Corneal curvature and central corneal thickness in eyeswith pseudoexfoliation syndromeIbrahim F. Hepsen, MD; Ramazan Yagcı, MD; Urgcan Keskin, MD

ABSTRACT • RÉSUMÉ

Background: Previous studies have reported different central corneal thickness (CCT) values in eyes withpseudoexfoliation syndrome (PXS).There has been no report about corneal curvature (CC) in PXS eyes. Theaim of this study was to investigate the CCT together with the CC in PXS eyes with and without glaucoma.

Methods: The CC (simulated keratometry 1 [SimK1], SimK2, and mean K) and CCT were prospectivelystudied by rotating a Scheimpflug camera in 72 PXS eyes and comparing them with 65 normal eyes.

Results: In all PXS eyes, the mean K was significantly steeper than in control eyes (44.1 [SD 1.5] diopters (D)vs. 43.2 [SD 2.0] D, p = 0.04).When PXS eyes were subcategorized, the mean K was significantly steeper inboth normotensive PXS eyes (44.6 [SD 1.6] D) and pseudoexfoliation glaucoma (PEXG) eyes (44.0 [SD 1.2]D) than normal eyes (43.2 [SD 2.0] D) (p = 0.001 and 0.02, respectively). In all PXS eyes, the mean CCT(546.4 [SD 39.6] μm) was not significantly different than the control eyes (542.9 [SD 32.2] μm) (p = 0.56).When PXS eyes were subcategorized, however, the CCT was significantly thinner in normotensive PXS eyesand significantly thicker in PEXG eyes.

Interpretation: It appears that PEXG eyes have a significantly steeper CC and higher CCT thannormal eyes.This could be of clinical significance because overestimation of true intraocularpressure may then occur in these eyes.

Contexte : Des études précédentes font état des diverses épaisseurs du centre de la cornée (ÉCC) dans desyeux atteints du syndrome de pseudoexfoliation (SPX), mais aucune ne traite de la courbure de la cornée(CC) des yeux atteints du SPX. Cette étude a pour objet d’examiner conjointement l’ÉCC et la CC desyeux atteints du SPX.

Méthodes : Il s’agit d’une étude prospective de la CC (par kératométrie simulée 1 [SimK1], SimK2 et Kmoyenne) et de l’ÉCC par rotation de la caméra Scheimpflug dans 72 cas de SPX comparativement à65 yeux normaux.

Résultats : Dans tous les yeux atteints du SPX, la K moyenne était significativement plus prononcée que dansles yeux témoins (44.1 dioptres (D) [ÉT 1.5] vs. 43.2 D [ÉT 2.0], p = 0.04). En classant par sous-catégoriesles yeux atteints du SPX, la K moyenne était significativement plus prononcée dans les yeux atteints du SPXnormotensifs (44.6 D [ÉT 1.6]) et ceux atteints de glaucome par pseudoexfoliation (GSPX) (44.0 D [ÉT 1.2]) que dans les yeux normaux (43.2 D [ÉT 2.0]) (p = 0.001 et 0.02, respectivement). Chez tous lesyeux atteints du SPX, la moyenne de l’ÉCC (546.4 [ÉT 39.6] μm) n’était pas significativement différente decelles des yeux témoins (542.9 [ÉT 32.2] μm) (p = 0.56).Toutefois, quand on répartissait les yeux atteintsdu SPX en sous-categories, l’ÉCC était significativement moindre dans les yeux normotensifs etsignificativement plus grande dans les yeux atteints du SPX.

Interprétation : Il semble que chez les yeux atteints du GSPX la courbure de la cornée est sig-nificativement plus prononcée et le centre de la cornée plus épais que chez les yeux normaux.Cela peut être important en clinique parce qu’il peut arriver qu’on surestime la véritablepression intraoculaire de ces yeux.

Goldmann applanation tonometry (GAT) is still theinternational gold standard for measurement of

intraocular pressure (IOP). The effect of central cornealthickness (CCT) on IOP readings using GAT has recentlybecome a topic of much interest. Patients with a diagno-sis of ocular hypertension are more likely to have anincreased CCT,1–3 whereas patients with a diagnosis ofnormal-tension glaucoma (NTG) may have a reduced

CCT.1,4 There are conflicting reports about the CCT inpatients with pseudoexfoliation syndrome (PXS) orpseudoexfoliation glaucoma (PEXG). Most studies5–8

have found similar CCT in PXS eyes and normal eyes,although some authors have reported lower9–11 orhigher12 values than the CCT of normal eyes.

The aim of this study was to investigate the CCT in PXSeyes with normal IOP, elevated IOP without glaucoma,

From the Department of Ophthalmology, School of Medicine, FatihUniversity, Ankara, Turkey

Originally received May 20, 2006. Revised Mar. 5, 2007Accepted for publication June 8, 2007Published online Sep. 13, 2007

Correspondence to: Ibrahim F. Hepsen, MD, Fatih University Hospital, EyeCenter, Alparslan Turkes C. No. 57, 06510 Emek, Ankara, Turkey; hepsenif@hotmail.com

This article has been peer-reviewed. Cet article a été évalué par les pairs.

Can J Ophthalmol 2007;42:677–80doi: 10.3129/can j ophthalmol.i07-145

Corneal curvature and thickness in PXS—Hepsen et al. 677

and those with glaucoma, with the corneal curvature (CC)representing another parameter possibly affecting IOPmeasurements, and to compare the results with those innormal eyes. To the best of our knowledge, there is noreport about corneal curvature in eyes with PXS or PEXG.

METHODS

The patients were recruited from the GlaucomaDepartment of T. Ozal Medical Center. The study fol-lowed the Declaration of Helsinki for research involvinghuman subjects and all participants provided informedconsent. The study included 72 volunteer patients withPXS with and without glaucoma and 65 control subjectswho had undergone preoperative evaluation for senilecataract. The patients were excluded if they had hadintraocular surgery or laser procedures, intraocular inflam-mation, angle closure on ocular history, severe or end-stageglaucoma damage, previous or present contact lens wear,early or suspect keratoconus on topography, post-menopausal hormonal substitution, and diabetic retinopa-thy. Prior to the pentacam examination, the study orcontrol group had not had ocular examination, such ascontact or noncontact tonometry or pupil dilatation, thatcould have influenced the results. As gonioscopy may inter-fere with the measurement of corneal curvature, the sub-jects did not undergo gonioscopic examination on the dayof corneal topography.

Pseudoexfoliation was diagnosed by the presence of pre-granular radial lines or typical white granular deposits onthe anterior lens surface and (or) at the pupillary margin.We studied the CC and CCT in patients with PXS andcompared them with the values in normal persons. Oneeligible eye of each participant was included for analysis.The 72 eyes with PXS were subcategorized into nor-motensive PXS, hypertensive PXS, and PXS with glau-coma. In patients with bilateral PXS together with elevatedIOP or glaucoma, the eye with the lower mean deviationon visual field testing (i.e., the better eye) was selected.Ocular hypertension was defined as an IOP greater than21 mm Hg with no glaucomatous defects on visual field

testing and normal appearance of the optic disc. Glaucomawas diagnosed by Humphrey 24-2 full threshold visualfield test, IOP measurement, and gonioscopic and opticnerve head examination with a 90 diopter (D) lens. PEXGpatients had open-angle, glaucomatous optic disc andvisual field changes, and IOP greater than 21 mm Hg.

Corneal topographic and pachymetric measurementswere taken by using a noncontact, noninvasive rotatingScheimpflug camera system (Pentacam, Oculus Inc.). Itenables cross-sectional images of the anterior eye to becaptured. Topographic analysis of the corneal front andback surfaces was based on the true elevation measure-ment from limbus to limbus. The front surfaces of thecentral cornea were selected for curvature analysis in anaxial (sagittal) representation map. Pachymetric measure-ments of the cornea were also calculated from the true ele-vation points of the corneal front and back surfaces. Theapex pachymetry readings were used for the pachymetryvalue of the central cornea.

Statistical evaluation was performed using the StatisticalPackage for the Social Sciences software (SPSS v. 12,Chicago, Ill.). The independent samples test (t test) was usedfor comparisons of means between 2 groups. A p value of<0.05 was considered significant. The Pearson correlation testwas used to evaluate any correlation between CC and CCT.

RESULTS

The mean age was 67.2 (range: 50–85) years in all PXSpatients and 61.9 (range: 42–78) years in normal subjects.In the study group, 25 patients had normal IOP, 22 hadhigh IOP, and the remaining 25 had mild to moderateglaucoma. Eleven patients had unilateral PXS.

The mean CC and CCT values obtained from normaland pseudoexfoliative eyes are shown in Table 1. ThePentacam Scheimpflug slit scan analysis revealed that theCC (simulated keratometry 1 [SimK1], SimK2, andmean K) was significantly steeper in all PXS eyes than incontrol eyes (Table 1). When PXS eyes were subcategorizedaccording to IOP and glaucomatous damage, the CC wassignificantly steeper in both normotensive PXS eyes and

Corneal curvature and thickness in PXS—Hepsen et al.

678 CAN J OPHTHALMOL—VOL. 42, NO. 5, 2007

Table 1—Demographic data, corneal curvature, and central corneal thickness values in normal eyes and eyes with pseudoexfoliation syndrome with normal IOP, elevated IOP, and glaucoma

Pseudoexfoliation eyes, mean (SD)

Characteristic Control, mean

(SD) Total p value Normal IOP p value OHT p value Glaucoma p value

SimK1 43.7 (2.0) 44.6 (1.6) 0.003* 45.2 (1.8) 0.001* 44.0 (1.6) 0.42* 44.6 (1.2) 0.01*

SimK2 42.7 (2.0) 43.6 (1.4) 0.007* 44.0 (1.4) 0.001* 43.2 (1.7) 0.29* 43.5 (1.2) 0.04*

Mean K 43.2 (2.0) 44.1 (1.5) 0.004* 44.6 (1.6) 0.001* 43.6 (1.6) 0.39,* 0.03† 44.0 (1.2) 0.02,* 0.13†

CCT (μm) (95% CI)

542.9 (32.2) (481–601)

546.4 (39.6) (427–634)

0.56* 528.1 (36.9) (427–598)

0.05* 550.4 (36.0) (482–613)

0.35,* 0.04† 561.3 (39.4) (497–634)

0.004,* 0.003†

Age 61.9 (1.2) 67.2 (8.4) 0.0001* 66.8 (7.9) 0.01* 64.7 (8.2) 0.18* 69.9 (8.6) 0.0001*

IOP‡ 14.8 (2.3) 21.9 (6.9) 0.0001* 14.1 (1.6) 0.91* 23.4 (1.9) 0.0001* 28.4 (5.3) 0.0001*

N (F/M) 65 (23/42) 72 (32/40) 25 (13/12) 22 (10/12) 25 (10/15) *Comparison with control eyes; statistical significance at p < 0.05 †Comparison with normotensive exfoliation eyes; statistical significance at p < 0.05.

‡With no medication.

Note: IOP, intraocular pressure; OHT, ocular hypertension; SimK1, simulated keratometry 1; CCT, central corneal thickness; CI, confidence interval; F/M, female/male.

PEXG eyes than in normal eyes (p = 0.001 and 0.02, respec-tively) (Table 1). Regarding CC, there was no significantdifference between hypertensive PXS eyes and normal eyes.

The mean CCT was 546.4 (SD 39.6) μm in all eyeswith PXS and 542.9 (SD 32.2) μm in normal subjects(p = 0.56). When PXS eyes were subcategorized, however,the CCT was significantly thinner in normotensive PXSthan in those with hypertensive PXS (p = 0.04), withPEXG (p = 0.003), and in normal eyes (p = 0.05)(Table 1). PEXG eyes also had a significantly higher CCTthan both normotensive PXS eyes (p = 0.003), and normaleyes (p = 0.04). There was no significant differencebetween normotensive PXS eyes and normal eyes. Table 1also indicates the statistically significant differencesbetween each of the PXS subgroups and normal eyes.

There was no significant correlation between CC(SimK1, SimK2, and mean K) and CCT in patients withall PXS eyes or in controls. In the study group, analysisrevealed values of r = –0.1 and p = 0.36 at the 0.05 levelbetween mean K and CCT; in the control group, r = +0.18and p = 0.1 at the 0.05 level between mean K and CCT.

INTERPRETATION

Accurate determination of IOP is an important factorin the diagnosis and follow-up of glaucoma. It is wellknown that CCT varies widely among individuals withdifferent types of glaucoma and those with normal eyes.Furthermore, CCT affects the IOP measurements that areobtained with applanation tonometry. Therefore, falsemeasurements of IOP due to increased or decreased CCTcan lead to misdiagnosis or mistreatment of glaucoma.

The present study examined CCT in PXS eyes withand without glaucoma together with CC, representinganother parameter possibly affecting IOP measurements.We found a significantly steeper corneal curvature in allPXS eyes as compared with normal eyes, whereas theCCT was not significantly different from controls.However, the CCT was significantly thinner in nor-motensive PXS eyes than both hypertensive PXS eyes andPEXG eyes. In eyes with PXS, the steeper corneas maylead to further error in the evaluation of GAT.

The GAT tends to overestimate IOP in the steepercornea. Although several authors have previously discussedthe effect of CC on GAT,13–16 to the best of our knowledgethe corneal steepness in PXS eyes has never been demon-strated before. Morad and associates4 found the CC to besimilar in patients with normal-tension glaucoma andprimary open-angle glaucoma (POAG) and in healthysubjects. In this study, the mean CC, which is normallyused in IOP studies if the astigmatism is below 1.5 D, was44.1 D in the PXS group and 43.2 D in the control group.As described in Schmidt’s original work13 and in its reviewby Whitacre and Stein,15 a steep curvature results in a falsehigh pressure reading because more force is used to flattenthe cornea. Mark14 calculated an elevation of 0.34 mm Hg

per increase in 1 D. Noticing the effect of CC on GAT,Gunvant et al.16 reported that for an increase of 1 mm ofmean CC there was a rise in IOP of 1.14 mm Hg measuredby the GAT. The effect that Mark found of corneal curvatureon IOP was much greater than the results of Gunvant et al;however, neither Mark nor Gunvant et al. concluded that itwas statistically significant. The results of the present studydid not confirm these findings. In line with the criteria men-tioned in Gunvant et al.’s study, the statistically significantvalues in this study (i.e., approximately 1 D differencebetween groups) would affect the true IOP in PXS patients.

Shimmyo et al.17 found a significant correlation betweenkeratometric power (both for K1 and K2) and CCT inhuman eyes, with the thicker cornea being flatter and thethinner cornea steeper. The CC in the present study wassignificantly different in patients with PXS and in normalsubjects. There was, however, no correlation betweenSimK1, SimK2, or mean K and CCT in PXS eyes, whichis not consistent with the study of Shimmyo et al.17

Reports in the literature regarding CCT in eyes with PXSand PEXG are controversial. Previous studies have reportedvalues lower than,9–11 higher than,12 or similar to5–8 the CCTof normal eyes (Table 2). In Aghaian et al.’s study,11 meanCCT in eyes with PEXG and POAG in a group of multi-ethnic glaucoma patients was significantly less than the CCTof normal eyes. Bechmann et al.9 determined CCT by usingoptical coherence tomography, and found that CCT was sig-nificantly lower in patients with low-tension glaucoma,PEXG, and POAG. Inoue et al.10 demonstrated that thecorneal endothelial cell density was significantly lower andthe central cornea was significantly thinner in eyes with PXSregardless of the presence of glaucoma. The mean valuesreported by Shah et al.,5 Ventura et al.,6 and Detorakis et al.7

of CCT in eyes with PXS and PEXG did not significantlydiffer from those of POAG, NTG, or control eyes. Herndonand colleagues2 also reported almost identical values of CCTin PEXG and POAG eyes. Puska et al.,12 however, foundCCT to be higher in normotensive eyes with early PXS thanin the fellow eyes without PXS.

We have found that the CCT value obtained for the eyes

Corneal curvature and thickness in PXS—Hepsen et al.

CAN J OPHTHALMOL—VOL. 42, NO. 5 2007 679

Table 2—Previously reported central corneal thickness values in eyes with pseudoexfoliation syndrome and pseudoexfoliation glaucoma

Central corneal thickness (Mean [SD], μm)

Investigators PXS/PEXG Normals p value Principle of

measurement

Aghaian et al.11 531 (7.3)* 552.9 (2.7) 0.02 USPBechmann et al.9

493* 530 <0.0001 OCT Detorakis et al.7

524.2 (21.3)† 531.4 (19.5) 0.21 USPInoue et al.10

529 (31)*† 547 (28) 0.03 USP Puska et al.12

528 (30)‡ 523 (32) <0.01 USPShah et al.5

530.7* 553.9 NS USP Ventura et al.6

500 (25)* 524 (25) NS OLCRYagci et al.8

526.3 (31.7)* 533.9 (29.3) 0.24 USP *In XFG eyes. †In PXS eyes. ‡In normotensive eyes. Note: PXS, pseudoexfoliation syndrome; PEXG, pseudoexfoliation glaucoma; USP, ultra-sonic pachymeter; OCT, optical coherence tomography; NS, not significant; OLCR, optical low-coherence reflectometry; XFG, exfoliative glaucoma.

with PXS was 546.4 (SD 39.6) μm and in the control group542.95 (SD 32.2) μm. These average values are high com-pared with values obtained by other methods, e.g., interfer-ometry, confocal microscopic scanning, or classical opticalmethods. The fact that all patients with PXS displayednormal CCT valus in the present study may thus reflect thefact that the study group included both the normotensiveand hypertensive eyes. The previously reported CCT valuesin eyes with PXS and PEXG and the tool of measurementare shown in Table 2. The CCT values in this study are con-sistent with or slightly higher than the findings of earlierstudies using ultrasonic pachymetry5,10–12 (Table 2).

In this study, the CCT was significantly thinner in eyeswith normotensive PXS compared with PXS eyes withelevated IOP and with glaucoma (Table 1). This findingis in accord with the results of Inoue et al.’s study.10 Theyreported that CCT was 534 (SD 37) μm in the PXS eyesin patients with glaucoma and 528 (SD 29) μm in thosewithout glaucoma. The greater CCT in PEXG eyes thannormotensive PXS eyes may be explained by decreasedendothelial cell density and decompensation of the barrierfunction of the endothelial cells in these eyes.10,12

Optical pachymetry and ultrasonic pachymetry mayshow differences in measurement results. Ehlers et al.18 usedoptical pachymetry and found that the GAT was most accu-rate when central corneal thickness was 520 μm. The CCTvalues measured by optic pachymetry seem to be lower thanthose measured by ultrasonic pachymetry,6,9 as shown inTable 2. In the present study, we used Pentacam to establishthe radius of CC and the mean CCT in a sample of PXSeyes compared with controls. There has been heightenedinterest in Scheimpflug noncontact pachymetry, particularlyin relation to refractive corneal surgery. The capability ofPentacam to measure CCT has been demonstrated recently.

Ultrasonic pachymetry is a widely used and reliable tech-nique to measure corneal thickness. However, this kind ofcontact examination has some problems. A probe has to beplaced perpendicularly on the center of the cornea. Ascorneal thickness increases peripherally, lateral displacementof the probe may cause elevated readings. Furthermore, it isdifficult to control the patient’s gaze during repeated meas-urements. So the placement of the probe is difficult toreproduce, and the applanation force may disturb the ante-rior reflecting surface by pushing away the precorneal tearfilm and by the thinning of the epithelium.9,19 Pentacam isa noninvasive, noncontact scanning method that uses amonitor-controlled rotating Scheimpflug camera.

In conclusion, we found that PXS eyes have a signifi-cantly steeper CC, and the CCT was significantly thinnerin normotensive PXS eyes and significantly thicker inPEXG eyes than normal eyes in patients of Turkish eth-nicity. However, the statistical significance of CCT or CCdoes not necessarily correlate with clinical significance.Therefore, further studies are needed to clarify the clini-cal importance of our findings and ascertain whetherthere is a correlation between CC and pseudoexfoliation.

REFERENCES

1. Copt RP, Thomas R, Mermoud A. Corneal thickness in ocularhypertension, primary open-angle glaucoma, and normaltension glaucoma. Arch Ophthalmol 1999;117:14–6.

2. Herndon LW, Choudhri SA, Cox T, Damji KF, Shields MB,Allingham RR. Central corneal thickness in normal, glaucoma-tous, and ocular hypertensive eyes. Arch Ophthalmol1997;115:1137–41.

3. Herman DC, Hodge DO, Bourne WM. Increased cornealthickness in patients with ocular hypertension. ArchOphthalmol 2001;119:334–6.

4. Morad Y, Sharon E, Hefetz L, Nemet P. Corneal thickness andcurvature in normal-tension glaucoma. Am J Ophthalmol1998;125:164–8.

5. Shah S, Chatterjee A, Mathai M, et al. Relationship betweencorneal thickness and measured intraocular pressure in a generalophthalmology clinic. Ophthalmology 1999;106:2154–60.

6. Ventura AC, Bohnke M, Mojon DS. Central corneal thicknessmeasurements in patients with normal tension glaucoma,primary open angle glaucoma, pseudoexfoliation glaucoma, orocular hypertension. Br J Ophthalmol 2001;85:792–5.

7. Detorakis ET, Koukoula S, Chrisohoou F, Konstas AG,Kozobolis VP. Central corneal mechanical sensitivity in pseu-doexfoliation syndrome. Cornea 2005;24:688–91.

8. Yagci R, Eksioglu U, Midillioglu I, Yalvac I, Altiparmak E,Duman S. Central corneal thickness in primary open angleglaucoma, pseudoexfoliative glaucoma, ocular hypertension,and normal population. Eur J Ophthalmol 2005;15:324–8.

9. Bechmann M, Thiel MJ, Roesen B, Ullrich S, Ulbig MW,Ludwig K. Central corneal thickness determined with opticalcoherence tomography in various types of glaucoma. Br JOphthalmol 2000;84:1233–7.

10. Inoue K, Okugawa K, Oshika T, Amano S. Morphologicalstudy of corneal endothelium and corneal thickness in pseudo-exfoliation syndrome. Jpn J Ophthalmol 2003;47:235–9.

11. Aghaian E, Choe JE, Lin S, Stamper RL. Central corneal thick-ness of Caucasians, Chinese, Hispanics, Filipinos, AfricanAmericans, and Japanese in a glaucoma clinic. Ophthalmology2004;111:2211–9.

12. Puska P, Vasara K, Harju M, Setala K. Corneal thickness andcorneal endothelium in normotensive subjects with unilateralexfoliation syndrome. Graefes Arch Clin Exp Ophthalmol2000;238:659–63.

13. Schmidt TA. The clinical application of the Goldmann appla-nation tonometer. Am J Ophthalmol 1960;49:967–78.

14. Mark HH. Corneal curvature in applanation tonometry. Am JOphthalmol 1973;76:223–4.

15. Whitacre MM, Stein R. Sources of error with use ofGoldmann-type tonometers. Surv Ophthalmol 1993;38:1–30.

16. Gunvant P, Baskaran M, Vijaya L, et al. Effect of corneal param-eters on measurements using the pulsatile ocular blood flowtonograph and Goldmann applanation tonometer. Br JOphthalmol 2004;88:518–22.

17. Shimmyo M, Ross AJ, Moy A, Mostafavi R. Intraocular pres-sure, Goldmann applanation tension, corneal thickness, andcorneal curvature in Caucasians, Asians, Hispanics, and AfricanAmericans. Am J Ophthalmol 2003;136:603–13.

18. Ehlers N, Bramsen T, Sperling S. Applanation tonometry andcentral corneal thickness. Acta Ophthalmol Copenh1975;53:34–43.

19. Nissen J, Hjortdal J, Ehlers N, Frost-Larsen K, Sorensen T.A clinical comparison of optical and ultrasonic pachometry.Acta Ophthalmol Copenh 1991;69:659–63.

Key words: central corneal thickness, corneal curvature, Scheimpflugcorneal topography, pseudoexfoliation syndrome

Corneal curvature and thickness in PXS—Hepsen et al.

680 CAN J OPHTHALMOL—VOL. 42, NO. 5, 2007