Stromal haze after combinedriboflavin-UVA corneal collagencross-linking in keratoconus: in vivoconfocal microscopic evaluation

ABSTRACT

The technique of corneal collagen cross-linking consists of photopo-lymerization of stromal fibres by the combined action of a photo-sensitizing substance (riboflavin or vitamin B2) and ultraviolet lightfrom a solid state UVA source. Photopolymerization increases therigidity of corneal collagen and its resistance to keratectasia. In thisreport we present two cases, studied through in vivo confocal micros-copy, with stage III keratoconus that developed stromal haze after thecross-linking treatment.

Key words: collagen cross-linking, dark microstriae, haze, in vivoconfocal microscopy, keratoconus, Vogt’s striae.

INTRODUCTION

The technique of corneal collagen cross-linking1 was developed inGermany in the 1990s by a research group at Dresden TechnicalUniversity and consists of photo-polymerization of stromal fibresby the combined action of a photosensitizing substance (riboflavinor vitamin B2) and ultraviolet light from a solid state UVA sourceincreasing the rigidity of corneal collagen and its resistance tokeratectasia. The aim was to slow down or arrest keratoconus pro-gression so as to delay or avoid recourse to penetrating keratoplasty.

Although cornea collagen cross-linking represents the firstapproach to stop or even to reduce the progression of keratoconus,the complication rate has not yet been exactly determined inliterature. Moreover, indications are not definitively fixed.

In this report we present two cases, studied through in vivoconfocal microscopy, with stage III keratoconus (according toKrumeich keratoconus clinical staging) that developed stromal hazeafter the cross-linking treatment.

CASE REPORTS

Between October 2004 and July 2006, 40 eyes of 39 patients, wereselected in the Department of Ophthalmology and Neurosurgery ofthe University of Siena.

All patients were affected by bilateral keratoconus. Thirty-fiveeyes were in Krumeich stage I-II and five in stage III.

Subjects under 18 and over 60 years of age, or patients withcorneal thickness’ less than 400 mm, with histories of herpetickeratitis and corneal scarring, severe eye dryness, current cornealinfections or concomitant autoimmune diseases were excluded.

The study design was approved by the Ethical Committee of theUniversity of Siena and the Siena University Medical Hospital, andwas conducted in accordance with the ethical standards set in the1964 Declaration of Helsinki, as renewed in 2000. All patients gavetheir written informed consent prior to their inclusion in the study.

Riboflavin-UVA-induced corneal cross-linking was performed inall patients by the same surgeon (AC).

Two cases with stage III keratoconus developed stromal hazeafter the cross-linking treatment in the postoperative course.

Preoperative best corrected spectacle visual acuity was logMAR0.70 in Case 1 and logMAR 0.5 in Case 2. Topographic Sim-Kaverage corneal curvatures were 58.6 diopters (D) and 50.94 D,respectively (CSO Eye Image System, Florence, Italy).

At the time of surgery, protocol included topical anaesthesia(4% lidocaine eye drops). After removing the corneal epithelium ina central circle of 9 mm diameter by means of a thin blunt metalspatula, the photosensitizing solution of 0.1% riboflavin – 20%dextran (Sooft, Italy) was administered 5 min before beginningirradiation, and every 5 min thereafter, for 30 min. The UVA sourcewas a solid state device (Exerion-Sas, Prato, Italy) consisting of twoUVA LEDs (370–10, 750 microwatt, Roithner Lasertechnic,Vienna) with a potentiometric voltage regulator (Exerion-Sas con-troller, Prato, Italy). Irradiated energy was checked by a UVA powermeter (Lasermate Q-Coherent, Santa Clara, CA, USA). Wave-length was 370 nm at a power of 3 mW/cm2 or 5.4 joule/cm2; dis-tance was 1 cm from corneal apex. After treatment the eyes weredressed with a soft therapeutic contact lens for 5 days. The patientswere medicated four times with topical antibiotic drops (ofloxacin),four times with non-steroid anti-inflammatory drug drops(diclofenac), and twice with mydriatic drops (cyclopentolate 1%).The soft contact lens was removed 5 days after treatment. Postop-erative follow up was performed as well as in vivo qualitative analysisof the corneal ultra-structural modifications, stromal wound healingand potential corneal haze effect in vivo after riboflavin-UVA-induced cross-linking of corneal collagen by HRT II system confo-cal microscopy (Rostock Cornea Module, Heidelberg, Germany) at1, 3 and 6 months. Confocal microscopy was also performed pre-operatively after the application of topical drops of cornealanaesthesia (oxybuprocaine chlorhydrate) and corneal gel(methylcellulose).

Six months after the cross-linking procedure best correctedspectacle visual acuity was logMAR 0.4 in the first patient andlogMAR 0.3 in the second.

Topographic Sim-K average corneal curvatures were 57.82 and49.46 D, respectively.

In the two cases, the preoperative in vivo confocal microscopyexaminations of the treated corneas revealed typical keratoconusfindings. Epithelium layers showed, together with an evident dis-tortion of the basal epithelium, areas with undefined cell borders(Fig. 1a). Subepithelial nerve plexus examination revealed the pres-ence of distorted hyper-reflective fibres in one case (Fig. 1b) and ofa pearl necklace shape in the other (Fig. 1c). The main stromalabnormalities were microstriae that appeared as multiple, thin, darklines in contrast with the brighter reflectivity of the stroma. In thedeep stromal layers (200–350 m) the striae were extracellular, andpresented different orientation and thickness with a reticularpattern (Fig. 2). This alteration was mainly detected in the deepstromal layers. The keratocyte nuclei were barely distinguishablewithin the edge of these linear structures. The endothelium wasnormal.

During the postoperative course, between the second and thethird month, the two cases developed stromal haze evident withbiomycroscopy (Fig. 3) and resistant to topical steroids.

In the two cases the stromal haze appeared unchanged at the6-month stage.

The postoperative confocal analysis confirmed the onset ofstromal scarring. At a depth of 170–200 mm the extracellular matrixgrew denser as the keratocytes population increased (Figs 3,4).This increase seemed compatible with intermediate stromalhaze.

580 Letters to the Editor

© 2007 The AuthorsJournal compilation © 2007 Royal Australian and New Zealand College of Ophthalmologists

DISCUSSION

Initial clinical trials2,3 indicated that the cross-linking technique wasa safe method for treating progressive keratoconus which cannot becorrected by contact lenses, making it possible to avoid or at leastdelay corneal transplant.

In our small case series, riboflavin-UVA-induced corneal cross-linking was performed in two patients with stage III progressivekeratoconus. During the postoperative course, between the secondand the third month, the two cases developed a stromal haze. The

postoperative best correct visual acuity data suggest that this hazedid not seem to impair patient vision.

In these two cases, the preoperative confocal microscopyshowed hypo-reflective stromal microstriae. As described in severalpapers,4,5 this finding is detectable in the anterior, intermediate andposterior stroma, with vertical, horizontal and oblique patterns, alsoin patients without clinical evidence of Vogt’s striae. Moreover inour confocal analysis the dark microstriae showed a predomi-nantly reticular pattern, not previously described in theliterature.

a cbFigure 1. Corneal layer modifica-tions in keratoconus by in vivo HRTII system confocal microscopy(400 ¥ 400 mm). Distortion of thebasal epithelium (a), abnormalcurved subepithelial nerve plexus (b)(Case 1) and pearl necklace shape ofthe nerves (c) (Case 2).

Figure 2. HRT II (400 ¥ 400 mm)stromal hypo-reflective dark bandsin keratoconus with predominantreticular patterns (Case 1, a,b) (Case2, c,d).

a

c

b

d

Letters to the Editor 581

© 2007 The AuthorsJournal compilation © 2007 Royal Australian and New Zealand College of Ophthalmologists

The reticular pattern could represent a confocal sign of anadvanced progressive keratoconus.

Our data suggest that the detection of reticular hypo-reflectivemicrostriae by in vivo confocal analysis, with or without Vogt’sstriae, could represent a relative contraindication to perform ariboflavin-UVA-induced corneal cross-linking. In fact in these casesthere is a high risk of the development of late stromal scarringduring the postoperative course. Although in our cases the hazedoes not seem to have affected the visual acuities, it is not possibleto exclude this complication without a longer follow up.

The routine preoperative execution of in vivo confocal micros-copy could help with patient selection, improving the inclusioncriteria and optimizing the results of a very promising techniquesuch as riboflavin-UVA-induced corneal cross-linking.

Cosimo Mazzotta PhD, Angelo Balestrazzi PhD,Stefano Baiocchi PhD, Claudio Traversi MD PhD and

Aldo Caporossi MDDepartment of Ophthalmology and Neurosurgery, Siena University,

Siena, ItalyReceived 20 October 2006; accepted 27 March 2007.

REFERENCES

1. Spoerl E, Huhle M, Seiler T. Induction of cross links in cornealtissue. Exp Eye Res 1998; 66: 97–103.

2. Caporossi A, Baiocchi S, Mazzotta C, Traversi C, Caporossi T.Parasurgical therapy for keratoconus by riboflavin-ultraviolettype A rays induced cross-linking of corneal collagen Prelimi-nary refractive results in an Italian study. J Cataract Refract Surg2006; 32: 837–45.

3. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A-induced collagen crosslinking for the treatment of keratoconus.Am J Ophthalmol 2003; 135: 620–7.

4. Hollingsworth JG, Efron N. Observations of banding patterns(Vogt striae) in keratoconus: a confocal microscopy study. Cornea2005; 24: 162–6.

5. Hollingsworth JG, Bonshek RE, Efron N. Correlation of theappearance of the keratoconic cornea in vivo by confocal micros-copy and in vitro by light microscopy. Cornea 2005; 24: 397–405.

Pseudoexfoliation syndrome: in vivoconfocal microscopy analysis

ABSTRACT

Pseudoexfoliation (PEX) syndrome is a common ocular disease thatalso affects the cornea. A case of clinical PEX syndrome, studied byin vivo corneal confocal microscopy is reported. The morphologicalanalysis of the confocal images demonstrated hyper-reflective depos-its and several dendritic cells in the basal epithelial layer. A fibrillarsubepithelial structure was also found.The endothelial layer showedcell anomalies (polymegathism and pleomorphism) and hyper-reflective small endothelial deposits. Confocal microscopy is an in vivoimaging method that may provide new information on corneal alter-ations in PEX, and detect early corneal features.

Key words: confocal microscopy, HRT II, keratopathy, pseudoex-foliation syndrome.

INTRODUCTION

Pseudoexfoliation (PEX) syndrome is a common ocular conditionwhich affects almost all the structures of the anterior segment of theeye. The hallmark of PEX syndrome is the appearance of whitishflecks on the anterior capsule of the crystalline lens.1 Depositionsare also observed at the level of the ciliary zonular apparatus, thejuxtacanalicular and trabecular tissue, and the corneal endothelium.These changes have been associated with cataract, chronic glau-coma, lens subluxation, pseudouveitis, keratopathy and retinal veinocclusion.2 Deposition of PEX material has been described asirregular clumps deposited on the corneal endothelium and alsoassociated with a decrease in endothelial cell count.1 In vivo confocalmicroscopy (IVCM) is a non-invasive examination technique thatallows early diagnosis at microstructural level of pathologic changesin the cornea. To our knowledge, no studies have been reported inthe literature about the examination of PEX patients by IVCM. Thisreport aims to describe the structural corneal features in a patientwith PEX syndrome using IVCM.

Figure 3. HRT II (400 ¥ 400 mm) corneal haze 6 months aftersurgery (Case 1).

Figure 4. Subclinical haze is detectable by in vivo HRT II confocalmicroscopy (400 ¥ 400 mm). Activated keratocytic nuclei (greenarrows) with increased stromal reflectance (red arrows) compatiblewith subclinical, microscopically detectable haze.

582 Letters to the Editor

© 2007 The AuthorsJournal compilation © 2007 Royal Australian and New Zealand College of Ophthalmologists