruptive effects on the corneal epithelial surface.18

The subepithelial nervous plexus remains visible throughout the observation times, and the initial shrinkage is then followed by a full regeneration of the fibers within the first 6 months, restoring the pretreatment state (Figs. 3 to 5). Thus, the TE procedure appears to be milder on the nerves than the epi-off procedure, which results instead in an immediate disappearance, owing to epithelial scraping, of subepithelial plexus and anterior mid-stromal nerve fibers.19,20

Keratocyte alterations are more evident in the anterior stroma. Although 30 min after treatment, the number of keratocytes was reduced owing to apoptosis, later examinations (starting at 6 months after the procedure) showed a repopulation of keratocytes with activated nuclei in the anterior stroma (Figs. 6 and 7). The presence of a hyper-reflective interkeratinocyte band (demarcation line) observed at a depth of about 110 μm (60 μm below the Bowman layer; Fig. 8) is an accepted indication that the cross-linking procedure has been effective.8,11 These findings partially differ from those observed after the epi-off procedure.19 In fact, if the timing of keratocyte repopulation remains similar in the two techniques, in the epi-off situation, keratocyte death is observed throughout a thicker stromal layer, up to a depth of 340 μm (including the epithelium),19 where a demarcation line becomes evident. Instead, in our epi-on procedure, keratocyte death was noted in the first third of the stromal layer, up to a depth of 110 μm, where the demarcation line was observed. It is indeed this limited penetration of TE-CXL that allows thinner corneas (between 360 and 400 μm) to be treated, posing no risk to the endothelium. This is likely due to the shielding effect of the epithelium during UV irradiation, which limits the penetration of UV-A rays, thus sparing the posterior part of the stroma. In fact, if the cornea is treated with TE riboflavin using the epi-on procedure, but irradiation is later performed after epithelium removal, the demarcation line occurs at a higher depth, at about 300 μm.21 It is worth noting that the anterior stroma has a higher concentration of keratocytes—which may secrete more collagen—than the posterior stroma,21 and therefore might contribute to most of the stabilization effect after the TE cross-linking procedure.

We detected no corneal haze that required a topical anti- inflammatory therapy, nor were damages to the endothelial layer observed (not shown), indicating that the procedure is safe.

Efficacy has been shown up to 18 months of observation, as evidenced by clinical improvement of the keratometric indexes of these TE-CXL–treated patients already reported in a previous publication.8 Further follow-up of these patients is in progress, to show efficacy also beyond 2 years of observation.

Last, but not least, because the method used in TE-CXL does not involve removal of the corneal epithelium, with its associated heavy discomfort, it could be safely performed in problematic patients and in the ambulatory setting.

Massimo Filippello

Clinica Di StefanoVelona

Via S. Euplio 16, 95124 Catania, Italy


Back to Top | Article Outline


MF and ES are part-time employees of the Sooft Company (Montegiorgio, Italy) that commercialize the transepithelial riboflavin (Ricrolin-TE) with the trometamol/EDTA enhancer. MF and ES are co-inventors and patent holders of this product. ES owns equity shares in Sooft Italia.

Back to Top | Article Outline


1. Spoerl E, Huhle M, Seiler T. Induction of cross-links in corneal tissue. Exp Eye Res 1998;66:97–103.
2. Spoerl E, Huhle M, Kasper M, Seiler T. [Increased rigidity of the cornea caused by intrastromal cross-linking.] Ophthalmologe 1997;94:902–6.
3. 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;29:1780–5.
4. Gkika M, Labiris G, Kozobolis V. Corneal collagen cross-linking using riboflavin and ultraviolet-A irradiation: a review of clinical and experimental studies. Int Ophthalmol 2011;31:309–19.
5. Kanellopoulos AJ, Binder PS. Collagen cross-linking (CCL) with sequential topography-guided PRK: a temporizing alternative for keratoconus to penetrating keratoplasty. Cornea 2007;26:891–5.
6. Ayres BD, Rapuano CJ. Excimer laser phototherapeutic keratectomy. Ocul Surf 2006;4:196–206.
7. Mazzotta C, Balestrazzi A, Traversi C, Baiocchi S, Caporossi T, Tommasi C, Caporossi A. Treatment of progressive keratoconus by riboflavin-UVA-induced cross-linking of corneal collagen: ultrastructural analysis by Heidelberg Retinal Tomograph II in vivo confocal microscopy in humans. Cornea 2007;26:390–7.
8. Filippello M, Stagni E, O'Brart D. Trans-epithelial corneal collagen cross-linking: a bilateral, prospective study. J Cataract Refract Surg 2012;38:283–91.
9. Alió JL, Shabayek MH. Corneal higher order aberrations: a method to grade keratoconus. J Refract Surg 2006;22:539–45.
10. Snibson GR. Collagen cross-linking: a new treatment paradigm in corneal disease—a review. Clin Experiment Ophthalmol 2010;38:141–53.
11. Caporossi A, Mazzotta C, Baiocchi S, Caporossi T. Long-term results of riboflavin ultraviolet a corneal collagen cross-linking for keratoconus in Italy: the Siena eye cross study. Am J Ophthalmol 2010;149:585–93.
12. 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.
13. Leccisotti A, Islam T. Transepithelial corneal collagen cross-linking in keratoconus. J Refract Surg 2010;26:942–8.
14. Chen W, Li Z, Hu J, Zhang Z, Chen L, Chen Y, Liu Z. Corneal alternations induced by topical application of benzalkonium chloride in rabbit. PLoS One 2011;6:e26103.
15. Kusano M, Uematsu M, Kumagami T, Sasaki H, Kitaoka T. Evaluation of acute corneal barrier change induced by topically applied preservatives using corneal transepithelial electric resistance in vivo. Cornea 2010;29:80–5.
16. Nakas M, Higashino S, Loewenstein WR. Uncoupling of an epithelial cell membrane junction by calcium-ion removal. Science 1966;151:89–91.
17. Scaletta LJ, MacCallum DK. A fine structural study of divalent cation-mediated epithelial union with connective tissue in human oral mucosa. Am J Anat 1972;133:431–53.
18. Mazzotta C, Traversi C, Baiocchi S, Caporossi O, Bovone C, Sparano MC, Balestrazzi A, Caporossi A. Corneal healing after riboflavin ultraviolet-A collagen cross-linking determined by confocal laser scanning microscopy in vivo: early and late modifications. Am J Ophthalmol 2008;146:527–33.
19. Liang H, Brignole-Baudouin F, Riancho L, Baudouin C. Reduced in vivo ocular surface toxicity with polyquad-preserved travoprost versus benzalkonium-preserved travoprost or latanoprost ophthalmic solutions. Ophthalmic Res 2012;48:89–101.
20. Al-Aqaba M, Calienno R, Fares U, Otri AM, Mastropasqua L, Nubile M, Dua HS. The effect of standard and transepithelial ultraviolet collagen cross-linking on human corneal nerves: an ex vivo study. Am J Ophthalmol 2012;153:258–66.
21. Ojeda JL, Ventosa JA, Piedra S. The three-dimensional microanatomy of the rabbit and human cornea. A chemical and mechanical microdissection-SEM approach. J Anat 2001;199:567–76.

keratoconus; cross-linking; transepithelial; riboflavin; confocal microscopy

© 2012 American Academy of Optometry