Effect of Corneal Collagen Cross-Linking onCorneal Innervation, Corneal Sensitivity,and Tear Function of Patients withKeratoconus

Georgios A. Kontadakis, MD, MSc, George D. Kymionis, MD, PhD, Vardhaman P. Kankariya, MD,Aristophanis I. Pallikaris, MSc

Purpose: To evaluate the effect of corneal collagen cross-linking (CXL) on corneal innervation, cornealsensitivity, and tear function in patients with keratoconus.

Design: Prospective, interventional case series.Participants: Twenty-four patients with bilateral keratoconus (30 eyes) who presented to the Institute of

Vision and Optics, University of Crete, from May 2008 to October 2008.Methods: Patients underwent CXL. Confocal microscopic analysis of corneal sub-basal nerve plexus (total

nerve length per image), corneal sensitivity (assessed with the Cochet–Bonnet esthesiometer), basic tearsecretion (assessed with Schirmer’s I test with anesthesia), and tear film stability (evaluated by means of tear filmbreak-up time [TFBUT]) were assessed preoperatively and at 1, 3, 6, 9, 12, 18, and 24 months postoperatively.

Main Outcome Measures: Comparisons between preoperative and each postoperative value of total nervelength per image, corneal sensitivity, Schirmer’s I test results, and TFBUT.

Results: Total nerve length per image and corneal sensitivity were significantly decreased until postopera-tive month 6 (for both parameters: P�0.05 paired-samples t test at 1, 3, and 6 months postoperatively). Totalnerve length per image tended to increase up to 2 years postoperatively, when it reached the preoperative level,but differences with the preoperative values after the sixth post-CXL month were insignificant. The results ofSchirmer’s I test and TFBUT had no statistically significant difference at any time point.

Conclusions: A transient decrease in corneal innervation and corneal sensitivity can be observed up to 6months after CXL. No significant effect of CXL could be detected on basic tear secretion and tear film stabilityin our group of patients.

Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed

in this article. Ophthalmology 2013;120:917–922 © 2013 by the American Academy of Ophthalmology.

boasdakaasisf

M

Ifc

Corneal surgical procedures have been associated with lossof corneal sensitivity and dry eye symptoms in severalstudies.1–13 Ablation of corneal tissue in refractive surgery,flap creation in LASIK, and clear corneal incisions in cat-aract surgery are procedures that affect corneal innervationand reduce corneal sensation.13–17 Loss of corneal sensationdiminishes the reflex that stimulates blinking and tear pro-duction (lacrimal and mucus secretion). Because of theseaspects, iatrogenic dry eye postsurgery is one of the maincomplications of corneal procedures. This affects patients’quality of life and may become severe in those with preop-erative dry eye symptoms.18

Corneal collagen cross-linking (CXL) with the combina-tion of riboflavin and ultraviolet A (UVA) radiation is anovel corneal surgical procedure to stabilize progressivekeratoconus.19–21 The method is based on the absorption ofUVA (370 nm) radiation by the cornea after the photosen-sitizer riboflavin is infused in the stroma. This procedure isconsidered to build up the bonds between the collagen

molecules and therefore increase corneal rigidity and sta- o

© 2013 by the American Academy of OphthalmologyPublished by Elsevier Inc.

ility.21–25 Corneal confocal microscopic studies have dem-nstrated a reduction in subepithelial nerve plexus densityfter CXL, which seems to return to the preoperative statuseveral months after the procedure.26,27 Reduction of nerveensity is expected to adversely affect corneal sensitivitynd tear function. This could be more prominent becauseeratoconic patients may already have tear film disordersnd reduced corneal sensitivity as a result of the abridgednd misshapen corneal nerves and the abnormal cornealhape in keratoconus.28–31 The purpose of the current studys to evaluate the alterations of corneal innervation, cornealensitivity, and tear function in patients undergoing CXLor the treatment of progressive keratoconus.

aterials and Methods

n this prospective interventional case series, we included subjectsrom a continuous cohort of patients with keratoconus who soughtonsultation at the Institute of Vision and Optics of the University

f Crete. Patients with progressive keratoconus, with corneal

917ISSN 0161-6420/13/$–see front matterhttp://dx.doi.org/10.1016/j.ophtha.2012.10.012

abmtl((wmcackenwivlqeIpNoaiis

CTAewr5wvUwcmEts

ai(dfccmfiws

SWSt

Ophthalmology Volume 120, Number 5, May 2013

thickness more than 400 �m, and with no other systemic or oculardisease were included in the study. Diagnosis of keratoconus wasbased on the axial topography map (iTrace, Tracey Technologies,Houston TX).32 Patients were followed at 6-month intervals, andkeratoconus was described as progressive when there was anincrease in the cone apex keratometry of 0.75 diopters or alterationof 0.75 diopters in the spherical equivalent refraction. On confir-mation of progression, we proposed CXL as a treatment for ker-atoconus to halt its progression. Patients were thoroughly coun-seled about the current clinical experience and possible outcomesand complications of the procedure. We obtained written informedconsent from all subjects according to the institutional guidelinesand the Declaration of Helsinki. Institutional review board/ethicscommittee approval was obtained.

Clinical EvaluationThe preoperative and postoperative evaluations included cornealtopography, uncorrected distance visual acuity and corrected dis-tance visual acuity assessment with manifest refraction, assess-ment of corneal sensitivity (Cochet–Bonnet esthesiometer) andtear film break-up time (TFBUT), Schirmer’s I test with anesthe-sia, slit-lamp biomicroscopy, applanation tonometry, ultrasoundpachymetry, and corneal confocal microscopy of the central cornea(Heidelberg Retina Tomography [HRT] II; Heidelberg Engineer-ing, Heidelberg, Germany). Examinations were conducted at eachvisit with the reported order.

We repeated patient evaluation at postoperative follow-up in-tervals of 1, 3, 6, 9, 12, 18, and 24 months. Patients were instructedto discontinue use of contact lenses at least 8 weeks before eachvisit to avoid possible effects on corneal topography and sensitiv-ity measurements.33

Corneal sensitivity was measured using the Cochet–Bonnetesthesiometer.34 The instrument consists of a nylon filament 60.0mm long and 0.12 mm in diameter. The force exerted by thefilament when it touches the cornea is inversely proportional to itslength. The same experienced observer (G.A.K.) conducted allmeasurements at the central cornea. The filament was movedtoward the cornea smoothly at a perpendicular angle. Contact wasdetected by slight bending in the filament. Subjects were asked tolook straight ahead and indicate when the stimulus was felt. Ifthere was no subject response to the first contact, the length of thefilament was decreased by 5.0 mm to increase its rigidity, and theprocedure was repeated until the subject reported sensation ofcorneal contact. The mean filament length, from a minimum of 3stimulus applications that produced a positive response from thesubject, was considered to be the corneal touch threshold.

We performed the standard TFBUT measurement to evaluatetear film stability. Moistened fluorescein strips were introduced tothe conjunctival sac with minimal stimulation and were undetectedby the patients. The subjects were then instructed to blink severaltimes for a few seconds to ensure adequate mixing of fluorescein.The time interval between the last complete blink and the appear-ance of the first corneal black spot in the stained tear film wasmeasured 3 times, and the mean value of the measurements wascalculated.

For the evaluation of basic tear secretion, the standard Schirm-er’s I test with topical anesthesia (proparacaine hydrochloride0.5%; Alcaine; Alcon Laboratories, Inc, Fort Worth, TX) wasperformed. The standardized strips of filter paper (Alcon Labora-tories, Inc) were placed in the inferior conjunctival fornix near thelateral canthus away from the cornea and left in place for 5 minuteswith the eyes closed. Readings were reported in millimeters ofwetting for 5 minutes.

Confocal microscopy was performed with a modified confocal

scanning laser ophthalmoscope (HRT II; Heidelberg Engineering) in f

918

ll patients. With the addition of the Rostock Cornea Module (Heidel-erg Engineering), the HRT II was converted into a confocal cornealicroscope that allowed the acquisition of 2-dimensional images of

he various layers of the cornea by sequentially scanning a 670-nmaser beam over the cornea. After the instillation of a local anesthetic1 drop of proparacaine hydrochloride 0.5%) and high-viscosity gelcarbomer 3.0 mg/g, Thilogel; Alcon Laboratories, Inc), patientsere asked to fixate using an external fixation target. The instru-ent objective was then brought into optical contact with the

ornea tissue by a disposable sterile polymethyl methacrylate cupnd a high-viscosity gel (Thilogel). Depth scans across the wholeornea were performed manually while an external electronic unitept track of the focal plane. The external interface of the polym-thyl methacrylate cup was taken as the reference point for thick-ess measurements. Images of the various layers of the corneaere acquired at the optical center of the cornea. The acquired

mages consisted of 384�384 pixels over a 400�400-�m field ofiew with a transversal resolution of approximately 2 �m and aongitudinal resolution of approximately 4 �m. Three highestuality images of the sub-basal nerve plexus were selected by anxperienced masked observer for each patient at each time point.mages were quantitatively analyzed with custom-made imagerocessing software in MATLAB (MATLAB, MathWorks Inc,atick, MA). During processing, the masked operator traced theutline of each nerve by using a simple point-and-click method,nd the software calculated the total nerve length of all nerves permage. The mean value of the total nerve length of the 3 selectedmages per patient per time point was calculated and used fortatistical analysis.

orneal Collagen Cross-Linking Surgicalechniquefter topical anesthesia with proparacaine hydrochloride 0.5%

yedrops was administered, the corneal epithelium was removedithin only an 8.0-mm diameter. Next, commercially available

iboflavin (riboflavin 0.1% isotonic eye drops with 20% dextran00, Medicross, Medio–Haus, Behrensbrook, Neudorf, Germany)as applied every 3 minutes for approximately 30 minutes. Ultra-iolet A irradiation was performed with a commercially availableVA system (UV-X, Peschke Meditrade, Nürnberg, Germany)ith Köhler illumination. Before treatment, the intended 3 mW/

m2 surface irradiance (370 nm, 5.4 J/cm2 surface dose after 30inutes) was calibrated using the UVA meter YK-34UV (Lutronlectronic Enterprise Co, Ltd, Taipei, Taiwan). During treatment,

he riboflavin solution was applied every 5 minutes to ensureaturation.

At the end of the procedure, a combination of steroid andntibiotic drop (Tobradex; Alcon Laboratories, Inc.) was admin-stered in all patients, and a silicon-hydrogel bandage contact lensLotrafilcon B, Air Optix; CIBA VISION, Duluth, GA; 14.0 mmiameter, 8.6 base curvature, Dk�140 barrers) was kept in place untilull corneal re-epithelialization occurred. After re-epithelialization, aourse of fluorometholone 0.1% eyedrops (FML, Allergan Pharam-euticals, Westport, Ireland) with weekly tapering was applied for 1onth, and patients were advised to use preservative-free arti-cial tears for the first 6 postoperative months. All proceduresere carried out at the Institute of Vision and Optics by the

ame surgeon (G.D.K.).

tatistical Analysise examined the distribution of continuous variables with the

hapiro–Wilk test for normality. Comparisons between preopera-ive and postoperative values at each time point were performed

or corneal sensitivity, Schirmer’s I test result, TFBUT, and nerve

fpm(

pnmuwm1iv

eudp

F

Kontadakis et al � Effect of CXL on Corneal Innervation, Sensitivity, and Tears

length. The differences between preoperative and postoperativevalues were assessed with the paired-samples t test if variables hada normal distribution and with the Wilcoxon signed-rank test if thevariables did not have a normal distribution. The Statistical Pack-age for the Social Sciences version 15.0 was used (SPSS Inc,Chicago, IL). A P value less than 0.05 was considered to bestatistically significant.

Results

Twenty-four patients (30 eyes) with keratoconus were included inthe study. The mean age at the time of procedure was 25�4.7years (range, 20–36 years).

The mean values of test results and P values for the comparisonof the preoperative and the postoperative result at each time pointare presented in Table 1. The total nerve length per image at eachpostoperative time point was compared with the preoperativevalues by means of paired-samples t test. The total nerve length perimage had a statistically significant reduction at the first postop-erative month, corresponding to complete corneal denervation formany of the patients. At the following postoperative months, thetotal nerve length per image tended to gradually increase until thesecond year postoperative visit when it was completely restored tothe mean preoperative values (Fig 1). Despite this, there were nostatistically significant differences from the preoperative valuessince the ninth postoperative visit. A statistically significant dif-

Table 1. Mean Values of Test Results an

Preoperative 1 Mo 3 Mos

Sensitivity 5.8 (5.0–6.0) 2.8 (0.5–5.5) 4.1 (1.0–6.0) 5.1P value �0.001 �0.001Schirmer’s I test 9.79�6.74 10.50�6.42 11.85�6.967 10P value 0.900 0.144TFBUT 5.78�2.87 6.17�4.70 6.43�4.31 5P value 0.833 0.662Nerve length 1697�670 163�371 663�788P value �0.001 �0.001

TFBUT � tear film break-up time.Mean values of test results and P values for comparison of preoperative(range), and other data are presented as mean � standard deviation. Comthe remaining data were compared by means of paired-samples t test.

Figure 1. Mean value of nerve length per image preoperatively and at each

postoperative visit (bars represent 95% CI). CI � confidence interval. p

erence of total nerve length per image between preoperative andostoperative time intervals was found at 1 month (P�0.001), 3onths (P�0.001), and 6 months (P � 0.001) postoperatively

Table 1).Corneal sensitivity was compared at each time point with the

reoperative value by using the Wilcoxon signed-rank test. Cor-eal sensitivity had a significant reduction at the first postoperativeonth visit with gradual restoration of the preoperative values

ntil postoperative month 9. Statistically significant differencesere found at postoperative month 1 (P�0.001), postoperativeonth 3 (P�0.001), and postoperative month 6 (P � 0.003). At 9,

2, 18, and 24 months postoperatively, the mean corneal sensitiv-ty was not significantly different compared with the preoperativealues (Table 1, Fig 2).

For the comparisons of Schirmer’s I test results and TFBUT atach time point postoperatively with the preoperative values, wesed the paired-samples t test. There was no statistically significantifference at any time point between the preoperative and theostoperative values of both tests (Table 1, Figs 3 and 4).

operative to Postoperative Comparisons

s 9 Mos 12 Mos 18 Mos 24 Mos

–6.0) 5.6 (4.0–6.0) 5.6 (5.0–6.0) 5.9 (5.5–6.0) 5.8 (5.5–6.0)3 0.581 0.290 0.129 0.6796.09 10.77�7.58 12.61�7.18 13.00�6.93 12.73�6.525 0.293 0.430 0.139 0.4102.34 5.92�3.58 6.53�2.97 6.00�1.57 7.42�1.972 0.453 0.136 0.629 0.619549 1422�920 1356�756 1468�798 1685�6011 0.175 0.214 0.370 0.834

ostoperative results at each time point. Sensitivity is presented as meanns of sensitivity were performed with the Wilcoxon signed-rank test, and

igure 2. Mean value of corneal sensitivity preoperatively and at each

d Pre

6 Mo

(2.00.00

.28�0.83

.04�0.11

919�0.00

and ppariso

ostoperative visit (bars represent 95% CI). CI � confidence interval.

919

rnc

cagcrpl

vctased

rfttstytntn

wfilTiaa

F

Ophthalmology Volume 120, Number 5, May 2013

Discussion

Various corneal surgical procedures have been associatedwith reduction of corneal sensitivity due to amputation orlaser ablation of corneal nerves.1,33 Corneal denervationplays a key role in the transient tear dysfunction that com-plicates the postoperative course of patients undergoingcorneal surgery.

Corneal collagen cross-linking is a novel corneal proce-dure used to biomechanically stabilize corneas with ectaticdisorders. The combined action of the photosensitizer ribo-flavin with the UVA radiation induces cross-linkage be-tween the collagen fibrils through the generation of reactiveoxygen species, leading to corneal stabilization.19–23 Stud-ies with qualitative evaluation of corneal confocal micros-copy images after CXL demonstrate that anterior stromalstructure, keratocyte density, and innervation are affectedpostoperatively.26,27

In this study, quantitative analysis of sub-basal nerveplexus confocal images after CXL revealed that cornealinnervation is significantly reduced at postoperative month1 and is gradually restored to almost similar preoperativelevels at postoperative month 9. Confocal images taken ateach time point at the central cornea demonstrate the nervedensity of the sub-basal plexus in this area, rather than theorigin and course of specific sub-basal nerve regeneration.Our study findings correlate with those of previous qualita-tive studies on corneal confocal microscopy analysis ofsub-basal nerve plexus in patients post-CXL. In a previousqualitative study, Kymionis et al26 reported that reinnerva-tion begins during the third month after CXL. They reportedthat, unlike in the first month, the nerves are visible at the3-month follow-up but are not yet restored to the preoper-ative density. In another qualitative study, Mazzotta et al27

reported a more rapid regeneration of subepithelial nervefibers that begins at 1 month and is almost complete 6months after the procedure, but does not return to thepreoperative structure until 2 years after the procedure. Inour quantitative study, the total nerve length per image

Figure 3. Mean result of Schirmer’s I test preoperatively and at eachpostoperative visit (bars represent 95% CI). CI � confidence interval.

tended to increase until postoperative month 24, whereas p

920

estoration was almost complete up to month 9 and totalerve length had already reached a nonsignificant differenceompared with the preoperative level.

In regard to sensitivity, our results demonstrate that CXLauses a significant transient hypoesthesia postoperatively,s expected because of the transient denervation. In ourroup of patients, we detected a prominent reduction inorneal sensitivity 1 month postoperatively that graduallyecovered until 9 months postoperatively. The majority ofatients had corneal sensation restored to the preoperativeevel at the 9-month follow-up visit and thereafter.

In a recent animal study of corneal sensitivity and inner-ation after CXL in rabbits, Xia et al35 demonstrated thatorneal sensitivity regained normal values 90 days after thereatment and corneal nerve fiber density appeared normalfter 180 days. In the current study, the time course wasimilar for sensitivity restoration and corneal nerve regen-ration; both took 9 months to restore to nonsignificantifferences compared with the preoperative values.

In regard to other refractive surgery procedures, studieseport variable periods of time for the restoration of sensoryunction and innervation after the procedure. In photorefrac-ive keratectomy (PRK) and Epi-LASIK, sensitivity returnso preoperative values approximately 3 to 6 months afterurgery. In LASIK, it takes 6 months to more than 1 year forhe restoration of corneal sensation.3–6,12 It takes up to 2ears in PRK and 5 years in LASIK for the sub-basal nerveso fully recover.14,15 The differences in time course ofeural rehabilitation among surgical techniques may be dueo differences in both the cause and the extent of the initialeural injury.14,36

Corneal denervation plays a causative role in dry eye,hich is one of the most recognized complications of re-

ractive surgery procedures. Patients report dryness andrritation, and the results of Schirmer’s I test and TFBUT areower than normal up to 3 to 6 months after the procedure.his finding is more prominent in patients post-LASIK than

n patients undergoing surface ablation techniques.7,9–11 Instudy of patients undergoing Epi-LASIK, Kalyvianaki et

l8 reported that there is no decrease in Schirmer’s I test

igure 4. Mean value of tear film break-up time preoperatively and at each

ostoperative visit (bars represent 95% CI). CI � confidence interval.

ttdnfOcaratbsHgBc

tpptdprc

R

1

Kontadakis et al � Effect of CXL on Corneal Innervation, Sensitivity, and Tears

values, but there is a small decrease in TFBUT in the firstpostoperative month. Nejima et al11 observed a decrease inSchirmer’s I test values but not in TFBUT in patientspost-PRK.11 In the current study, although sensitivity de-creases, the results demonstrate that basic tear secretion andTFBUT are not affected by CXL. The patients included inthe study were advised to use preservative-free artificialtears, which may have helped. On the other hand, thesensitivity of the peripheral unaffected part of the corneamay be sufficient to regulate basic tear secretion or theaberrant activity of the amputated corneal nerves may con-tribute to this.37

An important safety aspect in CXL is that corneal inner-vation, sensitivity, and tear function are already disturbed inpatients with keratoconus.28–31 Bleshoy38 reported that sen-sitivity is decreased in all corneal zones of patients withkeratoconus. This decrease depends on the severity of thecondition. Dogru et al28 reported that eyes with keratoconushave lower goblet cell density. Consistent with those find-ings, authors28 found that patients with keratoconus havesignificantly lower TFBUT values, although there is not adecrease in their Schirmer’s I test results. Tear film dys-function also depends on the severity of keratoconus. In ourgroup of patients, corneal sensitivity seemed to be withinnormal values, although it was not compared with a controlgroup of normal corneas. However, this is expected becausemost of the patients in the current study group were in theearly stages of keratoconus. Basic tear secretion seems to besimilar to that reported8,9 in studies of patients undergoingrefractive surgery, and the preoperative TFBUT of ourpatients was relatively low, consistent with the findings ofDogru et al.28 Reduced sensitivity after CXL in patientswith low tear break-up time theoretically could generate dryeye–related problems because of the decreased blinking rateand increased tear evaporation and exposure of cornealepithelium surface. The results of this study demonstratethat because TFBUT and Schirmer’s I test results were notaffected, dry eye does not seem to be a significant compli-cation of CXL in patients with keratoconus.

According to confocal microscopic studies of cornealinnervation in patients with keratoconus, the sub-basalnerves have abnormal patterns (thicker and more tortuous)and lower density than in normal subjects. The patternabnormality of corneal nerves has been correlated with theseverity of keratoconus.29–31 A causative role of these find-ings in the progression of keratoconus has been postulatedin the literature.39 Our results demonstrate that the sub-basalnerve plexus in keratoconus is able to return to the preop-erative status after almost complete denervation caused byCXL.

Contact lens use also affects corneal sensitivity.33 Tominimize any effect on measurements, we asked thepatients to discontinue contact lens use at least 8 weeksbefore each examination. Patients who use contact lensescontinuously after CXL could have a somewhat differenttime course of nerve regeneration and sensitivity resto-ration than demonstrated in this study. At the same time,reduction of sensitivity and the effect of CXL on cornealtopography may affect contact lens use in patients with

keratoconus.

A significant limitation of this study is that we used onlyhe Cochet–Bonnet esthesiometer to monitor corneal sensi-ivity. This instrument is a contact esthesiometer that pro-uces mechanical stimuli exploring the response of mecha-onociceptors. Consequently, recovery of corneal sensationor chemical and temperature stimuli is not monitored.ther instruments, such as gas esthesiometers, which use

arbon dioxide as an acidic stimulus, and warmed or cooledir for high and low temperature stimuli also evaluate theesponsiveness of polymodal and thermal receptors.40,41 Inddition, the Cochet–Bonnet esthesiometer lacks the abilityo detect subtle changes of sensitivity and hypersensitivityecause it has a lower threshold detection limit that isimilar to the normal threshold for most of the patients.owever, most of the studies evaluating the effect of sur-ical procedures on corneal sensitivity use the Cochet–onnet esthesiometer and consequently have somewhatomparable results.

In conclusion, the results of our study indicate that CXLransiently affects corneal innervation and sensation. Ouratients returned to preoperative sensitivity within the first 6ostoperative months, and their basic tear secretion test andear break-up time were not affected. Thus far, studiesemonstrate that CXL is a safe procedure, although com-lications similar to those of refractive surgery have beeneported.42,43 Iatrogenic dry eye does not seem to be a majoromplication in CXL.

eferences

1. Ang RT, Dartt DA, Tsubota K. Dry eye after refractive sur-gery. Curr Opin Ophthalmol 2001;12:318–22.

2. Kanellopoulos AJ, Pallikaris IG, Donnenfeld ED, et al. Com-parison of corneal sensation following photorefractive kera-tectomy and laser in situ keratomileusis. J Cataract RefractSurg 1997;23:34–8.

3. Matsui H, Kumano Y, Zushi I, et al. Corneal sensation aftercorrection of myopia by photorefractive keratectomy and laserin situ keratomileusis. J Cataract Refract Surg 2001;27:370–3.

4. Kim WS, Kim JS. Change in corneal sensitivity followinglaser in situ keratomileusis. J Cataract Refract Surg 1999;25:368–73.

5. Perez-Santonja JJ, Sakla HF, Cardona C, et al. Corneal sen-sitivity after photorefractive keratectomy and laser in situkeratomileusis for low myopia. Am J Ophthalmol 1999;127:497–504.

6. Nassaralla BA, McLeod SD, Nassaralla JJ Jr. Effect of myopicLASIK on human corneal sensitivity. Ophthalmology 2003;110:497–502.

7. Battat L, Macri A, Dursun D, Pflugfelder SC. Effects of laserin situ keratomileusis on tear production, clearance, and theocular surface. Ophthalmology 2001;108:1230–5.

8. Kalyvianaki MI, Katsanevaki VJ, Kavroulaki DS, et al. Com-parison of corneal sensitivity and tear function followingEpi-LASIK or laser in situ keratomileusis for myopia. Am JOphthalmol 2006;142:669–71.

9. Yu EY, Leung A, Rao S, Lam DS. Effect of laser in situkeratomileusis on tear stability. Ophthalmology 2000;107:2131–5.

0. Lee JB, Ryu CH, Kim JH, et al. Comparison of tear secretion

and tear film instability after photorefractive keratectomy and

921

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

Ophthalmology Volume 120, Number 5, May 2013

laser in situ keratomileusis. J Cataract Refract Surg 2000;26:1326–31.

11. Nejima R, Miyata K, Tanabe T, et al. Corneal barrier function,tear film stability, and corneal sensation after photorefractivekeratectomy and laser in situ keratomileusis. Am J Ophthal-mol 2005;139:64–71.

12. Matsui H, Kumano Y, Zushi I, et al. Corneal sensation aftercorrection of myopia by photorefractive keratectomy andlaser in situ keratomileusis. J Cataract Refract Surg 2001;27:370 –3.

13. Bragheeth MA, Dua HA. Corneal sensation after myopic andhyperopic LASIK: clinical and confocal microscopic study.Br J Ophthalmol 2005;89:580–5.

14. Erie JC, McLaren JW, Hodge DO, Bourne WM. Recovery ofcorneal subbasal nerve density after PRK and LASIK. Am JOphthalmol 2005;140:1059–64.

15. Calvillo MP, McLaren JW, Hodge DO, Bourne WM. Cornealreinnervation after LASIK: prospective 3-year longitudinalstudy. Invest Ophthalmol Visual Sci 2004;45:3991–6.

16. Linna TU, Vesaluoma MH, Perez-Santonja JJ, et al. Effect ofmyopic LASIK on corneal sensitivity and morphology ofsubbasal nerves. Invest Ophthalmol Visual Sci 2000;41:393–7.

17. Lee SJ, Kim JK, Seo KY, et al. Comparison of corneal nerveregeneration and sensitivity between LASIK and laser epithe-lial keratomileusis (LASEK). Am J Ophthalmol 2006;141:1009–15.

18. Konomi K, Chen LL, Tarko RS, et al. Preoperative character-istics and a potential mechanism of chronic dry eye afterLASIK. Invest Ophthalmol Visual Sci 2008;49:168–74.

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

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

21. McCall AS, Kraft S, Edelhauser HF et al. Mechanisms ofcorneal tissue cross-linking in response to treatment withtopical riboflavin and long-wavelength ultraviolet radiation(UVA). Invest Ophthalmol Vis Sci 2010;51:129–38.

22. Hafezi F, Kanellopoulos J, Wiltfang R, Seiler T. Cornealcollagen crosslinking with riboflavin and ultraviolet A to treatinduced keratectasia after laser in situ keratomileusis. J Cat-aract Refract Surg 2007;33:2035–40.

23. Raiskup-Wolf F, Hoyer A, Spoerl E, Pillunat LE. Collagencrosslinking with riboflavin and ultraviolet-A light inkeratoconus: long-term results. J Cataract Refract Surg 2008;34:796–801.

24. Wollensak G. Crosslinking treatment of progressive keratoconus:new hope. Curr Opin Ophthalmol 2006;17:356–60.

25. Kymionis G, Portaliou D. Corneal crosslinking with riboflavinand UVA for the treatment of keratoconus [letter]. J Cataract

Refract Surg 2007;33:1143–4, author reply 1144.

ogy Annual Meeting, October 24–27, 2009, San Francisco, California.

FTd

CGvC

922

6. Kymionis GD, Diakonis VF, Kalyvianaki M, et al. One-yearfollow-up of corneal confocal microscopy after corneal cross-linking in patients with post laser in situ keratosmileusisectasia and keratoconus. Am J Ophthalmol 2009;147:774–8.

7. Mazzotta C, Traversi C, Baiocchi S, et al. Corneal healingafter riboflavin ultraviolet-A collagen cross-linking deter-mined by confocal laser scanning microscopy in vivo: earlyand late modifications. Am J Ophthalmol 2008;146:527–33.

8. Dogru M, Karakaya H, Ozçetin H, et al. Tear function andocular surface changes in keratoconus. Ophthalmology 2003;110:1110–8.

9. Niederer RL, Perumal D, Sherwin T, McGhee CN. Laserscanning in vivo confocal microscopy reveals reduced inner-vation and reduction in cell density in all layers of the kera-toconic cornea. Invest Ophthalmol Vis Sci 2008;49:2964–70.

0. Teng CC. Electron microscope study of the pathology ofkeratoconus: I. Am J Ophthalmol 1963;55:18–47.

1. Mocan MC, Yilmaz PT, Irkec M, Orhan M. In vivo confocalmicroscopy for the evaluation of corneal microstructure inkeratoconus. Curr Eye Res 2008;33:933–9.

2. Rabinowitz YS. Keratoconus. Surv Ophthalmol 1998;42:297–319.

3. Martin XY, Safran AB. Corneal hypoesthesia. Surv Ophthal-mol 1988;c33:28–40.

4. Beuerman RW, McCulley JP. Comparative clinical assess-ment of corneal sensation with a new aesthesiometer. Am JOphthalmol 1978;86:812–5.

5. Xia Y, Chai X, Zhou C, Ren Q. Corneal nerve morphologyand sensitivity changes after ultraviolet A/riboflavin treat-ment. Exp Eye Res 2011;93:541–7.

6. Chan KY, Jarvelainen M, Chang JH, Edenfield MJ. A cryo-damage model for studying corneal nerve regeneration. InvestOphthalmol Visual Sci 1990;31:2008–21.

7. Belmonte C. Eye dryness sensations after refractive surgery:impaired tear secretion or “phantom” cornea? J Refract Surg2007;23:598–60.

8. Bleshoy H. Corneal sensitivity in keratoconus. J Br ContactLens Assoc 1986;1986:9–12.

9. Brookes NH, Loh I, Clover GM, et al. Involvement of cornealnerves in the progression of keratoconus. Exp Eye Res 2003;77:515–24.

0. Belmonte C, Aracil A, Acosta MC, et al. Nerves and sensa-tions from the eye surface. Ocul Surf 2004;2:248–53.

1. Belmonte C, Acosta MC, Schmelz M, Gallar J. Measurement ofcorneal sensitivity to mechanical and chemical stimulation with aCO2 esthesiometer. Invest Ophthalmol Vis Sci 1999;40:513–9.

2. Kymionis GD, Bouzoukis DI, Diakonis VF, et al. Diffuselamellar keratitis after corneal crosslinking in a patient withpost-laser in situ keratomileusis corneal ectasia. J CataractRefract Surg 2007;33:2135–7.

3. Kymionis GD, Portaliou DM, Bouzoukis DI, et al. Herpetickeratitis with iritis after corneal crosslinking with riboflavinand ultraviolet A for keratoconus. J Cataract Refract Surg

2007;33:1982–4.

Footnotes and Financial Disclosures

Originally received: February 11, 2012.Final revision: July 30, 2012.Accepted: October 3, 2012.Available online: January 21, 2013. Manuscript no. 2012-199.

Institute of Vision and Optics, University of Crete, Heraklion, Greece.

Partially presented as a poster at: the American Academy of Ophthalmol-

inancial Disclosure(s):he author(s) have no proprietary or commercial interest in any materialsiscussed in this article.

orrespondence:eorgios A. Kontadakis, MD, MSc, Institute of Vision and Optics, Uni-ersity of Crete School of Health Sciences, GR 71 003 Voutes, Heraklion,

rete, Greece. E-mail: kontadas@yahoo.com.