results of confocal microscopy examinations after collagen cross-linking with riboflavin and uva...

Fax +41 61 306 12 34E-Mail [email protected]

Original Paper

Ophthalmologica 2011;225:95–104 DOI: 10.1159/000319465

Results of Confocal Microscopy Examinations after Collagen Cross-Linking with Riboflavin and UVA Light in Patients with Progressive Keratoconus

Steffi Knappe Oliver Stachs Andrey Zhivov Marina Hovakimyan Rudolf Guthoff

Department of Ophthalmology, University of Rostock, Rostock , Germany

perreflective keratocyte nuclei. At about 6 months postop-eratively, the corneal stroma had virtually regained its nor-mal configuration. After therapy, confocal microscopy re-vealed that corneal endothelium was normal in terms of cell density and morphology at every time point. Conclusions:

Confocal in vivo laser scanning microscopy is an investiga-tive technique that permits reproducible visualization of structural changes in the cornea (epithelium, stroma and en-dothelium) following collagen cross-linking with riboflavin and UVA light. Once epithelial healing is complete, the epi-thelium and endothelium appear to be unaffected by the treatment. The most noteworthy structural changes, which are detected on confocal microscopy shortly after treat-ment, involve the anterior and middle corneal stroma. Over the course of time, up to 12 months postoperatively, these changes show a definite tendency to regress.

Copyright © 2010 S. Karger AG, Basel

Introduction

The term keratoconus, first coined by Friedrich von Ammon in the early decades of the 19th century [1] , de-scribes a noninflammatory degenerative disorder of the cornea. The condition is characterized by central thin-ning and cone-shaped bulging of the cornea [2, 3] .

Key Words

Keratoconus � Collagen cross-linking � Riboflavin � UVA irradiation � Activated keratocytes � In vivo confocal laser scanning microscopy

Abstract

Purpose: Keratoconus is a predominantly bilateral form of corneal degeneration that is associated with central thinning and cone-shaped bulging of the cornea usually accompa-nied by a progressive reduction in visual acuity. A recent therapeutic option is cross-linking, a procedure designed to prevent the progression of keratoconus by the photochemi-cal cross-linkage of collagen fibers. Patients and Methods: Eight eyes in 8 patients with progressive keratoconus were treated by the photochemical cross-linking method using ri-boflavin and UVA light. In addition to the usual ophthalmo-logical examinations, patients were examined pre- and post-operatively by confocal in vivo laser scanning microscopy. Follow-up examinations were performed at 2 weeks and at 2, 4, 6 and 12 months postoperatively. Results: Complete re-generation of corneal epithelium was detected by 2 weeks after therapy at the latest. The sub-basal nerve plexus could not be visualized by confocal microscopy after treatment. Immediately after treatment, the anterior corneal stroma had a honeycombed appearance but without the typical hy-

Received: June 2, 2009 Accepted after revision: January 22, 2010 Published online: September 24, 2010

Ophthalmologica

Steffi Knappe, MD Department of Ophthalmology, University of Rostock Doberaner Strasse 140 DE–18057 Rostock (Germany) Tel. +49 381 494 8581, Fax +49 381 494 8402, E-Mail SteffiKN74 @ gmx.de

© 2010 S. Karger AG, Basel0030–3755/11/2252–0095$38.00/0

Accessible online at:www.karger.com/oph

http://dx.doi.org/10.1159%2F000319465

Knappe/Stachs/Zhivov/Hovakimyan/Guthoff

Ophthalmologica 2011;225:95–10496

While the etiology of keratoconus remains unclear, the condition is known to have a genetic predisposition, and the inheritance pattern is autosomal dominant or au-tosomal recessive [4, 5] .

Enzyme defects are also thought to be present in the cornea, with increased expression of lysosomal and pro-teolytic enzymes and an imbalance of keratan sulphate and dermatan sulphate. It is further speculated that a dis-turbance of collagen molecules in terms of their degree of cross-linking is a key feature in the pathology [6] .

The incidence of keratoconus in the general popula-tion is estimated to be approximately 1 in 2,000 [7, 8] .

In the majority of cases, keratoconus becomes mani-fest bilaterally, often at the onset of puberty [9] . The earli-est symptoms often take the form of an increasing dete-rioration in visual acuity as a consequence of progressive myopia and/or increased astigmatism. Keratoconus has a variable prognosis. While the disease course may stabi-lize spontaneously, a rapid progression is noted in the ma-jority of cases [10, 11] .

To date, the treatment options available for keratoco-nus have included spectacle adjustment or fitting hard contact lenses [12] , insertion of intra-stromal plastic rings (known as intacs) [8] and lamellar or perforating kerato-plasty [8, 13] . The feature common to all these modalities is that they treat only the refractive and optical conse-quences of keratoconus rather than the underlying cause of the condition.

One therapeutic procedure that is designed to prevent or delay the progression of keratoconus is collagen cross-linking with riboflavin and UVA light [14–19] . This method is based on the photochemical cross-linking of collagen fibers in corneal stroma. The overall mechanical stability of the cornea is significantly increased by theresultant additional cross-linkages formed between col-lagen fibers, i.e. a corneal stiffening effect is produced without impairing corneal transparency in the process [20–24] .

In order to achieve the desired photo-oxidative effect, a corneal abrasion is produced followed by topical appli-cation of riboflavin and irradiation with UVA light.

Due to its high spatial resolution, confocal in vivo laser scanning microscopy is an investigative technique that enables corneal cell structures to be visualized rapidly and reproducibly with precise depth coordinates [25–29] .

The aim of the present study was to use confocal in vivo microscopy to analyze the time course of corneal structural changes following collagen cross-linking. A human cornea was investigated to assist the detailed in-terpretation of the confocal in vivo microscopy findings.

Patients and Methods

We performed collagen cross-linking with riboflavin and UVA irradiation on 8 eyes (8 patients; 2 women, 6 men; mean age: 33.6 8 13.9 years) with moderately pronounced and clinically detectable progressive keratoconus (keratoconus severity score 1–3) [30] . Patients were not eligible to participate if they already had advanced keratoconus with visible corneal scarring. Patients with a history of herpetic keratitis and corneal scarring, severe eye dryness, and current corneal infection or autoimmune disease were excluded.

Prior to collagen cross-linking, corneal thickness was at least 400 � m in all patients scheduled for treatment.

The following investigations were conducted preoperatively: best-corrected distant and near visual acuity, corneal topogra-phy, pachymetry (Pentacam, Oculus Optikgeräte GmbH, Wetz-lar, Germany), keratometry (Keratograph, Oculus Optikgeräte), aesthesiometry (Cochet-Bonnet Aesthesiometer L-12, Luneau, Chartres, France), slit-lamp and fundus examination, intraocu-lar pressure measurement and confocal in vivo laser scanning microscopy (Heidelberg Retina Tomograph II with the Rostock Cornea Module, HRT II + RCM, Heidelberg Engineering GmbH, Heidelberg, Germany). The same investigations were repeated at 2 weeks and at 2, 4, 6 and 12 months after cross-linking ther-apy.

The study was performed as part of a clinical research project approved by the Ethics Committee of the University of Rostock and was conducted in accordance with the latest revision of the Declaration of Helsinki.

Cross-Linking Treatment Cross-linking therapy was performed in an outpatient setting

under sterile conditions following local instillation anesthesia with proparacaine-POS 0.5% eye drops (Ursapharm GmbH & Co. KG, Saarbrücken, Germany). The cross-linking was performed according to the described protocol [21] .

Floxal � EDO � eye drops were prescribed 5 times daily during the postoperative period until contact lens removal after epithe-lial healing (approx. 5–7 days). When necessary, the patients also received paracetamol 500 mg tablets (Hexal AG, Holzkirchen, Germany) for analgesia. Contact lens removal was followed by topical therapy with Voltaren ophtha � sine AT (Novartis Pharma GmbH, Nürnberg, Germany) and Floxal EDO AT.

Confocal in vivo Laser Scanning Microscopy The confocal microscopy was performed as described previ-

ously [25–29] .

Histological Investigation In addition to confocal microscopy, we performed histological

investigations on one cross-linked human cornea and one human cornea that had not undergone cross-linking. The cross-linked cornea came from a keratoconus patient in whom collagen cross-linking therapy had been performed in the affected eye about 4 weeks prior to perforating keratoplasty. A non-cross-linked hu-man cornea from an enucleated choroid melanoma eye was used for the comparison of histological findings.

The corneas were briefly fixed in 4% buffered formalin at room temperature and processed for histology. After paraffinembedding, 4- � m sections were prepared, transferred to slides

Confocal Microscopy Examinations in Patients with Progressive Keratoconus

Ophthalmologica 2011;225:95–104 97

and routinely stained for hematoxylin/eosin and DAPI (4 � ,6-diamidino-2-phenylindole).

The HE- and DAPI-stained specimens were examined using a microscope (Leitz Aristoplan, Wetzlar, Germany) with a UV ex-citation filter for fluorescence microscopy.

Results

Clinical Over the 1-year period of the investigation, no adverse

effects or endothelial cell damage were noted in any of the patients treated.

Confocal in vivo Laser Scanning Microscopy The confocal in vivo microscopy images obtained in a

31-year-old patient may be seen as representative of the changes in corneal structure encountered in our popula-tion of keratoconus patients following collagen cross-linking therapy ( fig. 1–4 ).

Corneal Epithelium Preoperatively ( fig. 1 a, b), all layers of corneal epithe-

lium (superficial, intermediate and basal cells) displayed normal cell morphology.

a b

c d

e f

g h

Four

mon

ths

afte

r ope

ratio

nTw

o m

onth

saf

ter o

pera

tion

Two

wee

ksaf

ter o

pera

tion

Befo

re o

pera

tion

z = 10 μmz = 2 μm

z = 13 μmz = 4 μm

z = 16 μmz = 6 μm

z = 15 μmz = 7 μm

a b

c d

z = 43 μmz = 62 μm

z = 55 μmz = 52 μm

Fig. 1. Visualization of corneal epithelium by confocal in vivo mi-croscopy following collagen cross-linking in a 31-year-old male keratoconus patient. Distance from corneal surface (z-axis depth) is indicated on each image. Bar represents 50 � m. a , c , e , g Epi-thelial superficial cell layer. b , d , f , h Epithelial basal cell layer. Fig. 2. Confocal microscopy image of the junction region between Bowman’s membrane and the subbasal nerve plexus in a 31-year-old male keratoconus patient. Distance from corneal surface (z-axis depth) is indicated on each image. Bar represents 50 � m. a Before cross-linking, with branched and tortuous subbasal nerve fibers (arrows). b–d After cross-linking. b One month post-operatively. c Three months postoperatively (no evidence of nerve fibers). d Four months postoperatively, with evidence of nerve fi-bers (arrows). 1

2



Complete re-epithelialization had occurred in all pa-tients by the 2nd week after the intervention at the latest. Initially, compared with preoperative findings, confocal microscopy showed the epithelium to be characterized by cells of varying size and irregular shape ( fig. 1 c, d). Rapid normalization of cell morphology was observed over the further course of time ( fig. 1 e–h).

Subbasal Nerve Plexus – Bowman’s Membrane Preoperatively ( fig. 2 a, arrows), the subbasal nerve

plexus displayed the extremely tortuous and branched nerve fiber pattern typical of keratoconus.

Immediately after treatment and at 3 months postop-eratively ( fig. 2 b, c), no subbasal nerve plexus could be visualized in the central cornea. It was not until the 4th

Four

mon

ths

afte

r ope

ratio

nTw

o m

onth

saf

ter o

pera

tion

Two

wee

ksaf

ter o

pera

tion

Befo

re o

pera

tion

a b c d

e f g h

i j k l

m n o p

z = 143 μm z = 256 μm z = 450 μmz = 100 μm

z = 180 μm z = 248 μm z = 462 μmz = 85 μm

z = 168 μm z = 221 μm z = 457 μmz = 79 μm

z = 147 μm z = 253 μm z = 468 μmz = 90 μm

Fig. 3. Confocal images of corneal stroma at varying depths over time in a 31-year-old male keratoconus patient. Distance from corneal surface (z-axis depth) is indicated on each image. Bar represents 50 � m. a–d Before cross-linking. e–p After cross-linking.



postoperative month onwards ( fig. 2 d) that the first sub-basal nerve fibers again became visible on confocal mi-croscopy in the central cornea.

Corneal Stroma Preoperatively, characteristic hyperreflective kerato-

cyte nuclei were visualized in the anterior, middle and posterior corneal stroma ( fig. 3 a–d, 5 b). In addition, a re-ticular pattern of hypo-reflective micro-striae was some-times evident in the posterior corneal stroma ( fig. 3 d, ar-rows).

After collagen cross-linking, the most striking cor-neal changes were apparent in the corneal stroma. The

first follow-up confocal microscopy examination – per-formed after visible epithelial healing at about 2 weeks postoperatively – revealed a honeycombed or reticular hyperreflective stromal architecture, without visualiza-tion of the typical hyperreflective keratocyte nuclei ( fig. 3 e–g). This very striking ‘structural change’ was ev-ident throughout the anterior and middle corneal stro-ma, and it was noted that the hyperreflective character of this change diminished with increasing depth ( fig. 3 g). This stromal pattern was also still evident on confocal microscopy 2–3 months after treatment ( fig. 3 i–k, 5 b). From the 4th postoperative month onwards, the honey-combed plexus lost its hyperreflectivity, and the corneal

a b

c d

z = 467 μmz = 479 μm

z = 461 μmz = 458 μm

50 μm

50 μm10 μma b

Fig. 4. Images of corneal endothelium in a 31-year-old male keratoconus patient. Dis-tance from corneal surface (z-axis depth) is indicated on each image. Bar represents 50 � m. a Before cross-linking. b–d After cross-linking over time. b One month postoperatively. c Three months postop-eratively. d Six months postoperatively.

Fig. 5. Comparative images of the kerato-cyte network. a Electron micrograph of a normal cornea with keratocytes intercon-nected via their cell processes [48]. b Con-focal in vivo microscopy of the anterior corneal stroma 3 months after collagen cross-linking in a 31-year-old male kerato-conus patient; the image shows ‘activated keratocytes’ that form a network with their cellular processes; the intracellular sub-stance of the ‘activated keratocytes’ ap-pears hyperreflective.



stroma appeared to revert to its original state ( fig. 3 m–o). While the ‘plexus’ clearly lost its hyperreflectivity, the characteristic hyperreflective keratocyte nuclei ( fig. 3 n, o) reappeared, although they were reduced in number. Postoperatively, compared with the baseline preoperative findings ( fig. 3 d), the posterior corneal stroma was unchanged at each examination time point ( fig. 3 h, l, p).

Endothelium Corneal endothelial morphology ( fig. 4 a–d) was nor-

mally configured at each examination both preoperative-ly and postoperatively.

Histology Examination of the non-cross-linked cornea showed

visible keratocyte nuclei in all layers, i.e. in the anterior, middle and posterior stroma, both after HE staining and DAPI staining ( fig. 6 a, b). By comparison, it was particu-larly noteworthy in the cross-linked cornea that almost no keratocyte nuclei were detectable in the anterior and middle stroma at a depth of up to about 250–300 � m ( fig. 6 c, d). In the posterior stroma of the cross-linked cornea, the number and distribution of keratocyte nuclei were apparently consistent with those in the non-cross-linked cornea.

a

b

c

d

Fig. 6. Histological cross-section of a normal human cornea ( a , b ) and a cross-linked keratoconus cornea ( c , d ) following HE staining ( a , c ) and DAPI staining ( b , d ). a , b A normal cornea with four or five layers of epithe-lium with stained cell nuclei. Keratocyte nuclei are almost homogenously distributed throughout the entire (anterior, middle, posterior) stroma. The single layer of endothelium is intact throughout, and the endothelial cell nuclei are clearly stained. c , d A cornea with its intact four or five layers of epithelium with stained cell nu-clei. In contrast to the normal cornea, the anterior and middle corneal stroma (down to a depth of approxi-mately 250–300 � m) exhibits only very few keratocyte nuclei. Bar represents 50 � m.



Discussion

In vivo and in vitro experimental animal studies have shown that collagen cross-linking using riboflavin and UVA light produces an increase in biomechanical strength in the corneal stroma [17, 22] . The physicochem-ical properties of collagen tissue are modified in the pro-cess. One of the results of this treatment is to increase the diameter of collagen fibers in the anterior stroma by 12% and in the posterior stroma by 4.5% [19] .

The objective of collagen cross-linking is to reduce and/or arrest the progression of corneal ectasia. If this treatment is implemented in the early stage of the disease, the need for keratoplasty can be deferred or even avoided altogether. Because collagen cross-linking is a relatively rapid procedure that is low in complications and entails minimal inconvenience for the patient, it has assumed ever greater importance in the field of ophthalmology.

The depth to which corneal stroma is effectively treat-ed depends on the riboflavin concentration and on the intensity and duration of UVA irradiation. In routine practice, a riboflavin solution in a concentration of 0.1% and a light intensity of 3 mW/cm 2 are selected. Due to the absorption coefficient of riboflavin, the UVA effect is re-duced as stromal depth increases, in accordance with the Lambert-Beer law [31] ; in consequence, the cross-linking effect is not distributed homogenously throughout the thickness of the cornea. An investigation by Kohlhaas et al. [32] in enucleated porcine eyes has shown that the stiffening effect occurs only in the first third (ca. 200 � m) of the stroma.

In order for the desired therapeutic effect to occur, it is necessary prior to riboflavin application and UVA ir-radiation to produce an epithelial abrasion (ca. 8.0 mm in diameter); this is firstly because the epithelium would prevent sufficient riboflavin from entering the corneal stroma, and secondly because the epithelium enriched with riboflavin forms a protective barrier against UV light and absorbs a large proportion of the UVA energy [33] .

Corneal Epithelium Pain may develop postoperatively as a result of epithe-

lial abrasion; consequently, patients are provided with a therapeutic contact lens and, where necessary, with anal-gesics until epithelial healing is complete.

Complete epithelial regeneration was demonstrated by confocal in vivo microscopy 2 weeks postoperatively. It can be expected that the epithelium is closed already before the end of 2 weeks, but in this study no earlier con-

trol was done. Whereas the cells of the superficial, inter-mediate and basal layers appeared to be irregularly con-figured during the initial postoperative weeks, the cells rapidly regained their normal morphological appearance over the subsequent course of time. This finding suggests that the treatment has no effect on the morphology of the corneal epithelium after the regeneration phase.

Subbasal Nerve Plexus Confocal in vivo microscopy performed immediately

after treatment revealed that the subbasal nerve plexus with its branched and tortuous nerve fibers, which had been very easily identifiable preoperatively, could no lon-ger be visualized in the treated central corneal area. Only from the 3rd or 4th postoperative month onwards were isolated short nerve fiber structures again identifiable at the level of the subbasal nerve plexus, and these increased in number and size over the subsequent course of time. Because we had also investigated corneal sensitivity in our patients pre- and postoperatively, and because sensi-tivity was clearly reduced in all patients after treatment compared with preoperative baseline findings, it is likely that the subbasal nerve plexus in the treated area is de-stroyed as a result of therapy. The research team led by Mazzotta also arrived at this conclusion [34] . Neverthe-less, it is still unclear whether complete nerve fiber de-struction occurs as a result of mechanical epithelial abra-sion and of riboflavin/UVA treatment, or whether the nerve plexus simply cannot be visualized on confocal mi-croscopy during the initial postoperative months.

Corneal Stroma The most striking changes detected by confocal mi-

croscopy shortly after treatment involved the corneal stroma, and the anterior and middle stroma layers in par-ticular. The characteristic hyperreflective keratocyte nu-clei, such as those that were also detected in our kerato-conus patients before treatment, were replaced by large, strongly hyperreflective structures sometimes revealing a honeycombed/branched pattern. These changes exclu-sively involved the anterior and middle corneal stroma. When the appearance and distribution of these hyperre-flective structures were monitored during the subsequent postoperative period, it was found that they lost hyper-reflectivity as they ascended from deeper locations and became less prominent, whereas the characteristic hyper-reflective keratocyte nuclei returned. After 4 months, the corneal stromal seemed to return to its morphology base-line state because in the anterior stroma keratocyte nuclei were visible again. However, it was noteworthy that post-



operatively the number of keratocyte nuclei in the ante-rior stroma layers was clearly reduced. This finding may represent the evidence of postoperative keratocyte de-struction.

These postoperatively visible hyperreflective struc-tures in corneal stroma and their appearance on confocal microscopy have been reported previously, but their true significance and their pathological relevance following collagen cross-linking remain uncertain. This therefore raises questions as to the nature of the degenerative pro-cesses that unfold in the anterior stroma after therapy and as to how these hyperreflective structures should be interpreted. Similar findings have already been observed and described after refractive surgery procedures (e.g. LASIK, PRK) [34–36] .

It is known from previous histological studies in ex-perimental animals that riboflavin and UVA exposure in the treated corneal stroma initially lead to a loss of local keratocytes due to apoptosis down to a depth of 250–300 � m. Apoptotic cells display characteristic morphological criteria described by Kerr et al. [37] as long ago as 1972. These include loss of cell volume (cell shrinkage), and loss of contact with neighboring cells and with their mem-brane structures. The cell cytoplasm undergoes marked condensation. Over time, the cell membrane starts to pucker and so-called ‘apoptotic bodies’ develop [37] .

A study of our confocal microscopy images reveals lit-tle similarity with the described apoptotic cell changes. Their configuration most closely resembles that of a ker-atocyte. Keratocytes play an important role in maintain-ing corneal transparency and corneal stability and in the stroma of a healthy cornea they account for some 10–40% of stromal volume. In morphological terms, they are characterized by a flat, elongated shape comprising a cell body with numerous cell processes ( fig. 5 a), enabling them to interconnect in the stroma via gap junctions [38–40] . Examination of our confocal microscopy images suggests the presence of both cell bodies and multiple cell processes.

When a healthy cornea is viewed by confocal micros-copy, the cell borders and the cytoplasm of the keratocyte are not visible; only the cell nucleus is visualized as a hy-perreflective structure. The same confocal microscopy pattern was also found in our keratoconus patients prior to therapy with riboflavin and UVA. Because of the com-pletely altered postoperative appearance in the anterior and middle stroma, it may be assumed that collagen cross-linking exerts a reactive influence on the appear-ance of cellular components in the treated area. We sus-pect that the structures that we were able to identify by

confocal microscopy are ‘activated keratocytes’. Jester et al. [41] have suggested that keratocytes in an activated state modify the reflectivity of their cytoplasm. This change in cytoplasmic reflectivity is attributed to in-creased intracellular synthesis of proteins and enzymes, known as crystallins [41] . Although the precise signifi-cance of these crystallins is unknown, they appear to play a key role in tissue transparency. The synthesis of such crystallins increases the light-scattering properties of keratocytes, which means that on confocal laser-scan-ning microscopy the cell boundaries can be defined, and as a result the hyperreflective cell nucleus that is visible in the resting state can no longer be seen inside the cell.

Furthermore, activated keratocytes differ from non-activated keratocytes in terms of cell size. For example, an activated keratocyte can increase to approximately 10 times the size (or more) of its baseline state [42] . It is fur-ther known that an activated keratocyte is characterized by a prominent nucleolus and by a dilated endoplasmic reticulum.

If the activating stimulus is absent, an activated kera-tocyte reverts to its resting state, and its intracellular re-flectivity is diminished. This cellular behavior is mani-fested on in vivo CLSM in the reappearance of the cell nucleus. However, up until the 6th month postoperative-ly, there is a reduced number of keratocyte nuclei, a find-ing that is suggestive of keratocyte destruction. Previous histological studies have shown that just 24 h after ther-apy there is loss of keratocytes down as far as the deeper stromal layers [31, 43] . It is known that even simple epi-thelial abrasion leads to loss of keratocytes due to apop-tosis. However, this is limited to the anterior stroma with a maximal depth of 50–60 � m [44, 45] . After collagen cross-linking, the keratocyte destruction extends to far deeper stromal layers (down to 300 � m). This deeper-reaching cell destruction is explained by the fact that ri-boflavin – a photosensitizer – leads to an increase in UVA absorption, which in turn induces keratocyte apoptosis [31, 46] .

Histological Investigations To better understand our in vivo confocal microscopy

studies, we additionally conducted histological investiga-tions on one non-cross-linked and one cross-linked hu-man cornea. Approximately equal numbers of cell nuclei were visualized in the healthy cornea in the anterior, mid-dle and posterior stroma, whereas only very isolated ker-atocyte nuclei were found 4 weeks postoperatively in the treated, cross-linked cornea at a depth of approximately 250–300 � m. At the same time, the number and distribu-



tion of cell nuclei in the posterior stroma were roughly consistent with those in the untreated cornea. We did not anticipate this finding before starting the histological in-vestigations. If we assume that the hyper-reflective struc-tures, which we already noted on our confocal micros-copy images immediately after treatment, are ‘activated keratocytes’, then we would expect considerably more keratocytes at the treated deep stromal levels. These find-ings are most strongly suggestive of keratocyte loss, a phenomenon that has previously been reported by a number of authors [31, 43, 47] .

In conclusion, it is not possible at present to determine the nature of these highly unusual structures seen on confocal microscopy. In view of the high degree of mor-

phological similarity with keratocytes, we continue to be-lieve that they could be ‘activated keratocytes’. As to the future, further investigations, such as the assay of � -smooth muscle actin protein, may prove helpful in en-abling us to interpret our confocal microscopy findings even more accurately.

Acknowledgements

The authors wish to express their sincere thanks to Prof. Dr. Dr. med. Th. Seiler (Zürich) for supplying the cross-linked human cornea and to PD Dr. med. habil. F. Fankhauser (Dessau) for the loan of the UV-A irradiation lamp.

References

1 von Ammon FA: Klinische Darstellung der Krankheiten und Bildungsfehler des mensch-lichen Auges etc. 1838–1847; 4 Bde; Berlin.

2 Krachmer JH, Feder RS, Belin MW: Kerato-conus and related noninflammatory corneal thinning disorders. Surv Ophthalmol 1984; 28: 293–322.

3 Sharif KW, Casey TA, Coltart J: Prevalence of mitral valve prolapse in keratoconus pa-tients. J R Soc Med 1992; 85: 446–448.

4 Hammerstein W: Zur Genetik des Keratoko-nus. Albrecht von Graefes Arch Klin Exp Ophthalmol 1974; 190: 293–308.

5 Hallermann W, Wilson EJ: Genetische Be-trachtungen über den Keratokonus. Klin Monatsbl Augenheilkd 1977; 170: 906–908.

6 Cannon DJ, Foster CS: Collagen crosslink-ing in keratoconus. Invest Ophthalmol Vis Sci 1978; 17: 63–65.

7 Kennedy RH, Bourne WM, Dyer JA: A 48-year clinical and epidemiologic study of ker-atoconus. Am J Ophthalmol 1986; 101: 267–273.

8 Rabinowitz YS. Keratoconus. Surv Ophthal-mol 1998; 42: 297–319.

9 Lee LR, Hirst LW, Readshaw G: Clinical de-tection of unilateral keratoconus. Aust N Z J Ophthalmol 1995; 23: 129–133.

10 Zadnik K, Barr JT, Edrington TB, et al: Cor-neal scarring and vision in keratoconus: a baseline report from the Collaborative Lon-gitudinal Evaluation of Keratoconus (CLEK) Study. Cornea 2000; 19: 804–812.

11 Davis LJ, Schechtman KB, Wilson BS, et al: Longitudinal changes in visual acuity in ker-atoconus. Invest Ophthalmol Vis Sci 2006; 47: 489–500.

12 Garcia-Lledo M, Feinbaum C, Alio JL: Con-tact lens fitting in keratoconus. Compr Oph-thalmol Update 2006; 7: 47–52.

13 Tuft SJ, Moodaley LC, Gregory WM, et al: Prognostic factors for the progression of ker-atoconus. Ophthalmology 1994; 101: 439–447.

14 Wollensak J, Ihme A, Seiler T: Neue Befunde bei Keratokonus. Fortschr Ophthalmol 1987; 84: 28–32.

15 Spoerl E, Huhle M, Kasper M, Seiler T: Erhö-hung der Festigkeit der Hornhaut durch Quervernetzung. Ophthalmologe 1997; 94: 902–906.

16 Spoerl E, Huhle M, Seiler T: Induction of cross-links in corneal tissue. Exp Eye Res 1998; 66: 97–103.

17 Spoerl E, Schreiber J, Hellmund K, et al: Un-tersuchungen zur Verfestigung der Horn-haut am Kaninchen. Ophthalmologe 2000; 97: 203–206.

18 Wollensak G, Spörl E, Seiler T: Behandlung von Keratokonus durch Kollagenvernet-zung. Ophthalmologe 2003; 100: 44–49.

19 Wollensak G, Wilsch M, Spoerl E, Seiler T: Collagen fiber diameter in the rabbit cornea after collagen crosslinking by riboflavin/UVA. Cornea 2004; 23: 503–507.

20 Spoerl E, Seiler T: Techniques for stiffening the cornea. J Refract Surg 1999; 15: 711–713.

21 Wollensak G, Spoerl E, Seiler T: Riboflavin/ultraviolet-A-induced collagen cross-link-ing for the treatment of keratoconus. Am J Ophthalmol 2003; 135: 620–627.

22 Wollensak G, Spoerl E, Seiler T: Stress-strain measurements of human and porcine cor-nea after riboflavin/ultraviolet-A-induced crosslinking. J Cataract Refract Surg 2003; 29: 1780–1785.

23 Spoerl E, Wollensak G, Seiler T: Increased re-sistance of crosslinked cornea against enzy-matic digestion. Curr Eye Res 2004; 29: 35–40.

24 Spoerl E, Mrochen M, Sliney D, et al: Safety of UVA-riboflavin cross-linking of the cor-nea. Cornea 2007; 26: 385–389.

25 Stave J, Zinser G, Grümmer G, Guthoff R: Der modifizierte Heidelberg-Retina-Tomo-graph HRT. Erste Ergebnisse einer in-vivo-Darstellung von kornealen Strukturen. Ophthalmologe 2002; 99: 276–280.

26 Stachs O, Knappe S, Zhivov A, et al: 3D-kon-fokale Laser-Scanning-Mikroskopie derkornealen epithelialen Nervenstruktur. Klin Monatsbl Augenheilkd 2006; 223: 583–588.

27 Stachs O, Zhivov A, Kraak R, et al: In vivo three-dimensional confocal laser scanning microscopy of the epithelial nerve structure in the human cornea. Graefes Arch Clin Exp Ophthalmol 2007; 245: 569–575.

28 Zhivov A, Stachs O, Kraak R, et al: In vivo confocal microscopy of the ocular surface. Ocul Surf 2006; 4: 81–93.

29 Knappe S, Stachs O, Guthoff R: Korneale Veränderungen nach dem Tragen von Or-thokeratologie-Kontaktlinsen: eine Unter-suchung mit der konfokalen In-vivo-La-ser-Scanning-Mikroskopie. Ophthalmologe 2007; 104: 681–687.

30 McMahon T, Szczotka-Flynn L, Barr J, et al: A new method for grading the severity of keratoconus. The Keratoconus Servity Score (KSS). Cornea 2006; 25: 794–800.

31 Wollensak G, Spoerl E, Wilsch M, Seiler T: Keratocyte apoptosis after corneal collagen cross-linking using riboflavin/UVA treat-ment. Cornea 2004; 23: 43–49.

32 Kohlhaas M, Spoerl E, Schilde T, et al: Bio-mechanical evidence of the distribution of cross-links in corneas treated with ribofla-vin and ultraviolet A light. J Cataract Refract Surg 2006; 32: 279–283.



33 Podskochy A: Protective role of corneal epi-thelium against ultraviolet radiation dam-age. Acta Ophthalmol Scand 2004; 82: 714–717.

34 Mazzotta C, Balestrazzi A, Traversi C, et al: Treatment of progressive keratoconus by ri-boflavin-UVA-induced cross-linking of cor-neal collagen: ultrastructural analysis by Heidelberg Retinal Tomograph II in vivo confocal microscopy in humans. Cornea 2007; 26: 390–397.

35 Jester JV, Møller-Pedersen T, Huang J, et al: The cellular basis of corneal transparency: evidence for ‘corneal crystallins’. J Cell Sci 1999; 112: 613–622.

36 Erie JC, Patel SV, McLaren JW, et al: Kerato-cyte density in the human cornea after pho-torefractive keratectomy. Arch Ophthalmol 2003; 121: 770–776.

37 Kerr JF, Wyllie AH, Currie A: Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 1972; 26: 239–257.

38 Nishida T, Yasumoto K, Otori T, Desaki J: The network structure of corneal fibroblasts in the rat as revealed by scanning electron microscopy. Invest Ophthalmol Vis Sci 1988; 29: 1887–1890.

39 Müller LJ, Pels L, Vrensen GF: Novel aspects of the ultrastructural organization of human corneal keratocytes. Invest Ophthalmol Vis Sci 1995; 36: 2557–2567.

40 Hahnel C, Somodi S, Weiss DG, Guthoff RF: The keratocyte network of human cornea: a three-dimensional study using confocal la-ser scanning fluorescence microscopy. Cor-nea 2000; 19: 185–193.

41 Jester JV, Budge A, Fisher S, Huang J: Cor-neal keratocytes: phenotypic and species dif-ferences in abundant protein expression and in vitro light-scattering. Invest Ophthalmol Vis Sci 2005; 46: 2369–2378.

42 Ohno K, Mitooka K, Nelson LR, et al: Ker-atocyte activation and apoptosis in trans-planted human corneas in a xenograft mod-el. Invest Ophthalmol Vis Sci 2002; 43: 1025–1031.

43 Wollensak G, Iomdina E, Dittert DD, Herbst H: Wound healing in the rabbit cornea after corneal collagen cross-linking with ribofla-vin and UVA. Cornea 2007; 26: 600–605 .

44 Helena MC, Baerveldt F, Kim WJ, Wilson SE: Keratocyte apoptosis after corneal surgery. Invest Ophthalmol Vis Sci 1998; 39: 276–283.

45 Cho KS, Lee EH, Choi JS, Joo CK: Reactive oxygen species-induced apoptosis and ne-crosis in bovine corneal endothelial cells. In-vest Ophthalmol Vis Sci 1999; 40: 911–919.

46 Pitts DG, Cullen AP, Hacker PD: Ocular ef-fects of ultraviolet radiation from 295 to 365 nm. Invest Ophthalmol Vis Sci 1977; 16: 932–939.

47 Mazzotta C, Traversi C, Baiocchi S, et al: Conservative treatment of keratoconus by riboflavin-UVA-induced cross-linking of corneal collagen: qualitative investigation of corneal epithelium and subepithelial nerve plexus regeneration by in vivo HRT II system confocal microscopy in humans. Eur J Oph-thalmol 2006; 16: 530–535.

48 Krachmer JH, Mannis MJ, Holland EJ: Cor-nea: Fundamentals, Diagnosis and Manage-ment. Volume one. Second edition. Elsevier Mosby, Philadelphia, Edinburgh, London, New York, Oxford, St. Louis, Sydney, Toron-to, 2005.

results of confocal microscopy examinations after collagen cross-linking with riboflavin and uva...

Documents