Cornea1 Thickness and Curvature in Normal-tension Glaucoma

YAIR MORAD, MD, ELINOR SHARON, MD, LIMOR HEFETZ, MD, AND PINHAS NEMET, MD

l PURPOSE: To determine whether normal-ten-

sion glaucoma, defined as a condition in which glaucomatous optic nerve and visual field changes

exist without documentation of intraocular pres-

sure greater than 21 mm Hg or other apparent

cause for these changes, is overdiagnosed in pa-

tients with decreased central cornea1 thickness

and curvature.

l METHODS: Twenty-one patients with normal-

tension glaucoma were compared with 25 patients

with primary open-angle glaucoma and 27 age-

matched healthy subjects. Cornea1 thickness was

determined by ultrasonic pachymetry. Cornea1

curvature was determined using a keratometer.

Eyes with cornea1 pathology or previous intraocu-

lar surgery were excluded.

l RESULTS: Mean cornea1 thickness f SD in 21 eyes of 21 patients with normaLtension glaucoma

was 0.521 & 0.037 mm, significantly (P = .0028)

lower than in 25 eyes of 25 patients with primary

open-angle glaucoma (0.556 & 0.035 mm) and 27

eyes of 27 healthy subjects (0.555 + 0.034). Mean

cornea1 curvature in the three groups was not

appreciably different: 43.90 f 1.81 diopters, 43.66

rt 1.68 diopters, and 44.36 f. 1.13 diopters in the

patients with normal-tension glaucoma and pri-

mary open-angle glaucoma and the healthy subjects,

respectively.

l CONCLUSIONS: Cornea1 thickness is signifi-

cantly reduced in patients with normal-tension

Accepted for publication May 7, 1997. From the Department of Ophthalmology, Assaf-Harofe Medical Cen-

ter, Zerifin, and Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel.

Reprint requests to Yair Morad, MD, Department of Ophthalmology, Assaf-Harofe Medical Center, Zerifin 73000, Israel; fax: +972-8- 9779361; e-mail: yair-mo@green.co.il

glaucoma compared with patients with primary

open-angle glaucoma (P = .0028) and normal

subjects (P = .OO37). This may lead to underesti-

mation of intraocular pressure and misdiagnosis in

some of these patients. Cornea1 curvature was

similar in patients with normal-tension glaucoma

and primary open-angle glaucoma and in healthy

subjects.

G OLDMANN APPLANATION TONOMETRY’ IS BASED

on the Imbert-Fick law, which assumes that the surface of the cornea is perfectly elastic

and thin and therefore exerts no additional force against the applanating surface other than that of the internal pressure of the globe. In fact, this assumption may not be true in all cases. The relation between central cornea1 thickness and estimated intraocular pressure as measured by applanation tonometry was studied by a number of authors’j Ehlers and associ- ates’ compared intraocular pressure measured by to- nometry and manometry and found that thin corneas could lead to an underestimation of intraocular pressure by as much as 4.7 mm Hg. Variation in cornea1 curvature can also lead to errors in applana- tion tonometry, as reported by Schmidt,4 who showed that the flatter the cornea, the less the force required for applanation, leading to further underestimation of intraocular pressure.

Normal-tension glaucoma is defined as a condition with the following features: cupping of the optic nerve and visual field loss resembling that seen in other forms of glaucoma, no documented intraocular pressure greater than 21 mm Hg without treatment, and no other obvious or apparent cause for these changes.5 The present study was carried out to determine whether central cornea1 thickness and curvature play a clinically relevant role in differentiat-

164 0 AMERICAN JOIJRNN 0F OPHTHALMOLOGY I 998;125:164-I 68 FEBRUARY 1998

ing patients with normal-tension glaucoma from patients with primary open-angle glaucoma and healthy subjects.

SUBJECTS AND METHODS

THE STUDY INCLUDED THREE GROUPS. THE NORMAL-

tension glaucoma group included 21 patients diag- nosed in an outpatient glaucoma clinic as having normal-tension glaucoma based on the following criteria: glaucomatous optic nerve damage determined on ophthalmoscopic examination with slit-lamp bio- microscopy and a 90-diopter lens; intraocular pressure without treatment of 21 mm Hg or lower as recorded on a 24-hour diurnal pressure curve and on the patient’s entire medical record; normal gonioscopic examination; glaucomatous visual field on the Hum, phrey 24-2 static threshold perimeter; abnormal optic nerve topography by Heidelberg retina tomography; normal brain, orbits, and optic canals on computed tomography; and normal carotid Doppler echography+

The primary open-angle glaucoma group included 25 patients diagnosed as having primary open-angle glaucoma according to the following criteria: glauco- matous optic nerve damage determined on stereo- scopic examination with slit-lamp biomicroscopy and a 90-diopter lens; intraocular pressure without treat- ment greater than 21 mm Hg measured on at least two occasions; normal gonioscopic examination; glaucomatous visual field defects on the Humphrey 24-2 static threshold perimeter; and abnormal optic nerve topography by Heidelberg retina tomography.

The control group included 27 age-matched sub- jects who met the following criteria: no apparent glaucomatous damage to the optic disk on stereoscop- ic examination with slit-lamp biomicroscopy and a 90-diopter lens; intraocular pressure of 21 mm Hg or less measured on at least two occasions; normal visual field on the Humphrey 24-2 static threshold perime- ter; and normal optic nerve topography by Heidelberg retina tomography.

Eyes with cornea1 pathology or previous intraocular surgery were excluded. Eyes with myopia or hyperopia of more than 2.5 diopters or astigmatism of more than 1.0 diopter were also excluded because marked astig matism can induce an error in applanation readings.h

After obtaining institutional review board approval and informed consent, all subjects underwent a thorough ophthalmologic examination, including in- traocular pressure measurement by applanation to- nometry. Central cornea1 thickness was measured using an ultrasonic pachymeter. After applying a drop of benoxinate hydrochloride 0.4% for local anesthe- sia, the pachymeter tip was placed 1.5 mm temporal to the light reflex on the cornea, and three pachymetry readings were made. The average of three readings was recorded. Cornea1 curvature was measured by keratometer. Maximal and minimal curvature were measured in diopters, and the average curvature was calculated. All measurements were done by one of us (Y.M.).

Although both corneas were measured, only one eye of each patient or normal subject was selected for statistical analysis to prevent bias; usually this was the right eye, unless it had been excluded from the study. The two-tailed paired Student t test was used for statistical analysis.

RESULTS

AGE AND REFRACTION DID NOT DIFFER SIGNIFICANTLY

between groups (P = .24 to .85; Table 1). There was no significant difference between right and left eyes in cornea1 thickness or curvature (I’ = .75 to .91).

The measurements of cornea1 thickness and curva- ture are summarized in Table 2. Mean cornea1 thick- ness + SD was 0.521 + 0.037 mm and 0.556 +- 0.035 mm, respectively, in the normal-tension glaucoma and primary open-angle glaucoma groups (P = .0028). Mean cornea1 thickness in healthy subjects was 0.555 * 0.034 mm, significantly greater than that in patients with normal-tension glaucoma (P = .0037) and similar to that in patients with primary open-angle glaucoma (P = .97).

Mean cornea1 curvature was 43.90 +- 1.81 diopters, 43.66 * 1.68 diopters, and 44.36 ? 1.13 diopters, respectively, in the normal-tension glaucoma, primary open-angle glaucoma, and normal groups. The differ- ence between groups was not significantly different (P = .85 for patients with normal-tension glaucoma vs patients with primary open-angle glaucoma, P = .46 for patients with normal-tension glaucoma vs normal

VOL. 125, No. 2 CORNFAL THICKNESS AND CURVATURE IN NORMAN-TENSION GLAUCOMA 165

Mean + SD

Patients With Glaucoma

Variable

Normal-tension Primary Open-angle

(rl = 21) (n = 25)

Normal Subjects

(n = 27)

&iar) 6825 It 8,244 86.52 f 8.17 6433 f 12.47 Spherical equivalence (D) 0.28 IT 0.91 -0.42 ? 0.83 0.45 2 1.42

Patients With Glaucoma

Variable

Normal-tension (n = 21)

Primary Open-angle (n = 25)

Normal Subjects (n = 27)

Mean cornea1 curvature (D) 43.9 t 1.81 43.66 L 1.68 44.36 IT 1.13

subjects, and P = .24 for patients with primary open-angle glaucoma vs normal subjects).

DISCUSSION

GOLDMANN APPLANATION TONOMETRY IS ONE OF THE

most common methods of intraocular pressure mea- surement worldwide. Although regarded as accurate, numerous sources of error can critically influence pressure readings.’ Cornea1 thickness is one of the most important factors affecting cornea1 rigidity, which is a major source of errors in applanation tonometry.i Goldmann himself recognized that the cornea possesses a rigidity that opposes indentation. He assumed that the surface tension force created by the tear film drawing the tonometer tip in the opposite direction will equal that rigid force.’ Howev- er, as shown by Ehler and associates’ and Whitacre and associates,3 although this assumption may be true for a cornea of average thickness, cornea1 thicknesses vary widely in the healthy population (from 0.45 to 0.63 mm in the present study), and an extremely thick or thin cornea can respectively cause overesti- mation or underestimation of intraocular pressure.

The correlation between intraocular pressure as measured by applanation and central cornea1 thick-

ness was studied by a number of authors.2J Ehlers and associates’ found that the difference between appla- nation readings and the actual intraocular pressure measured by manometry was linearly correlated with cornea1 thickness. For the average values of cornea1 thickness (an average of 0.52 mm in their study), the difference between readings was clinically negligible; but extremely thin corneas caused underestimation and extremely thick corneas caused overestimation of intraocular pressure measurements. The authors’ ex- trapolated that a reduced cornea1 thickness of 0.45 mm could produce an underestimation of intraocular pressure by up to 4.7 mm Hg, whereas an increased cornea1 thickness of 0.59 mm could cause an overesti- mation of 5.2 mm Hg when the actual intraocular pressure is 20 mm Hg. These findings were repeated by Whitacre and associates,3 who found that in corneas thinner than 0.52 mm, an appreciable under- estimation of applanation tonometry was present in the entire range of intraocular pressure measured.

Errors in applanation tonometry can also be pro- duced by variation in cornea1 curvature. Mark* found a positive correlation between cornea1 curvature and intraocular pressure in a study of 400 eyes and calculated that an increase of 1 diopter of curvature was associated with an increase of 0.34 mm Hg by applanation tonometry.

166 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY i 998

According to our results, average cornea1 thickness is appreciably reduced in patients with normal-ten- sion glaucoma compared with patients with primary open-angle glaucoma and normal subjects. Our study is in agreement with that of Ehlers and Hansen,’ who compared results of optical pachymetry on seven patients with normal-tension glaucoma, 25 patients with primary open-angle glaucoma, and 76 normal subjects. They found central cornea1 thickness mark- edly reduced in the normal-tension glaucoma group (0.463 mm) compared with the primary open-angle glaucoma group (0.556 mm) and normal subjects (0.520 mm). In contrast, Tomlinson and Leighton” found no difference in cornea1 thickness in patients with normal-tension glaucoma, patients with primary open-angle glaucoma, and normal subjects matched for age, sex, and refraction.

Cornea1 curvature in our study was similar in patients with normal-tension glaucoma and primary open-angle glaucoma and in healthy subjects. Tomlinson and Leighton” found cornea1 curvature to be markedly reduced in individuals with normal- tension glaucoma compared with those with primary open-angle glaucoma and normal subjects (7.58 mm, 7.56 mm, and 7.83 mm, respectively). The different results of the present study and those of Tomlinson and Leighton can be explained by the fact that the latter study was performed almost 3 decades ago, when the diagnostic criteria for normal-tension glau- coma were less accurate, without the benefit of skull and orbital computed tomographic imaging and Doppler echography of the carotid arteries. The difference can also be attributed to the relatively small number of patients with normal-tension glauco- ma (11) examined in the study of Tomlinson and Leighton.

Our series included a larger number of patients than earlier studies did. Our diagnostic criteria ex- cluded the possibility of apparent damage to the optic nerve related to causes other than glaucoma. Average central cornea1 thickness values in our patients with

primary open-angle glaucoma and normal subjects were 0.556 I~I 0.035 mm and 0.555 + 0.034 mm, respectively-similar to those reported by Argus” (0.557 + 0.039 mm and 0.567 + 0.036 mm, respec-

tively) and Chisholm and associates” (0.55 I + 0,045 mm and 0.559 + 0.038 mm, respectively), both of whose studies also used ultrasonic pachymetry.

The importance of assessing cornea1 thickness in patients with glaucoma was stressed by Whitacre and associates,3 who stated that “cornea1 thickness mea- surement is necessary to interpret properly the results of applanation tonometry, particularly in eyes with thin corneas.” Pachymetry studies have shown that cornea1 thickness is considerably increased in pa- tients with ocular hypertension.“J3 It was suggested in one study” that up to 30% of ocular hypertensive patients with intraocular pressure measurements of less than 18 mm Hg were misdiagnosed because of the relatively high proportion of thick corneas in this population.

Although variation in cornea1 thickness could lead to clinically important errors in applanation tonome- try (+5.2 to -4.7 mm Hg),’ cornea1 curvature has only a minor effect on applanation readings (an elevation of 0.34 mm Hg per increase in 1 diopter, as calculated by Marks). This may explain our findings that cornea1 thickness and not cornea1 curvature can differentiate patients with normal-tension glaucoma from patients with primary open-angle glaucoma and normal subjects.

We conclude that reduced cornea1 thickness is common among patients with normal-tension glauco- ma. Some of these patients have corneas that are appreciably thinner than average, and this could lead to underestimation of intraocular pressure. When cornea1 thickness is accounted for, patients may be found to have intraocular pressure greater than 21 mm Hg and may then be more accurately classified as having primary open-angle glaucoma. Similar struc- tural differences might be found in other ocular elements of patients with normal-tension glaucoma, such as the optic nerve lamina cribrosa or sclera, or both. Thus, optic nerve damage could occur in lower levels of intraocular pressure because of a thinner and hence weaker lamina cribrosa. Further studies on larger cohorts of patients are needed to confirm these hypotheses.

REFERENCES

1. Coldmann H. Applanation tonometry. In: Newell FW, editor. Glaucoma: transactions of the second conference. New York: Josiah Macy, Jr, Foundation, 1957:167-220.

2. Ehlers N, Bramsen T, Sperling S. Applanation tonometry and central cornea1 thickness. Acta Ophthalmol 1975;53: 34-43.

VOL. I 25, No. 2 CORNEAL THICKNESS AND CURVATURE IN NORMAN-TENSION GLAUCOMA 167

3. Whitacre MM, Stein RA, Hassanein K. The effect of cornea1

thickness on applanation tonometry. Am J Ophthalmol 1993;115:592-596.

4. Schmidt TAF. The clinical application of the Goldmann

applanation tonometer. Am J Ophthalmol 1960;49: 967-978.

5. Levene RZ. Low tension glaucoma: a critical review and new material. Surv Ophthalmol 1980;24:62 l-63 1.

6. Schmidt TAF. Zur Applanationstonometrie an der spalt-

lampe. Ophthalmologica 1957;133:337-342. 7. Whitacre MM, Stein R. Sources of error with use of

Goldmann-type tonometers. Surv Ophthalmol 1993;38:

l-30. 8. Mark HH. Cornea1 curvature in applanation tonometry. Am

J Ophthalmol 1973;76:223-224.

9. Ehlers N, Hansen FK. Central cornea1 thickness in low

tension glaucoma. Acta Ophthalmol 1974;54:740-746. 10. Tomlinson A, Leighton DA. Ocular dimensions in low

tension glaucoma. Br J Ophthalmol 1972;56:97-101.

11. Argus WA. Ocular hypertension and central cornea1 thick-

ness. Ophthalmology 1995;102:1810-1812. 12. Chisholm IA, Drance SM, Chauhan BC. The glaucoma

suspect: differentiation of the future glaucomatous eye from

the non-glaucomatous suspect eye, I: ultrasonic measure- ments and eye-wall stress. Graefes Arch Clin Exp Ophthal- mol 1989;227:17-20.

13. Kruse Hansen F, Ehlers N. Elevated tonometer reading caused by thick cornea. Acta Ophthalmol 1971;49:775-778.

Authors Interactive@ We encourage questions and comments regarding this article via the Internet on Authors Interactive@ at http:// www.ajo.com/ Questions, comments, and author responses are posted.

168 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY I vva