Monitoring Glaucoma Progression with Visual Evoked Potentials of the Blue-Sensitive Pathway

Abstract

purpose. To determine the value of visual evoked potentials with blue-on-yellow pattern stimulation in follow-up of glaucoma.

methods. This prospective longitudinal concurrent study included a heterogeneous cohort of two groups, perimetric (n = 161) and preperimetric (n = 118), of patients with chronic open-angle glaucoma and 113 healthy control subjects. In the preperimetric glaucoma group, patients showed glaucomatous abnormalities of the optic disc, maximum intraocular pressure higher than 21 mm Hg, and unremarkable computerized visual field examination results. Patients underwent up to three VEP measurements with blue-on-yellow pattern stimulation, as well as qualitative and morphometric evaluation of color stereo optic disc photographs. Mean follow-up time between measurements was 24 months. VEP measurements were separately analyzed in preperimetric subjects, with and without progression of optic nerve damage. Progression of glaucoma was defined as increasing loss of neuroretinal rim.

results. A separate analysis of VEP peak times in patients in the preperimetric group, with and without progression of glaucomatous optic nerve damage, showed no significant difference at baseline but a significant prolongation (P = 0.01) in patients with progressive disease, 2 years before morphologic changes were evident. VEPs in patients with nonprogressive disease were statistically unchanged during the observation period. The perimetric group and both preperimetric groups showed significantly prolonged VEP peak times in comparison with the control group (P < 0.001).

conclusions. In addition to photographic evaluation to detect glaucomatous disc atrophy, the blue-on-yellow VEP may be an objective electrophysiological tool for monitoring patients with glaucoma, because peak times are significantly associated with progression of optic nerve damage.

The chronic open-angle glaucomas are associated with a continuing loss of optic nerve fibers and subsequent worsening of the quality of vision. To reduce the risk of ongoing loss of optic nerve substance, it is of utmost importance to monitor the progression of the diseases. Early knowledge of the progression of glaucomatous optic neuropathy may improve the ability to fight the disease, by intensifying treatment in those patients with signs of ongoing damage. The follow-up observation of patients with chronic open-angle glaucoma in clinical practice has so far been based on increased intraocular pressure, cupping of the optic nerve head, and visual field defects. Besides this classic triad, modern procedures using more sophisticated techniques promise to monitor progression of the damage of the retinal ganglion cells, their axons, and possibly, the nerve fibers. These are new perimetric techniques, ^{1 2} new imaging procedures for fundus evaluation (i.e., confocal scanning laser tomography; Heidelberg Retinal Tomograph [HRT]; Heidelberg Engineering, Heidelberg, Germany; and nerve fiber analysis, GDX; Laser Diagnostics, San Diego, CA ^{3 4 5} ), and possibly new psychophysical tests ^{6 7} or electrophysiological measurements. ^{8 9} The latter show the advantages of objectivity in comparison with psychophysical procedures and the ability to investigate various processing stages of the visual system separately. ^{10 11 12} Specifically, the blue-on-yellow pattern VEP is able to isolate the responses from the short-wavelength pathway, which may be a visual channel highly susceptible to glaucomatous damage. The selection of this stimulus is reminiscent of the results in psychophysical studies using blue-on-yellow stimuli, which reveal early field losses in patients with ocular hypertension before deficits occur in the conventional white-on-white perimetry. ²  

The best way to judge the predictive value of a procedure is with repeated measurements in a prospective follow-up study. The purpose of the present investigation was to compare VEP measurements in early and more advanced glaucoma and to study whether such repeated determinations can predict an increase in progressive damage to the optic nerve. The observation of an early alteration in functional abilities may improve the chance to fight the disease by intensifying medical treatment in those patients or by surgery. 

Subjects and Methods

Procedures

The study included 113 normal subjects, 118 patients with preperimetric open angle glaucoma, and 161 patients with perimetric open angle glaucoma (Table 1) . The preperimetric glaucoma group was additionally divided into patients with progression of glaucomatous optic disc atrophy and patients with no progression of glaucoma (Table 2) . The participants in the control group were recruited from the university administration staff. The patients in the preperimetric and perimetric glaucoma groups were referred by ophthalmologists for further diagnosis and follow-up of glaucoma. The study followed the tenets of the Declaration of Helsinki for research involving human subjects, and informed consent was obtained from all participants in the study. All individuals included in the study had open anterior chamber angles, clear optic media, and visual acuity of 20/25 or better. At the day of examination, intraocular pressure was 21 mm Hg or less in all individuals. Exclusion criteria included presence of all eye diseases other than glaucoma, diabetes mellitus, and a myopic refractive error exceeding −8 D. To reduce the influence of young and old age, all participants were between 31 and 67 years of age. Mean age in the normal group, the preperimetric glaucoma groups, and the perimetric glaucoma group did not vary significantly (Table 1) . In the normal eyes, the VEP amplitudes showed no significant age dependency, whereas the peak time increased by 5.7 ms over the age span observed in this study. Therefore, before statistical analysis was performed, the VEP peak times in all subjects were additionally age normalized by dividing each peak time by the equation of the linear regression of peak times to age and multiplying by the mean. For these calculations, the results in normal control subjects were used as the reference. 

All subjects underwent one or more visual field tests. Those subjects with a higher than 12% rate of false-positive or false-negative responses were excluded. If a single perimetric test was followed by a better retest, the earlier measurement was discarded. A perimetric glaucomatous visual field was defined as an G1 field (Octopus automated perimeter; Interzeag, Schlieren, Switzerland) with (1) at least three adjacent test points with a deviation of 5 dB or more and one test point with a deviation more than 10 dB lower than normal, (2) at least two adjacent test points with a deviation of 10 dB or more, (3) at least three adjacent test points with a deviation of 5 dB ore more abutting the nasal horizontal meridian, or (4) a mean visual field defect of more than 2.6 dB. For all eyes, 15° color stereo optic disc transparencies had been taken using a telecentric fundus camera (30° fundus camera, equipped with a 15° converter; Zeiss, Oberkochen, Germany). The disc slides were projected in a scale of 1 to 15. The outlines of the optic cup, optic disc, peripapillary scleral ring, and α and β zones of parapapillary atrophy were plotted and morphometrically analyzed. To obtain values in absolute size units (i.e., millimeters or square millimeters), the ocular and photographic magnification was corrected by using the Littmann method. ¹³ The optic cup was defined on the basis of contour and not of pallor. The border of the optic disc was identical with the inner side of the peripapillary scleral ring. Criteria for the diagnosis in all glaucomas were an open anterior chamber angle and glaucomatous changes of the optic nerve head, including an unusually small neuroretinal rim area in relation to the optic disc size and cup-to-disc ratios that were higher vertically than horizontally. ¹⁴  

In all patients included in the study, intraocular pressure was determined in a circadian curve, with measurements at 5 PM, 9 PM, 12 AM, 7 AM, and 12 PM. 

Normal Subjects

The normal subjects of the control group did not show any abnormality in the ophthalmic evaluation, including slit-lamp examination, tonometry, perimetry, and ophthalmoscopy. 

Preperimetric Glaucoma Groups

In the preperimetric glaucoma groups, maximum intraocular pressure was higher than 21 mm Hg, and patients showed glaucomatous abnormalities of the optic disc and localized or diffuse loss of the retinal nerve fiber layer. Findings in computerized visual field examinations (Octopus program G1; Interzeag) were unremarkable, and the standard indices (mean perimetric defect and corrected loss variance) were in the normal range. This preperimetric glaucoma group included 98 patients with primary open-angle glaucoma without any evident reason for the increased intraocular pressure, and 20 patients with open-angle glaucoma secondary to pigmentary glaucoma, pseudoexfoliation, or traumatic anterior chamber angle recession. 

Progressive and Nonprogressive Preperimetric Subgroups

To judge the prognostic value of the VEP, the preperimetric glaucoma group was divided into patients with progression of glaucomatous optic disc atrophy and patients with no progression of glaucoma (Fig. 1) . Progression of glaucoma was defined as a loss of neuroretinal rim, which could also be accompanied by an increase in parapapillary atrophy, a decrease in the visibility of the retinal nerve fiber layer, or optic disc hemorrhages. Perimetric results were not used as criteria for classification. All subjects in these preperimetric subgroups had three or more ophthalmic examinations, including VEPs and optic disc photography, with an interval of 24.0 ± 3.0 months. To evaluate the progression of glaucoma, the first slide and the newest slide of the optic nerve of the same eye were taken, mixed, and simultaneously projected in a random order. Two examiners (JBJ, WMB) jointly assessed qualitatively whether the two photographs differed. If a difference was detected and if the slide with the more marked optic nerve damage was the more recent photograph, the eye was considered to have progressive glaucomatous optic nerve damage. Of all pairs of optic disc photographs that were judged to differ from each other, approximately 5% to 10% were considered to show the smaller neuroretinal rim in the earlier photograph. These eyes were then classified as not showing a progression of glaucomatous optic nerve damage. After classification of the two preperimetric glaucoma subgroups, two earlier results of VEP measurements were analyzed in each subgroup for longitudinal assessment. Thirty-one patients, representing the group with progressive disease (progressive group), had had VEP measurements 2 and 4 years before the additional damage of the optic nerve head was observed. In 90% (28/31) of these patients perimetric results were always normal. Thus, in the progressive group, the subjects showed additional optic disc changes at the end of the observation period, whereas the perimetric results remained in the normal range in most patients. In the nonprogressive group, 30 patients had three consecutive VEP measurements in a temporal distance of 2 years each. Age, perimetry, and visual acuity did not differ between the progressive and nonprogressive groups. 

Perimetric Group

The perimetric group (n = 161) included 75 patients with primary open-angle glaucoma characterized by intraocular pressure higher than 21 mm Hg, 27 patients with open-angle glaucoma and elevated intraocular pressure measurements secondary to primary melanin-dispersion syndrome (pigmentary glaucoma) or pseudoexfoliation of the lens (pseudoexfoliative glaucoma), and 59 patients with normal-pressure glaucoma. For the diagnosis of normal-pressure glaucoma, all intraocular pressure measurements had to be less than 21 mm Hg without medication. In these latter patients ophthalmoscopy, medical history, and neuroradiologic, neurologic, and medical examinations did not reveal any reason other than glaucoma for the optic nerve damage. Twenty-five patients of this perimetric glaucoma group had two reexaminations after 2 years each. In this period, 5 of the 25 patients showed a progression of optic nerve damage, but alterations of perimetric mean defects were not statistically significant. All 25 subjects of this group entered follow-up analyses. 

VEP Recording

All patients and subjects included in the study underwent VEP determination with blue-on-yellow stimulation. Nearly all measurements were performed by one technician experienced with several thousand VEP examinations. This technician was masked to the group to which each individual study participant belonged. Recording was monopolar, from the inion referenced to the right ear lobe. The left ear lobe was grounded. After 10,000 times amplification (model EMP88, notch filter at 50 Hz; 3 dB points at 0.5 and 70 Hz; Pölzl, Munich, Germany) 75 sweeps were averaged in a personal computer. An analog-to-digital converter (model ME26; Meilhaus Electronic GmBH, Puchheim, Germany) with a 500-Hz sampling rate, 400-ms sweep time, and an artifact limiter was used. Each measurement was repeated at least twice. Traces contaminated by noise or α activities were discarded by the examiner. The electronically averaged mean of two of the most similar traces was used for statistical evaluation. 

Stimulation of the Short-Wavelength-Sensitive Pathway

A two-channel Maxwellian-view system for monocular stimulation equipped with a Xenon-arc lamp and a mechanical mirror system ¹⁵ was used. This stimulator allows much higher luminance and faster stimulus changes than those achieved with the commonly used video displays. A strong yellow (570 nm, 13,000 photopic troland [td]) homogeneous adaptation light suppressed the red and green cones, while a superimposed blue stripe pattern (0.9 cyc/deg, average luminance 330 photopic td, 460 nm) of the same extension stimulated the blue cones in the pattern onset (200 ms) and offset (500 ms) mode. Reliability and validity of this method have been described earlier. ¹⁶ Psychophysical and electrophysiological measurements with this stimulus setup ensured that the sensitivity curve was at its maximum in the blue, providing evidence that the responses were dominated by the activity of the blue-cone pathway. ¹⁷ The circular field size had a 30° diameter and crosshairs provided a central fixation mark. The subject positioned the head on a chin rest with the forehead against a headband. With the blue-on-yellow stimulus, a mainly negative VEP response is seen. Amplitude and peak time of the negative-onset response, analyzed as indicated in Figure 2 , was used in the statistical analyses. 

Statistics

Comparisons between groups were made using Student’s t-test. The level of significance was α = 0.05 (two-sided). In the normal control group, the statistical dependency of measurements from both eyes of the same subject was taken into account by using the mean measurements between the left and right eyes. In the glaucoma groups, one eye of each patient was studied. This was always the eye with the more advanced perimetric loss in the perimetric group. The right or left eye was chosen randomly in patients in the preperimetric group, if no progression had occurred. To determine the VEP time delay in a single patient, data were presented relative to the data in the baseline examination. Data from all subjects were used in cross-sectional analyses. Sensitivity and specificity were used to describe the diagnostic value of the procedures ¹⁸ in subgroups with different occurrence of the disease. Longitudinal data and paired statistics were recorded for patients with progressive or nonprogressive glaucoma (software package SPSSWIN, ver. 10; SPSS, Chicago, IL). 

Results

On the first determination, peak times were prolonged in all glaucoma groups in comparison with times in the normal group (Table 1) . Considering the amplitudes, only results of the perimetric group were reduced significantly in comparison with the control group. Because of this low significance, further analysis of amplitudes were omitted in this study. When the peak times of the nonprogressive group and the progressive group were compared at baseline and 2 years before the additional damage was morphologically evident, no statistically significant difference was found (Table 2) . In the follow-up examination, however, peak times were significantly associated with optic disc change. Mean and confidence interval of VEP peak times are presented in Figure 3 for all subjects’ first and repeated measurements. In the nonprogressive preperimetric and the perimetric groups, patients who underwent repeated VEPs showed no significant differences between reexaminations. In the progressive group, however, the third measurement (i.e., the measurement at the day the new damage was first seen) showed significant prolongation of the peak time in comparison with results of earlier tests. In addition, peak times measured 2 years before the progression was detected morphologically were significantly (P = 0.01) delayed in comparison with the times in the first determination (Fig. 3) . This is illustrated additionally in Figure 4 , which shows standardized data of all patients with stable disease and patients with preperimetric progressive disease: The more recent peak times in the progressive group (Fig. 4A) were mainly longer than the corresponding times in the first measurement, indicating a prolongation with increase of damage. In the stable group (Fig. 4B) , however, the follow-up peak times were both higher and lower than the first determination. 

In addition to longitudinal statistics, cross-sectional analysis was performed to judge the association of VEP peak times with progressive stages of the disease. All subjects studied were classified into subgroups according to the severity of visual field loss or optic nerve damage. VEP peak times showed a significant retardation with increase of glaucomatous defects. When the visual field criteria (Fig. 5A) were used, VEP peak times differentiated statistically between patients in the preperimetric group and patients with a perimetric mean defect exceeding 5 dB. Similarly, a statistically significant difference was found between patients with neuronal rim areas below 1.2 mm² and those with larger rim areas (Fig. 5B) . These findings are underlined by the analyses of sensitivities calculated at a pre-fixed specificity of 95% (Table 3) . For these calculations the same classifications as shown in Figure 5 were applied to judge the diagnostic value of the VEP in a single subgroup. The sensitivity ranged between 38.1% in preperimetric subjects and 81.3% in advanced glaucoma with perimetric mean defects exceeding 11.0 dB. 

Discussion

VEP has been described as a useful tool for diagnosis and monitoring of glaucomatous optic neuropathy. It is especially valid in analyzing and predicting visual properties in patients with severe visual impairments. ^{19 20 21 22} In addition, the VEP can be used as a prognostic tool in preoperative diagnosis. ^{23 24 25} In glaucoma, monitoring rests mainly on tonometry, optic disc analysis, and perimetry, whereas longitudinal studies by electrophysiological methods are less frequent. However, numerous investigators have described VEP amplitude reductions as well as peak time delays in patients with glaucomatous defects ^{16 26 27 28 29 30 31 32 33} and even in ocular hypertension. ^{34 35 36} In general, delays were more frequently observed. ^{34 37 38 39} New investigations with the multifocal technique ^{40 41} recommend the VEP for identification of preperimetric glaucoma or for measurement of visual field damage in perimetric glaucoma. Thus, results from cross-sectional studies may indicate a high value of such determinations for monitoring glaucoma. In the present study, patients with glaucoma were observed with conventional glaucoma procedures as well as with the blue-on-yellow VEP technique. Every subject of this study was tested with stimuli that are likely to favor the responses of the blue-on pathway, as described by Korth et al. ¹⁶ Under these conditions, the contribution of the red-and-green-sensitive pathway was sufficiently suppressed, and the responses were dominated by the activity of the blue-sensitive pathway. ¹⁷ The selection of this stimulus is based on the observation that in glaucoma and even in ocular hypertension, anomalies in blue color vision are found. ^{2 42 43 44 45 46}  

In the present longitudinal investigation, several patients in the preperimetric group showed an increase of optic disc cupping during the study period. Analysis of VEP results in these patients showed an increase of peak time in comparison with earlier measurements. VEPs measured 2 years before the additional damage occurred can be significantly prolonged in relation to a measurement obtained 2 years earlier. These results indicate that measurement of VEP peak time is most useful when comparisons are made within a given patient, relative to the baseline examination. Cross-sectionally, this observation is obscured by the normal variability ^{47 48 49} and limited reliability of VEP measurements, because peak time variation of approximately 10 ms between retests has to be regarded as normal. 

In addition to longitudinal inspection of the data, cross-sectional statistical analyses that included data from all subjects were performed. Such analyses revealed a high association between VEP peak times and perimetrically or morphometrically defined stages of glaucoma. The cross-sectional presentation in Figure 5A suggests that increasing perimetric defects are associated with increasing VEP delays. However, the longitudinal analysis illustrated in Figure 3 indicates that the 25 patients in the perimetric group did not show a prolongation of VEP peak time during the three follow-up examinations. This also seems to be in contrast to the 31 patients in the preperimetric progressive group, who showed an increasing VEP delay. To appreciate this different behavior in the latter two groups the perimetric group would have to be classified further into patients with papillometrically progressive and nonprogressive disease and analyzed separately. That most patients in this group (20 patients) were stable, whereas only five had progressive disease, may explain the missing peak time prolongation during follow-up in the perimetric group. Thus, the increasing peak time delays noted with increasing glaucomatous damage in the cross-sectional analysis (Fig. 5) suggest that an increase in VEP peak time in the perimetric progressive group would also appear in a follow-up analysis. However, this has to be verified by further studies. 

Besides the relatively high variability of electrophysiological measurements that is typical in all sensory determinations, there are other factors that could limit the value of VEP with the present stimulation. In a general population, the age of the patients may artificially increase the differences in the VEPs between the study groups, because of the age-related loss in the transparency of the optic media, the loss of optic nerve fibers, ¹⁴ and the dependence of the VEP on age, as already described. ⁵⁰ In the present study, care was taken that only subjects with clear lenses be included and that subjects be younger than 67 years, to reduce the influence of age-related yellow discoloration of the lens. To exclude a theoretical influence of age on the results of the study, we additionally normalized all measurements for age. 

In summary, the VEP with preferable stimulation of the short-wavelength pathway is one of the most sensitive electrophysiological tests and is significantly correlated with proceeding stages of glaucoma damage. The present data show that delay of peak time can even precede the clinical signs of progressing glaucoma. Because of intensive medical attendance in all study subjects, the conversion rate was low and time consuming. Further follow-up examinations in the patients converting from one group to another of the present study and other patients converting from preperimetric to perimetric glaucoma are necessary to confirm the present findings. Patients with ongoing defects during perimetry have to be studied to prove the present cross-sectional association of VEP peak time and perimetric damages in a long-term evaluation. The present findings suggest that VEP peak times that are prolonged in comparison with those in an earlier determination in the same eye may indicate progression of glaucomatous optic nerve damage. 

 Supported by Grant SFB 539 from the Deutsche Forschungsgemeinschaft, Bonn, Germany.

 Submitted for publication July 30, 2001; revised January 17, 2002; accepted January 29, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Composition of the Study Groups and Results of VEP Measurements Based on the First Determination

	Normal Subjects	Preperimetric Glaucoma	Perimetric Glaucoma
Number	113	118	161
Age (years)	51.2 ± 7.8	51.25 ± 8.6	53.0 ± 8.3
Women/men	53/60	46/72	91/70
Perimetric mean defect (dB)	0.37 ± 1.1	0.36 ± 1.2	6.9 ± 4.8
VEP
Amplitude (μV)	6.0 ± 2.7	5.4 ± 2.2	4.1 ± 2.3
t-Test		NS	6.2
Patients vs. normal subjects			P < 0.001
Peak times (ms)	115.2 ± 7.3	122.2 ± 10.3	134.0 ± 16.3
t-Test		5.8	11.5
Patients vs. normal subjects		P < 0.001	P < 0.001

 Amplitudes and peak times were compared between normal subjects and patients with glaucoma

Table 2.

 

VEP Peak Times in Progressive and Nonprogressive Preperimetric Glaucomas

Glaucoma Group	Determination 1			Determination 2			Determination 3
		Nonprogressive	Progressive	Nonprogressive	Progressive	Nonprogressive	Progressive
Patients (n)	30	31	30	31	30	31
VEP peak time (ms)	119.6 ± 8.5	122.5 ± 10.8	120.0 ± 10.7	125.9 ± 11.2	120.8 ± 9.8	130.8 ± 14.3
Unpaired t-test for comparison of groups	NS			NS			t-test 3.2 P = 0.002

*  Number of determinations

Figure 1.

 

Definition of progressive (n = 30) and nonprogressive (n = 31) glaucoma groups. Patients had three determinations with a time interval of 24.0 ± 3.0 months. All subjects in the progressive group had new additional optic disc damages at the third determination (square). At this time, three of these patients revealed additional visual field losses.

Figure 2.

 

Pattern onset-offset VEPs with blue-on-yellow stimulation. Three repeated measurements at a 2-year interval taken in an eye with preperimetric glaucoma with progression of disease. At the determination corresponding to the third trace ( ), a progression of optic nerve damage was detected morphologically. Arrows: peak time and amplitude measurement (main negativity versus preceding positivity). Onset at 0 ms, offset at 200 ms.

Figure 3.

 

VEP peak times in patients with glaucoma and control subjects (the mean with 95% confidence limits). All patients’ peak times were significantly prolonged in comparison with those in normal subjects. In the nonprogressive preperimetric and perimetric groups, eyes showed no significant differences between reexaminations. However, follow-up in the progressive group showed significant peak time prolongation. This happened 2 years before the progression of glaucomatous damage was morphologically evident. Large filled square: measurements corresponding to a morphologically evident increase in glaucomatous damage. *Significant differences in paired statistics: *P = 0.01, **P = 0.001, ***P < 0.001.

Figure 4.

 

Standardized peak times of pattern onset and offset VEPs in the progressive and nonprogressive preperimetric glaucoma groups. (A, ○) VEP results 4 years before an increase in morphologic damage occurred; (♦) VEP measurement 2 years before the progression of glaucomatous damage was morphologically evident. All subjects in the progressive group had a third VEP determination at the date at which a progression of optic nerve damage was detected (▪). The more recent data in the progressive group were mainly higher than the corresponding first values, pointing at a prolongation with increase of damage. (B) Analogously, the temporal distance between measurements was 2 years in the nonprogressive group (○ first, ♦ second, and ▪ third measurement). The data in this stable group were both higher and lower than the first determination.

Figure 5.

 

Mean data illustrate the increase in VEP peak times in glaucoma and the association with progressive stages of the disease. Patients were classified into subgroups according to the severity of their visual field defects (A) or optic nerve damage (B). Error bars: confidence interval. *P < 0.05, **P < 0.01, ***P < 0.001; Student’s t-test. (A) The preperimetric group had prolonged peak times in comparison to the control group. When the subjects were grouped by perimetric mean defects, VEP peak times were significantly longer in patients with mean defects exceeding 5 dB than in patients with preperimetric disease. (B) Similarly, statistically significant differences were found between patients with neuronal rim areas of less than 1.2 mm² and those with larger rim areas.

Table 3.

 

Diagnostic Value of VEP Assessment in Subgroups Classified According to the Severity of the Disease

	Sensitivity of VEP Peak Time
Progressive visual field losses*
No field losses (preperimetric subjects)	38, 1
<5	46, 2
5–7.9	61, 0
8–11	62, 5
>11	81, 3
Progressive optic nerve damage, †
>1.5	39, 5
1.5–1.2	37, 3
1.19–0.9	53, 4
0.89–0.6	65, 2
<0.6	75, 0

 Sensitivities as calculated for a defined specificity of 95%.

*  Classifications according to perimetric mean deficits (dB).

†  Classifications according to neuroretinal rim area (mm²).