corneal cross-linking as a treatment for keratoconus
TRANSCRIPT
Corneal Cross-Linking as a Treatment forKeratoconus
Four-Year Morphologic and Clinical Outcomes withRespect to Patient Age
Riccardo Vinciguerra, MD,1,2 Mario R. Romano, MD, PhD,1 Fabrizio I. Camesasca, MD,1
Claudio Azzolini, MD,2 Silvia Trazza, Orth,3 Emanuela Morenghi, PhD,1 Paolo Vinciguerra, MD1
Purpose: To report the 4-year outcomes of corneal cross-linking (CXL) for progressive keratoconus in apopulation of different age groups.
Design: Retrospective, single-center, nonrandomized clinical study.Participants: Four hundred consecutive eyes treated with corneal CXL for progressive keratoconus from
April 2006 through April 2010.Intervention: After removal of the epithelium, the cornea was irrigated for 30 minutes with a solution of 0.1%
riboflavin and 20% dextran, followed by irradiation with an ultraviolet A light of 3 mW/cm2 for 30 minutes.Main Outcome Measures: Best-corrected visual acuity (BCVA), sphere and cylinder refraction, corneal
topography, Scheimpflug tomography, and aberrometry were assessed at baseline and at 1, 6, 12, 24, 36, and48 months after corneal CXL treatment. The compiled data were stratified according to age (group A, youngerthan 18 years; group B, 18–29 years; group C, 30–39 years; and group D, older than 40 years).
Results: Comparative analysis included 400 eyes of 301 patients. Functional results showed a significantincrease in BCVA in group A by a mean reduction of –0.11 logarithm of the minimum angle of resolution (logMAR)after 12 months, in group B by a mean reduction of –0.31 logMAR after 36 months, in group C by a meanreduction of –0.33 logMAR after 36 months, and in group D by a mean reduction of –0.26 logMAR after 36months. Morphologic results showed an analogous regularization of corneal shape with a significant reductionof opposite sector index by a mean value of –0.53 at 12 months in group A, –1.14 at 36 months in group B, –1.10at 36 months in group C, and –0.55 at 12 months for group D. Optical quality improvement was demonstratedby a mean significant reduction of coma –1.52 �m after 12 months in group A, –1.58 �m after 24 months in groupB, –2.57 �m after 36 months for group C, and –0.25 �m after 36 months in group D.
Conclusions: Outcomes stratified by age indicate the efficacy of corneal CXL in stabilizing the progressionof ectatic disease in all age groups and improving the functional and morphologic parameters in select groups.Results indicated better functional and morphologic results in the population between 18 and 39 years of age.
Financial Disclosure(s): Proprietary or commercial disclosure may be found after the references.Ophthalmology 2013;120:908–916 © 2013 by the American Academy of Ophthalmology.
pocsb(adab
s
Keratoconus is a slowly progressive, asymmetric, bilateraldegenerative corneal disease that presently is the majorcause of corneal transplantation in Europe and is the secondmost common cause in the United States.1,2 The incidenceof keratoconus varies with the diagnostic parameters usedand is estimated to be between 1 in 2000 and 1 in 320persons.3,4 Usually, keratoconus is diagnosed in adoles-cence or childhood.5–7 The time of diagnosis is a negativeprognostic factor for increased risk of corneal transplantation,8
with younger patients tending to have a more aggressive pro-gression.9 Treatment strategies for pediatric and young patientsare an important concern for the physician because penetrating
keratoplasty in children is a highly challenging and demanding l908 © 2013 by the American Academy of OphthalmologyPublished by Elsevier Inc.
rocedure associated with a high risk of graft failure or failuref amblyopia treatment in clear grafts.10 Corneal collagenross-linking (CXL) is the only treatment that potentially canlow down or block the progression of ectatic disease.11–14 It isased on a photo-oxidative reaction, catalyzed by riboflavinB2 vitamin), that induces an increase in corneal stiffness15 andbiomechanical response16 that blocks the progression of theisease. Cross-linking induces many changes in collagen, fromn increased number of intrafibrillar and interfibrillar covalentonds to an increased resistance to enzymatic degradation.17,18
These changes in corneal biomechanics are known to be as-ociated with an improvement in visual acuity as well as morpho-
ogic and functional indices at the 12-month follow-up.13 ManyISSN 0161-6420/13/$–see front matterhttp://dx.doi.org/10.1016/j.ophtha.2012.10.023
dpdD2pgTiswpm
tpcci3pw
C
RdwhttrocwT2eipp4pSsic(lwmbt
wlgrd04
Vinciguerra et al � CXL as a Treatment for Keratoconus
reports in the literature have demonstrated the long-termresults with regard to the efficacy and safety of corneal CXLin the treatment of progressive keratoconus.11,19–24 The aimof this retrospective study was to evaluate the long-termoutcome of corneal CXL in a population of patients fromdifferent age groups.
Patients and Methods
Study Design
This was a retrospective, single-center, nonrandomized clinicalstudy.
Population
This retrospective study included all eyes that underwent cornealCXL from April 2006 through April 2010 at the Department ofOphthalmology, Istituto Clinico Humanitas, Rozzano, Milan, It-aly. The inclusion criteria for the treatment of corneal CXL weredocumentation of the progression of keratoconus, patients olderthan 9 years, signed informed consent, and corneal pachymetryresults of more than 400 �m.
The preoperative progression of keratoconus was demonstratedby at least 2 optical pachymetry results and corneal differentialtopography results obtained after approximately 6 months or ap-proximately 3 months for patients younger than 18 years and 1month for patients younger than 12 years. This difference inpreoperative evaluation time is supported by the evidence in theliterature that keratoconus in children is more aggressive than inadults.9 For this reason, pediatric patients were evaluated morefrequently to detect early progression. The parameters consideredto establish keratoconus progression always were proved withdifferential maps as change in curvature in the cone area of at least1.5 diopters obtained with an istantaneous map or as thinning ofmore than 20 �m in minimal pachymetry (depending on the timingof the different age groups). When progression was not demon-strated, follow-up in the pediatric group took place after 3, 6, and12 months to prove the stability.
Exclusion criteria were pachymetry results of less than 400 �m,a history of herpetic keratitis, dry eye, severe corneal infection, andconcomitant ocular or systemic autoimmune disease. Other exclu-sion criteria were a previous pregnancy or breastfeeding, thepresence of central or paracentral opacities, low compliance, andthe use of rigid contact lenses for more than 4 weeks before thebaseline evaluation.
The institutional review board ruled that approval was notrequired for this record review study. It was conducted accordinglyto the ethical standards set in the 1964 Declaration of Helsinki, asrevised in 2000. All patients provided informed consent.
During the preoperative and postoperative course (1, 6, 12, 24,36, and 48 months), the following parameters were assessed:best-corrected visual acuity (BCVA), slit-lamp biomicroscopy,Goldmann tonometry, dilated fundoscopy, corneal topography andaberrometry for the evaluation of low- and high-order aberrations(OPD II; Nidek, Gamagori, Japan), and optical tomography andpachymetry with a Pentacam (Oculus Optikgerate GmbH, Wetzlar,Germany).
Specifically, the Nidek OPD II was used to study topographicindices that provide data on the corneal shape, altered in kerato-conus. The surface asymmetry index (SAI) detects alterations incorneal regularity and is defined by the central weighted summa-tion of corneal power differences between corresponding points
180° apart on the mires. The opposite sector index (OSI) and aifferential sector index (DSI) divide the corneal surface into 8ie-like sectors with angles of 45°. The OSI gives the greatestifference in average power between opposite sectors, whereas theSI provides the greatest difference in average power between anysectors. Simulated keratometry 1 (SimK1) calculates the averageower on steepest meridian, and simulated keratometry 2 (SimK2)ives the average power of the meridian orthogonal to the steepest.he Nidek OPD II was also used to perform total (corneal and
nternal) wavefront analysis, but special attention was given in thistudy to corneal-generated aberrations. The Pentacam softwareas used to analyze anterior chamber and whole cornealachymetry, including central, superior, inferior, nasal, temporal,inimal, and apex pachymetry.
All preoperative and postoperative functional and morphologicests were performed in a manner identical to those in a previouslyublished clinical study.13 All data obtained were combined, andorrelation coefficients to age were assessed. No significant linearorrelation was found. For this reason, data were stratified accord-ng to decades of age: younger than 18 years, 18 to 29 years, 30 to9 years, and older than 40 years. Because of the small number ofatients between 18 and 19 years of age, this subgroup was mergedith the group 20 to 29 years of age.
orneal Cross-Linking Procedure
iboflavin ultraviolet A–induced corneal CXL was performed as aay-surgery procedure. It was conducted under topical anesthesiaith 2 applications of 4% lidocaine drops and 0.2% oxybuprocaineydrochloride. Thirty minutes before the procedure, pain medica-ion was administered and 2% pilocarpine drops were instilled inhe eye to be treated to reduce the amount of ultraviolet lighteaching the retina.13 The procedure was conducted under sterileperating conditions. After the lid speculum was applied, theorneal epithelium was removed in a central 9-mm–diameter areaith an Amoils brush (Amoils Brush Epithelial Scrubber; Visionechnology Co, Seoul, Korea). A solution of 0.1% riboflavin and0% dextran (Ricrolin; Sooft, Montegiorgio, Italy) was instilledvery minute for 30 minutes to irrigate the cornea fully. Duringrrigation, corneal thickness was tested using an ultrasoundachymeter (SP-2000; Tomey, Erlangen, Germany) to avoid ex-osure to ultraviolet light when pachymetry results were less than00 �m. When required, a hypo-osmolar 0.1% riboflavin solution,roduced by diluting 0.5% vitamin B2-riboflavin-5-phosphate (G.treuli & Co. AG, Uznach, Switzerland) with physiological saltolution (B. Braun Medical AG, Sempach, Switzerland), wasnstilled to promote corneal swelling. A 7.5-mm diameter of theentral cornea then was irradiated with an irradiance of 3 mW/cm2
UV-X System; Peschke Meditrade GmbH, Huenenberg, Switzer-and) for 30 minutes. During exposure time, the riboflavin solutionas again applied every 5 minutes. A calibrated ultraviolet Aeter (LaserMate-Q; Laser 2000, Wessling, Germany) was used
efore treatment to check the irradiance at a 1.0-cm distance. Bothopical anesthetics were added as needed during irradiation.
A soft therapeutic contact lens was applied until re-epithelializationas complete. After surgery, topical levofloxacin (Oftaquix; Tubi-
ux, Pahrama, Pomezia, Italy) and cyclopentolate (Ciclolux; Aller-an, Rome, Italy) was prescribed 4 times daily for 7 days and, afteremoval of contact lens, 0.15% dexamethasone 21-phosphaterops (Etacortilen; Sifi, Lavinaio, Italy) twice daily for 10 days and.15% sodium hyaluronate drops (BluYal; Sooft) 6 times daily for5 days were prescribed. In addition, all patients received oral
mino acid supplements (Aminoftal; Sooft) for 7 days.25909
ppoopftf1t
G
Sriadifcr
P�0.
Ophthalmology Volume 120, Number 5, May 2013
Statistical AnalysisData are expressed as number and percentage or mean�standard deviation, as appropriate. The comparison of variableswas performed on paired data to evaluate the improvement orrelapse of individual patients over time. Differences betweendata were evaluated with the Wilcoxon signed rank test, thechi-square test with Fisher correction when necessary, andanalysis of variance.
Variations of data over time are presented with box-and-whiskersplots (Figures 1–7). The graphs present the distribution of vari-ables: the box represents from the 25th to the 75th percentile, andthe internal line represents the median. The whiskers are limited byupper and lower adjacent values. Outside the whiskers, outliers canbe seen. A P value of less than 0.05 was considered significant.Statistical analysis was performed using Stata software version 11(Stata Corp., College Station, TX).
Results
Four hundred eyes of 301 patients were evaluated. The compara-tive functional analysis comprised the following: 49 eyes of
Figure 1. Box-and-whisker plot for simulated keratometry 1 (SimK1) shothe global population and (B) for the age-related subgroups. *P�0.05; **
Figure 2. Box-and-whisker plot for opposite sector index (OSI): showing d
population and (B) for the age-related subgroups. *P�0.05; **P�0.01. mos �910
atients between 9 and 17 years of age (12.25%), 185 eyes ofatients between 18 and 29 years of age (46.25%), 115 eyesf patients between 30 and 39 years of age (28.75%), and 51 eyesf patients older than 40 years of age (12.75%). The mean age ofatients was 29�10 years, and the sex distribution was 27.25%emale (109 eyes). The postoperative follow-up period was dis-ributed as follows: 53.5% of patients had a 1-year minimumollow-up, 12.75% of patients had a 2-year follow-up, and 5% and.75% of patients had a 3- and 4-year follow-up periods, respec-ively.
lobal Population
tructural Analysis. Topographic Results. Comparative SimK1esults showed a mean increase at month 1, followed by a signif-cant decrease at 6, 12, 24, 36, and 48 months of follow-up (seelso Fig 1A). The SimK2 results for the global population showedata similar to the SimK1 results, except for a nonsignificantmprovement at the 48-month follow-up visit. Comparative resultsor the topographic SAI, DSI, and OSI showed a mean increaseompared with baseline at month 1, followed by a significanteduction at all follow-up times (see Fig 2A).
differences between follow-up and preoperative values of SimK1 (A) of01. mos � months.
nces between follow-up and preoperative values of OSI (A) for the global
wing
iffere
months.
it(
raOatwrfetmgsSp(
5; **P
Vinciguerra et al � CXL as a Treatment for Keratoconus
Pachymetry Results. Pachymetry results for the global pop-ulation are summarized in Table 3 (available online at http://aaojournal.org). All pachymetry values were decreased signifi-cantly at month 1 compared with preoperative values. Allremaining follow-up values were compared with those from month1 to evaluate the previously reported increase in collagen thicknessafter corneal CXL.18,26 Minimal and apex pachymetry showedsignificant improvement at months 12 and 24 (Fig 3A); a trendtoward improvement also was reported at month 36 (P � 0.059 forboth minimal and apex pachymetry).
Functional Analysis. Visual acuity, refractive, and topo-graphic results are summarized in Table 1 (available online athttp://aaojournal.org).
Visual Acuity. After an initial increase at month 1 resultingfrom epithelial thickness change, the BCVA improved signifi-cantly at 6, 12, 24, 36, and 48 months of follow-up (Fig 4A).
Refractive Results. Except for an increase at month 1, com-parative spherical equivalents showed a significant reduction after12, 24, and 36 postoperative months (see also Fig 5A). The globalpopulation results for cylinder variation presented a significant
Figure 3. Box-and-whisker plot for minimal pachymetry showing differenthe global population and (B) for the age-related subgroups (B). *P�0.0
Figure 4. Box-and-whisker plot for best-corrected visual acuity (BCVA)
for the global population and (B) for the age-related subgroups. *P�0.05; **Pncrease at month 1 compared with baseline, followed by a returno a nonsignificant improvement at the other follow-up timesminimum P � 0.224).
Aberrometric Results. Aberrometric results are summa-ized in Table 2 (available online at http://aaojournal.org). Theberrometric results included fewer observations because thePD software is not always able to acquire the specific corneal
berrations in advanced keratoconus, limiting the calculation tootal aberrations. All comparisons between aberrometric dataere made with absolute values to evaluate possible aberromet-
ic shifts from negative to positive. The diameter of the wave-ront area evaluated was 6 mm and the order was up to theighth order of Zernike polynomials. Comparative results forotal aberrometry showed a significant improvement at allonths of follow-up to 48 months, excluding month 1. The
lobal population’s results for comatic aberration showed in-tead a significant reduction at all follow-up periods (Fig 6A).pherical aberration, after an initial increase at month 1, im-roved significantly at the 24- and 36-month follow-up visitsFig 7A).
etween follow-up and preoperative values of minimal pachimetry (A) of�0.01. mos � months.
ing differences between follow-up and preoperative values of BCVA (A)
ces b
show
�0.01. mos � months.911
rmasl
P
Sccsfma1m
m
P�0.
Ophthalmology Volume 120, Number 5, May 2013
Pediatric Population
Structure Analysis. Topographic Results. The analysis of thetopographic results for the pediatric population showed a signifi-cant increase of SimK1, SimK2, OSI, and DSI at month 1 com-pared with baseline. After month 1, the pediatric populationshowed significant SAI improvement at month 12. All other pa-rameters showed improvement that was not significant.
Pachymetry Results. After a significant reduction at month 1,minimal and apex pachymetry showed no significant differencescompared with 1 through 24 months of follow-up (Fig 3B).
Functional Analysis. Visual Acuity. A maximum of 2 yearsof follow-up was available for visual acuity results in the pediatricpopulation. Comparative results showed a significant improvementin BCVA at 6 and 12 postoperative months; after this time, theimprovement was not significant because of the small number ofdata (Fig 4B).
Refractive Results. In the pediatric subgroup, no significantimprovement in spherical equivalent was reported, except formonth 1 (Fig 5B). Comparative results for cylinder showed asignificant increase for this group at months 6 and 12.
Figure 5. Box-and-whisker plot for spherical equivalent showing differenthe global population and (B) for the age-related subgroups. *P�0.05; **
Figure 6. Box-and-whisker plot for coma aberration showing differences b
population and (B) for the age-related subgroups. *P�0.05; **P�0.01. mos �912
Aberrometric Results. Comparative results for total aber-ometry did not show any significant differences, whereas co-atic aberration showed a significant decrease at months 1, 6,
nd 12 (Fig 6B). Excluding a significant increase at month 1,pherical aberration exhibited no significant change from base-ine through 12 months of follow-up (Fig 7B).
opulation 18 to 29 Years of Age
tructural Analysis. Topographic Results. A significant in-rease in SimK1, SimK2, OSI, and DSI was observed at month 1ompared with baseline. After month 1, this group showed aignificant decrease in SimK1 at 12, 24, and 36 months ofollow-up (Fig 1B). Similarly, SimK2 decreased significantly atonths 6, 12, and 24. A significant decrease of SAI was observed
t all follow-up periods; OSI also decreased significantly at months2, 24, and 36 (Fig 2B), and DSI improved at 6, 12, 24, and 36onths of follow-up.
Pachymetry Results. After a significant reduction at month 1,inimal pachymetry showed a significant increase, compared with
tween follow-up and preoperative values of spherical equivalent (A) for01. mos � months.
n follow-up and preoperative values of coma aberration (A) for the global
ces be
etwee
months.
iw6c
P
Snwd(tam
mc3
g2
aNm
caSpN
CCpcbap
. mos
Vinciguerra et al � CXL as a Treatment for Keratoconus
month 1, after 6 and 12 months of follow-up (Fig 3B). Apexpachymetry showed a similar response in these follow-up periodsand an additional improvement at 24 months and 1 month, yet nosignificant improvement at month 36 (P � 0.07).
Functional Analysis. Visual Acuity. Again excludingmonth 1, comparative results for BCVA in the 18- to 29-year agebracket showed significant improvement for up to 36 months offollow-up (Fig 4B).
Refractive Results. Except in month 1, no significant in-crease in spherical equivalent or cylinder was reported for up to 48months of follow-up (Fig 5B).
Aberrometric Results. Comparative results for total aber-rometry showed a significant improvement after 6, 12, 24, and36 months. Comatic aberration showed a significant mean de-crease, compared with baseline, until the 24-month follow-upvisit (Fig 6B). Spherical aberration showed a significant meanincrease at month 1, followed by a nonsignificant decrease (at24 and 36 months of follow-up, P � 0.062 and P � 0.068,respectively; Fig 7B).
Population 30 to 39 Years of Age
Structural Analysis. Topographic Results. Compared withbaseline, SimK1, SimK2, SAI, OSI, and DSI all were increasedsignificantly at month 1. After this, topographic indices showedsignificant improvement in the following: SimK1 at months 6, 12,24, and 36 (Fig 1B); SimK2 at months 6 and 12; SAI at allfollow-up times; OSI at months 6, 12, 24, and 36 (Fig 2B); andDSI at months 6, 12, 24, and 36.
Pachymetry Results. After a significant reduction at month 1,minimal pachymetry increased, compared with month 1, at months12 and 24 (Fig 3B). Again, after a significant reduction at month1, apex pachymetry increased significantly, compared with month1, at months 12 and 24.
Functional Analysis. Visual Acuity. Comparative resultsshowed a significant improvement of BCVA at 12, 24, and 36postoperative months (Fig 4B); at 48 months, the improvementwas not significant because of the small number of data (P �0.068).
Refractive Results. Results of spherical equivalent, after aninitial worsening, showed a significant improvement at 12 and 24postoperative months (Fig 5B). Cylinder also showed, after an
Figure 7. Box-and-whisker plot for spherical aberration showing differenceglobal population and for the age-related subgroups. *P�0.05; **P�0.01
initial worsening, a significant decrease for this group at month 24. a
Aberrometric Results. Total aberrometry showed a signif-cant improvement at 6, 12, and 48 postoperative months. Comaas decreased significantly at months 1, 6, 12, 24, and 36 (FigB). Spherical aberration did not show a significant differenceompared with baseline in any follow-up time (Fig 7B).
opulation Older Than 40 Years of Age
tructural Analysis. Topographic Results. At month 1, a sig-ificant increase of SimK1 and SimK2 was observed comparedith baseline. After the first postoperative month, a significantecrease of SimK1 was seen at 6, 12, and 24 months of follow-upFig 1B). Only at month 24 did SimK2 decrease significatly. Theopographic index revealed a significant decrease in SAI and OSIt 12 months (Fig 2B); DSI showed a nonsignificant decrease atonth 12 (P � 0.057).
Pachymetry Results. After a significant reduction at month 1,inimal and apex pachymetry showed no significant difference,
ompared with month 1, for up to 48 months of follow-up (FigB).
Functional Analysis. Visual Acuity. The BCVA for thisroup showed a significant difference, compared with baseline, at4 months of follow-up (P � 0.034).
Refractive Results. Results for spherical equivalent revealedsignificant improvement after 24 months of follow-up (Fig 5B).o significant difference in cylinder was reported until the 48-onth follow-up visit.
Aberrometric Results. Total aberrometry showed a signifi-ant improvement after the first postoperative month. Comaticberration decreased significantly at months 1, 6, and 12 (Fig 6B).pherical aberration did not show any significant difference, com-ared with baseline, for up to 48 months of follow-up (Fig 7B).one of the patients required a second corneal CXL treatment.
omparison of 1-Year Resultsomparison after 1 year of follow-up was performed in the globalopulation and in the subgroups on an explorative basis. Thisomparison indicated that the group aged 30 to 39 years was theest responder compared with the other age groups. The compar-tive analysis for the parameters visual acuity (P � 0.059 withediatric group, P � 0.014 with the group 18 to 29 years of age,
een follow-up and preoperative values of spherical aberration (A) for the� months.
s betw
nd P � 0.071 with group older than 40 years), cylinder, and
913
ctchtdgs
P
Fsfirpp6iwiwmvt
P
FimiSodparsttd
P
Fpmivtitopctyp
Ophthalmology Volume 120, Number 5, May 2013
SimK1 (compared with the pediatric group, P � 0.006 and P �0.042, respectively) showed a significant difference.
Discussion
Corneal collagen CXL modifies the natural history of pro-gressive keratoconus by stabilizing the corneal tissue, pre-venting and slowing the progression of the ectatic dis-ease.11–14 The outcome of this procedure was stratified inthis retrospective study according to age. Because kera-toconus in children is more aggressive,9 pediatric patientswere evaluated more frequently than adults. The avail-ability of an extended 48-month follow-up allowed forextensive data collection aimed at defining how well thestability induced by corneal CXL in these patients ismaintained.
Global Population
After an initial worsening, there was a significant improvementin visual acuity for up to 48 months of follow-up. Parallel tothis improvement of visual acuity, significant morphologicchanges were demonstrated. In particular, a significant de-crease in spherical equivalent and a decrease in topographicparameters such as SimK1, SimK2, SAI, OSI, and DSI con-firmed data from previous reports demonstrating that cornealCXL can regularize corneal shape.11–13,27 The aberrometricresults included fewer observations because the OPD soft-ware cannot always detect specific corneal aberrations inadvanced keratoconus, thus limiting the study to the calcu-lation of total aberrometry. These data for total aberrometryshowed a significant decrease throughout the follow-upperiod, excluding month 1; conversely, the comatic aberra-tion decreased significantly through the follow-up periodfrom the first postoperative month. These findings suggestthat remodeling induced by corneal CXL and the conse-quent improvement in visual acuity not only is the result ofthe improvement of refraction and topographic parameters,but also of improvement of aberrations that can not beassessed with simple refractive indices. Finally, results forapex and minimum pachymetry showed a significant de-crease of corneal thickness in the first month of follow-up,followed by a significant increase. This initial drop wasshown to be the result of a decrease in interlamellar space,and the subsequent rise was shown to be the result of anincrease over time in the diameter of corneal collagenfibers.18,28 From the analysis, either the minimal or the apexpachymetry is not statistically different from the preopera-tive measurements after 36 months of follow-up.
These results suggest that corneal CXL induces a higherdegree of corneal stability, even beyond the estimated 2 to3 years it takes for corneal collagen turnover.29,30 Thiscould be explained by 3 hypotheses: a change in the char-acteristics of the keratocyte population, a slower turnover ofcross-linked fibrils, or that the corneal collagen turnover isslower than what was estimated. The first hypothesis couldbe supported by the evidence that stromal keratocytes inkeratoconus are associated with a patchy loss of CD34
immunoreaction, a marker of keratocytes. Six months after b914
orneal CXL, CD34 positivity is distributed regularlyhroughout the entire cornea.26 A pool of quiescent stemells may replace the apoptotic keratocytes. The secondypothesis is supported by Spoerl et al,31 who demonstratedhat cross-linked corneas are more resistant to collagen-egrading enzymes, leading to a slower turnover of colla-en. It is not clear, however, what could be the maximumtabilization time induced by corneal CXL.
ediatric Population
unctional results for the pediatric population showed aignificant improvement of BCVA at 6 and 12 months ofollow-up; after this time, the improvement was not signif-cant because of the small number of data. Morphologicesults for this group showed stability over time for allarameters (refractive, topographic, aberrometry, andachymetry), except for a significant increase in cylinder atand 12 months of follow-up. Comatic aberration showed
n the same period a significant improvement comparedith baseline. This evidence could explain the significant
mprovement of BCVA in the same follow-up period thatas not supported by other improvements except for co-atic aberration. These results differ from those of a pre-
iously published study that showed slightly better results inhe pediatric population.20
opulation 18 to 29 Years of Age
unctional results for this population showed a significantmprovement of BCVA at all follow-up periods, excludingonth 1. Morphologic results similarly showed significant
mprovement in topographic indices such as SimK1,imK2, SAI, OSI, and DSI up to a maximum of 36 monthsf follow-up. Aberrometric results indicated significant re-uctions in total aberration at up to 36 months of follow-up,arallel to a reduction in specific comatic and sphericalberration for up to 24 months of follow-up. Pachymetryesults showed a significant decrease at month 1 and sub-equently a significant increase compared with month 1;hese results are in line with other reports in the literaturehat demonstrated increased corneal thickness after an initialecrease resulting from collagen packing.18,28
opulation 30 to 39 Years of Age
unctional results for this group showed a significant im-rovement of BCVA at all follow-up visits, excludingonths 1 and 48. Morphologic results showed a parallel
mprovement of refractive, topographic, and aberrometricalues for up to 36 months of follow-up. Similar to the 18-o 29-year age group, pachymetry results showed a signif-cant increase compared with those from month 1. Again,hese results diverge from the age-related study by Capor-ssi et al20 that showed a poorer functional response inatients older than 27 years. The reason for this dissimilarityould be that the authors did not consider the subgroup 30o 39 years of age, but analyzed all patients older than 27ears. The improvement in the 30- to 39-year age group wasrobably not manifest in the statistical analysis of the study
ecause of the presence of patients older than 40 years.
enttdso
catam
R
1
1
1
1
1
1
Vinciguerra et al � CXL as a Treatment for Keratoconus
Population Older Than 40 Years
Despite the lack of functional improvement compared withbaseline, no patient had progression of disease for up to 48months of follow-up. Selected refractive, topographic, andaberrometric results (spherical equivalent, SAI, OSI, andcomatic aberration) did show improvement over time aftersurgery.
Our statistical analysis divided by age does not show anysignificant progression, according to the measured parame-ters, in any of the age groups at any time. Moreover,outcomes divided by age demonstrate the efficacy of cor-neal CXL in improving functional and morphologic param-eters in selected groups. The main finding suggests thatpatients between 18 and 39 years of age respond best tocorneal CXL therapy. This result is in contrast to a recentstudy that reports worse results in patients older than 27years.20
Although a lesser degree of success was observed in thepediatric patients and those older than 40 years comparedwith the other age groups, corneal CXL nevertheless wasable to stabilize the ectatic disease and to improve visualacuity in the pediatric group, as well as improve othersignificant morphologic changes in patients older than 40years. A more aggressive disease in the pediatric popula-tion9,20 probably resulted in the observed reduced therapeu-tic efficacy of corneal CXL.
In patients older than 40 years, the limited improvementof corneal CXL is probably the result of so-called age-related cross-links. These bonds, which are catalyzed spon-taneously or enzymatically from a reducible state, are morestable32 and, for this reason, less prone to the formation ofnew corneal CXL.
Altered collagen metabolism is at the basis of the patho-logic process in keratoconus. Cannon and Davison32 andCannon and Foster33 reported that in keratoconic corneas,the specific activity of reduced keratoconus corneal collagenis higher than in normal corneas but substantially lower thanin a fetus or child. As reported by the authors, collagen CXLbonds are reducible when first formed, but then, eitherspontaneously or by enzyme catalysis, this reducible char-acter is lost in normal tissue. In keratoconus, the persistenceof reducible cross-links in old tissue may be the result ofremodeling changes, continued slow turnover, or replace-ment synthesis, indicating that an abnormally high catabo-lism may be present in this disease. In a fetus and inchildren, the increase in reducible bonds is believed to bethe result of an increase in collagen synthesis.32,33 Thisincreased turnover of collagen fibers, even present in normalchildren compared with adults, could be at the root of moresevere disease and reduced improvement in children.
To the best of our knowledge, there are no other pub-lished procedures with better outcomes for the age group 18to 39 years. One study34 reported a better outcome inadolescents treated with corneal transplantation, whereasanother study showed worse outcomes in pediatric pa-tients.35 One report from 2008 showed good results ofintracorneal ring segments (INTACS) in all the evaluated
age groups.36This difference may be explained because of the differ-nt types of techniques used. In particular, the other tech-iques used for the treatment of keratoconus aim to replacehe ectatic tissue (transplantation) or to remodel it (refrac-ive surgery or INTACS). Corneal CXL uses a completelyifferent approach that aims to increase the biomechanicaltiffness of the tissue; this could lead to the difference inutcome.
In conclusion, this analysis suggests better results fororneal CXL treatment of keratoconus patients between 18nd 39 years of age. This information could be valuable forhe ophthalmic surgeon when selecting appropriate patientsnd predicting the outcomes of this procedure for the treat-ent of keratoconus.
eferences
1. Dobbins KR, Price FW Jr, Whitson WE. Trends in the indi-cations for penetrating keratoplasty in the midwestern UnitedStates. Cornea 2000;19:813–6.
2. Al-Yousuf N, Mavrikakis I, Mavrikakis E, Daya SM. Pene-trating keratoplasty: indications over a 10 year period. Br JOphthalmol 2004;88:998–1001.
3. Nielsen K, Hjortdal J, Aagaard Nohr E, Ehlers N. Incidenceand prevalence of keratoconus in Denmark. Acta OphthalmolScand 2007;85:890–2.
4. Steele TM, Fabinyi DC, Couper TA, Loughnan MS. Preva-lence of Orbscan II corneal abnormalities in relatives of pa-tients with keratoconus. Clin Experiment Ophthalmol 2008;36:824–30.
5. Hall KG. A comprehensive study of keratoconus. Br J PhysiolOpt 1963;20:215–56.
6. Rabinowitz YS, Rasheed K, Yang H, Elashoff J. Accuracy ofultrasonic pachymetry and videokeratography in detectingkeratoconus. J Cataract Refract Surg 1998;24:196–201.
7. Rabinowitz YS. Keratoconus. Surv Ophthalmol 1998;42:297–319.
8. Reeves SW, Stinnett S, Adelman RA, Afshari NA. Risk fac-tors for progression to penetrating keratoplasty in patients withkeratoconus. Am J Ophthalmol 2005;140:607–11.
9. Tuft SJ, Moodaley LC, Gregory WM, et al. Prognostic factorsfor the progression of keratoconus. Ophthalmology 1994;101:439–47.
0. Vanathi M, Panda A, Vengayil S, et al. Pediatric keratoplasty.Surv Ophthalmol 2009;54:245–71.
1. Caporossi A, Mazzotta C, Baiocchi S, Caporossi T. Long-term results of riboflavin ultraviolet A corneal collagencross-linking for keratoconus in Italy: the Siena Eye CrossStudy. Am J Ophthalmol 2010;149:585–93.
2. 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.
3. Vinciguerra P, Albe E, Trazza S, et al. Refractive, topo-graphic, tomographic, and aberrometric analysis of kerato-conic eyes undergoing corneal cross-linking. Ophthalmology2009;116:369–78.
4. 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.
5. Wollensak G, Spoerl E, Seiler T. Stress-strain measurements
of human and porcine corneas after riboflavin-ultraviolet-A-915
2
2
2
2
2
3
3
3
3
3
3
3
Ophthalmology Volume 120, Number 5, May 2013
induced cross-linking. J Cataract Refract Surg 2003;29:1780–5.
16. Spoerl E, Terai N, Scholz F, et al. Detection of biomechanicalchanges after corneal cross-linking using Ocular ResponseAnalyzer software. J Refract Surg 2011;27:452–7.
17. Wollensak G, Iomdina E. Biomechanical and histologicalchanges after corneal crosslinking with and without epithelialdebridement. J Cataract Refract Surg 2009;35:540–6.
18. Wollensak G, Wilsch M, Spoerl E, Seiler T. Collagen fiberdiameter in the rabbit cornea after collagen crosslinking byriboflavin/UVA. Cornea 2004;23:503–7.
19. Ashwin PT, McDonnell PJ. Collagen cross-linkage: a com-prehensive review and directions for future research. Br JOphthalmol 2010;94:965–70.
20. Caporossi A, Mazzotta C, Baiocchi S, et al. Age-related long-term functional results after riboflavin UV A corneal cross-linking. J Ophthalmol [serial online] 2011;2011:608041. Avail-able at: http://www.hindawi.com/journals/jop/2011/608041/.Accessed September 24, 2012.
21. Gkika M, Labiris G, Kozobolis V. Corneal collagen cross-linking using riboflavin and ultraviolet-A irradiation: a reviewof clinical and experimental studies. Int Ophthalmol 2011;31:309–19.
22. Hoyer A, Raiskup-Wolf F, Sporl E, Pillunat LE. Collagencross-linking with riboflavin and UVA light in keratoconus.Results from Dresden [in German]. Ophthalmologe 2009;106:133–40.
23. Kampik D, Koch M, Kampik K, Geerling G. Corneal riboflavin/UV-A collagen cross-linking (CXL) in keratoconus: two-year results [in German]. Klin Monbl Augenheilkd 2011;228:525–30.
24. Wollensak G, Iomdina E. Long-term biomechanical propertiesof rabbit cornea after photodynamic collagen crosslinking.
Acta Ophthalmol 2009;87:48–51.3 Studio Oculistico VinciEye, Milan, Italy.
PO
FT
P
CPR
916
5. Torres Munoz I, Grizzi F, Russo C, et al. The role of aminoacids in corneal stromal healing: a method for evaluatingcellular density and extracellular matrix distribution. J RefractSurg 2003;19(suppl):S227–30.
6. Mencucci R, Marini M, Paladini I, et al. Effects of riboflavin/UVA corneal cross-linking on keratocytes and collagen fibres inhuman cornea. Clin Experiment Ophthalmol 2010;38:49–56.
7. Vinciguerra P, Albe E, Trazza S, et al. Intraoperative andpostoperative effects of corneal collagen cross-linking on pro-gressive keratoconus. Arch Ophthalmol 2009;127:1258–65.
8. Mencucci R, Marini M, Gheri G, et al. Lectin binding innormal, keratoconus and cross-linked human corneas. ActaHistochem 2011;113:308–16.
9. Nishida T. Cornea. In: Krachmer JH, Mannis MJ, Holland EJ,eds. Cornea. Vol. 1: Basic Science. St. Louis, MO: Mosby;1997:3–28.
0. Smelser GK, Polack FM, Ozanics V. Persistence of donorcollagen in corneal transplants. Exp Eye Res 1965;4:349–54.
1. Spoerl E, Wollensak G, Seiler T. Increased resistance ofcrosslinked cornea against enzymatic digestion. Curr Eye Res2004;29:35–40.
2. Cannon DJ, Davison PF. Aging, and crosslinking in mamma-lian collagen. Exp Aging Res 1977;3:87–105.
3. Cannon DJ, Foster CS. Collagen crosslinking in keratoconus.Invest Ophthalmol Vis Sci 1978;17:63–5.
4. Lowe MT, Keane MC, Coster DJ, Williams KA. The outcomeof corneal transplantation in infants, children, and adolescents.Ophthalmology 2011;118:492–7.
5. Limaiem R, Chebil A, Baba A, et al. Pediatric penetratingkeratoplasty: indications and outcomes. Transplant Proc 2011;43:649–51.
6. Ertan A, Ozkilic E. Effect of age on outcomes in patients withkeratoconus treated by Intacs using a femtosecond laser. J
Refract Surg 2008;24:690–5.Footnotes and Financial Disclosures
Originally received: February 1, 2012.Final revision: October 7, 2012.Accepted: October 17, 2012.Available online: January 3, 2013. Manuscript no. 2012-143.1 Humanitas Clinical and Research Center, Rozzano, Italy.2 Department of Surgical and Morphological Sciences, University of In-subria, Circolo Hospital, Varese.
resentated at: American Academy of Ophthalmology Annual Meeting,ctober 2011, Orlando, Florida.
inancial Disclosure(s):he author(s) have made the following disclosure(s):
aolo Vinciguerra - Consultant - Nidek and Oculus.
orrespondence:aolo Vinciguerra, Istituto Clinico Humanitas, Via Manzoni 56, 20089
ozzano (Milan), Italy. E-mail: [email protected].