Correlation between Peripapillary Atrophy and Optic Nerve Damage in Normal-tension Glaucoma

Correlation between Peripapillary Atrophy and Optic Nerve Damage in N ormal--tension Glaucoma Ki Ho Park, MD,I Goji Tomita, MD, PhD,2 Shin Yih Liou, MD,2 Yoshiaki Kitazawa , MD, PhD 2 Purpose: To investigate the correlation between peripapillary atrophy and visual field defects as well as optic nerve head configurations in patients with normal-tension glaucoma (N:rG). Methods: Topographic measurements for peripapillary atrophy and optic nerve head using confocal scanning laser tomography and automated static threshold perimetry were performed on 102 eyes of 51 patients with NTG. Peripapillary atrophy was divided into (1) a central zone (zone Beta) with visible, large choroidal vessels and sclera, and (2) a peripheral zone (zone Alpha) with irregular hyper- and hypopigmentation. The area, angular extent around the disc, and radial extent of each zone were measured. Results: The area and extent of zone Beta increased significantly with increasing visual field defects expressed in terms of mean deviation, corrected pattern standard deviation, central visual field defects within 5° of fixation, and superior hemifield defects (r = 0.3770 ~ 0.5291, P < 0.01). The angular extent of zone Beta represented localized field defects better (r = 0.5217, P < 0.001) than diffuse field defects (r = -0.3770, P < 0.01). Zone Beta significantly correlated with optic nerve head topography. Intraindividual right-left-side differences of corrected pattern standard deviation showed the highest correlation with the side differences of zone Beta area (r = 0.6305, P < 0.001). The location of visual field defects correlated significantly with the location of peripapillary atrophy (chi-square = 9.0484, P = 0.011). Zone Alpha was not significantly correlated with visual field defects or optic nerve head configurations (P > 0.05). Conclusion: Peripapillary atrophy is significantly associated with functional and structural optic nerve damage in NTG. Ophthalmology 1996; 103:1899-1906

Peripapillary atrophy or crescent has been reported to be more frequently present and extensive in eyes with hightension glaucoma than in healthy eyes. l - 6·1t also has been suggested that there are significant correlations between Originally received: November 27, 1995. Revision accepted: July 17, 1996. I Department of Ophthalmology, College of Medicine, Seoul National University Hospital, Seoul, Korea. 2 Department of Ophthalmology, Gifu University School of Medicine, Gifu, Japan. None of the authors has any proprietary interest in the development or marketing the Heidelberg Retina Tomograph or any competing instrument. Reprint requests to Yoshiaki Kitazawa, MD, PhD, Department of Ophthalmology, Gifu University School of Medicine, 40 Tsukasa-machi, Gifu 500, Japan.

the extent or the location of peripapillary atrophy and visual field defects in high-tension glaucoma. 5•7• Only a few studies, however, have attempted to evaluate peripapillary atrophy in normal-tension glaucoma (NTG).9, lO Buus and Anderson9 reported that the extent and prevalence of peripapillary atrophy were significantly greater in NTG than in ocular hypertension. Jonas and Xu lO demonstrated that the area of peripapillary atrophy in NTG was significantly larger than that in normal eyes. However, to the best of our knowledge, no cross-sectional study has been done to determine the correlation between peripapillary atrophy and visual field defects as well as optic nerve head configuration in NTG. Recent studies have shown that peripapillary atrophy is a significant risk factor for progression of visual field

1899

Ophthalmology

Volume 103, Number 11, November 1996

damage in NTG. 11,12 These results prompted us to attempt to obtain precise measurements of peripapillary atrophy and determine whether there was any significant correlation between it and glaucomatous optic nerve damage in NTG, We developed a method to measure the area and extent of peripapillary atrophy using confocal scanning laser tomography, a new technique permitting three-dimensional evaluation of optic nerve head with a higher level of reproducibility than any other current method. 13 - 1? Using data obtained in a cross-sectional study of peripapillary and optic nerve head parameters, measured by confocal scanning laser tomography, and visual field defects, evaluated by automated static threshold perimetry, we investigated how closely peripapillary atrophy was related to functional and structural optic nerve damage in NTG.

Materials and Methods Criteria for Patient Selection We reviewed the medical records of 105 patients with a diagnosis of NTG according to the following criteria at the glaucoma clinic of Gifu University Hospital from December 1, 1993, to May 31, 1995: 1. Glaucomatous optic nerve head damage in both eyes; 2. At least one eye that met the following criteria for glaucomatous visual field defects as detected in two consecutive fields by the Humphrey field analyzer 630 (Zeiss-Humphrey, Inc, San Leandro, CA), using program 30-2, standard, and full-threshold strategy: A. Three or more adjacent points with P < 0.05 in a total deviation probability map, B. Two or more adjacent points with P < 0.01 in a total deviation probability map, or C. A difference of > 10 decibels (dB) across the nasal horizontal meridian at two or more adjacent points; 3. A reliability with fixation loss less than 20% and falsepositive and negative rates less than 15%; 4. A mean intraocular pressure (lOP) of 21 mmHg or less, based on measurements obtained at 2-hour intervals over 24 hours in both eyes; 5. The normal, open anterior chamber angle; 6. Results of neuroradiologic, rhinologic, and general medical examinations did not disclose any pathology responsible for optic nerve damage other than glaucoma; 7. No history of previous ocular surgery, including filtering procedures or laser treatments, in either eye. The "glaucomatous" optic nerve head damages were defined subjectively, ophthalmoscopically by two of us (GT and YK), if the optic nerve head appearances met all of the following criteria: (1) focal or generalized narrowing or disappearing of neuroretinal rim with enlarged amount of cupping or pallor, or eccentric cupping or pallor (that is, cupping enlarged more either superiorly or inferiorly than horizontally within the disc), corresponding to the visual field defects; (2) pallor does not extend 1900

over the cupping border. The existence of peripapillary atrophy was not taken into account on defining the glaucomatous optic nerve head damage. The judgments by two observers were performed independently. If either observer described a patient's optic disc as "non-glaucomatous," the patient was excluded. If any of 105 patients with NTG did not meet the following further inclusion criteria, they were excluded: (1) a refractive error in spherical equivalent less than -6.0 diopters (D); (2) a corrected visual acuity greater than 20/30; (3) a visual field test performed within 3 months of optic nerve head analysis; and (4) an optic nerve head image good enough to measure the area of peripapillary atrophy. Reasons for exclusion were (1) cataract (26 patients), with a best-corrected visual acuity of 20/30 or less; (2) high myopia (12 patients), with refractive power equal to or greater than -6.0 D; (3) poor quality of Heidelberg retina tomograph (HRT; version 1.11, Heidelberg Engineering, Heidelberg, Germany) images (10 patients); (4) macular lesion (5 patients); and (5) visual field not checked within 3 months of HRT (1 patient). A total of 51 patients with NTG were enrolled in this study.

Patients' Characteristics Of the 51 patients with NTG included in the study, 14 were men and 37 were women. The mean age of these patients was 57.8 ± 13.4 years (range, 20-77 years). Only one eye of each person was selected randomly for interindividual analysis. Visual field, refraction, and lOP characteristics of the randomly selected 51 eyes were as follows: the mean deviation (MD) and corrected pattern standard deviation (CPSD) averaged -6.00 ± 5.80 dB (range, -18.60 to 1.92 dB) and 7.59 ± 4.92 dB (range, 0.43 to 16.06 dB), respectively. Refraction (mean ± spherical equivalent) was -0.30 ± 2.26 D (range, -5.50 to 4.50 D). The mean, peak, and trough lOP of a 24-hour evaluation at 2-hour intervals was 14.3 ± 1.9 mmHg (range, 9.8-17.8 mmHg) , 17.1 ± 2.2 mmHg (range, 13.0-21.0 mmHg), and 11.6 ± 2.1 mmHg (range, 7.016.0 mmHg), respectively.

Measurements of Peripapillary Atrophy Peripapillary atrophy was divided into a central zone (zone Beta) characterized by chorioretinal atrophy with visible large choroidal vessels and sclera, and a peripheral zone (zone Alpha) with irregular hyper- and hypopigmentation. 4 - 6 The peripapillary scleral ring, a thin white band of tissue surrounding the optic nerve head, was not considered to be peripapillary atrophy. The morphometric analysis of peripapillary atrophy was done by a confocal scanning laser tomograph, the HRT. The mean topographic image of the optic nerve head based on three 10° X 10° measurements was used for the analysis. The standard reference plane was used for all measurements. The principle and methods of optic nerve head analysis using this instrument have been described in detail elsewhere. 15 •16 To evaluate the peripapillary atrophic area, a new con-

Park et al . Peripapillary Atrophy in Normal-tension Glaucoma tent, and radial extent of zone Beta were 5.8%, 3.6%, and 6.4%, respectively. The coefficient of variation for the area of zone Alpha was 14.9%, representing relatively lower reproducibility. The coefficients of variation for the area, angular extent, and radial extent of the total zone were 9.6%, 5.3%, and 4.6%, respectively. The mean standard deviation for the area, angular extent, and radial extent of zone Beta were 0.030 mm2 , 4.8°, and 19 j.lm, respectively. The mean standard deviation for the area of zone Alpha was 0.146 mm 2 . The mean standard deviations for the area, angular extent, and radial extent of the total zone were 0.154 mm2 , 13.7°, and 23 j.lm, respectively. Measurements of Intrapapillary and Visual Field Parameters

Figure 1. Measurement of p eripapillary atrophy (zone Beta) by drawing a new contour line (arrows) around the zone Beta using an image produced by a confocal scanning laser tomograph. The original contour line of the optic disc is indicated by arrowheads.

The HRT provides abundant information regarding a variety of parameters of the optic nerve head we refer to as "intrapapillary parameters." 14- 18 The intrapapiUary parameters analyzed in this study were rim area, rim volume, mean cup depth, mean retinal nerve fiber layer thickness at the disc margin, and cup shape 18 (i.e., third moment inside the disc, slope of the cup) . The visual field indices used to analyze correlation with peri- and intrapapillary parameters were mean deviation and corrected pattern standard deviation. The sum of central four values within 5° of fixation in the total devia-

tour line was drawn to delineate zone Beta (Fig 1), zone Alpha, and total zone (combined Beta and Alpha), respectively, on the image of optic nerve head taken by the HRT. Each area then was calculated by the HRT. Also measured was the angular extent around the optic disc and the largest radial extent of each zone (Fig 2). The former was measured in degrees from the angle made at the center of the disc. The location of the largest radial extent was recorded in terms of the clock hour around the disc, using a unit of 30 minutes. The location was defined as "inferior side of the disc" when it was located within the 3:00- to 8:30-clock positions around the disc in the right eye. It was defined as "superior side of the disc" when it was located within the 9:00- to 2:30-clock positions around the disc in the right eye. The mirror image was applied to the left eye. The measurements of peripapillary atrophy were performed by one of us (KHP) in a masked fashion without knowledge of the visual field data. However, for technical reasons, covering of the intrapapillary region was not available. Reproducibility of the Peripapillary Parameter Measurements For at least I-day intervals, three measurements were taken by the same examiner, who redrew the contour line around the peripapillary atrophy at each measurement in ten randomly selected eyes. Reproducibility was expressed as a coefficient of variation and mean standard deviation. The coefficients of variation for the area, angular ex-

Figure 2. The parame ters used for detailed measurements of peripapillary atrophy were area (A), angular extents (B) , and largest radial extent (C).

1901

Ophthalmology

Volume 103, Number 11, November 1996

tion plot was defined as the central visual field defect. The values within either the superior or inferior hemifield of the total deviation plot were added together, respectively, and the total values were defined as superior or inferior hemifield defects. The worse hernifield of an individual eye was determined by comparing the superior and inferior hemifield defects. Statistical Analysis Only one randomly selected eye per person was considered in the interindividual analysis of correlation between peripapillary, intrapapillary, and visual field parameters (51 eyes of 51 patients). For the intraindividual bilateral comparison, right -left-side differences between the parameters of both eyes were used in the correlation analysis (102 eyes of 51 patients). Correlation was expressed in terms of the Pearson correlation coefficient Rand P. Chi-square tests were performed to evaluate spatial correlation between peripapillary atrophy and visual field defects. For all significance tests, P < 0.05 (two-tailed) was used. All analyses were performed using SPSSIPC+, version 4.0 (SPSS, Inc, Chicago,IL).

Results Incidence, Area, and Extent of Peripapillary Atrophy Among 51 eyes of the patients studied with NTG, zone Beta was found in 43 (84.3%), zone Alpha in 46 (90.2%), and zone Beta or zone Alpha in 50 (98.0%). The area and extent of peripapillary atrophy are shown in Table 1. The mean area of zone Beta was 0.521 ± 0.412 mm2 (range, 0-1.781 mm 2). The angular extent of zone Beta ranged from 0° to 360° around die disc (mean, 141 ° ± 89°). The widest radial extent of zone Beta ranged from 0 to 0.821 mm (mean, 0.284 ± 0.177 mm). The mean Beta/disc-area ratio was 0.22 ± 0.18 (range, 00.73). The mean Beta/disc ratio (the ratio of radial extent to mean disc diameter) was 0.17 ± 0.11 (range, 0-0.58). The mean and maximum area of zone Alpha (0.677 and 1.893 mm2 , respectively) were larger than those of zone Beta. Correlation between Peripapillary and Visual Field Parameters All the peripapillary parameters of zone Beta correlated significantly (P < 0.01) with mean deviations, corrected pattern standard deviations, central visual field defects of four points within 5° of fixation, and superior hemifield defects (except for inferior hemifield defects) (Table 2). Among these, the Beta/disc-area ratios (ratios of zone Beta to disc area) showed the highest correlations with CPSDs (r = 0.5291, P < 0.001). The angular extent of zone Beta represented localized visual field defects better (r = 0.5217 and P < 0.001 for the CPSDs) than the

1902

Table 1. Area and Extent of Peripapillary Atrophy (n = 51 Eyes) Parameter Zone /3 Area (mm2 ) Angular extent (0) Radial extent (mm) /3/disc-area ratio /3/disc ratio* Zone a Area (mm Z) Total zanet Area (mm 2 ) Angular extent (0) Radial extent (mm) PPA/disc area ratio PPA/disc ratio*

Mean:±: SD

0.521 141 0.284 0.22 0.17

Range

0.412 89 0.177 0.18 0.11

0-1.781 0-360 0-0.821 0-0.73 0-0.58

0.677 :±: 0.397

0-1.893

1.198 237 0.476 0.52 0.28

0-2.884 0-360 0-1.085 0-1.25 0-0.76

:±: :±: :±: :±: :±:

:±: :±: :±: :±: :±:

0.621 96 0.190 0.26 0.12

= standard deviation; PPA = peripapillary atrophy. * Ratio of radial extent to mean disc diameter.

SD

t Zone a plus zone (3.

diffuse visual field defects (r = -0.3770 and P < 0.01 for the MDs), whereas the radial extent showed similar correlation coefficients for both (r = 0.4694 for the CPSDs and r = -0.4766 for the MDs). The area of zone Alpha was not significantly correlated with the visual field defects. The parameters of the total zone (combined Alpha and Beta) correlated with the visual field parameters with lower correlation coefficients and less significance than those of zone Beta (Table 2). Correlation between Peripapillary and Intrapapillary Parameters Of the zone Beta parameters, "angular extent" around the disc was the one that most highly correlated with the intrapapillary parameters (Table 3). The angular extent of zone Beta increased significantly with decreasing mean retinal nerve fiber layer thickness (r = 0.4229, P = 0.002), rim volume (r = -0.3977, P = 0.004), and rim area (r = -0.3402, P = 0.015) and with increasing slope of the cup (cup shape)18 (r = 0.4267, P = 0.002). The zone Alpha and most of the total zone parameters were not significantly correlated with the intrapapillary parameters (P > 0.05), with the exception of the angular extent of the total zone, which increased significantly with decreasing mean retinal nerve fiber layer thickness (P < 0.05). Correlation between Intrapapillary and Visual Field Parameters Of the intrapapillary parameters, only "rim area" was significantly correlated with the visual field parameters, including CPSD and superior hemifield defects (Table 4). The correlation coefficients and significance levels

Park et al . Peripapillary Atrophy in Normal-tension Glaucoma Table 2. Coefficients of Correlation between Peripapillary Parameters and Visual Field Parameters (n = 51 eyes) Coefficients of Correlation for Visual Field Parameters Peripapillary Parameter Zone f3 Area Angular extent Radial extent Beta/disc-area ratio Beta/disc ratio* Zone a Area Total zone Area Angular extent Radial extent

MD

CPSD

C4D

SHD

IHD

-0.3937t -0.3770t -004766* -Oo4373t -Oo4687t

0.5031* 0.5217+ Oo4694t 0.5291* Oo4618t

-0.3927t -0.3842t -Oo4037t -Oo4590t -Oo4128t

-0.3949t -Oo4123t -Oo4314t -004421 t -Oo4252t

-0.1234 0.0354 -0.2231 -0.0593 -0.1809

0.0229

0.0894

0.0366

0.0563

-0.0846

-0.2463 -0.0474 -0.3613t

0.3906t 0.2450 0.3619t

-0.2369 -0.0136 -0.3630t

-0.2258 -0.1082 -0.3071§

-0.1358 0.0947 -0.2362

MD = mean deviation; CPSD = corrected pattern standard deviation; C4D = sum of central four values in total deviation plot; SHD = sum of superior hemifield values in total deviation plot; IHD = sum of inferior hemifield values in total deviation plot.

* Ratio of radial extent to mean disc diameter. t P < 0.01. P < 0.001. § P < 0.05.

*

involved were lower than those for the peripapillary parameters and visual field parameters (Table 2). Correlation between Peripapillary Atrophy and Intraocular Pressure The peripapillary atrophy was not significantly correlated with the mean, peak, or trough lOPs measured at 24 hours (P> 0.05).

Intraindividual Correlations using Right -left-side Differences of Parameters Using both eyes (102 eyes) of 51 patients, we calculated the right-left-side difference for each parameter

and analyzed the correlations among the side differences of these parameters. The side differences of the zone Beta parameters correlated significantly with the side differences of the visual field and intrapapillary parameters (Table 5). The side difference of CPSD correlated with the side difference of zone Beta area with the highest correlation coefficient and significance (r = 0.6305, P < 0.001), and correlated significantly with the side difference of angular extent (r = 0.5715, P < 0.001), radial extent (r = 0.5388, P < 0.001), and Beta/disc-area ratio (r = 0.5049, P < 0.001). The side difference of the total zone area also correlated significantly with the side difference of CPSD (r = 0.5598, P < 0.001).

Table 3. Coefficients of Correlation between Peripapillary and Intrapapillary Parameters (n = 51 eyes) Coefficients of Correlation for Intrapapillary Parameters Peripapillary Parameter

Rim Area

Rim Volume

MCD

MRNFLT

Cup Shape

Zone f3 Area Angular extent Radial extent f3/disc-area ratio f3/disc ratio*

-0.2544 -0.3402t -0.2747 -0.3240t -0.2986t

-0.2456 -0.3977+ -0.1495 -0.2028 -0.1138

-0.0234 -0.0688 -0.0517 -0.0363 -0.0623

-0.2557

0.3512t 004267+ 0.3011 0.2954t 0.2424

= mean cup depth; MRNFLT = mean retinal nerve fiber * Ratio of radial extent to mean disc diameter.

MCD

-Oo4229t

-0.1187 -0.1446 -0.0442

layer thickness;

t P < 0.05.

*P < 0.01. 1903

Ophthalmology

Volume 103, Number 11, November 1996

Table 4. Coefficients of Correlation Between Intrapapillary Parameters and Visual Field Parameters (n = 51 eyes) Coefficients of Correlation for Visual Field Paramaters lntrapapillary Parameter

MD

CPSD

C4D

SHD

IHD

Rim area Rim volume Mean cup depth MRNFLT Cup shape

0.2370 0.1186 -0.0583 0.1123 -0.2381

-0.3340* -0.2219 0.0084 -0.2422 0.2085

0.2390 0.1103 -0.0912 0.1025 -0.2275

0.3160* 0.2472 -0.0845 0.2117 -0.2717

-0.2197 -0.2393 0.1681 -0.1381 0.1311

MO = mean deviation; CPSO = corrected pattern standard deviation; C40 = sum of central four values in total deviation plot; SHO = sum of superior hemifield values in total deviation plot; rHO = sum of inferior hemifield values in total deviation plot; MRNFLT = mean retinal nerve fiber layer thickness.

* p < 0.05. Spatial Correlation between Visual Field Defect and Peripapillary Atrophy Thirty-one eyes were worse in the superior hemifield, and 20 were worse in the inferior hemifield. Of the 31 eyes worse in the superior hemifield, 24 had their widest total zone at the corresponding inferior side of the disc, 6 at the superior side, and 1 had no area of peripapillary atrophy. Of the 20 eyes that were worse in the inferior hemifield, 12 had their widest total zone at the corresponding superior side of the disc, and eight at the inferior side. This spatial correlation was significant by the chi-square test (chi-square = 9.0484, P = 0.011) (Table 6). For zone Beta, this correlation was also significant (chi-square = 8.6106, P < 0.014).

Discussion While many previous studies have suggested that in eyes with high-tension glaucoma, peripapillary atrophy is more

frequent and extensive than in healthy eyes, and that the extent or the location of peripapillary atrophy is significantl y correlated with visual field defects,I-8 no detailed cross-sectional NTG study on the correlation between the peripapillary atrophy and the visual field as well as the optic nerve head configuration has been reported. Our study is the first in which peripapillary atrophy was measured, using confocal scanning laser tomography in patients with NTG, and their correlations with visual field defects and optic nerve head configurations analyzed. Our study demonstrates that the confocal scanning laser tomograph can detect the area and extent of zone Beta very clearly (Fig 1) and highly reproducibly, with no need for pupil dilation. 17 As such, the instrument proved useful in providing quick and detailed evaluations not only the optic nerve head but also of peripapillary atrophy. We found significant correlations between zone Beta peripapillary atrophy in NTG and visual field parameters,

Table 5. Coefficients of Correlation between Right-left-side Differences in Peripapillary Parameters and Side Differences in Other Parameters (n = 102 eyes) Side Differences in Peripapillary Parameters Zone (3 Area Angular extent Radial extent (3/disc-area ratio Zone (l' Area Total zone Area

Coefficients of Correlation for Side Differences in Other Parameters MD

CPSD

C4D

Rim Area

MRNFLT

-0.4648* -0.3648* -0.4598* -0.4361 *

0.6305t 0.5715t 0.5388t 0.S863t

-0.4364* -0.3881 * -0.3550+ -0.3947*

-0.4175* -0.4678* -0.3202+ -0.3707*

-0.4790* -0.4682* -0.3873* -0.4037*

-0.0924

0.1352

-0.2262

-0.1043

-0.1362

-0.4074*

0.5598t

-0.4690*

-0.3799*

-0.2729

MO = mean d eviation; CPSO = corrected pattern standard deviation; C40 deviation plot; MRNFLT = mean retinal nerve fiber layer thickness.

* p < 0.Ql. t p < 0.001. :j: P < 0.05.

1904

= sum of central

four values in total

Park et al . Peripapillary Atrophy in Normal-tension Glaucoma Table 6. Spatial Correlation between Visual Field Defects and Peripapillary Atrophy* Worse Hemifield in Total Deviation Location of Widest Total Zone

Superior ( eyes)

Inferior of disct Superior of disc:j: No total zone Total

24

* Chi-square = 9.0484;

6

1 31 P

Inferior ( eyes)

8

12 0 20

Total (eyes)

32 18 1 51

= 0.011.

t From the 3:00 to 8:30 o'clock position around the disc in the right eye when the location was expressed in 30-minute intervals. The mirror image was used for the left eye.

*

From the 9:00 to 2:30 o'clock position around the disc in the right eye. The mirror image was used for the left eye.

with higher coefficients and levels of significance, than for the intrapapillary parameters. Thus, zone Beta evaluation may provide more useful information about functional damage in NTG than intrapapillary parameters. The area and angular extent of zone Beta proved to be a better indicator of localized field defects (CPSD) than diffuse field defects (MD) (Tables 2 and 5). Peripapillary atrophy, especially in zone Beta, correlated significantly with central visual field defects. This may be a reflection of the pattern of visual field defects in NTG, which tend to be located close to the fixation point. 20.21 Thus, it suggested that peripapillary atrophy should be examined more carefully in patients in whom a visual field defect encroaches upon the fixation point. The significant correlation we found between peripapillary atrophy and superior hemifield defects, but not with inferior hemifield defects, could be related to our finding that the frequently affected portion in NTG was the superior hemifield,21 close to the fixation point, and the corresponding inferotemporal portion of the optic nerve head. 22 The reproducibility of zone Alpha measurements was not satisfactory because in some cases the contour of zone Alpha was not clear, even though the optic nerve head image was. Further, unlike that of zone Beta, the area of zone Alpha was not significantly correlated with either visual field defects or intrapapillary parameters. Such differences between zone Beta and Alpha also have been found in primary open-angle glaucoma. 5 In normal eyes, zone Alpha reportedly occurs in almost every eye, whereas zone Beta is present in only 10% to 15%.4,6 Thus, at least in NTG, zone Alpha measurements seem to be less meaningful than zone Beta measurements. As for the etiologic mechanism of peripapillary atrophy, Fantes and Anderson's22 histologic study suggested both congenital misalignment of tissues and acquired atrophy. Congenital peripapillary atrophy is found mainly in oblique implantation of the optic disc 23 or in myopia. We excluded patients with peripapillary atrophy due to high myopia (i.e., in whom the refractive error in spherical equivalent was less

than -6.0 D). However, clinically, there is no way to clearly differentiate congenital misalignment from acquired atrophy?2 Therefore, both types of peripapillary atrophy may have been included in our study. The result of our correlation analysis using intraindividual right -left-side differences suggested that the eye with more advanced visual field defects had larger peripapillary atrophy. Peripapillary atrophy might be a result of glaucomatous optic nerve damage. However, this would not explain the presence of peripapillary atrophy in normal eyes,4-6 or in patients with ocular hypertension 9 without glaucomatous cupping. In a retrospective review of 127 patients with glaucoma and 49 control subjects, Rockwood and Anderson 24 observed peripapillary abnormality in the form of retinal pigment epithelium changes in 21 % of the patients with glaucoma, who had progressive glaucomatous cupping, and in 4% of control subjects. Approximately four fifths of patients with glaucoma showed no evidence of significant peripapillary changes as glaucomatous damage progressed. Therefore, peripapillary atrophy is not the result of glaucomatous damage but, rather, is one of the risk factors for its development in the first place. There is considerable evidence that the peripapillary atrophy is a risk factor of glaucoma relatively less affected by the high lOP. Recent studies using regression analysis of survival data based on the Cox proportional hazard model found that peripapillary atrophy was one of the significant risk factors influencing progression of visual field damage in NTG. I1 ,12 Buus and Anderson9 reported a higher frequency (64%) and greater extent of peripapillary atrophy in NTG than in ocular hypertension (34%). The zone Beta in high-tension glaucoma is larger in eyes with relatively lower lOP than in those with higher IOP. 6 Using a high-tension glaucoma model in monkeys, Derick et al 25 reported that the presence of, or change in, peripapillary atrophy was not significantly associated with the IOPinduced optic nerve damage. In our current study ofNTG, we found no significant correlation between the parameters of peripapillary atrophy and lOPs (based on the mean, peak, and trough values in a 24-hour period). Thus, peripapillary atrophy may be a risk factor less affected by elevated lOP. In fluorescein angiography, zone Beta of peripapillary atrophy does not fluoresce in the choroidal filling phase. 26 Histologically, this area corresponded to areas of thinning or no choroid 22 and those without retinal pigment epithelium next to the disc. 22 ,27,28 Because the prelaminar portion of the optic nerve head receives its main blood supply from the peripapillary choroid via branches of short posterior ciliary arteries with characteristic sectoral distribution,29-33 the absence or dysfunction of that centripetal branch in the sector of peripapillary atrophy will cause ischemic optic nerve head damage in that segment. 3l ,33 Our study and previous reports have shown that visual field defects and optic nerve head damage associate with the location of peripapillary atrophyY Thus, peripapillary atrophy may represent a local vascular insufficiency in that area and in a corresponding segment of the optic

1905

Ophthalmology

Volume 103, Number 11, November 1996

nerve head,l,ll thereby acting as a local vascular risk factor for the development of optic nerve damage in NTG, Because we found that in NTG the peripapillary parameters were more closely related to visual field defects than the intrapapillary parameters, we conclude that in attempting to detect and follow glaucomatous nerve damage it is important to carefully examine peripapillary changes. Further study of peripapillary atrophy with prospective follow-up also is recommended. Owing to its high reproducibility and accuracy, we believe that confocal scanning laser tomography will be useful in this regard.

References 1. Primrose J, Early signs of the glaucomatous disc, Br J OphthalmoI1971;55:820-5, 2, Wilensky JT, Kolker AE. Peripapillary changes in glaucoma. Am J OphthalmoI1976;81:341-5, 3, Geijssen HC, Greve EL The spectrum of primary open angle glaucoma. I: Senile sclerotic glaucoma versus high tension glaucoma. Ophthalmic Surg 1987; 18:207-13. 4. Jonas JB, Nguyen XN, Gusek GC, Naumann GO. Parapapillary chorioretinal atrophy in normal and glaucoma eyes. I. Morphometric data, Invest Ophthalmol Vis Sci 1989; 30:908-18. 5, Jonas JB, Naumann GO. Parapapillary chorioretinal atrophy in normal and glaucoma eyes. II. Correlations, Invest Ophthalmol Vis Sci 1989;30:919-26. 6. Jonas JB, Fernandez MC, Naumann GO. Glaucomatous parapapillary atrophy, Occurrence and correlations. Arch Ophthalmol 1992; 110:214-22. 7. Anderson DR. Correlation of the peripapillary anatomy with the disc damage and field abnormalities in glaucoma. In: Greve EL, Heijl A, eds. Fifth International Visual Field Symposium, 1982. The Hague: Dr W. Junk, 1983;1-10 (Doc Ophthalmol Proc Ser; 35). 8. Heijl A, Samander C. Peripapillary atrophy and glaucomatous visual field defects. In: Heijl A, Greve EL, eds. Sixth International Visual Field Symposium, 1984. Dordrecht: Dr W. Junk, 1985;403-7 (Doc Ophthalmol Proc Ser; 42), 9. Buus DR, Anderson DR. Peripapillary crescents and halos in normal-tension glaucoma and ocular hypertension. Ophthalmology 1989;96:16-9. 10. Jonas JB, Xu L Parapapillary chorioretinal atrophy in normalpressure glaucoma, Am J Ophthalmol 1993; 115:501-5. 11. Geijssen He. Studies on Normal Pressure Glaucoma. New York: Kugler Publications, 1991. 12,Araie M, Sekine M, Suzuki Y, Koseki N. Factors contributing to the progression of visual field damage in eyes with normaltension glaucoma. Ophthalmology 1994; 101: 1440-4. 13. Dreher AW, Tso PC, Weinreb RN. Reproducibility of topographic measurements of the normal and glaucomatous optic nerve head with the laser tomographic scanner. Am J Ophthalmol1991; 111:221-9. 14. Mikelberg FS, Wijsman K, Schulzer M. Reproducibility of topographic parameters obtained with the Heidelberg Retina Tomograph. J Glaucoma 1993;2:101-3,

1906

15. Rohrschneider K, Burk RO, Volcker HE. Reproducibility of topometric data acquisition in normal and glaucomatous optic nerve heads with the laser tomographic scanner. Graefes Arch Clin Exp Ophthalmol 1993;231:457-64, 16. Rohrschneider K, Burk RO, Kruse FE, Volcker HE. Reproducibility of the optic nerve head topography with a new laser tomographic scanning device. Ophthalmology 1994; 101:1044-9, 17. Lusky M, Bosem ME, Weinreb RN. Reproducibility of optic nerve head topography measurements in eyes with undilated pupils. J Glaucoma 1993;2:104-9. 18. Brigatti L, Caprioli J. Correlation of visual field with scanning confocal laser optic disc measurements in glaucoma. Arch Ophthalmol 1995; 113: 1191-4, 19. Levene RZ. Low tension glaucoma: critical review and new material. Surv Ophthalmol 1980;24:621-64. 20. Phelps CD, Hayreh SS, Montague PR. Visual fields in lowtension glaucoma, primary open-angle glaucoma, and anterior ischemic optic neuropathy. In: Greve EL, Heijl A, eds. Fifth International Visual Field Symposium, 1982. The Hague: Dr W. Junk; 1983; 113-24 (Doc Ophthalmol Proc Ser; 35), 21. Caprioli J, Spaeth GL Comparison of the optic nerve head in high- and low-tension glaucoma, Arch Ophthalmol 1985; 103:1145-9, 22. Fantes FE, Anderson DR. Clinical histologic correlation of human peripapillary anatomy. Ophthalmology 1989;96:20-5. 23, Nevarez J, Rockwood EJ, Anderson DR. The configuration of peripapillary tissue in unilateral glaucoma. Arch Ophthalmol 1988; 106:901-3. 24. Rockwood EJ, Anderson DR. Acquired peripapillary changes and progression in glaucoma. Graefes Arch Clin Exp Ophthalmol 1988;226:510-5. 25. Derick RJ, Pasquale LR, Pease ME, Quigley HA. A clinical study of peripapillary crescents of the optic disc in chronic experimental glaucoma in monkey eyes, Arch Ophthalmol 1994; 112:846-50, 26. Hayreh SS, Walker WM, Fluorescent fundus photography in glaucoma. Am J Ophthalmol 1967;63:982-9. 27. Kubota T, Jonas JB, Naumann GO. Direct clinico-histological correlation of parapapillary chorioretinal atrophy, Br J Ophthalmol 1993;77:103-6. 28. Jonas JB, Konigsreuther KA, Naumann GO, Optic disc histomorphometry in normal eyes and eyes with secondary angle-closure glaucoma, II. Parapapillary region. Graefes Arch Clin Exp Ophthalmol 1992;230:134-9. 29. Lieberman MF, Maumenee AE, Green WR. Histologic studies of the vasculature of the anterior optic nerve. Am J Ophthalmol 1976;82:405-23. 30. Anderson DR, Braverman S. Reevaluation of optic disk vasculature. Am J OphthalmolI976;82:165-74. 31. Hayreh SS. Blood supply of the optic nerve head and its role in optic atrophy, glaucoma, and oedema of the optic disc. Br J Ophthalmol 1969;53:721-48. 32. Hayreh SS, Blood supply of the anterior optic nerve. In: Ritch R, Shields MB, Krupin T, eds. The Glaucomas. St. Louis: CV Mosby, 1989; vol 1, chap 5. 33. Hayreh SS. The 1994 Von Sallman Lecture, The optic nerve head circulation in health and disease. Exp Eye Res 1995;61:259-72.