Visual Field Defects in Patients With Normal-tension Glaucoma and Patients With High-tension Glaucoma John H. Zeiter, M.D., D o n g H. S h i n , M.D., Mark S. Juzych, M.D., Tim S. Jarvi, M.D., T h o m a s C. Spoor, M.D., and Fred Zwas, P h . D .
We compared the automated visual field test results of 24 patients with normal-tension glaucoma and 24 patients with high-tension glaucoma who were closely matched for the amount of visual field loss to determine any differences in the characteristics of visual field defects between the two groups. Patients were matched with a maximum allowable difference in mean deviation of 0.3 dB. Although the normal-tension group had a greater amount of focal visual field loss (pattern standard deviation), the difference was not statistically significant (P = .628). Additionally, there was no statistically significant difference in the amount of diffuse or focal visual field damage in the superior hemifields between the two groups; however, the patients with normal-tension glaucoma had a significantly greater amount of localized visual field loss in the inferior hemifield than the patients with high-tension glaucoma (P = .015). Our data support the hypothesis that a vascular mechanism may have a greater role in the pathogenesis of optic nerve damage and visual field loss in patients with normal-tension glaucoma than in patients with high-tension glaucoma.
3 I N C E the initial description of a patient with glaucomatous optic nerve damage and visual field loss in the absence of increased intraocular pressure by Von Graefe 1 in 1857, the relation ship between intraocular pressure, optic nerve damage, and visual field loss has aroused con-
Accepted for publication Sept. 16, 1992. From the Kresge Eye Institute, Wayne State University School of Medicine, Detroit, Michigan. This study was presented in part at the annual meeting of the Associa tion for Research in Vision and Ophthalmology, Sarasota, Florida, May 8, 1992. Reprint requests to Dong H. Shin, M.D., Kresge Eye Institute, 4717 St. Antoine Blvd., Detroit, MI 48201. 758
troversy. A better understanding of the patho genesis of high-tension and normal-tension glaucoma might result if one could show differ ences in the characteristics of optic nerve dam age and visual field defects between these two types. Unfortunately, there has been little agreement in published reports as to the simi larities and differences of these characteristics in patients with normal-tension glaucoma and patients with high-tension glaucoma. Some investigators have shown that visual field defects in patients with normal-tension glaucoma as compared to patients with hightension glaucoma tend to be of sudden onset, 2 closer to fixation,2'5 deeper and steeper, 25 more localized or less diffuse, 69 and more common in the superior hemifield.10·11 However, other in vestigators have found no differences between the visual field defects in patients with normaltension glaucoma and patients with high-ten sion glaucoma. 1215 The reasons for such contro versy appear to be caused by differences in study designs, including different types of perimetry, various definitions of normal-tension glaucoma regarding the highest recorded intra ocular pressure, different stages of optic nerve damage and visual field loss, and different methods used in statistical analysis of the data. We designed a study to compare the visual field defects of patients with normal-tension glaucoma and patients with high-tension glau coma by using strict criteria for defining nor mal-tension glaucoma and closely matching both groups with respect to the amount of visual field loss.
Patients and Methods We identified 24 patients with normal-ten sion glaucoma. Each patient had glaucomatous optic nerve damage, open anterior chamber angles on gonioscopy, a maximum recorded
©AMERICAN JOURNAL OF OPHTHALMOLOGY 114:758-763, DECEMBER, 1992
Vol. 114, No. 6
Visual Field Defects and Glaucoma
intraocular pressure that did not exceed 22 mm Hg by Goldmann applanation tonometry, a diurnal intraocular pressure curve done with out therapy with a minimum of 12 measure ments over at least a four-week period, a visual acuity of 20/30 or better, a pupil diameter of at least 3 mm, and a reliable visual field test (fixation losses less than 20% and false-positive and false-negative errors less than 33% accord ing to the manufacturer's criteria [Allergan Humphrey, San Leandro, California]) with re producible Aulhorn's stage 1, 2, or 3 visual field defects16 in at least one eye. In stage 1, relative defects are seen in the area of Bjerrum's scotoma. Stage 2 shows deep defects in the area of Bjerrum's scotoma, which form an incomplete arcuate scotoma, not yet connected to the blind spot. In stage 3, the arcuate defect is connected to the blind spot often with breakthrough into the nasal periphery. Cup/disk ratio was deter mined by examination of the dilated fundus performed by one glaucoma subspecialist by using contour to establish the vertical cup/disk ratio. Fundus examination for the cup/disk ratio determination was binocular by using a slit-lamp biomicroscope and a contact lens, Hruby lens, or a 78-diopter double aspheric lens. When binocular examination was not pos sible because of a poorly dilatable pupil, how ever, direct ophthalmoscopy with monocular clues such as disk vessel bending was used for determination of the vertical cup/disk ratio by contour. Patients were excluded if they had a history of trauma or any ocular diseases known to affect the visual field, such as diabetic retinopathy, macular degeneration, or a vascular occlusion. Also, patients with optic neuropathy other than glaucoma, local disk malformations, or neurologic or orbital abnormalities con firmed by computed tomography or serologie studies when applicable were excluded. Pa tients with advanced visual field loss (mean deviation s 20 dB) were also excluded from the study. If a patient had visual field loss in both eyes, then one eye was randomly selected by the toss of a coin. The 24 Humphrey visual fields from the 24 patients with normal-tension glaucoma were matched closely for extent of visual field dam age to 24 Humphrey visual fields of 24 patients with high-tension glaucoma, with a maximum allowable difference in mean deviation of 0.3 dB. The visual fields of patients with hightension glaucoma were selected from 162 visual field examinations of 62 patients with hightension glaucoma. All patients included in the
759
high-tension group had primary open-angle glaucoma with intraocular pressures of 26 mm Hg or greater before the initiation of therapy, glaucomatous optic nerve damage, open angles with mild to moderate pigment in the trabecular meshwork on gonioscopy, and reliable Humphrey visual field tests with Aulhorn's stage 1, 2, or 3 visual field defects. Patients with high-tension glaucoma with a visual acuity worse than 20/30, a pupil diameter less than 3 mm, or any other ocular condition known to influence the visual field were excluded. All visual field examinations were performed by trained technicians with a visual field analyzer (Allergan Humphrey 24-2 Threshold Program using Stimulus III, Allergan Humphrey, San Leandro, California). To determine if there was a difference in the frequency of visual field defects close to fixa tion between patients with normal-tension glaucoma and patients with high-tension glau coma, we counted the number of patients in each group with scotomas extending to within 6 degrees of central fixation. Patients were count ed if any one of the central four threshold values on the total deviation plot decreased 10 dB or more below age-corrected normal values, or if any of the central four threshold values along with contiguous values decreased 5 dB or more below age-corrected normal values. Because the normal-tension and high-tension groups of patients were closely matched for loss of overall sensitivity (mean deviation), we were able to determine a difference between the two groups with respect to localized visual field loss by comparing the pattern standard deviation. We then calculated the mean deviations and pattern standard deviations for the superior and inferior hemifields of all 48 patients. We attempted to find whether the superior or the inferior hemifield had more visual field loss both within each group (normal-tension and high-tension groups, respectively) and between each group (normal-tension vs high-tension groups). First, we compared the mean devia tions of the superior to the inferior hemifields to determine any differences in diffuse visual field loss within the normal-tension and hightension groups, respectively. Then we com pared the pattern standard deviations of the superior to the inferior hemifields to determine any differences in focal visual field loss within each of the two groups. Finally, we compared the superior and the inferior hemifields, respec tively, of the normal-tension group to those of the high-tension group with respect to both
760
AMERICAN JOURNAL OF OPHTHALMOLOGY
mean deviation and pattern standard deviation. We used the Wilcoxon signed rank test to compare mean deviation (matched) between patients with normal-tension glaucoma and pa tients with high-tension glaucoma. We used chi-square analysis with continuity correction to compare gender, race, number of patients in each group with vascular risk factors, and num ber of patients in each group with scotomas extending to within 6 degrees of central fixa tion. The Mann-Whitney U test was used for other unpaired statistical comparisons.
Results The mean deviations, cup/disk ratios, and other demographic data for the normal-tension and high-tension glaucoma groups were ana lyzed (Table 1). As expected, the mean devia tion values that were matched to within 0.3 dB in the selection process were almost identical in the two groups. As reported by other investiga tors, the normal-tension group had a younger mean age2 and a larger mean cup/disk ratio for a given amount of visual field loss 2,317 than the high-tension group. However, the differences were not statistically significant (P = .170 and P = .350, respectively). There were no significant differences in the gender, race, or systemic vascular risk factors between the two groups. We studied the differences in the number of patients with defects extending to within 6 TABLE 1 BACKGROUND VARIABLES FOR NORMAL-TENSION AND HIGH-TENSION GROUPS
VARIABLE
Mean deviation (dB) Cup/disk ratio Age (yrs) Gender ratio (male:female) Race ratio (black:white) Vascular risk factors*
PATIENTS WITH NORMAL-TENSION GLAUCOMA (N=24)
PATIENTS WITH HIGH-TENSION GLAUCOMA (N=24) P VALUE
9.16 ± 3.93 0.80 ± 0.10 67.5 ± 12.1 10:14
9.11 ± 3.89 0.75 ±0.15 71.6 ± 13.0 9:15
.475 .350 .170 .999
14:10
16:8
.766
9 (38%)
10 (42%)
.999
'Number of patients with cardiovascular disease or hyperten sion or both, or cardiovascular disease or diabetes mellitus, or both.
December, 1992
degrees of central fixation, in the amount of localized visual field loss (pattern standard deviation), and in the amounts of generalized and localized visual field loss in the upper and lower hemifields of the normal-tension and high-tension glaucoma groups (Table 2). We found no statistically significant difference in the number of patients with scotomas falling within 6 degrees of central fixation (P = .555). Although the normal-tension glaucoma group had a larger pattern standard deviation than the high-tension glaucoma group overall (7.37 ± 2.94 dB vs 6.94 ± 3.08 dB), this difference was not statistically significant (P = .628). When comparing the amount of superior to inferior hemifield involvement within the hightension glaucoma group, there was significantly more involvement of the superior than the inferior hemifield with respect to both mean deviation and pattern standard deviation (P = .004 and P = .0003, respectively). However, in the normal-tension glaucoma group, there was a tendency toward equal involvement of the two hemifields with no significant differences noted in either the mean deviation or the pat tern standard deviation (P = .918 and P = .350, TABLE 2 COMPARISON OF MEAN DEVIATION AND PATTERN STANDARD DEVIATION FOR NORMAL-TENSION AND HIGH-TENSION GROUPS PATIENTS WITH PATIENTS WITH NORMAL-TENSION HIGH-TENSION GLAUCOMA GLAUCOMA (N-24) (N=24) P VALUE
13 (54%) Scotomas within 16 (67%) .555 6 degrees of fixation* Pattern standard 7.37 ± 2.94 6.94 ± 3.08 .628 deviation (dB) Superior mean 10.16 ± 5.86* 12.58 ± 6.43* .180 deviation (dB) Inferior mean 9.55 ± 4.65* 7.52 ± 3.10* .167 deviation (dB) Superior pattern 7.28 ± 4.77* 9.29 ± 5.88" .177 standard deviation (dB) Inferior pattern 6.65 + 4.25* 4.17 + 1.86" .015 standard deviation (dB) »Number of patients with scotomas extending to within 6 degrees of central fixation. *P = .918. *P = .004. S P = .350. 'P = .0003.
Vol. 114, No. 6
Visual Field Defects and Glaucoma
respectively). There were no statistically signif icant differences between the normal-tension and high-tension glaucoma groups when com paring the amount of generalized visual field loss in the superior hemifields or the inferior hemifields (P = .180 and P = .167, respective ly). However, there was significantly more focal visual field loss located in the inferior hemifield in the normal-tension group than in the hightension group (P = .015), although there was no difference between the two groups in the amount of focal visual field loss located in the superior hemifield (P = .177).
Discussion Normal-tension glaucoma has been de scribed as having typical glaucomatous optic disk changes with resulting visual field loss, open angles, and intraocular pressures within the statistically normal range. Since the initial description of this entity, many questions have been raised concerning the mechanisms of op tic nerve damage in glaucoma, and whether there are different mechanisms involved in the pathogenesis of normal-tension vs high-ten sion glaucoma. For more than a decade, authors have been studying both similarities and dif ferences in the visual field defects between patients with normal-tension glaucoma and patients with high-tension glaucoma.2"16 A dif ferent pattern of visual field loss in these two populations raises the possibility of a different mechanism of optic nerve damage and might suggest a difference in the relative roles that vascular ischemia and intraocular pressure have in the pathogenesis of normal-tension and high-tension glaucoma. Unfortunately, there has been little agreement on what types of differences in visual field loss exist between these two groups of patients. Some investigators found that visual field defects were close to fixation much more com monly in patients with low-tension glaucoma than patients with high-tension glaucoma and that because of this, patients with low-tension glaucoma lost visual acuity earlier and more frequently than patients with high-tension glaucoma.2'4 Conversely, many other investiga tors have found no significant difference in the number of patients with normal-tension glau coma and patients with high-tension glaucoma with visual field defects near fixation.11·121518 In
761
our study, unlike the previously mentioned studies, the visual fields were closely matched for the amount of visual field loss in each group, and we found no significant difference in the number of patients with normal-tension glaucoma and patients with high-tension glau coma with scotomas extending to within 6 de grees of central fixation. Recently, Chauhan and associates 9 found that individuals with normal-tension glaucoma had greater areas of normal sensitivity and, there fore, more localized visual field defects than individuals with high-tension glaucoma. Previ ously, two separate studies had shown that patients with glaucoma with generalized visual field loss had a higher mean intraocular pres sure than those with more localized visual field loss.6·7 We investigated these findings by com paring the pattern standard deviation values (localized visual field loss) between normaltension and high-tension groups that had been closely matched for mean deviation (genera lized visual field loss). Although the normaltension group in our study had a greater amount of localized visual field loss than the high-tension group, this difference was not statistically significant. This finding is similar to that of Paul, Cohn, and Weber,19 who recently found no significant difference in the pattern standard deviation between patients with lowtension glaucoma (intraocular pressure ^ 2 1 mm Hg) and patients with high-tension glauco ma (intraocular pressure s 28 mm Hg) with similar mean deviations. We may not have found a significantly greater amount of localized visual field loss in patients with normal-tension glaucoma because the av erage mean deviation in our patients was high er than that of the patients in previous stud ies,6,7'9 which indicates slightly more advanced visual field loss in our groups. As visual field loss becomes more advanced, any differences in the amount of localized and diffuse damage between normal-tension and high-tension groups may become less. Additionally, the maximum intraocular pressure of the normaltension group and the minimum intraocular pressure of the high-tension group in our study were only 4 mm Hg apart, which may be re sponsible for some overlap of localized and diffuse damage between the two groups. In our study, we found a tendency toward equal involvement of the superior and inferior hemifields in the patients with normal-tension glaucoma. However, in the patients with high-
762
December, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
tension glaucoma, the superior hemifield was remarkably more involved than the inferior hemifield with respect to both generalized and localized visual field loss. This finding that patients with early to moderate primary openangle glaucoma have visual field defects that involve the superior hemifield to a greater ex tent than the inferior hemifield has been well documented. 20 ' 24 Additionally, many investigators have ad dressed the issue of whether there is any differ ence in the distribution of visual field defects over the superior and inferior hemifields be tween patients with normal-tension glaucoma and patients with high-tension glaucoma. 5101318 Although most studies found no differences between the two groups,10,12·'3 Gramer, Althaus and Leydhecker 5 found a preponderance of inferonasal visual field defects in patients with normal-tension glaucoma, whereas Greve and Geijssen11·18 found greater involvement of the superior hemifield in the normal-tension group as compared to the high-tension group. We found no statistically significant difference in the distribution of generalized visual field loss over the superior and inferior hemifields be tween the two groups. However, when we fo cused on localized visual field loss, we found that the normal-tension group had a signifi cantly greater amount of focal visual field de fects in the inferior hemifield. This finding of a preponderance of localized visual field loss in the inferior hemifield of patients with normal-tension glaucoma as com pared to patients with high-tension glaucoma may have important implications as to the rela tive role that vascular ischemia may have in optic nerve damage in patients with normaltension glaucoma. The most frequent visual field defect in nonarteritic anterior ischemie optic neuropathy is an altitudinal defect involv ing the inferior hemifield.13·28-26 Additionally, we have shown in previous studies that patients with primary open-angle glaucoma with diabe tes mellitus develop inferior visual field defects more frequently than patients with primary open-angle glaucoma without diabetes melli tus, speculating the possibility of a vascular contribution at the level of the optic nerve head in patients with glaucoma and diabetes. 27 · 28 Re cently, Gramer and Althaus 29 found that pa tients with primary open-angle glaucoma with low systolic blood pressure had more severe visual field defects in the lower hemifield com pared to the upper hemifield. Additionally, they found that the visual field defects in these patients with lower hemifield involvement
were twice as likely to deteriorate as those in patients with superior visual field loss.29 Several authors have suggested that vascular factors in addition to intraocular pressure have an important role in the pathogenesis of glau coma. 2932 Also, there appears to be an associa tion between normal-tension glaucoma and migraine headaches 33 as well as ocular vasospasm,84·35 which strongly suggests a possible vascular origin in this disease. Our findings of a greater amount of localized visual field loss in the inferior hemifield of patients with normaltension glaucoma as compared to patients with high-tension glaucoma is one more observation in support of the hypothesis that vascular ische mia may have a larger role in the pathogenesis of optic nerve damage and visual field loss in normal-tension glaucoma than in high-tension glaucoma. Although we found more patients with nor mal-tension glaucoma with localized visual field loss in the inferior hemifield, many pa tients in the high-tension group also had some localized defects in the inferior half of the visual field. This finding reminds us that there is a spectrum of visual field defects involving varying amounts of the inferior and superior hemifields between the two groups. Additional ly, when defining the two groups on the basis of intraocular pressure alone, one cannot control for other possible risk factors in the production of glaucoma. Also, even with the use of diurnal pressure curves in defining normal-tension glaucoma, there may still be some overlap of intraocular pressure between the two groups. These unavoidable circumstances along with variations in study designs and different stages of optic nerve damage and visual field loss help explain why there has been little agreement on the differences in visual field defects between patients with normal-tension glaucoma and pa tients with high-tension glaucoma. Additional studies using patients closely matched for the amount of visual field loss will help further define differences in visual field defects between patients with normal-tension glaucoma and patients with high-tension glau coma and provide more insight into the relative roles of intraocular pressure and vascular ische mia in the production of glaucoma.
References 1. Von Graefe, A.: Über die Iridectomie bei Glau coma und über den glaucomatosen Frozess. Graefes Arch. Clin. Exp. Ophthalmol. 3:456, 1857.
Vol. 114, No. 6
Visual Field Defects and Glaucoma
2. Levene, R. Z.: Low tension glaucoma. A critical review and new material. Surv. Ophthalmol. 24:621, 1980. 3. Hitchings, R. A., and Anderton, S. A.: A com parative study of visual field defects seen in patients with low-tension glaucoma and chronic simple glau coma. Br. J. Ophthalmol. 67:818, 1983. 4. Caprioli, J., and Spaeth, G. L.: Comparison of visual field defects in the low-tension glaucomas with those in the high-tension glaucomas. Am. J. Ophthalmol. 97:730, 1984. 5. Gramer, E., Althaus, G., and Leydhecker, W.: Localization and depth of glaucomatous visual field defects in relation to the size of the neuroretinal rim area of the disk in low tension glaucoma, glaucoma simplex, and pigmentary glaucoma. Clinical study with the Octopus 201 perimeter and the optic nerve head analyser. Klin. Monatsbl. Augenheilkd. 189: 190, 1986. 6. Glowazki, A., and Flammer, J.: Besteht ein Un terschied bei Patienten mit eher lokalisiertem und solchen mit eher diffusem Gesichtsfeldschaden? Klin. Monatsbl. Augenheilkd. 190:301, 1987. 7. Caprioli, J., Sears, M., and Miller, J. M.: Patterns of early visual field loss in open-angle glaucoma. Am. J. Ophthalmol. 103:512, 1987. 8. Drance, S. M., Douglas, G. R., Airaksinen, P. J., Schulzer, M., and Hitchings, R. A.: Diffuse visual field loss in chronic open-angle and low-tension glaucoma. Am. J. Ophthalmol. 104:577, 1987. 9. Chauhan, B. C , Drance, S. M., Douglas, G. R., and Johnson, C. A.: Visual field damage in normaltension glaucoma and high-tension glaucoma. Am. J. Ophthalmol. 108:636, 1989. 10. Drance, S. M.: The visual field of low tension glaucoma and shock-induced optic neuropathy. Arch. Ophthalmol. 95:1359, 1977. 11. Greve, E. L., and Geijssen, H. C : Comparison of glaucomatous visual field defects in patients with high and with low intraocular pressures. Doc. Oph thalmol. Proc. Ser. 35:101, 1982. 12. Motolko, M., Drance, S. M., and Douglas, G. R.: Visual field defects in low tension glaucoma. Arch. Ophthalmol. 100:1074, 1982. 13. Phelps, C. D., Hayreh, S. S., and Montague, P. R.: Visual fields in low-tension glaucoma, primary open-angle glaucoma, and anterior ischemie optic neuropathy. Doc. Ophthalmol. Proc. Ser. 35:113, 1982. 14. Lewis, R. A., Hayreh, S. S., and Phelps, C. D.: Optic disk and visual field correlations in primary open-angle and low-tension glaucoma. Am. J. Oph thalmol. 96:148, 1983. 15. King, D., Drance, S. M., Douglas, G., Schulzer, M., and Wijsman, K.: Comparison of visual field defects in normal-tension glaucoma and high-ten sion glaucoma. Am. J. Ophthalmol. 101:204, 1986. 16. Aulhorn, E.: Visual field defects in chronic glaucoma. In Heilman, K., and Richardson, K. T. (eds.): Glaucoma. Conceptions of a Disease. Pathogenesis, Diagnosis, Therapy. Philadelphia, W. B. Saunders, 1968, p. 157. 17. Caprioli, J., and Spaeth, G. L.: Comparison of the optic nerve head in high- and low-tension glau coma. Arch. Ophthalmol. 103:1145, 1985.
763
18. Geijssen, H. C : Visual field examination. In Geijssen, H. C. (ed.): Studies on Normal Pressure Glaucoma. Amstelveen, Kugler Publications, 1991, p. 153. 19. Paul, A. A., Cohn, H., and Weber, J.: Pattern standard deviation in low vs. high tension glaucoma. ARVO abstracts. Supplement to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1992, p. 1382. 20. Furuno, F., and Matsuo, H.: Early stage pro gression in glaucomatous visual field changes. Doc. Ophthalmol. Proc. Ser. 19:247, 1979. 21. Nicholas, S. P., and Werner, E. B.: Location of early glaucomatous visual field defects. Can. J. Oph thalmol. 15:131, 1980. 22. Hart, W. M., and Becker, B.: The onset and evolution of glaucomatous visual field defects. Oph thalmology 89:268, 1982. 23. Heijl, A., and Lundquist, L.: The frequency distribution of earliest glaucomatous visual field de fects documented by automatic perimetry. Acta Oph thalmol. 62:658, 1983. 24. Harrington, D. O., and Drake, M. V.: Glauco ma. In The Visual Fields. Text and Atlas of Clinical Perimetry, ed. 6. St. Louis, C. V. Mosby, 1991, p. 79. 25. Miller, G. B., and Smith, J. L.: Ischemie optic neuropathy. Am. J. Ophthalmol. 62:103, 1966. 26. Boghen, D. R., and Glaser, J. S.: Ischemie optic neuropathy. Brain 98:689, 1975. 27. Zeiter, J. H., Shin, D. H., and Baek, N. H.: Visual field defects in diabetic patients with primary open-angle glaucoma. Am. J. Ophthalmol. 111:581, 1991. 28. Zeiter, J. H., Shin, D. H., Baek, N. H., and Lee, M. K.: Patterns of visual field defects and disc rim loss in glaucoma patients with and without diabetes mellitus. ARVO abstracts. Supplement to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1991, p. 1105. 29. Gramer, E., and Althaus, G.: Risk-factors for deterioration of visual field defects in primary open angle glaucoma patients. ARVO abstracts. Supple ment to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1992, p. 1278. 30. Minckler, D. S., and Spaeth, G. L.: Optic nerve damage in glaucoma. Surv. Ophthalmol. 26:128, 1981. 31. Hayreh, S, S.: Inter-individual variation in blood supply of the optic nerve head. Doc. Ophthal mol. 59:217, 1985. 32. Anderson, D. R.: Glaucoma. The damage caused by pressure. XLVI Edward Jackson Memorial Lecture. Am. J. Ophthalmol. 108:485, 1989. 33. Phelps, C. D., and Corbett, J. T.: Migraine and low tension glaucoma. Invest. Ophthalmol. Vis. Sci. 26:1105, 1985. 34. Flammer, J., Guthauser, U., and Mahler, F.: Do ocular vasospasms help cause low tension glaucoma? Doc. Ophthalmol. Proc. Ser. 49:397, 1987. 35. Guthauser, U., Flammer, J., and Mahler, F.: The relationship between digital and ocular vasospasm. Graefes Arch. Clin. Exp. Ophthalmol. 266:223, 1988.
We compared the automated visual field test results of 24 patients with normal-tension glaucoma and 24 patients with high-tension glaucoma who were closely matched for the amount of visual field loss to determine any differences in the characteristics of visual field defects between the two groups. Patients were matched with a maximum allowable difference in mean deviation of 0.3 dB. Although the normal-tension group had a greater amount of focal visual field loss (pattern standard deviation), the difference was not statistically significant (P = .628). Additionally, there was no statistically significant difference in the amount of diffuse or focal visual field damage in the superior hemifields between the two groups; however, the patients with normal-tension glaucoma had a significantly greater amount of localized visual field loss in the inferior hemifield than the patients with high-tension glaucoma (P = .015). Our data support the hypothesis that a vascular mechanism may have a greater role in the pathogenesis of optic nerve damage and visual field loss in patients with normal-tension glaucoma than in patients with high-tension glaucoma.
3 I N C E the initial description of a patient with glaucomatous optic nerve damage and visual field loss in the absence of increased intraocular pressure by Von Graefe 1 in 1857, the relation ship between intraocular pressure, optic nerve damage, and visual field loss has aroused con-
Accepted for publication Sept. 16, 1992. From the Kresge Eye Institute, Wayne State University School of Medicine, Detroit, Michigan. This study was presented in part at the annual meeting of the Associa tion for Research in Vision and Ophthalmology, Sarasota, Florida, May 8, 1992. Reprint requests to Dong H. Shin, M.D., Kresge Eye Institute, 4717 St. Antoine Blvd., Detroit, MI 48201. 758
troversy. A better understanding of the patho genesis of high-tension and normal-tension glaucoma might result if one could show differ ences in the characteristics of optic nerve dam age and visual field defects between these two types. Unfortunately, there has been little agreement in published reports as to the simi larities and differences of these characteristics in patients with normal-tension glaucoma and patients with high-tension glaucoma. Some investigators have shown that visual field defects in patients with normal-tension glaucoma as compared to patients with hightension glaucoma tend to be of sudden onset, 2 closer to fixation,2'5 deeper and steeper, 25 more localized or less diffuse, 69 and more common in the superior hemifield.10·11 However, other in vestigators have found no differences between the visual field defects in patients with normaltension glaucoma and patients with high-ten sion glaucoma. 1215 The reasons for such contro versy appear to be caused by differences in study designs, including different types of perimetry, various definitions of normal-tension glaucoma regarding the highest recorded intra ocular pressure, different stages of optic nerve damage and visual field loss, and different methods used in statistical analysis of the data. We designed a study to compare the visual field defects of patients with normal-tension glaucoma and patients with high-tension glau coma by using strict criteria for defining nor mal-tension glaucoma and closely matching both groups with respect to the amount of visual field loss.
Patients and Methods We identified 24 patients with normal-ten sion glaucoma. Each patient had glaucomatous optic nerve damage, open anterior chamber angles on gonioscopy, a maximum recorded
©AMERICAN JOURNAL OF OPHTHALMOLOGY 114:758-763, DECEMBER, 1992
Vol. 114, No. 6
Visual Field Defects and Glaucoma
intraocular pressure that did not exceed 22 mm Hg by Goldmann applanation tonometry, a diurnal intraocular pressure curve done with out therapy with a minimum of 12 measure ments over at least a four-week period, a visual acuity of 20/30 or better, a pupil diameter of at least 3 mm, and a reliable visual field test (fixation losses less than 20% and false-positive and false-negative errors less than 33% accord ing to the manufacturer's criteria [Allergan Humphrey, San Leandro, California]) with re producible Aulhorn's stage 1, 2, or 3 visual field defects16 in at least one eye. In stage 1, relative defects are seen in the area of Bjerrum's scotoma. Stage 2 shows deep defects in the area of Bjerrum's scotoma, which form an incomplete arcuate scotoma, not yet connected to the blind spot. In stage 3, the arcuate defect is connected to the blind spot often with breakthrough into the nasal periphery. Cup/disk ratio was deter mined by examination of the dilated fundus performed by one glaucoma subspecialist by using contour to establish the vertical cup/disk ratio. Fundus examination for the cup/disk ratio determination was binocular by using a slit-lamp biomicroscope and a contact lens, Hruby lens, or a 78-diopter double aspheric lens. When binocular examination was not pos sible because of a poorly dilatable pupil, how ever, direct ophthalmoscopy with monocular clues such as disk vessel bending was used for determination of the vertical cup/disk ratio by contour. Patients were excluded if they had a history of trauma or any ocular diseases known to affect the visual field, such as diabetic retinopathy, macular degeneration, or a vascular occlusion. Also, patients with optic neuropathy other than glaucoma, local disk malformations, or neurologic or orbital abnormalities con firmed by computed tomography or serologie studies when applicable were excluded. Pa tients with advanced visual field loss (mean deviation s 20 dB) were also excluded from the study. If a patient had visual field loss in both eyes, then one eye was randomly selected by the toss of a coin. The 24 Humphrey visual fields from the 24 patients with normal-tension glaucoma were matched closely for extent of visual field dam age to 24 Humphrey visual fields of 24 patients with high-tension glaucoma, with a maximum allowable difference in mean deviation of 0.3 dB. The visual fields of patients with hightension glaucoma were selected from 162 visual field examinations of 62 patients with hightension glaucoma. All patients included in the
759
high-tension group had primary open-angle glaucoma with intraocular pressures of 26 mm Hg or greater before the initiation of therapy, glaucomatous optic nerve damage, open angles with mild to moderate pigment in the trabecular meshwork on gonioscopy, and reliable Humphrey visual field tests with Aulhorn's stage 1, 2, or 3 visual field defects. Patients with high-tension glaucoma with a visual acuity worse than 20/30, a pupil diameter less than 3 mm, or any other ocular condition known to influence the visual field were excluded. All visual field examinations were performed by trained technicians with a visual field analyzer (Allergan Humphrey 24-2 Threshold Program using Stimulus III, Allergan Humphrey, San Leandro, California). To determine if there was a difference in the frequency of visual field defects close to fixa tion between patients with normal-tension glaucoma and patients with high-tension glau coma, we counted the number of patients in each group with scotomas extending to within 6 degrees of central fixation. Patients were count ed if any one of the central four threshold values on the total deviation plot decreased 10 dB or more below age-corrected normal values, or if any of the central four threshold values along with contiguous values decreased 5 dB or more below age-corrected normal values. Because the normal-tension and high-tension groups of patients were closely matched for loss of overall sensitivity (mean deviation), we were able to determine a difference between the two groups with respect to localized visual field loss by comparing the pattern standard deviation. We then calculated the mean deviations and pattern standard deviations for the superior and inferior hemifields of all 48 patients. We attempted to find whether the superior or the inferior hemifield had more visual field loss both within each group (normal-tension and high-tension groups, respectively) and between each group (normal-tension vs high-tension groups). First, we compared the mean devia tions of the superior to the inferior hemifields to determine any differences in diffuse visual field loss within the normal-tension and hightension groups, respectively. Then we com pared the pattern standard deviations of the superior to the inferior hemifields to determine any differences in focal visual field loss within each of the two groups. Finally, we compared the superior and the inferior hemifields, respec tively, of the normal-tension group to those of the high-tension group with respect to both
760
AMERICAN JOURNAL OF OPHTHALMOLOGY
mean deviation and pattern standard deviation. We used the Wilcoxon signed rank test to compare mean deviation (matched) between patients with normal-tension glaucoma and pa tients with high-tension glaucoma. We used chi-square analysis with continuity correction to compare gender, race, number of patients in each group with vascular risk factors, and num ber of patients in each group with scotomas extending to within 6 degrees of central fixa tion. The Mann-Whitney U test was used for other unpaired statistical comparisons.
Results The mean deviations, cup/disk ratios, and other demographic data for the normal-tension and high-tension glaucoma groups were ana lyzed (Table 1). As expected, the mean devia tion values that were matched to within 0.3 dB in the selection process were almost identical in the two groups. As reported by other investiga tors, the normal-tension group had a younger mean age2 and a larger mean cup/disk ratio for a given amount of visual field loss 2,317 than the high-tension group. However, the differences were not statistically significant (P = .170 and P = .350, respectively). There were no significant differences in the gender, race, or systemic vascular risk factors between the two groups. We studied the differences in the number of patients with defects extending to within 6 TABLE 1 BACKGROUND VARIABLES FOR NORMAL-TENSION AND HIGH-TENSION GROUPS
VARIABLE
Mean deviation (dB) Cup/disk ratio Age (yrs) Gender ratio (male:female) Race ratio (black:white) Vascular risk factors*
PATIENTS WITH NORMAL-TENSION GLAUCOMA (N=24)
PATIENTS WITH HIGH-TENSION GLAUCOMA (N=24) P VALUE
9.16 ± 3.93 0.80 ± 0.10 67.5 ± 12.1 10:14
9.11 ± 3.89 0.75 ±0.15 71.6 ± 13.0 9:15
.475 .350 .170 .999
14:10
16:8
.766
9 (38%)
10 (42%)
.999
'Number of patients with cardiovascular disease or hyperten sion or both, or cardiovascular disease or diabetes mellitus, or both.
December, 1992
degrees of central fixation, in the amount of localized visual field loss (pattern standard deviation), and in the amounts of generalized and localized visual field loss in the upper and lower hemifields of the normal-tension and high-tension glaucoma groups (Table 2). We found no statistically significant difference in the number of patients with scotomas falling within 6 degrees of central fixation (P = .555). Although the normal-tension glaucoma group had a larger pattern standard deviation than the high-tension glaucoma group overall (7.37 ± 2.94 dB vs 6.94 ± 3.08 dB), this difference was not statistically significant (P = .628). When comparing the amount of superior to inferior hemifield involvement within the hightension glaucoma group, there was significantly more involvement of the superior than the inferior hemifield with respect to both mean deviation and pattern standard deviation (P = .004 and P = .0003, respectively). However, in the normal-tension glaucoma group, there was a tendency toward equal involvement of the two hemifields with no significant differences noted in either the mean deviation or the pat tern standard deviation (P = .918 and P = .350, TABLE 2 COMPARISON OF MEAN DEVIATION AND PATTERN STANDARD DEVIATION FOR NORMAL-TENSION AND HIGH-TENSION GROUPS PATIENTS WITH PATIENTS WITH NORMAL-TENSION HIGH-TENSION GLAUCOMA GLAUCOMA (N-24) (N=24) P VALUE
13 (54%) Scotomas within 16 (67%) .555 6 degrees of fixation* Pattern standard 7.37 ± 2.94 6.94 ± 3.08 .628 deviation (dB) Superior mean 10.16 ± 5.86* 12.58 ± 6.43* .180 deviation (dB) Inferior mean 9.55 ± 4.65* 7.52 ± 3.10* .167 deviation (dB) Superior pattern 7.28 ± 4.77* 9.29 ± 5.88" .177 standard deviation (dB) Inferior pattern 6.65 + 4.25* 4.17 + 1.86" .015 standard deviation (dB) »Number of patients with scotomas extending to within 6 degrees of central fixation. *P = .918. *P = .004. S P = .350. 'P = .0003.
Vol. 114, No. 6
Visual Field Defects and Glaucoma
respectively). There were no statistically signif icant differences between the normal-tension and high-tension glaucoma groups when com paring the amount of generalized visual field loss in the superior hemifields or the inferior hemifields (P = .180 and P = .167, respective ly). However, there was significantly more focal visual field loss located in the inferior hemifield in the normal-tension group than in the hightension group (P = .015), although there was no difference between the two groups in the amount of focal visual field loss located in the superior hemifield (P = .177).
Discussion Normal-tension glaucoma has been de scribed as having typical glaucomatous optic disk changes with resulting visual field loss, open angles, and intraocular pressures within the statistically normal range. Since the initial description of this entity, many questions have been raised concerning the mechanisms of op tic nerve damage in glaucoma, and whether there are different mechanisms involved in the pathogenesis of normal-tension vs high-ten sion glaucoma. For more than a decade, authors have been studying both similarities and dif ferences in the visual field defects between patients with normal-tension glaucoma and patients with high-tension glaucoma.2"16 A dif ferent pattern of visual field loss in these two populations raises the possibility of a different mechanism of optic nerve damage and might suggest a difference in the relative roles that vascular ischemia and intraocular pressure have in the pathogenesis of normal-tension and high-tension glaucoma. Unfortunately, there has been little agreement on what types of differences in visual field loss exist between these two groups of patients. Some investigators found that visual field defects were close to fixation much more com monly in patients with low-tension glaucoma than patients with high-tension glaucoma and that because of this, patients with low-tension glaucoma lost visual acuity earlier and more frequently than patients with high-tension glaucoma.2'4 Conversely, many other investiga tors have found no significant difference in the number of patients with normal-tension glau coma and patients with high-tension glaucoma with visual field defects near fixation.11·121518 In
761
our study, unlike the previously mentioned studies, the visual fields were closely matched for the amount of visual field loss in each group, and we found no significant difference in the number of patients with normal-tension glaucoma and patients with high-tension glau coma with scotomas extending to within 6 de grees of central fixation. Recently, Chauhan and associates 9 found that individuals with normal-tension glaucoma had greater areas of normal sensitivity and, there fore, more localized visual field defects than individuals with high-tension glaucoma. Previ ously, two separate studies had shown that patients with glaucoma with generalized visual field loss had a higher mean intraocular pres sure than those with more localized visual field loss.6·7 We investigated these findings by com paring the pattern standard deviation values (localized visual field loss) between normaltension and high-tension groups that had been closely matched for mean deviation (genera lized visual field loss). Although the normaltension group in our study had a greater amount of localized visual field loss than the high-tension group, this difference was not statistically significant. This finding is similar to that of Paul, Cohn, and Weber,19 who recently found no significant difference in the pattern standard deviation between patients with lowtension glaucoma (intraocular pressure ^ 2 1 mm Hg) and patients with high-tension glauco ma (intraocular pressure s 28 mm Hg) with similar mean deviations. We may not have found a significantly greater amount of localized visual field loss in patients with normal-tension glaucoma because the av erage mean deviation in our patients was high er than that of the patients in previous stud ies,6,7'9 which indicates slightly more advanced visual field loss in our groups. As visual field loss becomes more advanced, any differences in the amount of localized and diffuse damage between normal-tension and high-tension groups may become less. Additionally, the maximum intraocular pressure of the normaltension group and the minimum intraocular pressure of the high-tension group in our study were only 4 mm Hg apart, which may be re sponsible for some overlap of localized and diffuse damage between the two groups. In our study, we found a tendency toward equal involvement of the superior and inferior hemifields in the patients with normal-tension glaucoma. However, in the patients with high-
762
December, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
tension glaucoma, the superior hemifield was remarkably more involved than the inferior hemifield with respect to both generalized and localized visual field loss. This finding that patients with early to moderate primary openangle glaucoma have visual field defects that involve the superior hemifield to a greater ex tent than the inferior hemifield has been well documented. 20 ' 24 Additionally, many investigators have ad dressed the issue of whether there is any differ ence in the distribution of visual field defects over the superior and inferior hemifields be tween patients with normal-tension glaucoma and patients with high-tension glaucoma. 5101318 Although most studies found no differences between the two groups,10,12·'3 Gramer, Althaus and Leydhecker 5 found a preponderance of inferonasal visual field defects in patients with normal-tension glaucoma, whereas Greve and Geijssen11·18 found greater involvement of the superior hemifield in the normal-tension group as compared to the high-tension group. We found no statistically significant difference in the distribution of generalized visual field loss over the superior and inferior hemifields be tween the two groups. However, when we fo cused on localized visual field loss, we found that the normal-tension group had a signifi cantly greater amount of focal visual field de fects in the inferior hemifield. This finding of a preponderance of localized visual field loss in the inferior hemifield of patients with normal-tension glaucoma as com pared to patients with high-tension glaucoma may have important implications as to the rela tive role that vascular ischemia may have in optic nerve damage in patients with normaltension glaucoma. The most frequent visual field defect in nonarteritic anterior ischemie optic neuropathy is an altitudinal defect involv ing the inferior hemifield.13·28-26 Additionally, we have shown in previous studies that patients with primary open-angle glaucoma with diabe tes mellitus develop inferior visual field defects more frequently than patients with primary open-angle glaucoma without diabetes melli tus, speculating the possibility of a vascular contribution at the level of the optic nerve head in patients with glaucoma and diabetes. 27 · 28 Re cently, Gramer and Althaus 29 found that pa tients with primary open-angle glaucoma with low systolic blood pressure had more severe visual field defects in the lower hemifield com pared to the upper hemifield. Additionally, they found that the visual field defects in these patients with lower hemifield involvement
were twice as likely to deteriorate as those in patients with superior visual field loss.29 Several authors have suggested that vascular factors in addition to intraocular pressure have an important role in the pathogenesis of glau coma. 2932 Also, there appears to be an associa tion between normal-tension glaucoma and migraine headaches 33 as well as ocular vasospasm,84·35 which strongly suggests a possible vascular origin in this disease. Our findings of a greater amount of localized visual field loss in the inferior hemifield of patients with normaltension glaucoma as compared to patients with high-tension glaucoma is one more observation in support of the hypothesis that vascular ische mia may have a larger role in the pathogenesis of optic nerve damage and visual field loss in normal-tension glaucoma than in high-tension glaucoma. Although we found more patients with nor mal-tension glaucoma with localized visual field loss in the inferior hemifield, many pa tients in the high-tension group also had some localized defects in the inferior half of the visual field. This finding reminds us that there is a spectrum of visual field defects involving varying amounts of the inferior and superior hemifields between the two groups. Additional ly, when defining the two groups on the basis of intraocular pressure alone, one cannot control for other possible risk factors in the production of glaucoma. Also, even with the use of diurnal pressure curves in defining normal-tension glaucoma, there may still be some overlap of intraocular pressure between the two groups. These unavoidable circumstances along with variations in study designs and different stages of optic nerve damage and visual field loss help explain why there has been little agreement on the differences in visual field defects between patients with normal-tension glaucoma and pa tients with high-tension glaucoma. Additional studies using patients closely matched for the amount of visual field loss will help further define differences in visual field defects between patients with normal-tension glaucoma and patients with high-tension glau coma and provide more insight into the relative roles of intraocular pressure and vascular ische mia in the production of glaucoma.
References 1. Von Graefe, A.: Über die Iridectomie bei Glau coma und über den glaucomatosen Frozess. Graefes Arch. Clin. Exp. Ophthalmol. 3:456, 1857.
Vol. 114, No. 6
Visual Field Defects and Glaucoma
2. Levene, R. Z.: Low tension glaucoma. A critical review and new material. Surv. Ophthalmol. 24:621, 1980. 3. Hitchings, R. A., and Anderton, S. A.: A com parative study of visual field defects seen in patients with low-tension glaucoma and chronic simple glau coma. Br. J. Ophthalmol. 67:818, 1983. 4. Caprioli, J., and Spaeth, G. L.: Comparison of visual field defects in the low-tension glaucomas with those in the high-tension glaucomas. Am. J. Ophthalmol. 97:730, 1984. 5. Gramer, E., Althaus, G., and Leydhecker, W.: Localization and depth of glaucomatous visual field defects in relation to the size of the neuroretinal rim area of the disk in low tension glaucoma, glaucoma simplex, and pigmentary glaucoma. Clinical study with the Octopus 201 perimeter and the optic nerve head analyser. Klin. Monatsbl. Augenheilkd. 189: 190, 1986. 6. Glowazki, A., and Flammer, J.: Besteht ein Un terschied bei Patienten mit eher lokalisiertem und solchen mit eher diffusem Gesichtsfeldschaden? Klin. Monatsbl. Augenheilkd. 190:301, 1987. 7. Caprioli, J., Sears, M., and Miller, J. M.: Patterns of early visual field loss in open-angle glaucoma. Am. J. Ophthalmol. 103:512, 1987. 8. Drance, S. M., Douglas, G. R., Airaksinen, P. J., Schulzer, M., and Hitchings, R. A.: Diffuse visual field loss in chronic open-angle and low-tension glaucoma. Am. J. Ophthalmol. 104:577, 1987. 9. Chauhan, B. C , Drance, S. M., Douglas, G. R., and Johnson, C. A.: Visual field damage in normaltension glaucoma and high-tension glaucoma. Am. J. Ophthalmol. 108:636, 1989. 10. Drance, S. M.: The visual field of low tension glaucoma and shock-induced optic neuropathy. Arch. Ophthalmol. 95:1359, 1977. 11. Greve, E. L., and Geijssen, H. C : Comparison of glaucomatous visual field defects in patients with high and with low intraocular pressures. Doc. Oph thalmol. Proc. Ser. 35:101, 1982. 12. Motolko, M., Drance, S. M., and Douglas, G. R.: Visual field defects in low tension glaucoma. Arch. Ophthalmol. 100:1074, 1982. 13. Phelps, C. D., Hayreh, S. S., and Montague, P. R.: Visual fields in low-tension glaucoma, primary open-angle glaucoma, and anterior ischemie optic neuropathy. Doc. Ophthalmol. Proc. Ser. 35:113, 1982. 14. Lewis, R. A., Hayreh, S. S., and Phelps, C. D.: Optic disk and visual field correlations in primary open-angle and low-tension glaucoma. Am. J. Oph thalmol. 96:148, 1983. 15. King, D., Drance, S. M., Douglas, G., Schulzer, M., and Wijsman, K.: Comparison of visual field defects in normal-tension glaucoma and high-ten sion glaucoma. Am. J. Ophthalmol. 101:204, 1986. 16. Aulhorn, E.: Visual field defects in chronic glaucoma. In Heilman, K., and Richardson, K. T. (eds.): Glaucoma. Conceptions of a Disease. Pathogenesis, Diagnosis, Therapy. Philadelphia, W. B. Saunders, 1968, p. 157. 17. Caprioli, J., and Spaeth, G. L.: Comparison of the optic nerve head in high- and low-tension glau coma. Arch. Ophthalmol. 103:1145, 1985.
763
18. Geijssen, H. C : Visual field examination. In Geijssen, H. C. (ed.): Studies on Normal Pressure Glaucoma. Amstelveen, Kugler Publications, 1991, p. 153. 19. Paul, A. A., Cohn, H., and Weber, J.: Pattern standard deviation in low vs. high tension glaucoma. ARVO abstracts. Supplement to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1992, p. 1382. 20. Furuno, F., and Matsuo, H.: Early stage pro gression in glaucomatous visual field changes. Doc. Ophthalmol. Proc. Ser. 19:247, 1979. 21. Nicholas, S. P., and Werner, E. B.: Location of early glaucomatous visual field defects. Can. J. Oph thalmol. 15:131, 1980. 22. Hart, W. M., and Becker, B.: The onset and evolution of glaucomatous visual field defects. Oph thalmology 89:268, 1982. 23. Heijl, A., and Lundquist, L.: The frequency distribution of earliest glaucomatous visual field de fects documented by automatic perimetry. Acta Oph thalmol. 62:658, 1983. 24. Harrington, D. O., and Drake, M. V.: Glauco ma. In The Visual Fields. Text and Atlas of Clinical Perimetry, ed. 6. St. Louis, C. V. Mosby, 1991, p. 79. 25. Miller, G. B., and Smith, J. L.: Ischemie optic neuropathy. Am. J. Ophthalmol. 62:103, 1966. 26. Boghen, D. R., and Glaser, J. S.: Ischemie optic neuropathy. Brain 98:689, 1975. 27. Zeiter, J. H., Shin, D. H., and Baek, N. H.: Visual field defects in diabetic patients with primary open-angle glaucoma. Am. J. Ophthalmol. 111:581, 1991. 28. Zeiter, J. H., Shin, D. H., Baek, N. H., and Lee, M. K.: Patterns of visual field defects and disc rim loss in glaucoma patients with and without diabetes mellitus. ARVO abstracts. Supplement to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1991, p. 1105. 29. Gramer, E., and Althaus, G.: Risk-factors for deterioration of visual field defects in primary open angle glaucoma patients. ARVO abstracts. Supple ment to Invest. Ophthalmol. Vis. Sci. Philadelphia, J. B. Lippincott, 1992, p. 1278. 30. Minckler, D. S., and Spaeth, G. L.: Optic nerve damage in glaucoma. Surv. Ophthalmol. 26:128, 1981. 31. Hayreh, S, S.: Inter-individual variation in blood supply of the optic nerve head. Doc. Ophthal mol. 59:217, 1985. 32. Anderson, D. R.: Glaucoma. The damage caused by pressure. XLVI Edward Jackson Memorial Lecture. Am. J. Ophthalmol. 108:485, 1989. 33. Phelps, C. D., and Corbett, J. T.: Migraine and low tension glaucoma. Invest. Ophthalmol. Vis. Sci. 26:1105, 1985. 34. Flammer, J., Guthauser, U., and Mahler, F.: Do ocular vasospasms help cause low tension glaucoma? Doc. Ophthalmol. Proc. Ser. 49:397, 1987. 35. Guthauser, U., Flammer, J., and Mahler, F.: The relationship between digital and ocular vasospasm. Graefes Arch. Clin. Exp. Ophthalmol. 266:223, 1988.
