The Measurement of Corneal EpithelialThickness in Response to Hypoxia Using

Optical Coherence Tomography

JIANHUA WANG, BMED, MSC, DESMOND FONN, MOPTOM,TREFFORD L. SIMPSON, PHD, AND LYNDON JONES, PHD

● PURPOSE: To determine if corneal epithelial thicknessincreases in association with corneal edema induced bywearing soft contact lenses during eye closure.● DESIGN: Experimental study.● METHODS: One eye (randomly selected) of twentynoncontact lens wearers (10 males and 10 females, age35.6 � 9.6 years) was patched during 3 hours of softcontact lens (SCL) wear and the contralateral eye actedas a control. Corneal and epithelial thickness of both eyeswas measured before and after SCL wear using opticalcoherence tomography (OCT).● RESULTS: Immediately after contact lens removal, totalcorneal thickness was increased significantly by 13.8 �2.3% (mean � SD) compared with baseline (P < .0001,paired t test) and after 100 minutes was still 4.5 � 2.3%thicker than baseline (P < .0001, paired t test). Thecontrol eyes showed no change in total corneal thickness(P > .05, paired t test). Immediately after contact lensremoval, corneal epithelial thickness was increased by1.7 � 4.8%, but this change was not statistically signif-icant (P > .05, paired t test). Following contact lensremoval, epithelial thickness changed significantly (Re-peated measure analysis of variance [Re-ANOVA]:F(7,133) � 4.91, pH–F < 0.001) over the next 100minutes with thinning recorded at 60, 80, and 100minutes (P < .05, paired t test). There was no significantchange over time in epithelial thickness of the controleyes (Re-ANOVA: F(4, 76) � 0.91, pH–F � 0.464).● CONCLUSION: OCT demonstrated that corneal epithe-lial thickness does not increase in response to hypoxiafrom SCL wear and eye closure, in contrast to a signifi-cant increase in total corneal thickness. (Am J Oph-

thalmol 2002;133:315–319. © 2002 by Elsevier ScienceInc. All rights reserved.)

R EPORTS OF CORNEAL EPITHELIAL RESPONSE TO OXY-

gen deprivation are inconsistent in their conclu-sions. The corneal epithelium of excised rabbit eyes

has been shown to increase by 26–50 �m after oxygendeprivation.1–3 However, other animal studies have shownno significant increase in epithelial thickness in vitro4,5

and in vivo.6 Epithelial thinning has been found followingcontact lens wear in living monkey eyes.7 O’Leary andassociates8 found a nonsignificant increase of 1.7% inhuman corneal epithelial thickness when deprived ofoxygen for 2 hours using nitrogen gas in goggles.

Epithelial thickness has been measured using high-frequency ultrasonic pachymetry,9 confocal microscopy,10

and specular microscopy,11 all of which are contact andinvasive methods. Modified optical pachymetry has alsobeen used by some investigators.8,12 However, this tech-nique is more difficult to use when the epithelial lightscatter of the cornea increases as a result of hypoxia.4,6

Optical coherence tomography (OCT) is a new nonin-vasive and noncontact optical imaging modality, whichapplies low coherence interferometry and uses an imagemapping process to display high-resolution cross-sectionalimaging of the ocular microstructure. In ophthalmology,this instrument was originally designed to measure differ-ent layers of the retina.13,14 Recently, OCT has been usedto measure corneal and epithelial thickness in normalliving corneas. Izatt and associates15 used OCT to measurethe thickness of the total cornea and corneal epithelium inone normal eye and Maldonado and associates16 used OCTto measure the thickness of corneal cap and stromal bedafter laser in situ kertomileusis (LASIK). Feng and associ-ates17 reported measuring diurnal variation of cornealepithelial thickness by using OCT and we have shown thatthe instrument is highly reproducible for both cornealepithelial and total corneal thickness (2.7 and 2.8 �m,respectively) in 10 subjects measured on two differentoccasions.18

Accepted for publication Nov 19, 2001.Centre for Contact Lens Research, School of Optometry, University of

Waterloo, Waterloo, Ontario, Canada. This study was partly supported bya grant from the Canada Foundation for Innovation.

Proprietary interests: The authors have no proprietary interest in anymaterials or methods described within this article.

Reprint requests to Desmond Fonn, Centre for Contact Lens Research,School of Optometry, University of Waterloo, Waterloo, ON, Canada,N2L 3G1; fax: (519) 884 8769; e-mail: dfonn@sciborg.uwaterloo.ca

© 2002 BY ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED.0002-9394/02/$22.00 315PII S0002-9394(01)01382-4

The purpose of this study was to determine the epithelialresponse to hypoxia in the normal human cornea. We usedOCT to measure central corneal epithelial thickness andtotal central corneal thickness after 3 hours of hypoxiainduced by contact lens wear and eye closure.

METHODS

● SUBJECTS: Twenty subjects (10 males and 10 females,age 35.6 � 9.6 years) with no history of contact lens wearor any current ocular or systematic disease were recruitedfor this study. Informed consent was obtained from eachsubject after ethics approval was obtained from the Officeof Research Ethics, University of Waterloo. All subjectswere treated in accordance with the tenets of the Decla-ration of Helsinki. The relevant ocular biometric informa-tion is listed in Table 1.

● INSTRUMENTATION AND LENSES: An OCT system(Humphrey Instruments, Zeiss–Humphrey, San Leandro,CA) was used in this study to measure corneal thickness.Using the low-coherence interferometry technique, thisinstrument provides a single log reflectivity profile inwhich corneal thickness of the different layers can bedetermined by measuring the distance of log reflectivitypeaks from the various layers.16,18 A TV monitor is used tomonitor the eye position to ensure the central corneal areawas measured. A single line scan mode was used to measurethe corneal thickness with a scan length of 1.13 mm.There are 100 scanning points in each OCT measurement,therefore, the nominal lateral resolution is 11.3 �m. OCTcalibration was undertaken using the same procedure asthat described in a study conducted by Huang and associ-ates.19

In this study, we measured corneal thickness as thedistance between the first and last peaks and epithelialthickness as the distance between the first and secondpeaks, as described in other studies16 and as shown inFigure 1. In each OCT reflectivity profile, a peak (ampli-tude of light reflectivity) corresponds to an interfacebetween two different layers with different refractive indi-

ces (that is, the interface of air/tear layer). It is similar tothe peak detected by the ultrasonic pachometer (A scan)9

and confocal microscope.20 In the single reflectivity profilederived by a confocal microscope,20 high-frequency ultra-sound digital signal processing21 and partial coherenceinterferometry,22 the second peak was suggested to corre-spond to the basal nerve plexus by specular microscope20 orthe Bowman layer by partial coherence interferometry.22

In a study conducted by Feng and associates17 using OCT,they demonstrated that the second peak was absent afterthe epithelium was removed.

Lathe-cut, 38% water content, Hydroxyethyl-methacry-late (HEMA), thick (0.34–0.41 mm) soft contact lenseswere used in our study to induce corneal edema. Theselenses had an average oxygen transmissibility (Dk/t) of2.1 � 10�9.

● PROCEDURE: In a prospective experimental study, twostudy visits were scheduled after 10 AM to ensure thatcorneal edema induced by overnight eye closure haddissipated. All participants were asked to wake no laterthan 7 AM. At the initial visit, participants were screenedto determine their eligibility for the study. Study proce-dures were explained and informed consent was obtained.A biomicroscopic examination, automated refraction/keratometry, and measurements of both corneal epithelialthickness and total corneal thickness were conducted atthe initial screening visit. At the subsequent visit, thesebaseline measurements were obtained with OCT beforewearing a contact lens on a randomly selected eye. Afterlens insertion, the fit was checked to ensure adequatecorneal coverage and that no post lens debris was present.The contact lens wearing eye was then patched for 3 hourswith an ophthalmic eye patch, gauze, and surgical tape toensure complete eye closure. During the study period,participants were asked to remain at the Center for

TABLE 1. Subjects’ Biometric Data (Mean � SD)

Eyes Right Left

Corrected vision (logMar) �0.12 � 0.14 �0.14 � 0.11

K (horizontal) D 43.88 � 1.16 43.99 � 1.25

K (vertical) D 43.44 � 1.37 43.59 � 1.48

Rx (sphere ) D �1.04 � 1.70 �1.09 � 2.04

Rx (cylinder ) D �0.53 � 0.40 �0.68 � 0.49

K � curvature.

Rx � refraction. FIGURE 1. A single reflectivity profile (A) and a cross section(B) of the cornea can be displayed and used to measure thethickness of different layers. In the profile, the first peak (a)corresponds to the interface between air and the tear layer(corneal surface), the second peak (b) corresponds to thebasement membrane, and the last peak (c) corresponds to theendothelium.

AMERICAN JOURNAL OF OPHTHALMOLOGY316 MARCH 2002

Contact Lens Research (CCLR). Normal blinking wasallowed for the control eyes, but participants were askednot to close the control eye for longer than 2 minutes.After lens removal, measurements with OCT and pachym-etry were repeated six times at 20-minute intervals up to100 minutes. The control eye was measured at approxi-mately the same time as lens removal from the other eye,that is, at 60 minutes and 100 minutes.

● DATA ANALYSIS: Multiple central scan points fromOCT raw data were analyzed using a software programwhich was developed in the CCLR. Data analysis wasconducted using STATISTICA (StatSoft Inc., Tulsa,OK). Repeated measurement analysis of variance (Re-ANOVA) and paired t tests were used to determine ifthere was any significant (P � .05) change. Repeatabilitywas calculated as the standard deviation of the differencebetween measurements of the screening visit and baselinemeasurements at the initial study visit in all study subjects.

RESULTS

IN 40 EYES OF 20 NONCONTACT LENS WEARERS, THE STAN-

dard deviation of the difference between the two visitsbefore contact lens wear was 3.2 �m for corneal thicknessand 2.5 �m for epithelial thickness. There was no signif-icant (two-tailed t test, P � .05) difference betweenpatched eyes and control eyes in both corneal thicknessand epithelial thickness before contact lens insertion.

Immediately after contact lens removal, corneal thick-ness was increased significantly (P � .0001, paired t test)by 13.8 � 2.3% (mean � SD) compared with baseline andafter 100 minutes was still 4.5 � 2.3% thicker thanbaseline, as shown in Figure 2. The absolute micron values

are recorded in Table 2. Immediately after contact lensremoval, corneal epithelial thickness was increased by1.7 � 4.8%, but this change was not statistically significant(P � .05, paired t test). Following contact lens removal,the epithelium was thinned significantly (up to 3.0 �4.5%, P � .05, paired t test) over the next 100 minutes(Re-ANOVA: F(7, 133) � 4.91, pH–F � 0.001) as shown inFigure 3.

The thickness values of the control eyes are shown inTable 2 and the changes are plotted in Figures 2 and 3.There was no significant change in epithelial thickness(Re-ANOVA: F(4, 76) � 0.91, pH–F � 0.464) of the controleyes. However, total corneal thickness of the control eyeschanged significantly during the post lens wear period(Re-ANOVA: F4, 76 � 5.57, pH–F � 0.003) and this wasdue to thinning at 60 and 100 minutes (P � .05, paired ttest) compared with baseline.

DISCUSSION

IN OUR STUDY, AFTER 3 HOURS OF CONTACT LENS WEAR

and eye closure, OCT showed that epithelial thickness didnot increase in response to hypoxia as has been shownpreviously.4,6–8 This is in contrast to significant cornealswelling that developed.7,8 Immediately after lens removal,total corneal thickness had increased by 13.8% with OCT,which is within the range of 5%–15% of some previousreports23–26 that measured corneal swelling in response tocontact lens wear with eye closure. However, these studiesused optical pachymetry to measure corneal thickness.

Lambert and Klyce suggested that the absence of cornealepithelial thickness changes in response to hypoxia isbecause the living epithelium possibly contains somecompensatory regulatory mechanism dependent upon met-abolic energy or there may be radial swelling duringhypoxia.4 The absence of epithelial thickness increase wasalso found by Bergmanson.7 In the Bergmanson study,monkey corneas exhibited more intracellular and intercel-lular fluid and loss of cells, as well as cell flattening. Usingspecular microscopy, Wilson and Fatt6 noted the cornealepithelium of a living rabbit eye began to thin approxi-mately 80 minutes after hypoxia exposure.

Following lens removal, corneal and epithelial thicknessmeasurements continued as shown in Figure 2 and thetypical deswelling pattern of the cornea is shown. Duringcorneal deswelling, the epithelium also showed significantthinning (up to 3%) and corneal thinning was observed inthe control eyes as well. O’Leary8 has also reportedepithelial thinning during the corneal deswelling period.

Because this epithelial thinning effect has been consis-tently observed, we would speculate that this may be dueto an auto-regulation effect of corneal function in responseto lens wear. That is, the cornea continues to expel fluid orthins beyond its baseline (thinnest) thickness. It is alsopossible that the measurement of epithelial thickness that

FIGURE 2. Percentage change (mean � SE) of corneal thick-ness in patched eyes compared with control eyes in 20 subjects.Baseline refers to the measurements taken before soft contactlens (SCL) wear and eye closure. Removal point refers to themeasurements taken immediately after SCL removal following3 hours of lens wear. The control eyes were measured on threeoccasions following the baseline thickness.

MEASUREMENT OF CORNEAL EPITHELIAL THICKNESSVOL. 133, NO. 3 317

we recorded as “baseline” was, in fact, thicker thanbaseline because of diurnal variation, but this is highlyunlikely as the same effect was observed in all other studiesmentioned.

Corneal thinning of the control eyes may be due to asympathetic response as the contralateral eye deswells orthis may simply be due to diurnal changes. A sympatheticswelling response has been proposed previously by Fonnand his associates.24 They compared high oxygen transmis-sibility silicone hydrogel lenses with low oxygen transmis-sibility lenses and found that the control nonlens-wearingeyes swelled more when the lower transmissibility lens wasworn on the contralateral eyes than when the highertransmissibility lens was worn. Harris and Mandell27 alsofound a sympathetic swelling response in unilateralPMMA lens wearers. They proposed that the swellingresponse of nonlens-wearing eyes was due to an osmolarityeffect as a result of lacrimation. This contralateral thinning

effect of the cornea can also be due to an autoregulationmechanism following corneal deswelling, as suggestedabove, or it is simply part of diurnal variation of cornealthickness, which has been documented previously.28–30

Optical coherence tomography measurements includethe precorneal tear film thickness as suggested by Hitzen-berger and Drexler.31 Recent reports of the thickness oftear layer showed it is between 3–7 �m.32,33 However, theinclusion of the tear film thickness is unlikely to affect thechange in thickness results of the cornea or epithelium asit is a constant.

In conclusion, OCT is a reliable instrument for measur-ing corneal and epithelial thickness and for monitoringchanges of these structures. In our study, OCT demon-strated that corneal epithelial thickness does not increasein response to hypoxia from SCL wear and eye closure incontrast to a significant increase in total corneal thickness.An interesting finding in control eyes is that cornealthinning occurs during corneal deswelling of the contralat-eral eye and that the epithelium appears to thin followinglens wear and eye closure/hypoxia.

REFERENCES

1. Uniacke CA, Hill RM. The depletion course of epithelialglycogen with corneal anoxia. Arch Ophthalmol 1972;87:56–59.

2. Lowther GE, Hill RM. Recovery of the corneal epitheliumafter a period of anoxia. Am J Optom Arch Am Acad Optom1973;50:234–241.

3. Uniacke CA, Augsburger A, Hill RM. Epithelial swellingwith oxygen insufficiency. Am J Optom Arch Am AcadOptom 1971;48:565–568.

4. Lambert S, Klyce S. The origins of Sattler’s veil. Am JOphthalmol 1981;91:51–56.

5. Wilson GS, Fatt I, Freeman RD. Thickness changes in thestroma of an excised cornea during anoxia. Exp Eye Res1973;17:165–171.

TABLE 2. Corneal and Epithelial Thickness Measurements of the Experimental and Control Eyes After Contact Lens RemovalFollowed 3 Hours of Lens Wear

Baseline Removal 20 min 40 min 60 min 80 min 100 min

Cornea

Patched (mean) 560.3 637.8 627.2 618.7 605.7 594.6 585.6

SD 35.0 39.6 41.5 40.3 40.1 39.6 38.0

Control (mean) 562.1 561.7 - - 558.6 - 559.1

SD 31.9 32.6 - - 32.5 - 33.5

Epithelium

Patched (mean) 57.9 59.0 58.4 57.6 56.6 56.2 56.8

SD 1.8 2.6 2.6 2.4 2.1 1.9 1.6

Control (mean) 57.8 57.4 - - 57.4 - 57.6

SD 1.7 2.0 - - 2.1 - 2.3

SD � standard deviation.

FIGURE 3. Percentage change (mean � SE) of epithelialthickness of patched eyes compared with control eyes in 20subjects. Baseline refers to the measurements taken before softcontact lens (SCL) wear and eye closure. Removal point refersto the measurements taken immediately after SCL removalfollowing 3 hours of lens wear. The control eyes were measuredon three occasions following the baseline thickness.

AMERICAN JOURNAL OF OPHTHALMOLOGY318 MARCH 2002

6. Wilson G, Fatt I. Thickness of the corneal epithelium duringanoxia. Am J Optom Physiol Opt 1980;57:409–412.

7. Bergmanson JP, Ruben CM, Chu LW. Epithelial morpho-logical response to soft hydrogel contact lenses. Br J Oph-thalmol 1985;69:373–379.

8. O’Leary DJ, Wilson G, Henson DB. The effect of anoxia onthe human corneal epithelium. Am J Optom Physiol Opt.1981;58:472–476.

9. Reinstein DZ, Silverman RH, Trokel SL, Coleman DJ.Corneal pachymetric topography. Ophthalmol 1994;101:432–438.

10. Chan-Ling T, Pye DC. Pachometry: clinical and scientificapplications. In: Ruben M, Guillon M, editors. Contact lenspractice. London: Chapman & Hall, 1994:407–435.

11. Azen SP, Burg KA, Smith RE, Maguen E. A comparison ofthree methods for the measurement of corneal thickness.Invest Ophthalmol Vis Sci 1979;18:535–538.

12. Gauthier C, Holden B, Epstein D, Tengroth B, Fagerholm P,Hamberg-Nystrom H. Role of epithelial hyperplasia in re-gression following photorefractive keratectomy. Br J Oph-thalmol 1996;80:545–548.

13. Toth CA, Narayan DG, Boppart SA, Hee MR, Fujimoto JG,Birngruber R, Cain CP, DiCarlo CD, Roach WP. A com-parison of retinal morphology viewed by optical coherencetomography and by light microscopy. Arch Ophthalmol1997;115:1425–1428.

14. Bowd C, Weinreb RN, Williams JM, Zangwill LM. Theretinal nerve fiber layer thickness in ocular hypertensive,normal, and glaucomatous eyes with optical coherence to-mography. Arch Ophthalmol 2000;118:22–26.

15. Izatt JA, Hee MR, Swanson EA, Lin CP, Huang D, SchumanJS, Puliafito CA, Fujimoto JG. Micrometer-scale resolutionimaging of the anterior eye in vivo with optical coherencetomography. Arch Ophthalmol 1994;112:1584–1589.

16. Maldonado MJ, Ruiz-Oblitas L, Munuera JM, Aliseda D,Garcia-Layana A, Moreno-Montanes J. Optical coherencetomography evaluation of the corneal cap and stromal bedfeatures after laser in situ keratomileusis for high myopia andastigmatism. Ophthalmology 2000;107:81–87.

17. Feng Y, Varikooty J, Simpson T. Diurnal variation of cornealand corneal epithelial thickness measured using opticalcoherence tomography. Cornea 2001;20:480–483.

18. Fonn D, Wang JH, Simpson T. Topographical thickness ofthe epithelium and total cornea using optical coherencetomography. Invest Ophthalmol Vis Sci 2000;41:S675.

19. Huang Y, Cideciyan AV, Papastergiou GI, Banin E, Semple-

Rowland SL, Milam AH, Jacobson SG. Relation of opticalcoherence tomography to microanatomy in normal and rdchickens. Invest Ophthalmol Vis Sci 1998;39:2405–2416.

20. Li HF, Petroll WM, Moller-Pedersen T, Maurer JK, Ca-vanagh HD, Jester JV. Epithelial and corneal thicknessmeasurements by in vivo confocal microscopy through focus-ing (CMTF). Curr Eye Res 1997;16:214–221.

21. Reinstein DZ, Silverman RH, Rondeau MJ, Coleman DJ.Epithelial and corneal thickness measurements by high-frequency ultrasound digital signal processing. Ophthalmol1994;101:140–146.

22. Drexler W, Baumgartner A, Findl O, Hitzenberger CK,Sattmann H, Fercher AF. Submicrometer precision biometryof the anterior segment of the human eye. Invest Ophthal-mol Vis Sci 1997;38:1304–1313.

23. Holden B, McNally J, Mertz G, Swarbrick HA. Topograph-ical corneal oedema. Acta Ophthalmol 1985;63:684–691.

24. Fonn D, du Toit R, Simpson TL, Vega JA, Situ P, ChalmersRL. Sympathetic swelling response of the control eye to softlenses in the other eye. Invest Ophthalmol Vis Sci 1999;40:3116–3121.

25. Holden B. The ocular response to contact lens wear. OptomVis Sci 1989;66:717–733.

26. Herse P, Hooker B. Corneal edema recovery dynamics indiabetes: is the alloxan induced diabetic rabbit a usefulmodel? Invest Ophthalmol Vis Sci 1994;35:310–313.

27. Harris M, Mandell R. Contact lens adaptation: osmotictheory. Am J Ophthalmol 1969;46:196–202.

28. du Toit R, Vega J, Simpson T, Fonn D. Diurnal variation ofcorneal thickness, sensitivity, and temperature. Optom VisSci 1998;75(Suppl):S82.

29. Harper C, Boulton M, Bennett D, Marcyniuk B, Jarvis-EvansJ, Tullo AB, Ridgway A. Diurnal variations in humancorneal thickness. Br J Ophthalmol 1996;80:1068–1072.

30. Mertz G. Overnight swelling of the living human cornea.J Am Optom Assoc 1980;51:211–213.

31. Hitzenberger CK, Drexler W, Fercher AF. Measurement ofcorneal thickness by laser Doppler interferometry. InvestOphthalmol Vis Sci 1992;33:98–103.

32. King-Smith PE, Fink BA, Fogt N. Three interferometricmethods for measuring the thickness of layers of the tear film.Optom Vis Sci 1998;76:19–32.

33. King-Smith PE, Fink BA, Fogt N, Nichols KK, Hill RM,Wilson GS. The thickness of the human precorneal tear film:evidence from reflection spectra. Invest Ophthalmol Vis Sci2000;41:3348–3359.

MEASUREMENT OF CORNEAL EPITHELIAL THICKNESSVOL. 133, NO. 3 319