- Email: [email protected]
Factors associated with optic disc hemorrhages in glaucoma
Factors Associated with Optic Disc Hemorrhages in Glaucoma Adael S. Soares, MD,1,3 Paul H. Artes, PhD,1 Pantelis Andreou, PhD,2 Raymond P. Leblanc, MD,1 Balwantray C. Chauhan, PhD,1 Marcelo T. Nicolela, MD1 Purpose: To evaluate factors associated with optic disc hemorrhages (ODHs) in patients with open-angle glaucoma. Design: Cohort study. Participants: One hundred thirty-seven patients with open-angle glaucoma, with an initial mean age of 60.0 years (standard deviation [SD], 11.0 years) followed up for a mean of 9.5 years (SD, 5.0 years). Methods: The association between ODHs and various patient-related variables (diabetes, systemic hypertension, heart disease, hypercholesterolemia, migraine, hypothyroidism, use of platelet antiaggregant agents) and eye-related variables (mean and range of intraocular pressure, refractive error, and severity of disease) was investigated using multivariate time-to-event analyses in patients with open-angle glaucoma. To determine the influence of the level of intraocular pressure (IOP) on the occurrence of ODHs, we compared the IOP at the time the first ODH was noted with the mean IOP of the previous 3 visits. As a control, a similar analysis was performed on the same eyes using a randomly selected visit before the occurrence of the first ODH. Main Outcome Measurements: Patient-related and eye-related variables associated with ODHs and the comparative level of IOP in which these ODHs were observed. Results: Fifty eyes of 38 patients (28%) had 1 or more ODHs during the follow-up period. The ODHs were associated with presence of diabetes (hazard ratio, 4.43; 95% confidence interval [CI], 1.8 –10.50; P ⫽ 0.001) and use of aspirin (hazard ratio, 2.30; 95% CI, 1.2– 4.6; P ⫽ 0.019). The IOP at the time of the first ODH was, on average, 1.4 mmHg lower than the mean IOP of the 3 previous visits (95% CI, ⫺2.2 to ⫺0.6 mmHg; P⬍0.001), whereas in the control analysis, the respective value was 0.3 mmHg lower (95% CI, ⫺1.0 to 0.5; P ⫽ 0.410). Conclusions: Optic disc hemorrhages were associated with diabetes and aspirin use and were observed at relatively lower IOP during follow-up. Ophthalmology 2004;111:1653–1657 © 2004 by the American Academy of Ophthalmology.
Glaucoma is characterized by progressive loss of retinal ganglion cells and their axons, leading to the characteristic changes of the optic disc and loss of visual field. Elevated intraocular pressure (IOP) is the most important risk factor for the development of glaucoma.1,2 Other significant risk factors include increasing age,3 black race,3 family history of glaucoma,4,5 and myopia.6 Although some studies found an association between glaucoma and certain diseases such
Originally received: October 20, 2003. Accepted: March 1, 2004.
Manuscript no. 230698.
1
Department of Ophthalmology and Visual Sciences, Dalhousie University, Halifax, Canada.
2
Department of Community Health and Epidemiology, Dalhousie University, Halifax, Canada.
3
Department of Ophthalmology, Federal University of Sa˜o Paulo, Sa˜o Paulo, Brazil. Presented at: Association for Research in Vision and Ophthalmology Annual Meeting, May, 2003; Fort Lauderdale, Florida. Supported by the Canadian Institutes of Health Research, Ottawa, Canada (grant no.: MOP-11357). Correspondence to Marcelo T. Nicolela, MD, Department of Ophthalmology, Dalhousie University, 1278 Tower Road, Halifax, Nova Scotia, Canada B3H 2Y9. E-mail: [email protected].
© 2004 by the American Academy of Ophthalmology Published by Elsevier Inc.
as diabetes,7 systemic hypertension,8 hypothyroidism,9 and migraine,10 others could not confirm these findings.11–14 Optic disc hemorrhages (ODHs) frequently are observed in patients with glaucoma,15 but they also can occur infrequently in healthy persons.16 Optic disc hemorrhages usually persist from 2 to 35 weeks, with an average of 11 weeks.17 They are observed more often in patients with normal-tension glaucoma as opposed to those patients with elevated IOP (20%17 to 46%18 and 2%19 to 37%,20 respectively). Optic disc hemorrhages frequently are associated with glaucoma progression and are considered a risk factor for the deterioration of the disease.15 After a mean follow-up of 9 years, Rasker et al21 reported that 80% to 89% of glaucoma patients with ODHs progressed, as opposed to 32% of patients who did not have ODHs. Two recently published randomized clinical trials, the Collaborative Normal Tension Glaucoma Study22 and the Early Manifest Glaucoma Trial,23 also showed higher rates of progression in glaucoma patients with ODHs at baseline or in patients with more frequent ODHs during follow-up, respectively. Despite its importance in glaucoma, the association between ODHs and systemic diseases, the use of anticoagulant medication, and intrinsic eye factors such as IOP, refractive ISSN 0161-6420/04/$–see front matter doi:10.1016/j.ophtha.2004.03.023
1653
Ophthalmology Volume 111, Number 9, September 2004 error, and glaucoma severity has not been fully elucidated. We have noted clinically that treated patients who have ODHs tend to have well-controlled IOP, leading to the hypothesis that these hemorrhages are observed more commonly at lower IOP levels. The purpose of this cohort study was to test this hypothesis and to evaluate additional factors associated with ODHs in patients with open-angle glaucoma followed up for several years.
Patients and Methods Patients From 1992 through 1995, patients with open-glaucoma were enrolled into a prospective study at our center, which was approved by the Ethics Committee of the Queen Elizabeth II Health Sciences Centre. Inclusion criteria were: (1) best-corrected visual acuity of at least 20/30; (2) glaucomatous optic disc damage, such as notching or progressive thinning of the neuroretinal rim; (3) mild to moderate visual field defect, with mean deviation better than ⫺10 dB using the 30-to-2 full-threshold strategy program of the Humphrey Field Analyzer (Carl Zeiss Meditech, Dublin, CA); and (4) the presence of open angles by gonioscopy. Exclusion criteria were: (1) concomitant ocular disease; (2) refractive error exceeding 6 diopters (D) equivalent sphere or 2.5 D of astigmatism; (3) aphakia; (4) systemic disease or systemic medication known to affect the visual field. Patients with diabetes and systemic hypertension without obvious diabetic or hypertensive retinopathy could be enrolled into the study.
Procedures and Study Variables For this present study, we analyzed the data starting with the patient’s first visit to the glaucoma clinic, which could have occurred before enrollment. Most of the patients were followed up during the entire period by one of the authors (RPL). During each visit, the presence of ODHs was investigated, and, if present, their location was depicted in a diagram of the optic disc. An ODH was considered new if it occurred at a different location from a previous one, or if it recurred at least 3 months later if in the same location.17 Information regarding the following variables was collected during each study visit: (1) patient-related variables: presence of diabetes, systemic hypertension, heart disease, hypercholesterolemia, migraine, and use of anticoagulant medication (platelet antiaggregant and cumarinic agents); (2) eye-related variables: IOP, ocular refractive error, and severity of visual field and optic disc damage, as defined below. Information on systemic diseases was based on the patient’s self-report and was confirmed according to the class of medication used. All data from visits during the first 3 months after laser or surgical procedure for glaucoma were excluded from the analyses. Because all patients taking platelet antiaggregant medication were using aspirin, the agent name instead of the class of the systemic medication is used herein. After enrollment, patients were examined every 4 to 6 months. At every visit, a standard automated visual field examination was performed, using the Humphrey Field Analyzer, program 30-to-2, full-threshold strategy. At the same visits, a computerized examination of the optic disc using scanning laser tomography (Heidelberg Retina Tomograph; Heidelberg Engineering, Dossenheim, Germany) also was performed. For our analyses, the severity of the visual field damage was expressed as the mean deviation, whereas the severity of the optic disc damage was expressed as the rim area deviation. The latter variable measures the number of standard
1654
deviations by which the log-transformed area of the neuroretinal rim deviates from the expected optic disc size-corrected normal value. Visual field and optic disc status were interpolated using linear regression for patients who had additional clinic visits at which visual field or scanning laser tomography tests, or both, were not performed.
Statistical Analysis Survival time distribution function analyses using Wilcoxon and log-rank tests (multivariate analysis, forward stepwise approach) were used to identify variables associated with the elapsed follow-up time before the first ODH (time to hemorrhage). For the patient-related variables, the date of the first ODH was used, whether it occurred in the right or left eye. For eye-related variables, right and left eyes of each patient were treated as independent observations while adjusting for the effect of those patientrelated variables that had been found significant. Variables identified with P⬍0.1 then were included in a multivariate Cox’s proportional hazard model. To investigate the hypothesis that ODHs are observed at lower levels of IOP, we calculated the mean differences in the IOP between the visit in which the first ODH was observed and the average IOP of the respective previous 3 visits. As a control analysis, we calculated the mean difference in the IOP between a randomly selected visit before the first ODH and the mean IOP of the previous 3 visits on the same eyes. We chose to compare the IOP at the time of the first ODH with the previous IOP measurements because the detection of an ODH could lead the attending physician to initiate a more aggressive IOP-lowering therapy, therefore affecting the IOP on subsequent visits. The statistical analyses were performed using SPSS software, version 10.1 (SPSS, Inc., Chicago, IL), and SAS software, version 8 (SAS Institute, Cary, NC). Statistical significance was assumed at P⬍0.05.
Results Our sample contained 137 open-angle glaucoma patients, who had a total of 3612 visits for this study. The mean age at the first visit was 60.0 years (standard deviation, 11.0 years; range, 34 – 81 years), whereas the mean follow-up was 9.5 years (standard deviation, 5.0 years). Demographic and baseline data of the study patients are shown in Table 1. Optic disc hemorrhages were observed in 38 patients (28%), of whom 12 patients (32%) had bilateral ODHs. Twenty eyes (40%) had more than 1 ODH (range, 2– 8 ODHs). In total, 98 ODH episodes were observed, with 61 (62%) in the inferotemporal quadrant, 32 (33%) in the superotemporal quadrant, 3 (3%) in the nasosuperior quadrant, and 2 (2%) in the inferotemporal quadrant of the optic disc. In a univariate analyses, statistically significant differences between patients with and without ODH were observed in the proportion of patients using aspirin (P ⫽ 0.009) and in follow-up time (P ⫽ 0.030), although the number of visits per year of follow-up was similar in the 2 groups (P ⫽ 0.682; Table 2). Of the 33 patients using aspirin, 30 (91%) were using a dosage of 80 mg/day and 3 (9%) were using 325 mg/day. Because only 3 patients were using cumarinic agents, this variable was not included in the multivariate survival analyses, although 2 of these patients had ODHs during the follow-up. Cox’s proportional hazard model using the variables selected in the survival analyses showed that diabetes and use of aspirin were significantly associated with ODHs, with hazard ratios of 4.43
Soares et al 䡠 Optic Disc Hemorrhages in Open-Angle Glaucoma Table 1. Demographic and Baseline Data of the 137 Patients Male* Female* Diabetes* Systemic hypertension* Heart disease* Hypothyroidism* Migraine* Hypercholesterolemia* Aspirin use* Warfarin use* Refractive error (diopters)† RAD (arbitrary units)† MD (decibels)† Mean IOP during follow-up (mmHg)† Range IOP during follow-up (mmHg)†
65 (47) 72 (53) 17 (12.4) 71 (51.8) 14 (10.2) 14 (10.2) 22 (16.0) 17 (12.4) 33 (24.1) 3 (2.2) ⫺0.7 ⫾ 2.4 ⫺2.64 ⫾ 2.4 ⫺4.5 ⫾ 3.9 17.6 ⫾ 2.7 12.4 ⫾ 6.3
IOP ⫽ intraocular pressure; MD ⫽ mean deviation; RAD ⫽ rim area deviation. *n (%). † Mean ⫾ standard deviation.
(95% confidence interval [CI], 1.80 –10.50) and 2.30 (95% CI, 1.20 – 4.60), respectively (Table 3). The IOP at the time of the first observed ODH was significantly lower than the average IOP of the previous 3 visits (mean difference, ⫺1.4 mmHg; 95% CI, ⫺0.6 to ⫺2.2 mmHg; P⬍0.001). In the control analysis performed with the same eyes, the IOP at the randomly chosen visit before the first observed ODH was not significantly lower than the average IOP at the previous 3 visits (mean difference, ⫺0.3 mmHg; 95% CI, ⫺1.0 to 0.5 mmHg; P ⫽ 0.410).
Table 2. Univariate Analysis of Demographic, Systemic, and Ocular Variables
Male Female Age (yrs) Follow-up (yrs) No. of visits per year Refractive error (diopters) Diabetes Heart disease Systemic hypertension Hypothyroidism Migraine Hypercholesterolemia Aspirin use Cumarinic drugs use RAD (arbitrary units) MD (decibels) Mean IOP (mmHg) Range IOP (mmHg)
Patients with Optic Disc Hemorrhage (n ⴝ 38)
Patients without Optic Disc Hemorrhage (n ⴝ 99)
16 (42) 22 (58) 60.0 ⫾ 10.0 11.0 ⫾ 5.0 3.0 ⫾ 1.0 ⫺0.6 ⫾ 2.4 7 (18) 3 (8) 16 (42) 5 (13) 9 (24) 3 (8) 15 (39) 2 (5) ⫺2.9 ⫾ 2.2 ⫺5.0 ⫾ 4.1 17.4 ⫾ 2.7 12.4 ⫾ 5.0
49 (49) 50 (51) 60.0 ⫾ 11.0 9.0 ⫾ 5.0 3.0 ⫾ 1.0 ⫺1.1 ⫾ 2.4 10 (10) 11 (11) 55 (55) 9 (9) 13 (13) 14 (14) 18 (18) 1 (1) ⫺2.4 ⫾ 2.4 ⫺4.3 ⫾ 3.8 17.6 ⫾ 2.6 12.1 ⫾ 6.1
P Value 0.280 0.637 0.030 0.682 0.301 0.186 0.580 0.129 0.482 0.132 0.321 0.009 0.128 0.102 0.355 0.464 0.710
IOP ⫽ intraocular pressure; MD ⫽ mean deviation; RAD ⫽ rim area deviation. Variables with units: mean ⫾ standard deviation, independent-samples t test; otherwise n (%), chi-square test.
Table 3. Cox’s Proportional Hazard Model
Age Diabetes Aspirin Hypothyroidism Hypercholesterolemia
Hazard Ratio (95% Confidence Interval)
P Value
0.97 (0.93–1.00) 4.43 (1.80–10.50) 2.30 (1.20–4.60) 2.61 (0.96–7.03) 0.34 (0.09–1.18)
0.053 0.001 0.019 0.058 0.091
Variables with P⬍0.1 on Wilcoxon or log-rank survival analyses, or both, were included in the Cox’s proportional hazard model.
Discussion Various mechanisms have been proposed to explain the cause of ODHs in glaucoma. Drance and Begg24 suggested that ODHs result from an ischemic microinfarction at the optic disc, whereas Quigley et al25 suggested that structural changes at the level of the lamina cribrosa may cause mechanical rupture of small blood vessels. Although the precise cause of ODHs is still unknown, their prognostic importance for the progression of glaucoma has been well established.15,18,19,21,22,26,27 It is not clear, however, if the hemorrhages themselves are harmful to glaucoma or whether they are only a sign of an active disease process that will lead to progression. Another possibility explaining the prognostic relevance of ODHs is that they occur more frequently in subgroups of patients with glaucoma who may be at higher risk for progression. Previous studies have suggested that ODHs occur more often in patients with focal optic disc loss, and it is possible that these patients are more prone to progress than patients with other types of optic disc damage.28 Most ODHs occur at the poles of the optic disc, specifically at the superotemporal and inferotemporal locations,16,18,19,27,29 –31 corresponding to areas of the optic disc that are most susceptible to glaucomatous damage.32 In the present study, we also found that most ODHs (95%) were detected in the superotemporal and inferotemporal quadrants of the optic disc. In the present study, the use of aspirin was associated with ODHs. Aspirin prevents thromboxane A2 production by cyclooxygenase inhibition. Thromboxane A2 promotes platelet aggregation, one of the steps in the coagulation cascade.33 Aspirin therefore either could increase the risk of developing ODHs, could lead to larger ODHs that may take longer to absorb and consequently are more likely to be detected during periodic examinations, or both. An alternative explanation is that patients using aspirin are more likely to have vascular diseases that ultimately could be related to the presence of ODH. In a large screening study, Grodum et al34 recently reported a positive association between ODHs and aspirin use. However, that study did not distinguish between patients with glaucoma and healthy subjects. In another screening study, Healey et al16 did not observe a relationship between ODH and aspirin use in either normal subjects or patients with glaucoma. In the present longitu-
1655
Ophthalmology Volume 111, Number 9, September 2004 dinal study, patients were examined frequently, at regular intervals, and over several years. This study, therefore, may be better positioned to unveil risk factors associated with the occurrence of ODHs than a cross-sectional study design. As far as we are aware, this is the first longitudinal study showing an association between aspirin use and ODHs. It is possible that use of other anticoagulant medications, such as cumarinic agents, also is associated to the occurrence of ODHs. This study, however, was underpowered to show any possible association, because only 3 patients were using cumarinic agents (2 of them experienced ODHs during follow-up). In our study, patients with diabetes also had a higher risk of developing ODHs. A number of studies have reported a similar finding,16,27,31,35 whereas others19,34,36 could not confirm such an association. It is plausible that the vasculopathy associated with diabetes37 causes ischemic changes in the optic disc, renders the small optic nerve vessels more vulnerable to mechanical rupture, or both. We excluded patients with obvious diabetic retinopathy, including retinal hemorrhages and exudates. If glaucoma patients with diabetic retinopathy have an even higher incidence of ODHs than those without obvious diabetic retinopathy, the association found in our study between ODHs and diabetes could be even higher. Similar to diabetes, the effect of systemic hypertension on the occurrence of ODHs is controversial. Although some investigators reported a link between ODHs and systemic hypertension,16,36,38 other studies, including ours, did not find such an association.19,34,35 The mean and range of IOP measurements during follow-up were not significantly different between glaucoma patients with and without ODHs. Our study did show, however, that the IOP at the time the first ODH was observed was significantly lower than that of previous visits. This finding was not observed in the control analysis of the same eyes using randomly selected visits without ODHs, confirming our previous clinical impression that ODHs tend to be detected in examinations when the IOP is at a relatively lower level during a patient’s follow-up. We can speculate that a reduction in IOP may cause transient mechanical stresses on the central retinal vein as it exits the eye by forward movement of the lamina cribrosa leading to increased venous resistance and, perhaps, increasing the risk of ODHs. Another possibility is that at lower levels of IOP, the pressure differences across the vessel walls are relatively greater than at higher IOP, which could cause larger and more frequent ODHs in predisposed eyes.39 Although the magnitude of IOP difference between the time the ODH was observed and the previous visits was relatively small, it is possible that the IOP may have been even lower at the time the ODH actually occurred. In conclusion, our study showed that open-angle glaucoma patients with diabetes and those using aspirin had a higher risk of having ODHs. Optic disc hemorrhages were observed more frequently at relatively lower levels of IOP throughout patient follow-up. These findings may be relevant to the management of glaucoma patients, particularly in view of the important association between ODHs and glaucoma deterioration. It is important to emphasize that our study design permits us to reveal only an association
1656
between aspirin use and the detection of ODHs in glaucoma. Based on the evidence thus far, it does not suggest that the use of aspirin is harmful in glaucoma. Further work should be undertaken to evaluate if aspirin use, as well as presence of diabetes, predisposes glaucoma patients, either directly or indirectly, to a higher risk of glaucoma progression. Acknowledgments. The authors thank Dr David F. GarwayHeath (Moorfields Eye Hospital, London, England) for providing part of the scanning laser tomography normal database used to calculate rim area deviation in our analysis.
References 1. Leske MC, Connell AM, Wu SY, et al. Risk factors for open-angle glaucoma. The Barbados Eye Study. Arch Ophthalmol 1995;113:918 –24. 2. Sommer A, Tielsch JM, Katz J, et al. Relationship between intraocular pressure and primary open angle glaucoma among white and black Americans. The Baltimore Eye Survey. Arch Ophthalmol 1991;109:1090 –5. 3. Tielsch JM, Sommer A, Katz J, et al. Racial variations in the prevalence of primary open-angle glaucoma. The Baltimore Eye Survey. JAMA 1991;266:369 –74. 4. Tielsch JM, Katz J, Sommer A, et al. Family history and risk of primary open angle glaucoma. The Baltimore Eye Survey. Arch Ophthalmol 1994;112:69 –73. 5. Wolfs RC, Klaver CC, Ramrattan RS, et al. Genetic risk of primary open-angle glaucoma. Population-based familial aggregation study. Arch Ophthalmol 1998;116:1640 –5. 6. Mitchell P, Hourihan F, Sandbach J, Wang JJ. The relationship between glaucoma and myopia: the Blue Mountains Eye Study. Ophthalmology 1999;106:2010 –5. 7. Mitchell P, Smith W, Chey T, Healey PR. Open-angle glaucoma and diabetes: the Blue Mountains Eye Study, Australia. Ophthalmology 1997;104:712– 8. 8. Dielemans I, Vingerling JR, Algra D, et al. Primary openangle glaucoma, intraocular pressure, and systemic blood pressure in the general elderly population. The Rotterdam Study. Ophthalmology 1995;102:54 – 60. 9. Smith KD, Tevaarwerk GJ, Allen LH. Reversal of poorly controlled glaucoma on diagnosis and treatment of hypothyroidism. Can J Ophthalmol 1992;27:345–7. 10. Wang JJ, Mitchell P, Smith W. Is there an association between migraine headache and open-angle glaucoma? Findings from the Blue Mountains Eye Study. Ophthalmology 1997;104: 1714 –9. 11. Tielsch JM, Katz J, Quigley HA, et al. Diabetes, intraocular pressure, and primary open-angle glaucoma in the Baltimore Eye Survey. Ophthalmology 1995;102:48 –53. 12. Leske MC, Wu SY, Nemesure B, Hennis A. Incident openangle glaucoma and blood pressure. Arch Ophthalmol 2002; 120:954 –9. 13. Karadimas P, Bouzas EA, Topouzis F, et al. Hypothyroidism and glaucoma. A study of 100 hypothyroid patients. Am J Ophthalmol 2001;131:126 – 8. 14. Klein BE, Klein R, Meuer SM, Goetz LA. Migraine headache and its association with open-angle glaucoma: the Beaver Dam Eye Study. Invest Ophthalmol Vis Sci 1993;34:3024 –7. 15. Drance SM. Disc hemorrhages in the glaucomas. Surv Ophthalmol 1989;33:331–7. 16. Healey PR, Mitchell P, Smith W, Wang JJ. Optic disc hemorrhages in a population with and without signs of glaucoma. Ophthalmology 1998;105:216 –23.
Soares et al 䡠 Optic Disc Hemorrhages in Open-Angle Glaucoma 17. Kitazawa Y, Shirato S, Yamamoto T. Optic disc hemorrhage in low-tension glaucoma. Ophthalmology 1986;93:853–7. 18. Ishida K, Yamamoto T, Sugiyama K, Kitazawa Y. Disk hemorrhage is a significantly negative prognostic factor in normaltension glaucoma. Am J Ophthalmol 2000;129:707–14. 19. Diehl DL, Quigley HA, Miller NR, et al. Prevalence and significance of optic disc hemorrhage in a longitudinal study of glaucoma. Arch Ophthalmol 1990;108:545–50. 20. Susanna R, Drance SM, Douglas GR. Disc hemorrhages in patients with elevated intraocular pressure. Occurrence with and without field changes. Arch Ophthalmol 1979;97:284 –5. 21. Rasker MT, van den Enden A, Bakker D, Hoyng PF. Deterioration of visual fields in patients with glaucoma with and without optic disc hemorrhages. Arch Ophthalmol 1997;115: 1257– 62. 22. Drance S, Anderson DR, Schulzer M, Collaborative NormalTension Glaucoma Study Group. Risk factors for progression of visual field abnormalities in normal-tension glaucoma. Am J Ophthalmol 2001;131:699 –708. 23. Leske MC, Heijl A, Hussein M, et al, Early Manifest Glaucoma Trial Group. Factors for glaucoma progression and the effect of treatment: the Early Manifest Glaucoma Trial. Arch Ophthalmol 2003;121:48 –56. 24. Drance SM, Begg IS. Sector haemorrhage—a probable acute ischaemic disc change in chronic simple glaucoma. Can J Ophthalmol 1970;5:137– 41. 25. Quigley HA, Addicks EM, Green WR, Maumenee AE. Optic nerve damage in human glaucoma. II. The site of injury and susceptibility to damage. Arch Ophthalmol 1981;99:635– 49. 26. Daugeliene L, Yamamoto T, Kitazawa Y. Risk factors for visual field damage progression in normal-tension glaucoma eyes. Graefes Arch Clin Exp Ophthalmol 1999;237:105– 8. 27. Shihab ZM, Lee PF, Hay P. The significance of disc hemorrhage in open-angle glaucoma. Ophthalmology 1982;89:211–3.
28. Nicolela MT, Drance SM. Various glaucomatous optic nerve appearances: clinical correlations. Ophthalmology 1996;103: 640 –9. 29. Airaksinen PJ, Mustonen E, Alanko HI. Optic disc hemorrhages. Analysis of stereophotographs and clinical data of 112 patients. Arch Ophthalmol 1981;99:1795– 801. 30. Liou SY, Sugiyama K, Uchida H, et al. Morphometric characteristics of optic disk with disk hemorrhage in normaltension glaucoma. Am J Ophthalmol 2001;132:618 –25. 31. Poinoosawmy D, Gloster J, Nagasubramanian S, Hitchings RA. Association between optic disc haemorrhages in glaucoma and abnormal glucose tolerance. Br J Ophthalmol 1986; 70:599 – 602. 32. Quigley HA. Neuronal death in glaucoma. Prog Retin Eye Res 1999;18:39 –57. 33. Brenner G. Pharmacology. Philadelphia: WB Saunders; 2000: 160 – 8. 34. Grodum K, Heijl A, Bengtsson B. Optic disc hemorrhages and generalized vascular disease. J Glaucoma 2002;11:226 –30. 35. Tuulonen A, Takamoto T, Wu DC, Schwartz B. Optic disk cupping and pallor measurements of patients with a disk hemorrhage. Am J Ophthalmol 1987;103:505–11. 36. Kottler MS, Drance SM. Studies of hemorrhage on the optic disc. Can J Ophthalmol 1976;11:102–5. 37. Cooper ME, Bonnet F, Oldfield M, Jandeleit-Dahm K. Mechanisms of diabetic vasculopathy: an overview. Am J Hypertens 2001;14:475– 86. 38. Drance SM, Fairclough M, Butler DM, Kottler MS. The importance of disc hemorrhage in the prognosis of chronic open angle glaucoma. Arch Ophthalmol 1977;95:226 – 8. 39. Sonnsjo B, Dokmo Y, Krakau T. Disc haemorrhages, precursors of open angle glaucoma. Prog Retin Eye Res 2002;21: 35–56.
1657
Glaucoma is characterized by progressive loss of retinal ganglion cells and their axons, leading to the characteristic changes of the optic disc and loss of visual field. Elevated intraocular pressure (IOP) is the most important risk factor for the development of glaucoma.1,2 Other significant risk factors include increasing age,3 black race,3 family history of glaucoma,4,5 and myopia.6 Although some studies found an association between glaucoma and certain diseases such
Originally received: October 20, 2003. Accepted: March 1, 2004.
Manuscript no. 230698.
1
Department of Ophthalmology and Visual Sciences, Dalhousie University, Halifax, Canada.
2
Department of Community Health and Epidemiology, Dalhousie University, Halifax, Canada.
3
Department of Ophthalmology, Federal University of Sa˜o Paulo, Sa˜o Paulo, Brazil. Presented at: Association for Research in Vision and Ophthalmology Annual Meeting, May, 2003; Fort Lauderdale, Florida. Supported by the Canadian Institutes of Health Research, Ottawa, Canada (grant no.: MOP-11357). Correspondence to Marcelo T. Nicolela, MD, Department of Ophthalmology, Dalhousie University, 1278 Tower Road, Halifax, Nova Scotia, Canada B3H 2Y9. E-mail: [email protected].
© 2004 by the American Academy of Ophthalmology Published by Elsevier Inc.
as diabetes,7 systemic hypertension,8 hypothyroidism,9 and migraine,10 others could not confirm these findings.11–14 Optic disc hemorrhages (ODHs) frequently are observed in patients with glaucoma,15 but they also can occur infrequently in healthy persons.16 Optic disc hemorrhages usually persist from 2 to 35 weeks, with an average of 11 weeks.17 They are observed more often in patients with normal-tension glaucoma as opposed to those patients with elevated IOP (20%17 to 46%18 and 2%19 to 37%,20 respectively). Optic disc hemorrhages frequently are associated with glaucoma progression and are considered a risk factor for the deterioration of the disease.15 After a mean follow-up of 9 years, Rasker et al21 reported that 80% to 89% of glaucoma patients with ODHs progressed, as opposed to 32% of patients who did not have ODHs. Two recently published randomized clinical trials, the Collaborative Normal Tension Glaucoma Study22 and the Early Manifest Glaucoma Trial,23 also showed higher rates of progression in glaucoma patients with ODHs at baseline or in patients with more frequent ODHs during follow-up, respectively. Despite its importance in glaucoma, the association between ODHs and systemic diseases, the use of anticoagulant medication, and intrinsic eye factors such as IOP, refractive ISSN 0161-6420/04/$–see front matter doi:10.1016/j.ophtha.2004.03.023
1653
Ophthalmology Volume 111, Number 9, September 2004 error, and glaucoma severity has not been fully elucidated. We have noted clinically that treated patients who have ODHs tend to have well-controlled IOP, leading to the hypothesis that these hemorrhages are observed more commonly at lower IOP levels. The purpose of this cohort study was to test this hypothesis and to evaluate additional factors associated with ODHs in patients with open-angle glaucoma followed up for several years.
Patients and Methods Patients From 1992 through 1995, patients with open-glaucoma were enrolled into a prospective study at our center, which was approved by the Ethics Committee of the Queen Elizabeth II Health Sciences Centre. Inclusion criteria were: (1) best-corrected visual acuity of at least 20/30; (2) glaucomatous optic disc damage, such as notching or progressive thinning of the neuroretinal rim; (3) mild to moderate visual field defect, with mean deviation better than ⫺10 dB using the 30-to-2 full-threshold strategy program of the Humphrey Field Analyzer (Carl Zeiss Meditech, Dublin, CA); and (4) the presence of open angles by gonioscopy. Exclusion criteria were: (1) concomitant ocular disease; (2) refractive error exceeding 6 diopters (D) equivalent sphere or 2.5 D of astigmatism; (3) aphakia; (4) systemic disease or systemic medication known to affect the visual field. Patients with diabetes and systemic hypertension without obvious diabetic or hypertensive retinopathy could be enrolled into the study.
Procedures and Study Variables For this present study, we analyzed the data starting with the patient’s first visit to the glaucoma clinic, which could have occurred before enrollment. Most of the patients were followed up during the entire period by one of the authors (RPL). During each visit, the presence of ODHs was investigated, and, if present, their location was depicted in a diagram of the optic disc. An ODH was considered new if it occurred at a different location from a previous one, or if it recurred at least 3 months later if in the same location.17 Information regarding the following variables was collected during each study visit: (1) patient-related variables: presence of diabetes, systemic hypertension, heart disease, hypercholesterolemia, migraine, and use of anticoagulant medication (platelet antiaggregant and cumarinic agents); (2) eye-related variables: IOP, ocular refractive error, and severity of visual field and optic disc damage, as defined below. Information on systemic diseases was based on the patient’s self-report and was confirmed according to the class of medication used. All data from visits during the first 3 months after laser or surgical procedure for glaucoma were excluded from the analyses. Because all patients taking platelet antiaggregant medication were using aspirin, the agent name instead of the class of the systemic medication is used herein. After enrollment, patients were examined every 4 to 6 months. At every visit, a standard automated visual field examination was performed, using the Humphrey Field Analyzer, program 30-to-2, full-threshold strategy. At the same visits, a computerized examination of the optic disc using scanning laser tomography (Heidelberg Retina Tomograph; Heidelberg Engineering, Dossenheim, Germany) also was performed. For our analyses, the severity of the visual field damage was expressed as the mean deviation, whereas the severity of the optic disc damage was expressed as the rim area deviation. The latter variable measures the number of standard
1654
deviations by which the log-transformed area of the neuroretinal rim deviates from the expected optic disc size-corrected normal value. Visual field and optic disc status were interpolated using linear regression for patients who had additional clinic visits at which visual field or scanning laser tomography tests, or both, were not performed.
Statistical Analysis Survival time distribution function analyses using Wilcoxon and log-rank tests (multivariate analysis, forward stepwise approach) were used to identify variables associated with the elapsed follow-up time before the first ODH (time to hemorrhage). For the patient-related variables, the date of the first ODH was used, whether it occurred in the right or left eye. For eye-related variables, right and left eyes of each patient were treated as independent observations while adjusting for the effect of those patientrelated variables that had been found significant. Variables identified with P⬍0.1 then were included in a multivariate Cox’s proportional hazard model. To investigate the hypothesis that ODHs are observed at lower levels of IOP, we calculated the mean differences in the IOP between the visit in which the first ODH was observed and the average IOP of the respective previous 3 visits. As a control analysis, we calculated the mean difference in the IOP between a randomly selected visit before the first ODH and the mean IOP of the previous 3 visits on the same eyes. We chose to compare the IOP at the time of the first ODH with the previous IOP measurements because the detection of an ODH could lead the attending physician to initiate a more aggressive IOP-lowering therapy, therefore affecting the IOP on subsequent visits. The statistical analyses were performed using SPSS software, version 10.1 (SPSS, Inc., Chicago, IL), and SAS software, version 8 (SAS Institute, Cary, NC). Statistical significance was assumed at P⬍0.05.
Results Our sample contained 137 open-angle glaucoma patients, who had a total of 3612 visits for this study. The mean age at the first visit was 60.0 years (standard deviation, 11.0 years; range, 34 – 81 years), whereas the mean follow-up was 9.5 years (standard deviation, 5.0 years). Demographic and baseline data of the study patients are shown in Table 1. Optic disc hemorrhages were observed in 38 patients (28%), of whom 12 patients (32%) had bilateral ODHs. Twenty eyes (40%) had more than 1 ODH (range, 2– 8 ODHs). In total, 98 ODH episodes were observed, with 61 (62%) in the inferotemporal quadrant, 32 (33%) in the superotemporal quadrant, 3 (3%) in the nasosuperior quadrant, and 2 (2%) in the inferotemporal quadrant of the optic disc. In a univariate analyses, statistically significant differences between patients with and without ODH were observed in the proportion of patients using aspirin (P ⫽ 0.009) and in follow-up time (P ⫽ 0.030), although the number of visits per year of follow-up was similar in the 2 groups (P ⫽ 0.682; Table 2). Of the 33 patients using aspirin, 30 (91%) were using a dosage of 80 mg/day and 3 (9%) were using 325 mg/day. Because only 3 patients were using cumarinic agents, this variable was not included in the multivariate survival analyses, although 2 of these patients had ODHs during the follow-up. Cox’s proportional hazard model using the variables selected in the survival analyses showed that diabetes and use of aspirin were significantly associated with ODHs, with hazard ratios of 4.43
Soares et al 䡠 Optic Disc Hemorrhages in Open-Angle Glaucoma Table 1. Demographic and Baseline Data of the 137 Patients Male* Female* Diabetes* Systemic hypertension* Heart disease* Hypothyroidism* Migraine* Hypercholesterolemia* Aspirin use* Warfarin use* Refractive error (diopters)† RAD (arbitrary units)† MD (decibels)† Mean IOP during follow-up (mmHg)† Range IOP during follow-up (mmHg)†
65 (47) 72 (53) 17 (12.4) 71 (51.8) 14 (10.2) 14 (10.2) 22 (16.0) 17 (12.4) 33 (24.1) 3 (2.2) ⫺0.7 ⫾ 2.4 ⫺2.64 ⫾ 2.4 ⫺4.5 ⫾ 3.9 17.6 ⫾ 2.7 12.4 ⫾ 6.3
IOP ⫽ intraocular pressure; MD ⫽ mean deviation; RAD ⫽ rim area deviation. *n (%). † Mean ⫾ standard deviation.
(95% confidence interval [CI], 1.80 –10.50) and 2.30 (95% CI, 1.20 – 4.60), respectively (Table 3). The IOP at the time of the first observed ODH was significantly lower than the average IOP of the previous 3 visits (mean difference, ⫺1.4 mmHg; 95% CI, ⫺0.6 to ⫺2.2 mmHg; P⬍0.001). In the control analysis performed with the same eyes, the IOP at the randomly chosen visit before the first observed ODH was not significantly lower than the average IOP at the previous 3 visits (mean difference, ⫺0.3 mmHg; 95% CI, ⫺1.0 to 0.5 mmHg; P ⫽ 0.410).
Table 2. Univariate Analysis of Demographic, Systemic, and Ocular Variables
Male Female Age (yrs) Follow-up (yrs) No. of visits per year Refractive error (diopters) Diabetes Heart disease Systemic hypertension Hypothyroidism Migraine Hypercholesterolemia Aspirin use Cumarinic drugs use RAD (arbitrary units) MD (decibels) Mean IOP (mmHg) Range IOP (mmHg)
Patients with Optic Disc Hemorrhage (n ⴝ 38)
Patients without Optic Disc Hemorrhage (n ⴝ 99)
16 (42) 22 (58) 60.0 ⫾ 10.0 11.0 ⫾ 5.0 3.0 ⫾ 1.0 ⫺0.6 ⫾ 2.4 7 (18) 3 (8) 16 (42) 5 (13) 9 (24) 3 (8) 15 (39) 2 (5) ⫺2.9 ⫾ 2.2 ⫺5.0 ⫾ 4.1 17.4 ⫾ 2.7 12.4 ⫾ 5.0
49 (49) 50 (51) 60.0 ⫾ 11.0 9.0 ⫾ 5.0 3.0 ⫾ 1.0 ⫺1.1 ⫾ 2.4 10 (10) 11 (11) 55 (55) 9 (9) 13 (13) 14 (14) 18 (18) 1 (1) ⫺2.4 ⫾ 2.4 ⫺4.3 ⫾ 3.8 17.6 ⫾ 2.6 12.1 ⫾ 6.1
P Value 0.280 0.637 0.030 0.682 0.301 0.186 0.580 0.129 0.482 0.132 0.321 0.009 0.128 0.102 0.355 0.464 0.710
IOP ⫽ intraocular pressure; MD ⫽ mean deviation; RAD ⫽ rim area deviation. Variables with units: mean ⫾ standard deviation, independent-samples t test; otherwise n (%), chi-square test.
Table 3. Cox’s Proportional Hazard Model
Age Diabetes Aspirin Hypothyroidism Hypercholesterolemia
Hazard Ratio (95% Confidence Interval)
P Value
0.97 (0.93–1.00) 4.43 (1.80–10.50) 2.30 (1.20–4.60) 2.61 (0.96–7.03) 0.34 (0.09–1.18)
0.053 0.001 0.019 0.058 0.091
Variables with P⬍0.1 on Wilcoxon or log-rank survival analyses, or both, were included in the Cox’s proportional hazard model.
Discussion Various mechanisms have been proposed to explain the cause of ODHs in glaucoma. Drance and Begg24 suggested that ODHs result from an ischemic microinfarction at the optic disc, whereas Quigley et al25 suggested that structural changes at the level of the lamina cribrosa may cause mechanical rupture of small blood vessels. Although the precise cause of ODHs is still unknown, their prognostic importance for the progression of glaucoma has been well established.15,18,19,21,22,26,27 It is not clear, however, if the hemorrhages themselves are harmful to glaucoma or whether they are only a sign of an active disease process that will lead to progression. Another possibility explaining the prognostic relevance of ODHs is that they occur more frequently in subgroups of patients with glaucoma who may be at higher risk for progression. Previous studies have suggested that ODHs occur more often in patients with focal optic disc loss, and it is possible that these patients are more prone to progress than patients with other types of optic disc damage.28 Most ODHs occur at the poles of the optic disc, specifically at the superotemporal and inferotemporal locations,16,18,19,27,29 –31 corresponding to areas of the optic disc that are most susceptible to glaucomatous damage.32 In the present study, we also found that most ODHs (95%) were detected in the superotemporal and inferotemporal quadrants of the optic disc. In the present study, the use of aspirin was associated with ODHs. Aspirin prevents thromboxane A2 production by cyclooxygenase inhibition. Thromboxane A2 promotes platelet aggregation, one of the steps in the coagulation cascade.33 Aspirin therefore either could increase the risk of developing ODHs, could lead to larger ODHs that may take longer to absorb and consequently are more likely to be detected during periodic examinations, or both. An alternative explanation is that patients using aspirin are more likely to have vascular diseases that ultimately could be related to the presence of ODH. In a large screening study, Grodum et al34 recently reported a positive association between ODHs and aspirin use. However, that study did not distinguish between patients with glaucoma and healthy subjects. In another screening study, Healey et al16 did not observe a relationship between ODH and aspirin use in either normal subjects or patients with glaucoma. In the present longitu-
1655
Ophthalmology Volume 111, Number 9, September 2004 dinal study, patients were examined frequently, at regular intervals, and over several years. This study, therefore, may be better positioned to unveil risk factors associated with the occurrence of ODHs than a cross-sectional study design. As far as we are aware, this is the first longitudinal study showing an association between aspirin use and ODHs. It is possible that use of other anticoagulant medications, such as cumarinic agents, also is associated to the occurrence of ODHs. This study, however, was underpowered to show any possible association, because only 3 patients were using cumarinic agents (2 of them experienced ODHs during follow-up). In our study, patients with diabetes also had a higher risk of developing ODHs. A number of studies have reported a similar finding,16,27,31,35 whereas others19,34,36 could not confirm such an association. It is plausible that the vasculopathy associated with diabetes37 causes ischemic changes in the optic disc, renders the small optic nerve vessels more vulnerable to mechanical rupture, or both. We excluded patients with obvious diabetic retinopathy, including retinal hemorrhages and exudates. If glaucoma patients with diabetic retinopathy have an even higher incidence of ODHs than those without obvious diabetic retinopathy, the association found in our study between ODHs and diabetes could be even higher. Similar to diabetes, the effect of systemic hypertension on the occurrence of ODHs is controversial. Although some investigators reported a link between ODHs and systemic hypertension,16,36,38 other studies, including ours, did not find such an association.19,34,35 The mean and range of IOP measurements during follow-up were not significantly different between glaucoma patients with and without ODHs. Our study did show, however, that the IOP at the time the first ODH was observed was significantly lower than that of previous visits. This finding was not observed in the control analysis of the same eyes using randomly selected visits without ODHs, confirming our previous clinical impression that ODHs tend to be detected in examinations when the IOP is at a relatively lower level during a patient’s follow-up. We can speculate that a reduction in IOP may cause transient mechanical stresses on the central retinal vein as it exits the eye by forward movement of the lamina cribrosa leading to increased venous resistance and, perhaps, increasing the risk of ODHs. Another possibility is that at lower levels of IOP, the pressure differences across the vessel walls are relatively greater than at higher IOP, which could cause larger and more frequent ODHs in predisposed eyes.39 Although the magnitude of IOP difference between the time the ODH was observed and the previous visits was relatively small, it is possible that the IOP may have been even lower at the time the ODH actually occurred. In conclusion, our study showed that open-angle glaucoma patients with diabetes and those using aspirin had a higher risk of having ODHs. Optic disc hemorrhages were observed more frequently at relatively lower levels of IOP throughout patient follow-up. These findings may be relevant to the management of glaucoma patients, particularly in view of the important association between ODHs and glaucoma deterioration. It is important to emphasize that our study design permits us to reveal only an association
1656
between aspirin use and the detection of ODHs in glaucoma. Based on the evidence thus far, it does not suggest that the use of aspirin is harmful in glaucoma. Further work should be undertaken to evaluate if aspirin use, as well as presence of diabetes, predisposes glaucoma patients, either directly or indirectly, to a higher risk of glaucoma progression. Acknowledgments. The authors thank Dr David F. GarwayHeath (Moorfields Eye Hospital, London, England) for providing part of the scanning laser tomography normal database used to calculate rim area deviation in our analysis.
References 1. Leske MC, Connell AM, Wu SY, et al. Risk factors for open-angle glaucoma. The Barbados Eye Study. Arch Ophthalmol 1995;113:918 –24. 2. Sommer A, Tielsch JM, Katz J, et al. Relationship between intraocular pressure and primary open angle glaucoma among white and black Americans. The Baltimore Eye Survey. Arch Ophthalmol 1991;109:1090 –5. 3. Tielsch JM, Sommer A, Katz J, et al. Racial variations in the prevalence of primary open-angle glaucoma. The Baltimore Eye Survey. JAMA 1991;266:369 –74. 4. Tielsch JM, Katz J, Sommer A, et al. Family history and risk of primary open angle glaucoma. The Baltimore Eye Survey. Arch Ophthalmol 1994;112:69 –73. 5. Wolfs RC, Klaver CC, Ramrattan RS, et al. Genetic risk of primary open-angle glaucoma. Population-based familial aggregation study. Arch Ophthalmol 1998;116:1640 –5. 6. Mitchell P, Hourihan F, Sandbach J, Wang JJ. The relationship between glaucoma and myopia: the Blue Mountains Eye Study. Ophthalmology 1999;106:2010 –5. 7. Mitchell P, Smith W, Chey T, Healey PR. Open-angle glaucoma and diabetes: the Blue Mountains Eye Study, Australia. Ophthalmology 1997;104:712– 8. 8. Dielemans I, Vingerling JR, Algra D, et al. Primary openangle glaucoma, intraocular pressure, and systemic blood pressure in the general elderly population. The Rotterdam Study. Ophthalmology 1995;102:54 – 60. 9. Smith KD, Tevaarwerk GJ, Allen LH. Reversal of poorly controlled glaucoma on diagnosis and treatment of hypothyroidism. Can J Ophthalmol 1992;27:345–7. 10. Wang JJ, Mitchell P, Smith W. Is there an association between migraine headache and open-angle glaucoma? Findings from the Blue Mountains Eye Study. Ophthalmology 1997;104: 1714 –9. 11. Tielsch JM, Katz J, Quigley HA, et al. Diabetes, intraocular pressure, and primary open-angle glaucoma in the Baltimore Eye Survey. Ophthalmology 1995;102:48 –53. 12. Leske MC, Wu SY, Nemesure B, Hennis A. Incident openangle glaucoma and blood pressure. Arch Ophthalmol 2002; 120:954 –9. 13. Karadimas P, Bouzas EA, Topouzis F, et al. Hypothyroidism and glaucoma. A study of 100 hypothyroid patients. Am J Ophthalmol 2001;131:126 – 8. 14. Klein BE, Klein R, Meuer SM, Goetz LA. Migraine headache and its association with open-angle glaucoma: the Beaver Dam Eye Study. Invest Ophthalmol Vis Sci 1993;34:3024 –7. 15. Drance SM. Disc hemorrhages in the glaucomas. Surv Ophthalmol 1989;33:331–7. 16. Healey PR, Mitchell P, Smith W, Wang JJ. Optic disc hemorrhages in a population with and without signs of glaucoma. Ophthalmology 1998;105:216 –23.
Soares et al 䡠 Optic Disc Hemorrhages in Open-Angle Glaucoma 17. Kitazawa Y, Shirato S, Yamamoto T. Optic disc hemorrhage in low-tension glaucoma. Ophthalmology 1986;93:853–7. 18. Ishida K, Yamamoto T, Sugiyama K, Kitazawa Y. Disk hemorrhage is a significantly negative prognostic factor in normaltension glaucoma. Am J Ophthalmol 2000;129:707–14. 19. Diehl DL, Quigley HA, Miller NR, et al. Prevalence and significance of optic disc hemorrhage in a longitudinal study of glaucoma. Arch Ophthalmol 1990;108:545–50. 20. Susanna R, Drance SM, Douglas GR. Disc hemorrhages in patients with elevated intraocular pressure. Occurrence with and without field changes. Arch Ophthalmol 1979;97:284 –5. 21. Rasker MT, van den Enden A, Bakker D, Hoyng PF. Deterioration of visual fields in patients with glaucoma with and without optic disc hemorrhages. Arch Ophthalmol 1997;115: 1257– 62. 22. Drance S, Anderson DR, Schulzer M, Collaborative NormalTension Glaucoma Study Group. Risk factors for progression of visual field abnormalities in normal-tension glaucoma. Am J Ophthalmol 2001;131:699 –708. 23. Leske MC, Heijl A, Hussein M, et al, Early Manifest Glaucoma Trial Group. Factors for glaucoma progression and the effect of treatment: the Early Manifest Glaucoma Trial. Arch Ophthalmol 2003;121:48 –56. 24. Drance SM, Begg IS. Sector haemorrhage—a probable acute ischaemic disc change in chronic simple glaucoma. Can J Ophthalmol 1970;5:137– 41. 25. Quigley HA, Addicks EM, Green WR, Maumenee AE. Optic nerve damage in human glaucoma. II. The site of injury and susceptibility to damage. Arch Ophthalmol 1981;99:635– 49. 26. Daugeliene L, Yamamoto T, Kitazawa Y. Risk factors for visual field damage progression in normal-tension glaucoma eyes. Graefes Arch Clin Exp Ophthalmol 1999;237:105– 8. 27. Shihab ZM, Lee PF, Hay P. The significance of disc hemorrhage in open-angle glaucoma. Ophthalmology 1982;89:211–3.
28. Nicolela MT, Drance SM. Various glaucomatous optic nerve appearances: clinical correlations. Ophthalmology 1996;103: 640 –9. 29. Airaksinen PJ, Mustonen E, Alanko HI. Optic disc hemorrhages. Analysis of stereophotographs and clinical data of 112 patients. Arch Ophthalmol 1981;99:1795– 801. 30. Liou SY, Sugiyama K, Uchida H, et al. Morphometric characteristics of optic disk with disk hemorrhage in normaltension glaucoma. Am J Ophthalmol 2001;132:618 –25. 31. Poinoosawmy D, Gloster J, Nagasubramanian S, Hitchings RA. Association between optic disc haemorrhages in glaucoma and abnormal glucose tolerance. Br J Ophthalmol 1986; 70:599 – 602. 32. Quigley HA. Neuronal death in glaucoma. Prog Retin Eye Res 1999;18:39 –57. 33. Brenner G. Pharmacology. Philadelphia: WB Saunders; 2000: 160 – 8. 34. Grodum K, Heijl A, Bengtsson B. Optic disc hemorrhages and generalized vascular disease. J Glaucoma 2002;11:226 –30. 35. Tuulonen A, Takamoto T, Wu DC, Schwartz B. Optic disk cupping and pallor measurements of patients with a disk hemorrhage. Am J Ophthalmol 1987;103:505–11. 36. Kottler MS, Drance SM. Studies of hemorrhage on the optic disc. Can J Ophthalmol 1976;11:102–5. 37. Cooper ME, Bonnet F, Oldfield M, Jandeleit-Dahm K. Mechanisms of diabetic vasculopathy: an overview. Am J Hypertens 2001;14:475– 86. 38. Drance SM, Fairclough M, Butler DM, Kottler MS. The importance of disc hemorrhage in the prognosis of chronic open angle glaucoma. Arch Ophthalmol 1977;95:226 – 8. 39. Sonnsjo B, Dokmo Y, Krakau T. Disc haemorrhages, precursors of open angle glaucoma. Prog Retin Eye Res 2002;21: 35–56.
1657