Introduction

In addition to intraocular pressure (IOP), vascular factorsare important in the pathogenesis of open-angle glaucoma,

including both normal-tension glaucoma (NTG) and high-tension glaucoma (HTG).1–3 In contrast to HTG, in whichIOP plays a more significant role, factors other than IOPmay have greater weight in the pathogenesis or progressionof NTG. Cerebrovascular diseases,4 myocardial ischemia,5

and migraine6–8 are more frequent in patients with NTGthan in those with HTG or in normal subjects, findingswhich are compatible with this hypothesis. Previous reportssuggest that the pattern of visual field damage is somewhatdifferent between NTG and HTG.9–15 It is still unknown,however, how the IOP-dependent factors and the IOP-

Jpn J Ophthalmol 2004;48:340–344 DOI 10.1007/s10384-004-0072-0© Japanese Ophthalmological Society 2004

LABORATORY INVESTIGATION

Visual Field Damage in Normal-tension Glaucoma Patients With or WithoutIschemic Changes in Cerebral Magnetic Resonance Imaging

Junko Suzuki1, Atsuo Tomidokoro1, Makoto Araie1, Goji Tomita1,Junkichi Yamagami1, Toshiyuki Okubo2, and Tomohiko Masumoto3

1 Department of Ophthalmology, University of Tokyo School of Medicine, Tokyo, Japan;2 Department of Radiology, Yamanashi University, Yamanashi, Japan; 3 Department of Radiology,

University of Tokyo School of Medicine, Tokyo, Japan

Abstract

Purpose: To compare the pattern of visual field damage between normal-tension glaucoma (NTG)patients with signs indicative of ischemic changes and those NTG patients without signs of ischemicchanges, using brain magnetic resonance imaging (MRI), in a single center, cross-sectional study.

Methods: In 94 consecutive NTG patients who were younger than 61 years old, brain MRI images wereobtained using fluid-attenuated inversion recovery pulse sequences. The presence of signs indicative ofischemic changes in brain MRI images was decided separately by two neuroradiologists masked to thediagnosis and stage of glaucoma. Visual field testing was performed using the 30-2 program of theHumphrey Visual Field Analyzer. Between the patients with signs indicative of ischemic changes in brainMRI (ischemic group) and those without MRI signs (nonischemic group), total deviation (TD) at eachtest point less the average of TDs of the 30-2 program ([TD � TDmean])—was compared at each testpoint.

Results: Signs indicative of ischemic changes in brain MRI were found in 32 of the 94 patients (34.0%).Age, blood pressure, refraction, intraocular pressure, the average of TDs, mean deviation, and correctedpattern standard deviation were not significantly different between the ischemic (N � 32) and nonis-chemic (N � 62) groups (P � 0.2). [TD � TDmean] in the ischemic group was significantly smaller thanthat in the nonischemic group at 6 nonedge contiguous test points in the inferior pericentral to nasalfield (P � 0.005–0.047).

Conclusion: NTG patients with signs indicative of ischemic changes in brain MRI had a relatively deeperdepression in the inferior pericentral visual field. Jpn J Ophthalmol 2004;48:340–344 © Japanese Ophthalmological Society 2004

Key Words: brain magnetic resonance imaging, high-tension glaucoma, ischemic change, normal-tension glaucoma, open-angle glaucoma

Received: August 20, 2003 / Accepted: February 10, 2004Correspondence and reprint requests to: Makoto Araie, Department

of Ophthalmology, University of Tokyo School of Medicine, 7-3-1Hongo, Bunkyo-ku, Tokyo 113-8655, Japan e-mail: araie-tky@umin.ac.jp

J. SUZUKI ET AL. 341BRAIN MRI IN NTG PATIENTS

independent factors contribute to these differencesbetween NTG and HTG.

Ischemic changes in brain magnetic resonance images(MRI) are more common in NTG patients than in controlsubjects,16,17 suggesting that vascular insufficiency in thecentral nervous system has some relation to the pathogen-esis of NTG. It can be hypothesized that vascular factors forthe pathogenesis of open-angle glaucoma may be moreimportant in NTG patients with ischemic changes in thecentral nervous system rather than in those without. As aconsequence, comparisons between NTG patients withbrain ischemic changes and those without might clarify thecontribution of vascular factors to the clinical features ofopen-angle glaucoma. In the present study, NTG patientsunderwent brain MRI using fluid-attenuated inversionrecovery (FLAIR) pulse sequences, which provided highsensitivity to the detection of ischemic lesions in the centralnervous system,18–21 and the visual field results were com-pared between the NTG patients in whom signs indicativeof ischemic changes were detected in the brain MRI andthose in whom such signs were not detected.

We conducted a single center, cross-sectional study tocompare the visual field test results between the NTGpatients with signs indicative of ischemic changes in brainMRI and those without such signs.

Subjects and Methods

Ninety-four consecutive NTG patients (49 male and 45female) who met the inclusion criteria below were enrolledin this cross-sectional study between January 1998 and February 1999 at the Department of Ophthalmology, Uni-versity of Tokyo Graduate School of Medicine (Tokyo,Japan). The inclusion criteria were as follows: (1) bilateralNTG diagnosed prior to this study; (2) younger than 61years; (3) previously capable of reliable visual field testing(i.e, �20% fixation errors and �33% false-positive andfalse-negative errors) with the 30-2 program of HumphreyVisual Field Analyzer (Humphrey-Zeiss., San Leandro, CA,USA); (4) no refractive error of 6 diopters or more, fundusabnormalities, or cataract or other media opacity in eithereye; and (5) no systemic hypertension necessitating medicaltreatment, diabetes mellitus, cardiovascular disorders, orcerebral ischemic diseases. Diagnosis of NTG was madeaccording to typical glaucomatous findings in disc appear-ance and corresponding visual field damage in eyes with consistently normal IOP and the absence of any contributing ocular or systemic disorders. Normal IOP was defined as IOP that never exceeded 21mmHg without ocular hypotensive medication during the follow-up period, including during 24-hour fluctuation testing. Thetenets of the Declaration of Helsinki were followed withregard to study subjects. Informed consent was obtained forthis research project from all patients in accordance withthe ethics approval granted by the institutional reviewboard of the University of Tokyo Graduate School of Medicine.

In the 94 patients, visual field testing and brain MRI were performed within 2 months of each other. Brain MRIimages were obtained using a 1.5-T system (MagnetomVision, Siemens, Erlangen, Germany) with a FLAIR pulsesequence system (repetition time, 9000ms; inversion time,2200ms). A spin-echo sequence was quickly read using theturbo system (Siemens) with an echo length of 7 and aneffective echo time of 119ms. Fifteen axial sections (thick-ness, 5mm; intersection gap, 2.5mm) covering almost theentire brain were obtained. Two neuroradiologists (T.O.,T.M.), who were masked to the diagnosis and stage of glau-coma in each patient, separately reviewed the MRI imagesto evaluate the presence of signs indicative of ischemicchanges. The images showed single or multiple small hyper-intense lesions with a diameter of 3mm or greater22–24 andexpanded brain infarct lesions in the cerebral white matteror basal ganglia. Only patients identified by both radiolo-gists as having such signs were classified in the ischemicgroup. Otherwise, patients were classified as being in thenonischemic group.

The visual field was tested with the 30-2 program usingthe trial lens chosen according to the instrument guide andbased on the patient’s age and refractive error. At the timeof visual field testing, medical therapies, including glaucomaeye drops, were not changed from the patient’s dailyregimen. If a reliable result was not obtained, the test wasrepeated. An eye was chosen for analysis by using a tableof random numbers. A random number was given to eachpatient. If the final digit of the number was even, the datafor his or her right eye were analyzed, and vice versa. Themirror images of visual field data were used for left eyes.The 74 test points obtained by the 30-2 program were num-bered for the current analysis (Fig. 1). In order to highlight

Figure 1. Numbering of test points of the Humphrey Visual Field Analyzer 30-2 program.

342 Jpn J OphthalmolVol 48: 340–344, 2004

local (i.e, point by point) depression compared to a givenamount of overall visual field damage, [TD � TDmean] wasused, where TD is total deviation, the difference in decibelsbetween the patient’s threshold and the age-correctednormal reference value at each test point (STATPAC 2,Humphrey-Zeiss) and TDmean is the average of the TDs atthe 74 test points. Between the ischemic and nonischemicgroups, [TD � TDmean] was compared point by point.

Results

Single or multiple small hyperintense lesions 3mm orgreater in diameter were found in the brain MRI scans in32 of the 94 patients (34.0%), while no large ischemiclesions, which could possibly relate to clinical symptoms ofcerebral infarction, were observed in the current patients.There was no significant difference between the ischemicand nonischemic groups for the following indicators: theratio of male to female (�2 test); ratio of right eye to left eye(P � 1.0), age (unpaired t test); systolic blood pressure; dias-tolic blood pressure; refraction; IOP; mean deviation (MD);TDmean, and corrected pattern standard deviation (Table 1).No patients had a history of intraocular surgery in thecurrent study population.

[TD � TDmean] in the ischemic group was significantlysmallers on disk (more negative) than that in the nonis-chemic group at 6 nonedge contiguous test points: #41 (P �0.047), #48 (P � 0.042), #49 (P � 0.016), #50 (P � 0.038),#51 (P � 0.005), and #52 (P � 0.040) (Fig. 2).

Discussion

In the present study, 32 (34.0%) of the 94 NTG patients(age, 50.6 � 6.3 years) had slight or mild ischemic signs, such

as small hyperintense lesions in cerebral white matter orbasal ganglia, in brain MRI with FLAIR pulse sequences.Stroman et al.16 reported diffuse cerebral ischemic changesin 8 (40.0%) of 20 NTG patients (age, 73 � 11 years), whichwas significantly greater than the prevalence in age-matched subjects with normal ocular findings (1 of 20 sub-jects, i.e, 5%; age, 67 � 8 years). Ong et al.17 reported thatthe extent of cerebral infarcts in NTG patients (N � 10) wasapproximately double that in age-matched healthy subjects(N � 10), although they did not describe the prevalence ofischemic lesions. The prevalence of ischemic findings inbrain MRI in the current NTG patients is similar to theresults of Stroma et al.,16 although the patients were muchyounger in the current study (51 � 6 years vs. 73 � 11 years).These different results might be due to differences in thesensitivity to intracranial ischemic changes, which is betterin MRI with FLAIR pulse sequences, as used in the currentstudy, than in conventional T1- or T2-weighted imaging.18–21

Although no study, to our knowledge, has previouslyevaluated the prevalence of asymptomatic ischemic lesionsin the brain using FLAIR pulse sequence, some studiesreported the prevalence of ischemic lesions in neurologi-cally normal subjects evaluated with T1- or T2 weightedimaging. Price et al.24 reported that brain infarct lesions (�3mm in diameter) were found in 961 (28%) of 3397 subjectsaged 60 or older without known prior stroke; and Maeshimaet al.25 found focal ischemic lesions (�5mm) in 21 (25%) of84 normal Japanese middle-aged subjects aged 40–59 (48.7� 5.5, mean � SD) years. Thus, although we did not obtain

Figure 2. Comparison of [TD � TDmean] values between the ischemicand nonischemic groups. The upper number in the square indicatesmean value for the ischemic group; the middle number that for the non-ischemic group; and the lower number shows the P value for the dif-ference between the two values above (unpaired t test). Shaded squaresindicate test points at which [TD � TDmean] in the ischemic group weresignificantly lower than those in the nonischemic group (P � 0.05).TD,total deviation.

Table 1. Patient characteristics

Ischemic Nonischemic P

Number of patients 32 62Male/female 18/14 31/31 0.664Right/left 17/15 32/30 1.000Age (year) 50.8 � 6.3 50.7 � 6.5 0.907Systolic blood pressure 125.7 � 12.6 126.2 � 15.4 0.875

(mmHg)Diastolic blood pressure 74.5 � 10.6 77.5 � 13.2 0.269

(mmHg)Refraction (diopter) �2.5 � 2.4 �2.2 � 2.5 0.563IOP (mmHg) 15.9 � 2.6 15.3 � 2.1 0.278Mean deviation (decibel) �7.0 � 5.7 �6.8 � 6.4 0.863TDmean (decibel) �6.6 � 5.6 �6.7 � 6.3 0.952Corrected pattern standard 7.6 � 4.0 7.5 � 5.2 0.993

deviation

Data were expressed as mean � SD.Ischemic, eyes of patients with ischemic changes in brain magnetic

resonance imaging (MRI). Nonischemic, eyes of patients with normalbrain MRI study. P, P value for differences between ischemic and non-ischemic groups (�2 test or unpaired t test); IOP, intraocular pressure;TDmean, the average of the total deviation at the 74 test points.

J. SUZUKI ET AL. 343BRAIN MRI IN NTG PATIENTS

the control data in normal subjects without glaucoma in thecurrent study, the currently estimated prevalence (34.0%)of ischemic lesions (�3mm) in NTG subjects aged �60years may be somewhat higher than the expected preva-lence in normal Japanese middle-aged subjects, even afterconsideration of the higher sensitivity of the FLAIR pulsesequence.

There is no clear evidence that circulatory insufficiencyin the brain directly relates to the pathogenesis of glauco-matous optic neuropathy. Some previous studies, however,have concluded that cerebral ischemic changes are moreoften or more diffusely observed with MRI in NTG patientsthan in normal age-matched subjects.16,17 More recently,Harris et al.26 reported decreased cerebrovascular bloodflow velocity and vasoreactivity in 15 open-angle glaucomapatients than in 16 age-matched normal subjects. Thecurrent results showing that the presence of cerebralischemic lesions had some association with the pattern ofvisual field damage in NTG patients may also suggest somerelationship between open-angle glaucoma, especiallyNTG, and cerebral blood flow.

Comparison of the ischemic and nonischemic groupsindicated that averages of age, systemic blood pressure,refraction, IOP, MD, TDmean, and corrected pattern standarddeviation were very similar between the groups (Table 1),suggesting that the systemic and ocular background forglaucoma and the stages of visual field damage did notdiffer.There was interindividual variation in stages of visualfield damage, however, in each of the groups, indicated bythe relatively large standard deviations in MD. A precisematching of MDs between the 2 groups was very difficultbecause of the limited numbers of patients included in thecurrent study. To exclude the influence of interindividualvariation as much as possible, [TD � TDmean] was used,where TDmean is the average of TDs at 74 test points of the30-2 program. [TD � TDmean] thus indicates the localdepression at each test point compared with a given amountof overall visual field damage.14 [TD � TDmean] was signifi-cantly more negative in the ischemic group than in the non-ischemic group at 6 nonedge contiguous test points in theinferior pericentral to nasal field (Fig. 2).Although none theP values (0.005–0.047) was small enough to completelyexclude the possibility of a false positive error due to mul-tiple comparisons (74 test points), the finding that these 6nonedge points were located contiguous to the nerve fiberbundle suggests the clinical significance of the currentresults. In a clinical context, reliable fields can be consideredto have a localized glaucomatous defect if a cluster of 3 ormore nonedge test points show abnormal values (P �0.05).27 When TD, instead of [TD � TDmean], was comparedbetween the 2 groups in a point-by-point manner, a signifi-cant difference was found at only one test point in the infe-rior pericentral area (#51 in Fig. 1), with a P value of 0.024.Although this P value at only 1 test point was not statisti-cally significant in consideration of the multiple compar-isons at 70 test points, this finding was compatible with thecurrent result obtained using [TD � TDmean]. MD could beused as an alternative to TDmean to normalize the interindi-

vidual variation of the overall visual field damage.When thesame analyses were performed using [TD � MD] insteadof [TD � TDmean], virtually equivalent results were obtained(data not shown).

In the current study, the NTG patients in the ischemicgroup showed a tendency toward equal involvement of thesuperior and inferior hemifields, while in the patients in thenonischemic group, the superior hemifield was remarkablymore involved than the inferior hemifield, resulting in sig-nificant differences between the 2 groups in the inferiorhemifield (Fig. 2). Such a pattern of visual field defectsshown in the nonischemic group, with deeper damage in thesuperior hemifield, is comparable to previously reportedfindings observed in common NTG patients in comparisonwith HTG patients.14,15 Therefore, the significant differencesin the inferior hemifield between the ischemic and non-ischemic groups may indicate relatively deeper damage inthe inferior hemifield in the ischemic group.

The inferior field is reportedly more likely to have suf-fered in HTG patients with diabetes mellitus,28,29 and themost frequent visual field defect in nonarteritic anteriorischemic optic neuropathy is an altitudinal defect involvingthe inferior hemifield.30,31 In the current study, because thestudy subjects were limited to NTG patients and no sig-nificant intergroup difference existed in the backgroundfactors, the influence of IOP and other systemic abnormal-ities on the results should be minor.Therefore, the influenceof non-IOP-dependent factors is thought to be more sensi-tively represented. Thus, the present findings suggest thatthe non-IOP-dependent factors, possibly including vascularfactors, are more likely to be involved in the damage in theinferior pericentral field of NTG.

There are regional differences in the fenestration orpores of the lamina cribrosa. The superior and inferiorregions have larger pores, thinner connective tissue, and lessglial cell support than the nasal and temporal portions.32–35

Moreover, it has been suggested that the inferior portion ofthe optic disc might be relatively more vulnerable to IOP-dependent factors.2 In the present study, the inferior peri-central area in the visual field was relatively more depressedin NTG patients with signs indicative of ischemic changeson brain MRI compared with those without such signs. Thisfinding is consistent with the hypothesis that a part of thesuperior portion of the optic nerve head (ONH) is morevulnerable to non-IOP-dependent damaging factors, whichmay include vascular insufficiency. Since the inferior peri-central visual field is relatively more important to quality oflife,36 the present findings may have clinical implications,especially in the late stage of NTG.

In conclusion, the present study indicates that NTGpatients with signs indicative of ischemic changes in brainMRI have a relatively deeper depression in the inferiorpericentral area than those NTG patients without signs ofischemic changes. This finding suggests that the inferiorpericentral visual field or the corresponding portion of the ONH is relatively more vulnerable to non-IOP-dependent factors, possibly including vascular factors, inNTG patients.

344 Jpn J OphthalmolVol 48: 340–344, 2004

References

1. Minckler DS, Spaeth GL. Optic nerve damage in glaucoma. SurvOphthalmol 1981;26:128–148.

2. Fechtner RD, Weinreb RN. Mechanisms of optic nerve damage in primary open-angle glaucoma. Surv Ophthalmol 1994;39:23–42.

3. Flammer J. The vascular concept of glaucoma. Surv Ophthalmol1994;38 (Suppl):S3–S6.

4. Feldman F, Sweeney V, Drance S. Cerebro-vascular studies inchronic simple glaucoma. Can J Ophthalmol 1969;4:358–364.

5. Kaiser HJ, Flammer J, Burckhardt D. Silent myocardial ischemia inglaucoma patients. Ophthalmologica 1993;207:6–7.

6. Corbett JJ, Phelps CD, Eslinger P, Montague PR. The neurologicevaluation of patients with low-tension glaucoma. Invest Ophthal-mol Vis Sci 1985;26:1101–1104.

7. Phelps CD, Corbett JJ. Migraine and low-tension glaucoma.A case–control study. Invest Ophthalmol Vis Sci 1985;26:1105–1108.

8. Cursiefen C, Wisse M, Cursiefen S, Junemann A, Martus P, KorthM. Migraine and tension headache in high-pressure and normal-pressure glaucoma. Am J Ophthalmol 2000;129:102–104.

9. Hitchings RA, Anderton SA. A comparative study of visual fielddefects seen in patients with low-tension glaucoma and chronicsimple glaucoma. Br J Ophthalmol 1983;67:818–821.

10. Caprioli J, Spaeth GL. Comparison of visual field defects in thelow-tension glaucomas with those in the high-tension glaucomas.Am J Ophthalmol 1984;97:730–737.

11. Gramer E, Althaus G, Leydhecker W. Topography and progressionof visual field damage in low-tension glaucoma, open-angle glau-coma and pigmentary glaucoma with the program Delta of theOctopus Perimeter 201. A clinical study. In: Greve E, Heijl A,editors. Seventh International Visual field Symposium, 1986.Dordrecht: Martinus Nijhoff/Dr W Junk; 1987. p. 349–363.

12. Chauhan BC, Drance SM, Douglas GR, Johnson CA. Visual fielddamage in normal-tension and high-tension glaucoma. Am J Ophthalmol 1989;108:636–642.

13. Zeiter JH, Shin DH, Juzych MS, Jarvi TS, Spoor TC, Zwas F.Visual field defects in patients with normal-tension glaucoma andpatients with high-tension glaucoma. Am J Ophthalmol 1992;114:758–763.

14. Araie M, Yamagami J, Suziki Y. Visual field defects in normal-tension and high-tension glaucoma. Ophthalmology 1993;100:1808–1814.

15. Koseki N, Araie M, Suzuki Y, Yamagami J. Visual field damageproximal to fixation in normal- and high-tension glaucoma eyes.Jpn J Ophthalmol 1995;39:274–283.

16. Stroman GA, Stewart WC, Golnik KC, Cure JK, Olinger RE. Mag-netic resonance imaging in patients with low-tension glaucoma.Arch Ophthalmol 1995;113:168–172.

17. Ong K, Farinelli A, Billson F, Houang M, Stern M. Comparativestudy of brain magnetic resonance imaging findings in patients withlow-tension glaucoma and control subjects. Ophthalmology 1995;102:1632–1638.

18. De Coene B, Hajnal JV, Gatehouse P, et al. MR of the brain usingfluid-attenuated inversion recovery (FLAIR) pulse sequences.AJNR Am J Neuroradiol 1992;13:1555–1564.

19. Taoka T, Iwasaki S, Nakagawa H, et al. Fast fluid-attenuated inver-sion recovery (FAST-FLAIR) of ischemic lesions in the brain: com-parison with T2-weighted turbo SE. Radiat Med 1996;14:127–131.

20. Okubo T, Hayashi N, Shirouzu I, et al. Detection of brain metasta-sis: comparison of Turbo-FLAIR imaging, T2-weighted imagingand double-dose gadolinium-enhanced MR imaging. Radiat Med1998;16:273–281.

21. Okuda T, Korogi Y, Shigematsu Y, et al. Brain lesions: when shouldfluid-attenuated inversion-recovery sequences be used in MR eval-uation? Radiology 1999;212:793–798.

22. Kobayashi S, Okada K, Koide H, Bokura H. Yamaguchi S. Subcor-tical silent brain infarction as a risk factor for clinical stroke. Stroke1997;28:1932–1939.

23. Watanabe N, Imai Y, Nagai K, et al. Nocturnal blood pressure andsilent cerebrovascular lesions in elderly Japanese. Stroke 1996;27:1319–1327.

24. Price TR, Manolio TA, Kronmal RA, et al. Silent brain infarctionon magnetic resonance imaging and neurological abnormalities in community-dwelling older adults. The Cardiovascular HealthStudy. CHS Collaborative Research Group. Stroke 1997;28:1158–1164.

25. Maeshima S, Moriwaki H, Ozaki F, Okita R, Yamaga H, UeyoshiA. Silent cerebral infarction and cognitive function in middle-agedneurologically healthy subjects. Acta Neurol Scand 2002;105:179–184.

26. Harris A, Zarfati D, Zalish M, et al. Reduced cerebrovascular bloodflow velocities and vasoreactivity in open-angle glaucoma. Am JOphthalmol 2003;135:144–147.

27. Anderson D, Patella V. Interpretation of a single field. Automatedstatic perimetry, Second edition. St. Louis, MO: Mosby; 1999. p.121–190.

28. Zeiter JH, Shin DH, Baek NH. Visual field defects in diabeticpatients with primary open-angle glaucoma. Am J Ophthalmol1991;111:581–584.

29. Zeiter JH, Shin DH. Diabetes in primary open-angle glaucomapatients with inferior visual field defects. Graefes Arch Clin ExpOphthalmol 1994;232:205–210.

30. Miller GR, Smith JL. Ischemic optic neuropathy. Am J Ophthal-mol 1966;62:103–115.

31. Boghen DR, Glaser JS. Ischaemic optic neuropathy. The clinicalprofile and history. Brain 1975;98:689–708.

32. Quigley HA, Addicks EM. Regional differences in the structure ofthe lamina cribrosa and their relation to glaucomatous optic nervedamage. Arch Ophthalmol 1981;99:137–143.

33. Radius RL, Gonzales M.Anatomy of the lamina cribrosa in humaneyes. Arch Ophthalmol 1981;99:2159–2162.

34. Radius RL. Regional specificity in anatomy at the lamina cribrosa.Arch Ophthalmol 1981;99:478–480.

35. Jonas JB, Mardin CY, Schlotzer-Schrehardt U, Naumann GO.Morphometry of the human lamina cribrosa surface. Invest Ophthalmol Vis Sci 1991;32:401–405.

36. Esterman B. Grid for scoring visual fields. I. Tangent screen. ArchOphthalmol 1967;77:780–786.