Peripapillary Atrophy in Primary

Angle-closure Glaucoma: A Comparative

Study With Primary Open-angle Glaucoma

HIDEYA UCHIDA, MD, TETSUYA YAMAMOTO, MD, GOJI TOMITA, MD,

AND YOSHIAKI KITAZAWA, MD

c PURPOSE: To investigate the characteristics of peripap- mean deviation (P 5 .0001) in primary open-angle
illary atrophy in primary angle-closure glaucoma and to glaucoma.
compare peripapillary atrophy in primary angle-closure c CONCLUSIONS: The peripapillary atrophy in primary
glaucoma with that in primary open-angle glaucoma. angle-closure glaucoma has a different relationship to
c METHODS: Twenty-one eyes (of 21 patients) with the structural and functional optic disk changes than
chronic primary angle-closure glaucoma that had no that in primary open-angle glaucoma. Different
evidence of an acute attack or combined-mechanism mechanisms seem to be involved in the development of
glaucoma and 31 eyes (31 patients) with primary open- the optic disk damage in the two types of glaucoma.
angle glaucoma were enrolled in this study. The cup-to- (Am J Ophthal-mol 1999;127:121±128. © 1999 by
disk area ratio and the visual ®eld mean deviation Elsevier Science Inc. All rights reserved.)
were matched in both groups. The prevalence of
peripapillary atrophy and peripapillary atrophy-to-
disk area ratio of eyes with primary angle-closure
glaucoma and eyes with primary open-angle glaucoma
P ERIPAPILLARY ATROPHY IS MORE FREQUENTLY OB-served
and is greater in patients with primary open-angle glaucoma and
normal-tension glaucoma than in normal individuals.1±7 Several
were compared. The cor-relation between the investigators have dem-
peripapillary atrophy area and the cup-to-disk area onstrated a signi®cant correlation between the degree of
ratio or the mean deviation was also evaluated. peripapillary atrophy and optic disk damage and between
c RESULTS: Eight (38%) of 21 eyes with primary angle- the location of peripapillary atrophy and the location of
closure glaucoma and 21 (68%) of 31 eyes with primary optic disk damage and visual ®eld defects.5±7 The area of
open-angle glaucoma had peripapillary atrophy. The peri- peripapillary atrophy, particularly that of the central zone
papillary atrophy-to-disk area ratio averaged 0.16 in betaÐ characterized by marked chorioretinal atrophy with
primary angle-closure glaucoma and 0.41 in primary visible large choroidal vessels and scleraÐmight correlate
open-angle glaucoma. There were statistically signi®cant with the neuroretinal rim area, cup-to-disk ratio, nerve
differences regarding the prevalence of peripapillary at- ®ber layer score, and visual ®eld loss in glaucoma.5
rophy (P 5 .048) and peripapillary atrophy-to-disk area Corre-lations between peripapillary atrophy and functional
ratio (P 5 .005) between the two groups. There was no and structural optic disk damage are also found in eyes
signi®cant correlation between the peripapillary atrophy with normal-tension glaucoma by means of confocal
area and the cup-to-disk area ratio or the mean deviation scanning laser ophthalmoscopy.8
in primary angle-closure glaucoma. In contrast, the peri- Although the increased intraocular pressure plays an
papillary atrophy-to-disk area ratio signi®cantly corre- important role in the development of optic disk damage in
lated with the cup-to-disk area ratio (P 5 .003) and the eyes with primary open-angle glaucoma and primary
angle-closure glaucoma, the pathophysiologic process of
the optic nerve damage may not be identical in the two
Accepted for publication Aug 28, 1998. clinical entities. In a substantial proportion of eyes with
From the Department of Ophthalmology, Gifu University School of primary open-angle glaucoma, vascular factors are alleg-
Medicine, Gifu, Japan.
This study was supported by Grant-in-Aid for Scienti®c Research (B) edly responsible for the optic disk damage.9 In contrast,
08457461 from the Ministry of Education, Science, and Culture of the optic disk damage in eyes with chronic primary angle-
Japanese Government. closure glaucoma that had no evidence of an acute attack
Reprint requests to Hideya Uchida, MD, Gifu University School of
Medicine, 40 Tsukasa-machi Gifu, Gifu 500-8705, Japan; fax: 81 58 265 or combined mechanism is considered more likely to be
9012. caused primarily by the increased intraocular pressure.

0002-9394/99/$20.00 © 1999 BY ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED. 121

PII S0002-9394(98)00318-3
 The cause of peripapillary atrophy is not clear. Vascular California), no history of other ocular or intracranial
de®ciency or ischemia of peripapillary choroidal circula- diseases affecting the optic disk or the visual ®eld, and
tion is, however, regarded as one of the most important availability of a clear image by laser scanning tomography.
risk factors for peripapillary atrophy from previous When both eyes met the criteria, one eye from each patient
¯uoresce-in10,11 or indocyanine green12 angiographic and was randomly selected for the statistical analysis. Twenty-
histolog-ic13,14 studies. Therefore, we designed this study one patients with primary angle-closure glaucoma were
to investigate whether the characteristics of peripapillary selected from our patient database.
atrophy in primary angle-closure glaucoma differ from Thirty-one primary open-angle glaucoma cases, which
those in primary open-angle glaucoma. Because the clini- were matched to the primary angle-closure glaucoma cases
cal ®ndings of optic disk damage and visual ®eld change by visual ®eld mean deviation of the visual ®eld indices,
in chronic primary angle-closure glaucoma are identical to cup-to-disk area ratio, and pretreatment intraocular pres-
those in primary open-angle glaucoma and normal-tension sure, were selected by means of the following criteria: a
glaucoma, a study of peripapillary atrophy in primary history of increased intraocular pressure exceeding 21 mm
angle-closure glaucoma could help not only to clarify the Hg on two or more occasions with or without treatment;
role peripapillary atrophy plays in the development of normal open angle, glaucomatous optic nerve head change
optic disk damage, but also to shed light on the pathogen- determined by direct ophthalmoscopy; reliable glaucoma-
esis of peripapillary atrophy. tous visual ®eld defects with the full threshold Central 30-
Although the characteristics of peripapillary atrophy and 2 program of the Humphrey Field Analyzer 630; no
the differences in its prevalence have been studied history of other ocular and intracranial diseases affecting
extensively in normal-tension glaucoma and high-tension the optic disk or the visual ®eld; and availability of a clear
glaucoma,15±17 to the best of our knowledge, no studies image by laser scanning tomography.
have been conducted in primary angle-closure glaucoma. Glaucomatous visual ®eld defects in the primary angle-
This is undoubtedly because it is dif®cult to obtain a high- closure glaucoma and primary open-angle glaucoma pa-
quality optic disk image in the presence of a miotic pupil, tients were de®ned as follows: reliable ®xation (,20%
which often cannot be dilated because of posterior
®xation loss) and visual ®eld results (,15% false-positive
synechiae.
and false-negative responses), typical focal paracentral
The recent development of confocal scanning laser tomog-
scotomas, nasal steps, arcuate scotomas, or combinations
raphy enables us to obtain an adequate image of the optic
of these corresponding to optic disk change, two or more
nerve head and the peripapillary area through an undilated
contiguous points with a signi®cant loss in the superior or
pupil in most eyes. The device has several advantages, such
as the short time to obtain an image, high level of inferior arcuate areas (P # .01) compared with perimeter-
reproducibil-ity, and ability to evaluate many variables of the stored age-matched control subjects, three or more contig-
optic disk and the retinal nerve ®ber layer. Furthermore, uous points with a signi®cant loss in the superior or
recently developed software enables us to measure the area of inferior arcuate areas (P # .05) or a 10-dB difference
across the nasal horizontal midline in two or more adjacent
peripap-illary atrophy.8,18 In the current study, we
locations, and results outside normal limits on the
investigated the characteristics of peripapillary atrophy in
Glaucoma Hemi-®eld Test (STATPAC 2 software), which
primary angle-closure glaucoma and in primary open-angle
glaucoma by means of laser scanning tomography. divides both the upper and lower hemi®elds into ®ve
zones containing clusters of points that correspond to
nerve ®ber layer distribution.
All patients were diagnosed and followed up at the
METHODS Glaucoma Clinic of Gifu University Hospital, Gifu, Japan.
INCLUDED IN THE STUDY WERE PATIENTS WITH PRIMARY angle-
The study protocol was approved by the Ethical Review
closure glaucoma who met the following criteria: a history of Committee of Gifu University, and tenets of the Declara-
increased intraocular pressure (.21 mm Hg) on at least two tion of Helsinki were followed. Verbal informed consent
occasions with or without glaucoma treatment, existence of was obtained from all patients after the study was thor-
peripheral anterior synechiae on gonioscopy, no history of an oughly explained to them.
acute attack of primary angle-closure glaucoma, no ocular The backgrounds of the study population are shown in
signs of an acute attack of primary angle-closure glaucoma Table 1. The average and SD of the age and refractive
including the appearance of the iris and lens, no evidence of error were 69 6 10.6 years (range, 41 to 88 years) and 11.2
combined-mechanism glau-coma, glaucomatous optic nerve 6 1.3 diopters (range, 21.0 to 13.0 diopters), respectively,
head changes of various degrees by direct ophthalmoscopy, for the patients with primary angle-closure glaucoma, and
presence of reproduc-ible, reliable glaucomatous visual ®eld 59 6 10.7 years (range, 42 to 83 years) and 20.8 6 2.5
defects with the full threshold, Central 30-2 program of the diopters (range, 28.5 to 3.0 diopters), respec-tively, for the
Humphrey Field Analyzer 630 (Carl Zeiss-Humphrey patients with primary open-angle glaucoma.
Systems, Dublin, The optic nerve head and the peripapillary region

122 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 1999

 TABLE 1. Characteristics of the Study Population

Average 6 SD

PACG (21 Eyes) POAG (31 Eyes) Matched POAG* (23 Eyes) P Value

Age (yrs) 69 6 10.6 59 6 10.7 61 6 10.3 .003 [.02]²

Refractive error (D) 1.2 6 1.3 20.8 6 2.5 0.3 6 1.4 .003 [NS]
Mean deviation (dB) 212.19 6 9.24 210.20 6 8.89 29.6 6 9.1 NS [NS]
Cup-to-disk area ratio 0.56 6 0.23 0.51 6 0.14 0.53 6 0.13 NS [NS]
Pretreatment IOP (mm Hg) 25.9 6 7.4 24.9 6 7.2 23.4 6 2.8 NS [NS]

IOP 5 intraocular pressure; NS 5 not signi®cant; PACG 5 primary angle-closure glaucoma; POAG
5 primary open-angle glaucoma.
*Matched with eyes with PACG for refractive error.
²Values in brackets are comparisons between the PACG and matched POAG groups. Mann-

Whitney U test.

FIGURE 1. Measurement of peripapillary atrophy with the Heidelberg Retina Tomograph. After a new contour line (arrows) is
drawn to encircle the peripapillary atrophy, the Heidelberg Retina Tomograph calculates the peripapillary area.

images were scanned three times with a Heidelberg Retina line was drawn to encircle the peripapillary atrophy (Figure
Tomograph (Heidelberg Engineering GmbH, Heidelberg, 1),8,18 the peripapillary atrophy area was automatically
Germany). Images were obtained mainly with a 10 3 10 calculated with the program Atrophy Zone Analysis of the
degree retinal ®eld, and a 15 3 15 degree retinal ®eld was Heidelberg Retina Tomograph. The peripapillary atrophy-to-
used when the peripapillary atrophy area protruded from disk area ratio (normalized peripapillary atrophy area) was
the 10 3 10 degree retinal ®eld in the Heidelberg Retina used for statistical analysis. The intraobserver and
Tomograph. The mean of the three topographic images interobserver (H.U. and G.T.) coef®cients of variation of
was used for further analyses. measurements for peripapillary atrophy-to-disk area ratio
We de®ned peripapillary atrophy as a peripapillary area were calculated on the basis of randomly chosen sets of 10
that consisted of a zone with chorioretinal atrophy and eyes measured twice on separate days. The average and SD of
visible large choroidal vessels and an outlining zone with the coef®cients of variation of intraobserver measure-ments
irregular retinal pigment epithelium. After a new contour of the two observers were 7.7% 6 4.5% (H.U.) and

VOL. 127, NO. 2 PERIPAPILLARY ATROPHY IN PACG 123

 TABLE 2. Comparison of Intrapapillary and Peripapillary Regions Between PACG and POAG

Matched POAG*
PACG (21 Eyes) POAG (31 Eyes) (23 Eyes) P Value

Disk area (mm2) (average 6 SD) 2.51 6 0.38 2.44 6 0.41 2.65 6 0.35 NS [NS]²
Cup-to-disk area ratio 0.56 6 0.20 0.51 6 0.14 0.53 6 0.13 NS [NS]
(average 6 SD)
Peripapillary atrophy-to-disk area 0.16 6 0.24 0.41 6 0.32 0.43 6 0.37 .005 [.006]
ratio (average 6 SD)
Prevalence of peripapillary atrophy 8 (38) 21 (68) 16 (70) .048 [.068]
(no. [%])

NS 5 not signi®cant; PACG 5 primary angle-closure glaucoma; POAG 5 primary open-angle

glaucoma.
*Matched with eyes with PACG for refractive error.
²Values in brackets are comparisons between the PACG and matched POAG groups. All values

are by Mann-Whitney U test except for prevalence of peripapillary atrophy (Fisher exact test).

8.2% 6 7.4% (G.T.). The average and SD of the coef®- RESULTS

cients of variation of interobserver measurement was 9.0%
6 11.5%. THERE WAS A STATISTICALLY SIGNIFICANT DIFFERENCE IN
Because there is a reported association between refrac- age and refractive error between eyes with primary angle-
tive error and peripapillary atrophy,19 from the original 31 closure glaucoma and those with primary open-angle
eyes with primary open-angle glaucoma, we selected a glaucoma (P 5 .003 each). The average and SD of the
subgroup of 23 eyes matched with 21 eyes with primary visual ®eld mean deviation was 212.19 6 9.24 dB for eyes
angle-closure glaucoma for refractive error as well as with primary angle-closure glaucoma and 210.20 6 8.89
visual ®eld mean deviation, cup-to-disk area ratio, and dB for eyes with primary open-angle glaucoma. The aver-
pretreat-ment intraocular pressure, and compared the two age and SD of cup-to-disk area ratio was 0.56 6 0.23 for
groups as well. The average and SD of the age and eyes with primary angle-closure glaucoma and 0.51 6 0.14
refractive error of the refractive error±matched patients for eyes with primary open-angle glaucoma. The average
with primary open-angle glaucoma were 61 6 10.3 years and SD of pretreatment intraocular pressure was 25.9 6 7.4
(range, 42 to 83 years) and 10.3 6 1.4 diopters (range, 22.3 mm Hg for eyes with primary angle-closure glaucoma and
to 3.0 diopters), respectively (Table 1). 24.9 6 7.2 mm Hg for eyes with primary open-angle
The prevalence of peripapillary atrophy was compared glaucoma. There were no statistically signi®cant differ-
by means of the Fisher exact test on the two glaucoma ences in the visual ®eld mean deviation, optic disk area,
types. Average peripapillary atrophy-to-disk area ratio was cup-to-disk area ratio, or pretreatment intraocular pressure
also compared by the Mann-Whitney U test. The corre- between the two groups. In the refractive error±matched
lation between the peripapillary atrophy-to-disk area ratio group, there was no signi®cant difference in any factor but
and the cup-to-disk area ratio or the mean deviation of the age between eyes with primary angle-closure glaucoma
visual ®eld indices were estimated in both groups by the and those with primary open-angle glaucoma (Table 1).
Spearman correlation coef®cient. In eyes without peripap- Eight (38%) of 21 eyes with primary angle-closure
illary atrophy, the peripapillary atrophy-to-disk area ratio glaucoma and 21 (68%) of 31 eyes with primary open-angle
was rated as zero. Finally, we performed discriminant glaucoma had peripapillary atrophy. The prevalence of
analysis to investigate factors that contributed to the peripapillary atrophy was barely signi®cantly higher in
discrimination between primary angle-closure glaucoma primary open-angle glaucoma than in primary angle-closure
and primary open-angle glaucoma. Age, refractive error, glaucoma (P 5 .048, Fisher exact test). The average and SD of
mean deviation, cup-to-disk area ratio, pretreatment in- the peripapillary atrophy-to-disk area ratio was 0.16 6 0.24
traocular pressure, and peripapillary atrophy-to-disk area for eyes with primary angle-closure glaucoma and 0.41 6 0.32
were used as the explanatory variables. A software for eyes with primary open-angle glaucoma. The peripapillary
package (SPSS for Windows, version 6.0; SPSS Inc, atrophy-to-disk area ratio of eyes with primary open-angle
Chicago, Illinois) was used to perform the statistical glaucoma was signif-icantly larger than that of eyes with
analyses. A P value of less than .05 was considered a primary angle-closure glaucoma (P 5 .005) (Table 2). No
statistically signif-icant difference. signi®cant correlation

124 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 1999

 FIGURE 4. The correlation between the peripapillary
FIGURE 2. There was no statistically signi®cant correlation
atrophy-to-disk area ratio and the cup-to-disk area ratio in
between the peripapillary atrophy-to-disk area ratio and the
primary open-angle glaucoma. There was a statistically
cup-to-disk area ratio in primary angle-closure glaucoma.
signi®cant cor-relation between them (r 5 .52, P 5 .003).

FIGURE 5. The correlation between the peripapillary atrophy-

FIGURE 3. There was no statistically signi®cant correlation to-disk area ratio and the visual ®eld mean deviation in primary
between the peripapillary atrophy-to-disk area ratio and the open-angle glaucoma. There was a statistically signi®cant cor-
visual ®eld mean deviation in primary angle-closure glaucoma. relation between them (r 5 2.61, P < .001).

was found between the peripapillary atrophy-to-disk area discriminant factor in differentiating primary angle-closure
ratio and the cup-to-disk area ratio or visual ®eld mean glaucoma from primary open-angle glaucoma (Table 3).
deviation in the primary angle-closure glaucoma group In the refractive error±matched groupsÐ21 eyes with
(Figures 2 and 3). There was, however, a statistically primary angle-closure glaucoma and 23 eyes with primary
signi®cant correlation between the peripapillary atrophy- open-angle glaucoma matched for refractive error, visual
to-disk area ratio and the cup-to-disk area ratio (r 5 .52, P ®eld mean deviation, cup-to-disk area ratio, and pretreat-
5 .003) and the mean deviation (r 5 2.61, P , .001) in the ment intraocular pressureÐthe prevalence of peripapillary
primary open-angle glaucoma group (Figures 4 and 5). atrophy in eyes with primary open-angle glaucoma (70%)
The discriminant analysis identi®ed only peripapillary was higher than that in eyes with primary angle-closure
atrophy-to-disk area ratio (P 5 .015) as a signi®cant glaucoma (38%), with a borderline signi®cance (P 5 .068,

VOL. 127, NO. 2 PERIPAPILLARY ATROPHY IN PACG 125

 Although correlation between the area of peripapillary
TABLE 3. Discriminant Factors for Differentiating PACG atrophy calculated on the basis of the Heidelberg Retina
From POAG (Discriminant Analysis) Tomograph images and those evaluated by histologic
analysis has not yet been fully clari®ed, the Heidelberg
Factor P Value Odds Ratio
Retina Tomograph is a suitable device in the current
Age .32 NS studyÐit uses a diode laser beam (wavelength, 670 nm) as
Refractive error .15 NS the laser source, which generally penetrates to the fundus,
Cup-to-disk area ratio .62 NS even in eyes with a small pupil, and the program incorpo-
Peripapillary atrophy-to-disk area ratio .02 1.54*
rated into the Heidelberg Retina Tomograph is essentially
Mean deviation .08 NS
the automation of the planimetric measurements of peri-
Corrected pattern standard deviation .30 NS
Pretreatment IOP .93 NS
papillary atrophy and the disk area, which were used by
previous investigators to morphometrically determine peri-
IOP 5 intraocular pressure; NS 5 not signi®cant; PACG 5 papillary atrophy.
primary angle-closure glaucoma; POAG 5 primary open-angle We compared the characteristics of peripapillary atro-phy
glaucoma. in eyes with primary angle-closure glaucoma with those in
*Odds ratio per 0.1 change in peripapillary atrophy-to-disk eyes with primary open-angle glaucoma, both of which were
area ratio.
matched for the mean deviation of the visual ®eld indices, the
optic disk area, and the cup-to-disk area ratio. We found a
barely signi®cantly lower prevalence of peripapillary atrophy
Fisher exact test). The peripapillary atrophy-to-disk area
in eyes with primary angle-closure glaucoma (38%)
ratio in the refractive error±matched eyes with primary
compared with eyes with primary open-angle glaucoma
open-angle glaucoma was signi®cantly larger than that in
(68%). The prevalence of peripapillary atrophy in primary
eyes with primary angle-closure glaucoma (P 5 .006;
angle-closure glaucoma is similar to that in eyes with ocular
Table 2). In the refractive error±matched eyes with
primary open-angle glaucoma, there was a statistically hypertension17,19,20 and in normal eyes.19 The average
signi®cant correlation between the peripapillary atrophy- peripapillary atrophy-to-disk area ratio in eyes with primary
to-disk area ratio and the cup-to-disk area ratio (r 5 .59, P open-angle glaucoma was signi®cantly larger than in eyes
5 .003) and the mean deviation (r 5 2.51, P 5 .014). with primary angle-closure glaucoma. We also found a
statistically signi®cant correlation be-tween the peripapillary
atrophy-to-disk area ratio and the cup-to-disk area ratio and
DISCUSSION the visual ®eld mean deviation in eyes with primary open-
angle glaucoma. The results are consistent with those of
THE CHARACTERISTICS OF PERIPAPILLARY ATROPHY HAVE been previous studies.3±7 In contrast, we failed to ®nd a
studied by several groups of investigators in different types of signi®cant correlation between the peripap-illary atrophy-to-
glaucoma. Caprioli and Spaeth15 reported no statistically disk area ratio and the intrapapillary variables or the visual
signi®cant differences in the prevalence of peripapillary ®eld mean deviation in eyes with primary angle-closure
atrophy between high-tension and low-ten-sion glaucoma. glaucoma. One reason for this result might be the small
Jonas and Xu16 reported that peripapillary atrophy zones were number of eyes with peripapillary atrophy in primary angle-
not signi®cantly different between normal-tension glaucoma closure glaucoma.
and primary open-angle glau-coma. In contrast, Buus and Peripapillary atrophy is known to have various causes.
Anderson17 reported that peripapillary atrophy is more Some eyes with peripapillary atrophy have been associated
frequent in eyes with normal-tension glaucoma and that eyes with congenital factors, others with acquired changes (for
with normal-tension glau-coma have larger area of example, stretching of the eyeball, especially in severe
peripapillary atrophy than eyes with ocular hypertension. myopia, age-related structural changes, acquired changes
Kasner and associates19 reported that eyes with ocular in glaucoma), or both.21,22 Peripapillary atrophy has been
hypertension have a signi®cantly lower prevalence of observed in 20% to 50% of normal eyes of white sub-
peripapillary atrophy than normal eyes. Many investigations jects.20,23 In an earlier study, we found peripapillary atro-
of peripapillary atrophy have been performed in eyes with
phy in 25% of Japanese subjects.24 On the basis of
open-angle glaucoma and in normal individuals. The
characteristics of peripapillary atrophy in eyes with primary previous studies, other characteristics of peripapillary
angle-closure glaucoma, however, have been studied little, if atrophy can be summarized as follows:
at all. One reason might be the dif®culty in obtaining high- (1) Change of peripapillary atrophy is associated with
quality photo-graphs with an ordinary fundus camera. In the progressive glaucomatous change in some cases.20,25
present study, we used Heidelberg Retina Tomograph images (2) Peripapillary atrophy area is correlated with age. 5
and software to measure the area of peripapillary atrophy. (3) Peripapillary atrophy is observed more frequently in
myopic eyes than in hyperopic eyes,19 and the area

126 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 1999

 of peripapillary atrophy is also associated with the evidence of an acute attack or combined-mechanism
axial length.26 glaucoma. Therefore, the optic disk change seen in eyes
(4) Peripapillary atrophy represents a decrease in peri- with primary angle-closure glaucoma might be caused
papillary choroidal circulation in glaucoma. primarily by increased intraocular pressure. In contrast, a
Fluorescein angiographic studies showed that impaired variety of factors contribute to the optic disk damage in
choroidal circulation corresponds to the area of peripapil- chronic open-angle glaucoma. Mechanical and vascular
lary atrophy in eyes with open-angle glaucoma.10,11 A factors are now widely discussed in relation to the patho-
recent indocyanine green angiographic study12 demon- genesis of optic nerve damage in chronic open-angle
strated a higher prevalence of areas of hypo¯uorescence in glaucoma.9 Since the peripapillary atrophy is associated
the peripapillary region of glaucomatous eyes compared with decreased vascularity in the peripapillary choroidal
with normal control eyes, which might be caused by an circulation in glaucoma, the different characteristics of
absence of choriocapillaris. Recent histologic studies13,14 peripapillary atrophy between the two groups seem to
demonstrated a de®ciency of choroidal capillary vessels suggest a pathophysiologic difference. The changes in the
corresponding to peripapillary atrophy. optic disk and the peripapillary area are more closely
In the present studyÐin which the visual ®eld change and associated in eyes with primary open-angle glaucoma than
the cup-to-disk area ratio of eyes with primary angle-closure in those with primary angle-closure glaucoma. Further-
glaucoma were matched with those in eyes with primary more, the discriminant analysis indicated only peripapil-
open-angle glaucomaÐpatients with primary an-gle-closure lary atrophy-to-disk area ratio as a discriminant factor that
glaucoma were older than those with primary open-angle distinguishes primary open-angle glaucoma from primary
glaucoma. Thus, the difference in age does not explain the angle-closure glaucoma. Even though the sample size in
smaller area of peripapillary atrophy in eyes with primary the present study is small, the difference in the character-
angle-closure glaucoma. It is possible that the observed istics of peripapillary atrophy between primary angle-
difference in the characteristics of peripapillary atrophy closure glaucoma and primary open-angle glaucoma may
between primary angle-closure glaucoma and pri-mary open- be considered to re¯ect the difference in the pathophysi-
angle glaucoma can be attributed to the bio-metric difference ologic processes of the glaucomatous changes in the two
between the two groups. Eyes with primary angle-closure types of glaucoma.
glaucoma a priori have different biometric features than In summary, this cross-sectional study con®rmed an
normal eyes and eyes with open-angle glaucoma. They also association between peripapillary atrophy and an intrapap-
have a shorter axial length and a smaller corneal diameter and illary variable (cup-to-disk area ratio) and a visual ®eld
are typically associated with hyperopia.27 Therefore, in the index (mean deviation) in eyes with primary open-angle
present study, we did not attempt to match the refractive error glaucoma. However, we were unable to ®nd a signi®cant
between the two groups initially; consequently, the eyes with association between peripapillary atrophy and the glauco-
primary angle-closure glaucoma tended to be more hyperopic matous damage in primary angle-closure glaucoma. The
than those with primary open-angle glaucoma. Then, as the results suggest a difference in the pathophysiologic
second step, we performed additional analyses using process of glaucomatous optic neuropathy between the two
refractive error±matched groups of eyes with primary open- sub-groups.
angle glaucoma and primary angle-closure glaucoma to
diminish biometric factors related to peripapillary atrophy.
Even in the refractive error±matched eyes with primary open- REFERENCES
angle glaucoma, the peripapillary atrophy-to-disk area ratio
was statistically signi®cantly larger than that in eyes with 1. Primrose J. Early sign of the glaucomatous disc. Br J Oph-
primary angle-closure glaucoma, and there was a statisti-cally thalmol 1971;55:820 ± 825.
signi®cant correlation between the peripapillary at-rophy-to- 2. Wilensky JT, Kolker AE. Peripapillary changes in glaucoma.
Am J Ophthalmol 1976;81:341±345.
disk area ratio and the cup-to-disk area ratio and the visual 3. Jonas JB, Nguyen XN, Gusek GC, Naumann GO. Parapap-
®eld mean deviation. The prevalence of peri-papillary illary chorioretinal atrophy in normal and glaucoma eyes.
atrophy in primary angle-closure glaucoma (38%) was also Invest Ophthalmol Vis Sci 1989;30:908 ±918.
less than that of the refractive error±matched eyes with 4. Jonas JB, Fernandez MC, Naumann GO. Glaucomatous
primary open-angle glaucoma (70%); however, the difference parapapillary atrophy: occurrence and correlations. Arch
Ophthalmol 1992;110:214 ±222.
was not statistically signi®cant (P 5 .068).
5. Jonas JB, Naumann GO. Parapapillary chorioretinal atrophy
in normal and glaucoma eyes, II: correlations. Invest Oph-
The difference in the characteristics of peripapillary thalmol Vis Sci 1989;30:919 ±926.
atrophy between eyes with primary angle-closure glaucoma 6. Anderson DR. Correlation of the peripapillary damage with
and those with primary open-angle glaucoma might be the disk anatomy and ®eld abnormalities in glaucoma. Doc
Ophthalmol 1983;35:1±10.
attributable to the difference in causes between the two 7. Heijl A, Samander C. Peripapillary atrophy and glaucoma-tous
glaucoma subtypes. In this study, we enrolled patients with visual ®eld defects. Doc Ophthalmol 1985;42:403± 407.
only chronic primary angle-closure glaucoma that had no 8. Park KH, Tomita G, Liou SY, Kitazawa Y. Correlation

VOL. 127, NO. 2 PERIPAPILLARY ATROPHY IN PACG 127

 between peripapillary atrophy and optic nerve damage in optic disc hemorrhage with retinal nerve ®ber layer defect
normal tension glaucoma. Ophthalmology 1996;103:1899 ± and peripapillary atrophy in normal-tension glaucoma. Oph-
1906. thalmology 1997;104:1926 ±1933.
9. Schulzer M, Drance SM, Carter CJ, et al. Biostatistical 19. Kasner O, Feuer WJ, Anderson DR. Possibly reduced prev-
evidence for two distinct chronic open-angle glaucoma alence of peripapillary crescents in ocular hypertension. Can
populations. Br J Ophthalmol 1990;74:196 ±200. J Ophthalmol 1989;24:211±215.
10. Hayreh SS, Walker WM. Fluorescent fundus photography in 20. Airaksinen PA, Tuulonen JA, Alanko HI, et al. Change of
glaucoma. Am J Ophthalmol 1967;63:982±989. peripapillary atrophy in glaucoma. In Kreiglstein GK, ed.
11. Raitta C, Sarmela T. Fluorescein angiography of the optic Glaucoma update III. New York: Springer-Verlag NY Inc,
disc and peripapillary area in chronic glaucoma. Acta Oph- 1987:97±102.
thalmol 1970;48:303±308. 21. Hogan MJ, Alvarado JA, Waddel JE. Histology of the
12. O'Brart DP, de Souza Lima M, Bartsch DU, Freeman W, human eye: an atlas and textbook. Philadelphia: WB
Weinreb RN. Indocyanine green angiography of the peripap- Saunders Co, 1971:530.
illary region in glaucomatous eyes by confocal scanning laser
22. Nevarez J, Rockwood EJ, Anderson DR. The con®guration
ophthalmoscopy. Am J Ophthalmol 1997;123:657± 666.
of peripapillary tissue in unilateral glaucoma. Arch
13. Fantes FE, Anderson DR. Clinical histologic correlation of
Ophthalmol 1988;106:901±903.
human peripapillary anatomy. Ophthalmology 1989;96:20 ±
23. Quigley HA, Katz J, Derick RJ, et al. An evaluation of optic
25.
disc and nerve ®ber layer examination in monitoring glau-
14. Kubota T, Schlotzer-Schrehardt UM, Naumann GOH,
Kohno T, Inomata H. The ultrastructure of parapapillary coma progression. Ophthalmology 1992;99:19 ±28.
chorioretinal atrophy in eyes with secondary angle closure 24. Hayakawa T, Sugiyama K, Tomita G, et al. Correlation of
glaucoma. Graefes Arch Clin Exp Ophthalmol 1996;234: the peripapillary atrophy area with optic disc cupping and
351±358. disc hemorrhage. J Glaucoma. Forthcoming.
15. Caprioli J, Spaeth GL. Comparison of the optic nerve head in 25. Rockwood EJ, Anderson DR. Acquired peripapillary
high- and low-tension glaucoma. Arch Ophthalmol 1985; changes and progression in glaucoma. Graefes Arch Clin
103:1145±1149. Exp Oph-thalmol 1988;226:510 ±515.
16. Jonas JB, Xu L. Parapapillary chorioretinal atrophy in nor- 26. Kubota T, Jonas J, Naumann GOH. Direct clinico-histolog-
mal-pressure glaucoma. Am J Ophthalmol 1992;115:501± ical correlation of parapapillary chorioretinal atrophy. Br J
505. Ophthalmol 1993;73:103±106.
17. Buus DR, Anderson DR. Peripapillary crescents and halos in 27. Ritch R, Lowe RF. Angle-closure glaucoma: mechanisms
normal-tension glaucoma and ocular hypertension. Ophthal- and epidemiology. In: Ritch R, Shields MB, Krupin T,
mology 1989;96:16 ±19. editors. The glaucomas, 2nd ed. Volume 2. St Louis: Mosby,
18. Sugiyama K, Tomita G, Kitazawa Y, et al. The association of 1996:801± 819.

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

128 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 1999