Correlation of Pattern Electroretinogram with Optic Disc Cup Shape in Ocular Hypertension

Abstract

purpose. To evaluate the correlation of pattern electroretinogram (PERG), an index of inner retinal function, with confocal scanning laser (CSLO) optic disc structural parameters in ocular hypertension (OHT).

methods. Thirty-four patients with OHT, normal white-on-white (Humphrey 30-2) perimetry, and normal clinical optic discs were examined with PERG and CSLO disc analysis. Two groups of normal subjects (n = 38 and 18, for PERG and CSLO, respectively) and a group of 12 patients with early open-angle glaucoma (EOAG) were also tested. Pattern electroretinogram amplitudes were measured in response to sinusoidal gratings of variable spatial frequency (0.58–5.8 cycles/degree), modulated in counter–phase at 7.5 Hz. Morphometric optic disc parameters were obtained by the Heidelberg Retina Tomograph (HRT), either globally or from predefined disc sectors. In addition to standard parameters, the cup shape measure, an index of depth variation and steepness of the cup walls, was determined.

results. In individual OHT patients, PERG amplitudes at 2.6 cycles/degree were negatively correlated with cup shape measures (r =− 0.43, P < 0.01) obtained from analysis of the inferotemporal (IT) sector. No significant correlations were found for the other parameters. On average, the cup shape measures derived from IT sector or global analysis were significantly (P < 0.01) worse, and closer to the measures of EOAG patients, in OHT patients with abnormal PERG compared with those with normal PERGs. The cup shape measure displayed a low sensitivity (20%) and a high specificity (100%) in predicting PERG abnormalities in individual OHT patients.

conclusions. The results indicate that in OHT there is a significant although weak correlation between PERG amplitude and the shape of the optic disc cup, suggesting a parallel involvement of both function and morphology. Combined PERG and optic disc cup structural analysis is of potential diagnostic value to detect early damage to optic nerve head in individual OHT patients.

In glaucoma, structural changes to the optic nerve head are usually associated with deterioration of function. ^{1 2 3} In ocular hypertension (OHT), morphologic signs of early optic nerve damage usually precede the onset of typical visual field loss, although subtle functional deficits may be revealed by refined psychophysical and electrophysiological techniques. ⁴ Morphologic optic disc abnormalities have been described in OHT by computer-assisted planimetric analysis of the optic disc (see, for example, ^{Ref. 5)} and, more recently, by confocal scanning laser ophthalmoscopy (CSLO), ^{6 7} which has provided three-dimensional analysis of the optic disc structure. Reproducibility and potential advantages of the CSLO tomography over other techniques have been described elsewhere. ^{8 9 10 11} Functional deficits in OHT have been reported by measuring luminance and chromatic contrast sensitivities, ^{12 13} motion detection perimetry, ¹⁴ blue-on-yellow perimetry, ^{15 16} and luminance and chromatic pattern electroretinograms (PERGs). ^{13 17 18}  

Although the correlation between CSLO and functional losses has been well documented in glaucomatous eyes (see, for example, ^{Refs. 1} and 3), the same relationship has not been clearly established in OHT eyes. An approach to evaluate the correlation is to compare CSLO parameters of the optic disc with a sensitive test of retinal ganglion cell function (i.e., the PERG). ^{19 20} The PERG has been reported to be abnormal in a substantial proportion of OHT eyes ^{13 17 18 21 22 23} and of predictive value for the development of early field losses in OHT eyes. ^{24 25} A quantitative association between losses in CSLO and PERG measurements would indicate that structural and functional damage to the optic nerve develops in parallel in early stages of disease. Lack of correlation, on the other hand, would mean that a certain amount of structural damage is necessary for functional deficits to become detectable (as previously suggested ²⁶ ) or that functional deficits may precede structural damage. Clinically, evaluating the relationship between structural and functional damage in OHT could help to better delineate the boundaries between healthy and pathologic optic nerve heads in glaucoma-risk eyes. The present study was designed to evaluate the correlation between CSLO optic disc and PERG measurements in a cohort of OHT patients with normal conventional white-on-white perimetry. Morphometric and functional results of OHT eyes were also compared with those obtained from normal control subjects or patients with early, clinically-defined, glaucoma. 

Methods

Subjects

Thirty-four patients with a diagnosis of OHT (intraocular pressure, IOP > 21 mm Hg on two or more separate occasions, normal optic disc appearance, normal Goldmann and Humphrey 30-2 threshold test perimetry, best corrected visual acuity ≥20/20) were examined with both PERG and CSLO tomography. Normal appearance of the optic disc, on routine stereoscopic examination with slit-lamp biomicroscopy and 78-diopter (D) lens, was defined as a vertical cup-to-disc diameter ratio less than or equal to 0.5, with no asymmetry (≥0.2, unexplained by side differences in disc size), excavation, thinning of the rim, notching, hemorrhages, nerve fiber layer defects, or parapapillary atrophy. Two independent groups of normal subjects (n = 38 and 18 evaluated by PERG and CSLO, respectively) and a group of 12 early open-angle glaucoma (EOAG) patients were also tested. Diagnosis of EOAG was established on the basis of an elevated IOP (>21 mm Hg on two separate occasions), an open angle, the presence of abnormal white-on-white automated perimetry (mean deviation ≤ 10 dB on Humphrey 30-2) with a typical reproducible defect, and glaucomatous optic disc, evaluated by slit-lamp biomicroscopy, with a cup-to-disc ratio greater than 0.6 (or an interocular cup-to-disc ratio asymmetry greater than or equal to 0.2) and one or more of the above-listed disc abnormalities. Clinical and demographic characteristics of the study population are summarized in Table 1 . Age, sex, and refractive error distributions of control subjects were comparable to those of OHT patients. The patients enrolled in the study were recruited from a larger cohort of OHT and EOAG patients evaluated at the Glaucoma Service of the Institute of Ophthalmology, Catholic University (Rome, Italy). Additional inclusion criteria were reliable visual fields (at least two 30-2 threshold tests within 1 month) and clear CSLO images of the optic nerve head with a definable disc margin. PERG and visual field and CSLO analyses were obtained for each patient within 2 months. Cases with refractive error of less than −5.00 or more than +4.00 D spherical equivalent, astigmatism of more than ±1.00 D, presence of disorders affecting the optic disc or visual field, low perimetric reliability, ²⁷ or poor CSLO images (average variability > 25 μm) were excluded. At the time of testing none of the patients was under treatment with ocular hypotensive drugs. Informed consent was obtained from every subject or patient after the procedures used in the study were fully explained. The research followed the tenets of the Declaration of Helsinki. 

Apparatus and Procedure

Electroretinograms were recorded according to a previously published technique. ²³ Briefly, stimuli were vertical sinusoidal gratings of variable spatial frequency (0.58, 0.88, 1.3, 1.7, 2.6, and 5.8 cycles/degree), modulated in counter–phase at 7.5 Hz and electronically generated on a high-resolution TV monitor (contrast: 90%; mean luminance: 80 candela [cd]/m²; field size: 14° [width] × 24° [height]). Subjects were fixated at the center of the stimulating field with natural pupils, and size was measured. No differences in pupil size were observed between patients and control subjects or between the two patient groups. Electroretinograms were recorded by a Ag/AgCl electrode taped on the skin of the lower eyelid. A similar electrode, placed over the eyelid of the contralateral unstimulated eye was used as reference. ^{17 19 23} Responses were amplified (100,000), filtered (1–30 Hz), sampled with a resolution of 12 bits and averaged (400 events) with automatic artifact rejection. Two replications were obtained for each record to verify reproducibility. The amplitude (in microvolts) of the Fourier analyzed response 2nd harmonic was measured and plotted as a function of spatial frequency. 

CSLO tomography of the optic disc was performed by the Heidelberg Retina Tomograph (HRT; Heidelberg Engineering GmbH, Heidelberg, Germany), by analyzing the mean of three 10° topographic images for each eye, according to Weinreb et al. ²⁸ Details of this instrument and its reproducibility have been published. ^{10 11 28 29 30} Test-retest variability of the three measurements of each point, expressed by the average of the standard deviations of the topographic values of each pixel in the three images, was 13.47 ± 4.39 μm (range, 7.3 to 23.38 μm) for the normal group, 12.48 ± 4.43 μm (range, 6.7 to 24.25μ m) for the OHT group, and 14.37 ± 4.13 μm (range, 7.31 to 23.01 μm) for the EOAG group. Each mean topography image was automatically corrected for horizontal and vertical tilt. ³¹ The margin of the optic disc was manually drawn on the image as a contour line around the inner edge of the peripapillary scleral ring of Elschnig, using a computer mouse system by a trained operator (TS). Two axial boundaries, the curved surface and the reference plane, were used by the 2.01 HRT software to generate the optic nerve head measurements. ³¹ Most of two- and three-dimensional data (e.g., cup and rim area, cup and rim volume) were obtained with respect to the standard reference plane, placed by the current software 50 μm posterior to the mean height of the disc margin contour line at the papillomacular bundle, more precisely in a temporal segment between 350° and 356°. ³² Other topographic optic disc parameters (e.g., cup shape measure and maximal cup depth) were automatically measured relative to the curved surface. This surface is bound by the disc contour line and follows the height variation of the retinal surface along the contour line, whereas the height of its center equals the mean height of the optic disc margin; all connecting lines from the center to a boundary point are straight lines. ³¹ For each optic disc, the following morphometric parameters ³ were evaluated either globally or for the predefined HRT disc sectors: disc area, cup area, cup-to-disc area ratio, rim area, cup volume, rim volume, maximal cup depth, and cup shape measure. The cup shape measure is a measure of the skewness of the frequency distribution of depth values of disc cupping. ^{31 33 34} It summarizes in numerical terms the structure of the cup, taking into account both depth variation and steepness of the cup walls. ^{1 2} Unlike other structural cup or rim parameters, it is independent of reference plane. ³¹ The parameter has a negative value for a flat or nearly flat excavation and turns to positive values if the slope at the edges of the excavation increases. ³⁴ In normal eyes cup shape measure is typically negative, whereas glaucomatous eyes tend to be less negative or positive. Magnification error was automatically corrected by using patients’ keratometry readings. The optic disc sectors included in the analysis were: superotemporal (ST, 45°), superonasal (SN, 45°), nasal (N, 90°), inferonasal (IN, 45°), inferotemporal (IT, 45°), and temporal (T, 90°). Disc sector analysis was dictated by previous clinical and histopathologic findings ^{35 36 37} that early glaucomatous optic nerve damage is often detectable at specific sectors of the disc, including mainly superior and inferior poles. 

Statistical Analysis

Pattern electroretinogram and HRT data from only one eye, randomly selected, per subject or patient were included in the analysis. Between-group comparisons (including normal, OHT, and EOAG eyes) were performed by univariate and multivariate ANOVAs, with multiple contrasts. A P < 0.05 was considered significant. Correlations between PERG amplitudes (dependent variables) and HRT parameters (predictor variables) derived from OHT eyes were performed by multiple linear regression analyses, which evaluated the effect of a single variable while correcting for the effects of the disc size, given its well-known influence on the other disc parameters. ³⁸ Given the large number of variables analyzed, a conservative two-tailed P ≤ 0.01 was adopted. Because no a priori assumption could have been made on the presence of a linear relationship between PERG amplitudes and HRT parameters in individual OHT eyes, correlations were also evaluated by a different approach. For each PERG variable (i.e., amplitudes at the different spatial frequencies), normative values (mean and 95% confidence limits) were established. Individual OHT eyes’ PERGs were considered abnormal if their amplitudes were below the lower 95% confidence limits at one or more spatial frequencies. OHT eyes were then subdivided into those with normal or abnormal PERGs, and the HRT parameters derived from the two groups were compared by multivariate and univariate ANOVAs. A P < 0.05 was considered significant. To determine the sensitivity and specificity of a single morphometric parameter in predicting corresponding PERG losses, the incidence of HRT parameters abnormalities in individual OHT eyes was determined (taking as reference values the normal 95% confidence limits established for discs of different sizes, ^{38 39} i.e., having disc areas in the ranges 1–2 and 2–3 mm², see also the Results section) and compared with that of PERG abnormalities by 2 × 2 tables. For the purposes of this analysis, limited to OHT eyes with a normal visual field, the PERG was considered as a functional“ gold standard”. It should be noted, however, that the technique, as the HRT analysis, is still an experimental procedure and cannot be considered as a gold standard in clinical terms. 

Results

In Table 2 the mean (±SD) PERG amplitudes at the different spatial frequencies are reported for normal, OHT, and EOAG eyes. Mean amplitudes of OHT eyes were significantly lower [multivariate F(6,65): 3.3, P = 0.01] than those of normal eyes, with greatest losses at medium spatial frequencies (1.3–2.6 cycles/degree). Comparison between OHT and EOAG eyes also showed that in the latter PERG amplitudes at 1.3 cycles/degree were on average smaller[ univariate F(1,39): 4.2, P < 0.05] than those observed in the former. Among individual OHT eyes, PERG amplitudes were found to be abnormal (i.e., lower than 95% confidence limits at one or more spatial frequencies, see the Methods section) in 15 eyes (44%). Table 3 reports the mean values (±SD) of HRT parameters for the different groups of the study population. Data from global and sectorial analyses are reported. Among the various cup or rim parameters, only those known as the most sensitive and specific for glaucoma detection ² (i.e., the cup-to-disc area ratio, the neuroretinal rim area, and the cup shape measure) are reported. The mean values of optic disc sizes are also included in Table 3 . The average values of cup-to-disc area ratio and cup shape measure (derived from global disc analysis) differed significantly between normal and OHT eyes [cup-to-disc area ratio: univariate F(1,48): 146.3; cup shape measure: univariate F(1,48): 91.7; P < 0.01] as well as between normal and EOAG eyes (cup-to-disc area ratio: univariate F(1,26): 61.4; cup shape measure: univariate F(1,26): 20.4; P < 0.01). Rim area did not differ between normal and OHT eyes but did so (P <0.05) between normal and EOAG eyes [univariate F(1,26): 13]. HRT parameters of individual OHT eyes were compared with the 95% confidence limits established for normal eyes with disc areas in the ranges 1 to 2 and 2 to 3 mm². Cup-to-disc area ratio was abnormal in 14 (41.2%), rim area in 8 (23.5%), and cup shape measure in 3 (8.8%) of 34 OHT patients. 

In individual OHT eyes, significant PERG abnormalities at low and medium spatial frequencies tended to be associated with more positive values of cup shape measure. These trends were observed when the parameter was derived either from the global analysis or the analyses of temporal disc sectors. Figure 1 shows the PERG and HRT results obtained from the right eyes of three representative patients: an OHT with normal PERG, an OHT with abnormal PERG, and an EOAG patient. In Figure 1A , PERG amplitudes of the patients are plotted as a function of spatial frequency. The continuous and dotted lines in the plots indicate the mean and the lower 95% confidence limits, respectively, established for the different amplitudes in normal subjects. Figure 1B shows the HRT reflectance images of the optic discs from each patient and the corresponding z-profiles taken along oblique meridians (indicated by the thick lines) crossing both SN and IT sectors. Each cupping profile has been selected as the most representative (i.e., that closest to the average) of all the profiles at the IT sector. It can be noted that the steepness of the cup profile is substantially greater in the OHT eye with abnormal PERG compared with that of the OHT eye with normal PERG, although similar to that of the EOAG eye. The differences in cup profiles between OHT patients with normal or abnormally reduced PERGs are quantified by differences in cup shape measures, which were −0.24 and− 0.11 in the OHT eyes with normal and abnormal PERGs, respectively. The cup shape measure in the EOAG eye was −0.07. 

Figure 2 shows a scattergram of PERG amplitudes recorded individually from OHT eyes at 2.6 cycles/degree and plotted as a function of the corresponding cup shape measures at the IT sector. The negative correlation was statistically significant (multiple regression: r = −0.43, coefficient for cup shape measure: −1.03, P < 0.01). No other significant correlations were found between PERG amplitudes and HRT measures in OHT eyes. 

Figure 3 shows the frequency distribution histograms of the three HRT parameters (i.e., cup-to-disc area ratio, rim area, and cup shape measure) gathered from the sectorial analysis (IT sector) in the different groups of the study population. For OHT eyes, the data are separately plotted for the group with normal PERG amplitude (n = 19) and for that with abnormally reduced PERGs (n = 15). It can be seen that in OHT eyes with abnormal PERGs, but not in those with normal PERGs, the distribution of cup shape measure tends to be more similar to those of EOAG eyes than to those of normal subjects. The distributions of cup-to-disc area ratio and neuroretinal rim area do not show similar differences across the patient group. A multivariate ANOVA, including simultaneously cup-to-disc area ratio, rim area, and cup shape measure as dependent variables, showed significant differences between OHT eyes with normal and abnormal PERGs[ multivariate F(3,29): 4.6, P = 0.01]. Univariate F tests, however, reached the significance level only for the cup shape measures [F(1,31): 10.44, P < 0.01]. The difference in cup shape measure distributions between OHT eyes with normal and those with abnormal PERGs was also observed in the global disc analysis[ not shown, univariate F(1,31): 12.04, P < 0.01]. 

In Table 4 incidences of HRT parameter and PERG abnormalities in individual OHT patients are compared by 2 × 2 tables. The functional parameter PERG was considered in this analysis the gold standard of reference (see also the Methods section). It can be noted that all three HRT parameters (cup-to-disc area ratio, rim area, and cup shape measure, derived from global disc analysis) displayed a low sensitivity and a relatively high specificity in predicting corresponding PERG abnormalities. Among the parameters, the highest specificity and positive predictive value was displayed by cup shape measure. 

Discussion

The goal of the present study was to evaluate in eyes with OHT, normal conventional automated perimetry, and normal clinical appearance of the optic nerve head, the relationship between early PERG losses and structural optic disc parameters derived from CSLO analysis. Analysis of correlations between PERG and HRT parameters showed that PERG amplitudes at medium spatial frequencies (i.e., responses previously reported as specifically vulnerable in OHT ^{17 23} ) tended to decrease significantly as the cup shape measure of individual eyes moved toward the abnormal range (>−0.10).² This was demonstrable when amplitudes were compared with cup shape measures calculated separately for the IT disc sector, which has been found to be the most frequently affected in early glaucoma. ^{35 37} The correlation did not attain statistical significance when the structural parameter was derived from global disc analysis. However, a significant difference in the distributions of parameter values was found when eyes with normal or abnormal PERGs were compared, with the latter showing average values closer to those found in a glaucomatous population. 

In the past, only a few studies have evaluated the correlation between PERG and disc morphometry in OHT. ^{40 41} None of these found, in cross-sectional evaluations, a significant association between PERG amplitudes or latencies and disc parameters. Longitudinal evaluations provided contrasting results. In a 6-month follow-up study, Bach and Funk ⁴⁰ found that PERG amplitude losses in glaucoma suspect eyes were significantly correlated with progressive rim area losses. In contrast, in a longer longitudinal study, Bömer et al. ⁴¹ reported a poor value of the PERG in predicting the worsening of morphometric parameters in glaucoma suspects. None of the previous studies evaluated the relationship between PERG and the shape of the optic disc cup, expressed quantitatively by the cup shape measure. 

Recent clinical evidence ² indicates that the cup shape measure can be used with high diagnostic precision to discriminate between normal and glaucomatous eyes. Studies evaluating in glaucomatous eyes the relationship between visual field loss and structural damage to the optic nerve found that the cup shape measure showed the strongest correlations with mean deviation or corrected pattern SD values obtained from white-on-white ^{1 2 3} or blue-on-yellow ³ perimetries. Taken together, these previous findings support the suggestion that an abnormality in the cup shape measure reflects indirectly glaucomatous damage to retinal ganglion cells and optic nerve axons. ¹ This is in agreement with histologic findings, ⁴² demonstrating that morphologic changes in the lamina cribrosa are correlated with neural loss in open-angle glaucoma, and clinical belief ³⁶ that increased cupping of the disc is a manifestation of glaucomatous neural damage. It should be emphasized, however, that the amount of correlation we found was rather weak (r = −0.43), with the structural parameter accounting for no more than 20% of the PERG variance. Comparable results were found in the previous studies correlating perimetric sensitivities with HRT parameters when only the populations of OHT and EOAG patients were considered. ³ The weakness of the association may have different causes that are detailed below (see the next paragraph) and indicates that in the clinical setting a full characterization of the status of the optic nerve head in OHT requires both functional tests and morphologic disc analysis. 

A subset of the OHT eyes evaluated in this study showed significant abnormalities of PERG, HRT parameters, or both. PERG alterations tended to be more frequent than those of HRT parameters. This higher sensitivity of the PERG over structural disc parameters in OHT may have different but not mutually exclusive explanations. First, PERG losses most probably reflect a functional/histologic damage to the optic nerve fibers, which may develop before the occurrence of changes in the cup structure. Recent clinical observations ⁴³ showing that PERG amplitude in OHT eyes was inversely correlated with the thickness of the peripapillary nerve fiber layer as measured by OCT imaging, lend support to this hypothesis. Second, the normal values for most HRT parameters are strongly dependent on the size and shape of the optic disc (see for example ^{Ref. 38)} , resulting in an increased variability and low clinical sensitivity in borderline cases. Interestingly, cup-to-disc area ratio, rim area, and cup shape measure displayed a low sensitivity but a relatively high specificity in predicting PERG losses in individual eyes. The highest predictive value was shown by the cup shape measure. This raises the possibility that when used in combination PERG and cup shape measure could help in defining the limits between normal and pathologic optic discs, strengthening an otherwise uncertain diagnosis of optic nerve damage in individual OHT eyes. Clearly, only longitudinal studies evaluating the rate of development of field losses in different subcategories of OHT eyes will clarify the clinical relevance of the present findings. 

In summary, the results of this study show that in OHT there is a significant, although weak, correlation between the PERG, an index of inner retinal function, and optic disc structure. In individual eyes, abnormalities in the shape of optic disc cup may be highly predictive for the presence of ERG losses. These data suggest a parallel involvement of both structure and function in OHT and a potential clinical value of combined PERG and CSLO optic disc analysis in detecting eyes at increased risk for glaucoma damage. 

 Reprint requests: Benedetto Falsini, Istituto di Oftalmologia, Università Cattolica S. Cuore, Lgo F. Vito 1, 00168 Rome, Italy.

 Submitted for publication November 11, 1998; revised February 22, 1999; accepted April 12, 1999.

 Proprietary interest category: N.

 

Table 1.

 

Table 1.

 

Demographic and Clinical Characteristics of the Study Population

	Normal for PERG amplitudes (n = 38)	Normal for HRT parameters (n = 18)	OHT (n = 34)	EOAG (n = 12)
Age (years)	38± 8*	38.22± 8.94	37.88± 13.76	40.25± 15.09
	(24–54)	(23–53)	(20–59)	(22–61)
Sex (M/F)	19/19	10/8	18/16	8/4
Refractive error (diopters spherical equivalent)	(−3.50− +3.50)	(−2.00−+1.50)	(−3.50−+1.50)	(−2.50−+2.00)
Mean IOP	14± 1	14.6± 1.2	22.05± 3.29	24.67± 3.02
(mm Hg)	(12–16)	(12–18)	(19–28)	(20–28)
Maximum IOP			25.13± 6.81	25± 4.03
(mm Hg)			(23–28)	(24–30)

*  Means ± standard deviation (range).

Table 2.

 

Table 2.

 

PERG Results in Normal, OHT and EOAG Eyes

Spatial Frequency (cycles/degree)	Normal (n = 38)	OHT (n = 34)	EOAG (n = 12)
0.58	0.55 (0.06)*	0.41 (0.21)	0.31 (0.15)
0.88	0.66 (0.06)	0.47 (0.23)	0.41 (0.16)
1.3	0.85 (0.13)	0.57 (0.28)	0.41 (0.13)
1.7	1.05 (0.25)	0.61 (0.30)	0.48 (0.16)
2.6	0.66 (0.12)	0.58 (0.28)	0.47 (0.28)
5.8	0.48 (0.12)	0.41 (0.18)	0.33 (0.14)

*  Means (standard deviation).

Table 3.

 

Table 3.

 

HRT Results in Normal Subjects, OHT and EOAG Patients

	Normal (n = 18)	OHT pooled (n = 34)	OHT (normal PERG) (n = 19)	OHT (abnormal PERG) (n = 15)	EOAG (n = 12)
DISC AREA
Total	1.98 (0.35)*	2.21 (0.33)	2.26 (0.32)	2.14 (0.35)	2.18 (0.29)
Supero-temporal	0.26 (0.05)	0.29 (0.05)	0.30 (0.04)	0.28 (0.06)	0.28 (0.04)
Supero-nasal	0.25 (0.05)	0.28 (0.04)	0.29 (0.04)	0.27 (0.05)	0.28 (0.04)
Nasal	0.48 (0.09)	0.54 (0.09)	0.54 (0.09)	0.52 (0.08)	0.53 (0.08)
Infero-nasal	0.24 (0.05)	0.27 (0.05)	0.28 (0.04)	0.26 (0.05)	0.27 (0.04)
Infero-temporal	0.28 (0.06)	0.30 (0.05)	0.31 (0.05)	0.29 (0.05)	0.29 (0.04)
Temporal	0.48 (0.09)	0.53 (0.09)	0.54 (0.09)	0.52 (0.08)	0.53 (0.07)
CUP/DISC AREA RATIO
Total	0.19 (0.07)	0.29 (0.13)	0.28 (0.12)	0.31 (0.14)	0.36 (0.15)
Supero-temporal	0.22 (0.11)	0.32 (0.17)	0.31 (0.16)	0.34 (0.17)	0.41 (0.20)
Supero-nasal	0.09 (0.08)	0.24 (0.16)	0.23 (0.17)	0.25 (0.16)	0.31 (0.17)
Nasal	0.06 (0.07)	0.19 (0.14)	0.19 (0.14)	0.19 (0.15)	0.24 (0.17)
Infero-nasal	0.06 (0.07)	0.16 (0.11)	0.16 (0.11)	0.16 (0.12)	0.24 (0.15)
Infero-temporal	0.17 (0.12)	0.28 (0.13)	0.26 (0.12)	0.30 (0.14)	0.37 (0.13)
Temporal	0.43 (0.10)	0.48 (0.15)	0.46 (0.14)	0.50 (0.17)	0.55 (0.16)
RIM AREA
Total	1.58 (0.21)	1.55 (0.27)	1.60 (0.22)	1.48 (0.32)	1.36 (0.25)
Supero-temporal	0.20 (0.03)	0.19 (0.04)	0.20 (0.04)	0.18 (0.05)	0.17 (0.06)
Supero-nasal	0.23 (0.03)	0.21 (0.04)	0.22 (0.04)	0.20 (0.04)	0.19 (0.06)
Nasal	0.45 (0.06)	0.43 (0.08)	0.44 (0.08)	0.42 (0.09)	0.39 (0.07)
Infero-nasal	0.23 (0.04)	0.23 (0.03)	0.23 (0.03)	0.22 (0.04)	0.20 (0.04)
Infero-temporal	0.23 (0.06)	0.21 (0.05)	0.23 (0.04)	0.20 (0.05)	0.18 (0.02)
Temporal	0.27 (0.06)	0.27 (0.08)	0.28 (0.07)	0.25 (0.09)	0.23 (0.06)
CUP SHAPE MEASURE
Total	−0.26 (0.08)	−0.18 (0.08)	−0.21 (0.07)	−0.15 (0.08)	−0.15 (0.07)
Supero-temporal	−0.21 (0.11)	−0.10 (0.14)	−0.13 (0.12)	−0.06 (0.15)	−0.07 (0.12)
Supero-nasal	−0.27 (0.14)	−0.16 (0.11)	−0.18 (0.12)	−0.14 (0.11)	−0.07 (0.24)
Nasal	−0.30 (0.15)	−0.20 (0.14)	−0.23 (0.13)	−0.17 (0.13)	−0.20 (0.06)
Infero-nasal	−0.31 (0.18)	−0.16 (0.12)	−0.18 (0.13)	−0.14 (0.12)	−0.11 (0.11)
Infero-temporal	−0.19 (0.13)	−0.15 (0.11)	−0.19 (0.10)	−0.10 (0.10)	−0.09 (0.08)
Temporal	−0.13 (0.10)	−0.09 (0.08)	−0.11 (0.07)	−0.06 (0.08)	−0.08 (0.04)

*  Means (standard deviation).

Figure 1.

 

(A) Plots of PERG amplitudes at different spatial frequencies obtained from three representative patients’ eyes (from top to bottom): an OHT patient with normal PERG, an OHT with abnormal PERG, and an EOAG patient. In each plot, the continuous and dotted lines indicate the normal mean values and 95% confidence limits, respectively, established for the different response amplitudes. (B) HRT reflectance images of the optic discs of the same eyes whose PERGs are reported in (A). To the right of each image, the corresponding z-profiles of the cupping (taken along oblique meridians and crossing both SN and IT sectors) are also shown. Each cupping profile has been selected as that closest to the average of all profiles at the IT sector. Dashed lines in the z-profiles plots indicate the level of the reference plane.

Figure 2.

 

Pattern electroretinogram amplitudes, recorded from individual OHT eyes at 2.6 cycles/degree, plotted as a function of corresponding cup shape measures obtained from analysis of IT disc sector (multiple regression with cup shape measure and disc area as dependent variables: r = −0.43, coefficient cup shape measure =− 1.03, P < 0.01).

Figure 3.

 

Frequency distribution histograms of three HRT parameters (cup-to-disc area ratio, rim area, and cup shape measure) gathered from the IT disc sector analysis in the different groups of the study population. For OHT eyes, data are separately plotted for the group with normal PERG amplitudes (n = 19) and for that with abnormally reduced PERGs (n = 15). N^o, number.

Table 4.

 

Table 4.

 

Sensitivity and Specificity of HRT Parameter Abnormalities in Predicting PERG Abnormalities in Individual OHT Eyes

	Abnormal PERG (n = 15)	Normal PERG (n = 19)
Abnormal CDA* (n = 14)	8	6	Sensitivity: 53.3% Specificity: 68.4%
Normal CDA (n = 20)	7	13	Positive predictive value: 57.1%
Abnormal RA, † (n = 8)	5	3	Sensitivity: 33.3% Specificity: 84.2%
Normal RA (n = 26)	10	16	Positive predictive value: 62.5%
Abnormal CSM, ‡ (n = 3)	3	0	Sensitivity: 20% Specificity: 100%
Normal CSM (n = 31)	12	19	Positive predictive value: 100%

*  CDA = cup-to-disc area ratio;

†  RA = rim area;

‡  CSM = cup shape measure.