Investigations into a Vascular Etiology for Low-tension Glaucoma CEDRIC J. CARTER, MB, BS, FRCP, DONALD E. BROOKS, PhD, D. LYNN DOYLE, MD, FRCS, STEPHEN M. DRANCE, MD, OC
Abstract: Increased intraocular pressure is accepted as a primary etiologic factor for the atrophy of the optic nerve head and visual field defects of hightension glaucoma. Other factors must be present to explain these findings in low-tension glaucoma. One of the current theories is that low-tension glaucoma is the result of decreased optic nerve perfusion on the basis of vascular disease or other factors such as altered blood viscosity. This study compared the noninvasive vascular profiles, coagulation tests, and rheological profiles of 46 consecutive cases of low-tension glaucoma with 69 similarly unselected cases of high-tension glaucoma and 47 age-matched controls. Despite the multifactorial approach and the use of previously validated objective tests, no significant group differences were detected with any of the above investigations. If vascular disease is important in the etiology of low-tension glaucoma, then it must be localized or vasospastic since this study does not support the concept of a generalized vascular etiology, either of an atheromatous or hyperviscous nature, for the genesis of low-tension glaucoma. Ophthalmology 1990; 97:49-55
Glaucoma is a condition characterized by excavation and atrophy of the optic nerve head and characteristic visual field defects. The condition was first recognized in the last century and, until recently, was thought to be almost invariably associated with increased intraocular pressure (lOP). It was hypothesized that increased lOP either decreased optic nerve perfusion, giving rise to the degenerative optic nerve changes, or produced mechanical pressure damage to the axons of the nerve head. 1,2 The development of applanation tonometry confirmed earlier suggestions that not all glaucomatous optic atrophy was associated with increased lOPs. This raised the possibility that either there were two separate diseases (i.e., hightension glaucoma and low-tension glaucoma) with entirely Originally received: June 5, 1989. Revision accepted: July 24, 1989. From the Departments of Pathology, Vascular Surgery, and Ophthalmology, University of British Columbia, Vancouver. Supported by a grant from the B. C. Health Care Research Foundation. Reprint requests to Cedric J. Carter, MB, Department of Laboratory Medicine, 2211 Westbrook Mall, Vancouver, Canada V6T 2B5.
different etiologies that happened to have the same pathologic appearances or that increased lOP was one of several risk factors for an underlying pathologic process. 3 This controversy remains unresolved. Some epidemiologic studies of low-tension glaucoma suggest that patients with low-tension glaucoma have an excess of ischemic cerebral and cardiovascular disease and that this may explain the ischemic optic nerve degeneration. 2 Further epidemiologic evidence for an ischemic etiology includes the age dependency oflow-tension glaucoma. Some series also have shown statistical associations with the presence of diabetes and vascular hypotension. An earlier study of 29 confirmed cases of low-tension glaucoma showed that these patients, relative to agematched controls, had positive histories of major vascular crises, including myocardial infarction and perioperative hypotension. 3 In addition, these patients appeared to have increased platelet adhesion and increased prestress euglobulin lysis times. The latter is a measure of fibrinolytic activity and is believed to be a reflection of free tissue plasminogen activator levels. This relatively small study did not include a control group of patients with hightension glaucoma.
49
OPHTHALMOLOGY
•
JANUARY 1990
Subsequent investigators looked at other factors known to be associated with more generalized arterial disease. Whole blood and plasma viscosity, which have been demonstrated to be abnormal in patients with myocardial infarction, angina, and intermittent claudication, were studied in glaucoma patients. 4 A large European study compared 83 patients diagnosed as having low-tension glaucoma with 23 patients with high-tension glaucoma and 50 control patients. s Higher whole blood and plasma viscosity values were shown in the patients with low-tension glaucoma relative to controls. In addition, plasma viscosity was higher in low-tension glaucoma patients relative to high-tension glaucoma cases. It is not clear how patients were selected for this study since there were many more low-tension glaucoma patients than high-tension glaucoma patients. If this were a true random selection of patients, the latter should have been far more common. Another unexplained feature of this study was the 25% incidence of ischemic vascular disease in the 50 control patients compared to the absence of ischemic vascular disease in the 23 high-tension glaucoma patients. This suggests that high-tension glaucoma in some way protects against vascular disease. From a biologic point of view, this is unlikely. Despite the methodologic limitations in some of the above studies, the overall conclusions are in keeping with a vascular contribution to the development of low-tension glaucoma. It is now possible to take other approaches to the documentation and demonstration of vascular disease. Since the earlier reports, major developments have occurred in noninvasive vascular diagnostic techniques. Measurements that can be made include segmental Doppler arterial pressures, strain gauge plethysmography for smaller vessels and digits, and duplex ultrasound examination of larger arteries. 6 None of these techniques has been applied systematically to patients with glaucoma. An alternative approach to the demonstration of generalized vascular disease is to look for the presence of various blood markers that have been associated with acute arterial events and more generalized atherosclerosis. In population surveys, elevated levels of fibrinogen, factor VII, and factor VIII were associated with an increased incidence of ischemic heart disease. 7 Another blood finding associated with generalized vascular disease is an increase in the euglobulin lysis time (ELT). In its current form, the time taken for a euglobulin fraction plasma clot to lyse is determined before the application of ischemic stress to the upper limbs. The normal population shows a shortening of the lysis time following ischemic stress to the forearm. Patients with a predisposition to, or who have developed vascular disease often show a relative decrease in the poststress shortening of the ELT. 8.9 Other investigators looked at markers of platelet activation. Platelet deposition is thought to be a fundamental event in the development of generalized atheroma. to In addition, in cerebrovascular disease, platelet microemboli have been demonstrated to cause a variety of ophthalmologic and cerebrovascular conditions. II Platelet microemboli are not easily demonstrated; however, it is possible to demonstrate an increase in markers of platelet 50
•
VOLUME 97
•
NUMBER 1
release, such as raised levels of beta thromboglobulin (BTG) in generalized vascular disease and in various ophthalmologic conditions including retinal vein occlusions. 12- 14 Another method that has been used in the investigation of suspected platelet-mediated vascular disease is to look for spontaneous platelet aggregation. This phenomenon has been reported in a variety of vascular diseases, including primary thrombocythemia, transient ischemic attacks, stroke, acute myocardial infarction, and peripheral vascular disease. ls . 16 When this technique was applied to a large group of patients with low-tension glaucoma and classic high-tension glaucoma, an association between spontaneous platelet aggregation and low-tension glaucoma was observed. 17 Defining the etiology of low-tension glaucoma is of therapeutic importance. If a clear relationship exists between vascular disease and low-tension glaucoma, then a case can be made for clinical trials of medications that have been shown to be effective in other forms of cerebrovascular disease (e.g., the use of an antiplatelet agent such as acetylsalicyclic acid).18 This study assembled a consecutive group of patients with high-tension and low-tension glaucoma together with an age- and sex-matched cohort with no evidence ofglaucoma. All patients and subjects underwent a complete noninvasive vascular assessment using segmental Doppler pressures and strain gauge plethysmography. Whole blood and plasma viscosities, antithrombin III, factor VII, factor VIII, and fibrinogen levels were measured. The vascular fibrinolytic status was examined by pre- and poststress ELT. Evidence for platelet activation was examined by measuring 24-hour urinary BTG excretion. Through the use of a variety of diagnostic modalities, including peripheral vascular testing, coagulation profiles, and rheological measurements, we investigated whether it was possible to show vascular differences between the glaucomatous and nonglaucomatous population.
MATERIALS AND METHODS PATIENTS AND STUDY SUBJECfS
One hundred sixty-two subjects were recruited. Sixtynine patients had been diagnosed as having primary openangle, high-tension glaucoma, with characteristic field defects, funduscopic appearances, and lOPs in excess of 21 mmHg. Forty-six were low-tension glaucoma patients as defined by field defects of classic glaucoma type, open anterior chamber angles, no history of uveitis or other ocular disease, and lOPs consistently 21 mmHg or less, including diurnal testing. Both groups of glaucoma patients were consecutive patients presenting for evaluation at the University of British Columbia glaucoma clinics. The 47 control subjects came from two sources. The majority were spouses of glaucoma patients. These subjects were chosen because there was no question of consanguinity and they would have been exposed to similar environmental factors. The remaining controls were patients admitted for elective cataract surgery in whom glaucoma
CARTER et al
•
ETIOLOGY OF LOW-TENSION GLAUCOMA
had been excluded. Because the underlying etiology for glaucoma is not known, no attempt was made to match control subjects with the patients for factors other than smoking and age. This avoided the problem of overmatching. 19 Informed consent was obtained for this research project from all patients and control subjects in accordance with the ethics approval granted by the University of British Columbia's Ethics Approval Committee.
surements were taken, and brachial to ankle pressure rations were calculated. These measurements were taken before and after exercise. The exercise regimen consisted of treadmill walking using a 10-minute Bruce protocol or until ischemic pain intervened. Small arterial vessels were evaluated by mercury strain gauge plethysmography pressure measurements of the fingers and toes using a Medsonic SPG 16 plethysmograph. 24
OPHTHALMOLOGIC TESTING
All control subjects had ocular evaluations that excluded the presence of high-tension or low-tension glaucoma. The glaucoma patients had diurnal tension studies and perimetry repeated. All laboratory and vascular investigations apart from the initial ophthalmologic documentation were performed without knowledge of the patient's disease category. Blind testing was done to avoid observer bias. 20 COAGULATION AND HEMATOLOGY TESTING
All subjects had a full peripheral blood count performed with various coagulation assays. Coagulant factor VII levels were done using a prothrombin time assay (Simplastin Automated, Organon Teknika, Durham, NC) and factor VII deficient plasma (Pacific Hemostasis, Ventura, CA). Coagulant factor VIII levels were determined by a onestage activated partial thromboplastin time (APTT) -based assay (Automated APTT, Organon Teknika, Durham, NC) and factor VIII deficient plasma (Pacific Hemostasis, Ventura, CA). Fibrinogen levels were determined using a thrombin clotting assay (American Dade, AHS, Puerto Rico). Antithrombin levels (ATIII) were done spectrophotometrically (Dupont ACA III, Dupont, Wilmington, DE). Euglobulin lysis times were performed both pre- and poststress. 21 The stress consisted of inflating a sphygmomanometer cuff on the upper arm to 90 mmHg and maintaining this pressure for 5 minutes. All blood samples for the coagulation assays were put into a standard buffered 3.8% sodium citrate anticoagulant (Becton Dickinson, Rutherford, NJ). A 24-hour urine specimen was taken and a BTG assay (Amersham, Arlington Heights, IL) performed using a modification of the standard plasma assay. 22 BIOCHEMICAL TESTING
Total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol and triglyceride levels were tested using spectrophotometric assays (Kodak Ektachem, Rochester, NY). Additional investigations included hemoglobin Ale levels to document biochemical evidence of glucose intolerance (Bio-Rad, Hercules, CA). VASCULAR TESTING
All subjects underwent noninvasive vascular testing. To assess the status of the larger arteries, segmental arterial pressures were determined using occlusive cuffs at two levels on the lower limb and one level on the upper limb using a Medsonic bi-directional DopplerY Direct mea-
RHEOLOGY TESTING
Plasma viscosity was determined using a CannonManning semi-micro type viscometer. Whole blood viscosity was examined using a Contraves LS-2 Couette viscometer with a concentric cylinder measuring system that had a stationary inner cylinder and a rotating outer cylinder. Measurements were done at 3rC, with nominal shear rates ranging from 0.031 S-l to 124.0 S-l and hematocrits ranging from 0.25% to 0.5% including the patient's own venous hematocrit. These tests produced a series of viscosity readings that were incorporated into an analytical procedure that generated a constitutive function that best described the rheological behavior of whole blood. 25 This function describes the shear rate and hematocrit dependence of viscosity in terms of four numerical parameters (Xt, X 2 , X 3 , and X 4 ) whose values completely describe the flow behavior of the sample within the operating limits used. These four parameters are independent of shear rate and hematocrit and can be used to numerically describe the rheological behavior of any given blood sample. The values of the four parameters determined for any sample are then compared statistically with various other individuals and clinical groupings to provide an objective comparison of the flow behavior of blood. STATISTICAL ANALYSIS
Analysis of quantitative data was made by comparison of variances. Because there were three patient groups, oneway analysis of variance was performed with the NewmanKeuls correction for comparing group means. 26 Analysis for the presence of vascular disease was made on the basis of absolute arterial readings and on the criteria of patients being designated as normal or abnormal according to previously determined threshold values of normalcy. The latter results were analyzed on the basis of 2 X K contingency table using the chi-square statistic. 27
RESULTS The ages of the comparative groups, male to female ratios, and the percentage of smokers are shown in Table 1. The three groups appear to be comparable with respect to mean age, age distribution, and smoking habits. Males appear to be slightly overrepresented in the high-tension glaucoma group. Since these were consecutive and unselected patients, there are no obvious explanations for this observation. 51
OPHTHALMOLOGY
•
JANUARY 1990
LTG
Controls
Age (yrs) Range
66 37-86
64 42-84
66 66-85
Male to female ratio
42:27
16:30
20:27
33
25
35
Percentage of smokers HTG
=
high-tension glaucoma; LTG
=
VOLUME 97
•
NUMBER 1
cholesterol ratios, and hemoglobin Ale levels are summarized in Table 3. No statistically significant differences were present.
Table 1. Patient Characteristics HTG
•
VASCULAR TESTS
Segmental arterial pressures. Segmental arterial pressures are expressed as ankle to brachial ratios. The group means are shown in Table 4. An alternative analysis of these data is to determine the proportion of patients with ratios below 1.0 (Table 4). No statistically significant differences were demonstrated among the three groups using either analytical approach. Strain gauge plethysmography. The results of the strain gauge pressures in the fingers and toes relative to brachial and tibial pressures are shown in Table 5. The first section of the table shows the mean pressure drops occurring in the brachial and digital arteries and tibial and toe arteries, respectively. The second section of the table shows the number of subjects with brachial-digital and tibial-toe arterial pressure decreases of greater than 40 mmHg and 70 mmHg, respectively. No statistically significant differences were detected after analysis by analysis of variance in the first section and by the chi-square statistic in the second section.
low-tension glaucoma.
COAGULATION STUDIES
Coagulation studies are summarized in Table 2. The various coagulation tests for each experimental group were compared using one-way analysis of variance with a Newman-Keuls multiple comparisons test. No statistically significant differences among the mean values were demonstrated. BIOCHEMICAL TESTING
Values obtained for cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, high-density lipoprotein cholesterol to total
Table 2. Coagulation Studies Mean ± 1 SO
LTG
HTG 1.148 1.110 3.075 2.87 418.7 282.8 36.4 1.04
Factor VII (U/ml) Factor Vlll:c (U/ml) Fibrinogen (g/I) BTG at 24 hrs Prestress EL T (mins) Poststress EL T (mins) Percent decrease' AT III (U/ml) SO = standard deviation; HTG III; EL T = euglobulin lysis time.
=
, Calculated with the following equation:
(pr~post) X re
1.158 ± 0.219 1.084 ± 0.385 3.126 ± 0.671 2.79 ± 0.27 441.8 ± 217.8 308.6 ± 243.5 35.8 ± 24.6 1.07 ± 0.13
± 0.219 ± 0.385 ± 0.671 ± 0.29 ± 23.5 ± 219.2 ± 24.6 ± 0.15
high-tension glaucoma; LTG
Controls
low-tension glaucoma; BTG
=
1.139 1.435 3.318 2.62 540.5 389.8 30.6 1.08
± 0.358 ± 0.431 ± 0.731 ± 0.25 ± 201.5 ± 242.7 ± 27.0 ± 0.12
beta thromboglobulin; AT III
=
=
antithrombin
100.
Table 3. Biochemical Tests Mean ± SO
LTG
HTG Total cholesterol (mmol/I) HOL cholesterol (mmol/I) LOL cholesterol (mmol/I) Triglycerides (mmol/I) HOL cholesterol/total cholesterol Hemoglobin Ale SO = standard deviation; HTG LOL = low-density lipoprotein.
52
=
5.80 1.86 3.84 1.67 0.21 0.053
± ± ± ± ± ±
high-tension glaucoma; LTG
0.88 0.31 0.90 1.07 0.07 0.005 =
5.89 1.202 3.39 1.90 0.21 0.060
± ± ± ± ± ±
low-tension glaucoma; HOL
Controls
1.29 0.42 1.51 1.35 0.08 0.037 =
5.64 1.31 3.55 1.72 0.24 0.054
high-density lipoprotein;
± ± ± ± ± ±
1.28 0.33 1.02 0.94 0.07 0.009
CARTER et al
•
ETIOLOGY OF LOW·TENSION GLAUCOMA
Table 4. Arterial Testing by Segmental Doppler
LTG
HTG
Ankle to brachial ratios (mean ±1 SD) Subjects with ratio 40 mmHg) Ankle/toe (>70 mmHg)
%8 2%8
%8 2%8
211a6
%6 1%6
1%1
11a6
%1 1°/31
71a1
HTG = high-tension glaucoma; LTG = low-tension glaucoma; R = right; L = left; SD = standard deviation. Table 6. Viscosity Testing-Constitutive Function
Parameter (mean ± 1 SD)
X1 X2 X3
X4
Plasma viscosity (mean + 1 SD)
HTG (n = 69)
LTG (n
2.356 ± 1.4 5.1 ± 1.1 0.53 ± 0.05 0.003 1.27 ± 0.6
2.16 ± 1.5 4.6 ± 1.3 0.51 ± 0.12 -0.003 1.29 ± 0.9
=
45)
Controls (n
=
47)
2.21 ± 1.5 5.22 ± 1.0 0.53 ± 0.06 0.001 1.27 ± 0.8
HTG = high-tension glaucoma; LTG = low-tension glaucoma; SD = standard deviation. RHEOLOGY
The comparative values of the four discriminant parameters derived from whole blood viscometry and the plasma viscosity values are shown in Table 6. No significant differences were demonstrated among the group means. For comparison with other studies that have not used the constitutive function approach to the characterization of blood flow properties, the whole blood viscosity is shown at the nominal shear rates of 19.7 and 124 S-1 at native hematocrit (Table 7). It should be noted that because blood is a non-Newtonian fluid, these are the nominal shear rates as determined by the instrument setting rather than the actual shear rates occurring in the test system. Nominal shear rates have been used by previous investigators.
DISCUSSION The exact pathogenesis of the glaucomas remains unanswered. It is clear that an elevated lOP is an important
risk factor that also has a causal relationship to glaucomatous damage. There are, however, many optic nerves that do not appear to be damaged by prolonged increased lOP. In other cases, marked and progressive damage occurs despite a statistically normal or even low lOP. Other factors must be present, and a comparison oflow-tension glaucoma with high-tension glaucoma is an obvious area of study to demonstrate such factors. Vascular etiologies have been implicated in the development of the glaucomas. The vascular factors could be organic or vasospastic or combinations of these. In other organic vascular pathologies, noninvasive diagnostic techniques and various blood markers, including elevated fibrinogen, factor VII, factor VIII, euglobulin lysis times, and markers of platelet activation have shown abnormalities. Iflow-tension glaucoma has an organic vascular basis like transient ischemic attack, myocardial infarction, stroke, and peripheral ~~s cular disease, one would expect diagnostic abnormahties ip patients with low-tension glaucoma. Because there is no direct method for definitively measuring in vivo the blood supply provided by ~he ophthalmic artery and its distal branches, documentatIOn 53
OPHTHALMOLOGY
•
JANUARY 1990
Table 7. Viscosity Testing Mean ± SO Shear Rate
HTG (n = 69)
LTG (n = 46)
Controls (n = 63)
Peak 19.7 S·l Steady state 124 S-l
10.09 ± 1.8 8.85 ± 1.6 5.87 ± 0.7
9.78 ± 2.0 8.47 ± 1.8 5.73 ± 0.9
9.34 ± 1.8 8.21 ± 1.6 5.59 ± 0.7
SO = standard deviation; HTG LTG = low-tension glaucoma.
=
high-tension glaucoma;
of vascular factors in low-tension glaucoma must be by indirect investigations such as we have undertaken. Compared with previous publications, this is one of the more extensive studies. Care was taken to avoid bias in selection of patients and controls. An obligate consequence of hospital or specific referral center-based studies is the generation of false associations. For example, if one condition initiates a hospital visit, then nonspecific laboratory and investigational findings will detect other, but not necessarily directly related, conditions. The selection of appropriate control groups helps prevent this source of bias. Two control groups were selected for this study. The first control group comprised nonglaucomatous subjects who were selected to allow an age match. The second control group consisted of carefully documented consecutive cases of high-tension glaucoma. By chance, they happened to be of similar age to the low-tension glaucoma cohort. Selection to permit a sex match was not possible since in the glaucoma patients there was a different sex distribution between the high- and low-tension glaucoma cases. As such, with the exception of sex, the experimental groups show no evidence of major imbalance or any obvious selection artifacts. The underlying rationale of this study was that if there was a major vascular component to low-tension glaucoma, one should be able to demonstrate this by techniques that had already been used to implicate a vascular etiology in low-tension glaucoma or other defined vascular pathologies. Because increased lOP could be an additional risk factor for a primary glaucomatous predisposition, as exemplified by low-tension glaucoma, a high-tension glaucoma group was an important comparison group. If the above hypothesis was correct, then the high-tension glaucoma test results might be expected to fall between the low-tension glaucoma and control results. The similarity of the vascular laboratory results between the comparative groups excludes major differences in overall vascular pathology but cannot exclude the presence of some highly localized process. Earlier studies relied on clinical history to determine vascular status. There could be a problem with the relative sensitivity of the objective vascular tests used in this study, but these same tests have been validated in most of the cardiovascular conditions previously described as being associated with low-tension glaucoma.6.23.24 The coagulation studies were not informative. Due to problems of test reproducibility, we were not able to repeat 'i4
•
VOLUME 97
•
NUMBER 1
the earlier studies that examined spontaneous platelet aggregation and platelet adhesion, but we were able to measure clotting factors VII and VIII and fibrinogen. Recent studies have confirmed that raised levels offactor VII and fibrinogen are present in people with atherosclerosis.z s The magnitude of these differences are in the region of 10% and should have been detectable in a study of this size. Similarly, no significant group differences were observed in 24-hour urinary BTG excretion. All these negative findings are in keeping with the results of the direct vascular testing. In view of the close association between lipids and vascular disease, a full lipid screen was performed, but, again, no group differences were demonstrated. The final series of investigations were the studies on plasma and whole blood viscosity. Two previous studies implicated increased blood viscosity in the genesis oflowtension glaucoma. We were unable to confirm these findings. It should be noted that our method of analyzing whole blood viscosity is fundamentally different from that of previous investigators, and there is no data on the comparative sensitivities of various rheological techniques for the investigation of vascular disease. However, data obtained using a more traditional approach (Table 7) did not show any distinction among the groups. Although differences exist between this study and earlier studies in terms of specific methods, the study was balanced and had appropriate controls and adequate statistical power. The lack of any positive findings for incrp :4sed atherosclerosis or abnormal rheology is important because if there is any vascular basis for low-tension glaucoma, then it is mediated through other factors such as vasospasticity. In summary, this study failed to demonstrate group differences for the markers of atherosclerotic vascular disease among patients with high-tension and low-tension glaucoma and nonglaucomatous controls. Similarly, no rheologic differences were detected. Using the above methods and statistical analysis, this study does not support an organic vascular pathology for low-tension glaucoma.
REFERENCES 1. Grehn F, Prost M. Function of retinal nerve fibers depends on perfusion pressure: neurophysiologic investigations during acute intraocular pressure elevation. Invest Ophthalmol Vis Sci 1983; 24:347-53. 2. Spaeth GL. Fluorescein angiography: its contributions towards understanding the mechanisms of visual loss in glaucoma. Trans Am Ophthalmol Soc 1975; 73:491-553. 3. Orance SM, Sweeney VP, Morgan RW, Feldman F. Studies of factors involved in the production of low tension glaucoma. Arch Ophthalmol 1973; 89:457-65. 4. Trope GE, Salinas RG, Glynn M. Blood viscosity in primary openangle glaucoma. Can J Ophthalmol1987; 22:202-4. 5. Klaver JHJ, Greve EL, Goslinga H, et al. Blood and plasma viscosity measurements in patients with glaucoma. Br J Ophthalmol 1985; 69: 765-70. 6. Barnes RW. Noninvasive diagnostic techniques in peripheral vascular disease. Am Heart J 1979; 97:241-58. 7. Meade TW, North WRS, Chakrabarti R, et al. Haemostatic function and cardiovascular death: early results of a prospective study. Lancet 1980; 1: 1050-4.
CARTER et al
•
ETIOLOGY OF LOW-TENSION GLAUCOMA
8. Balkau-Ulutsin S. Fibrinolytic system in atherosclerosis. Semin Thromb Hemost 1986; 12:91-101 . 9 . Hamsten A, De Faire U, Walldius G, et al. Plasminogen activator inhibitor in plasma: risk factor for recurrent myocardial infarcation. Lancet 1987; 2:3-9. 10. Ross R. George Lyman Duff Memorial Lecture. Atheroscerosis: a problem of the biology of arterial wall cells and their interactions with blood components. Arteriosclerosis 1981 ; 1:293-311. 11 . Russell RWR . Observations on the retinal blood-vessels in monocular blindness. Lancet 1961 ; 2:1 422-8. 12. Cella G, Zahavi J, de Haas HA, Kakkar W ./Hhromboglobulin, platelet production time and platelet function in vascular disease. Br J Haematol 1979; 43:127-36. 13. Dodson PM, Westwick J, Marks G, et al. /3-thromboglobulin and platelet factor 4 levels in retinal vein occlusion. Br J Ophthalmol 1983; 67: 1430-46. 14. Borsey DQ, Prowse CV, Gray RS, et al. Platelet and coagulation factors in proliferative diabetic retinopathy. J Clin Patho11984; 37:659-64. 15. Lowe GOO, Reavey MM, Johnston RV , et al. Increased platelet aggregates in vascular and non-vascular illness: correlation with plasma fibrinogen and effect of ancrod . Thromb Res 1979; 14:377-86. 16. Wu KK , Hoak JC. Spontaneous platelet aggregation in arterial insufficiency: mechanisms and implications. Thrombos Haemost 1976; 35:702-11 . 17. Hoyng PFJ, Greve EL, Frederikse K , et aI. Platelet aggregation and glaucoma. Doc Ophthalmol1985; 61 :167- 73.
18. The Canadian Cooperative Stroke Study Group. A randomized trial of aspirin and sulphinpyrazone in threatened stroke. N Engl J Med 1978; 299:53-9. 19. Miettinen OS. Matching and design efficiency in retrospective studies. Am J Epidemiol 1970; 91: 111-22. 20. Sackett DL. Bias in analytic research. J Chron Dis 1979; 32:51-63. 21. latridis SG, Ferguson JH. Active Hageman factor: a plasma lysokinase of the human fibrinolytic system. J Clin Invest 1962; 41 :1277-87. 22. Dawes J, Smith RC, Pepper OS. The release, distribution and clearance of human /3-thromboglobulin and platelet factor 4. Thromb Res 1978; 12:851-61. 23. Knox RA, Strandness DE. Ultrasound techniques for evaluation of lower extremity arterial occlusion . Semin Ultrasound 1981 ; 4:264-75. 24. Sumner DS. Mercury strain-gauge plethysmography. In: Bemstein EF, ed. Non-invasive Diagnostic Techniques in Vascular Disease, 2nd ed. SI. Louis: CV Mosby, 1982; 117-35. 25. Easthope PL, Brooks DE. A comparison of rheological constitutive functions for whole human blood. Biorheology 1980; 17:235-47. 26. Armitage P. Statistical Methods in Medical Research. Oxford: Blackwell, 1977; 189-268. 27. Colton T. Statistics in Medicine. Boston: Little, Brown and Company, 1974; 151-88. 28. Meade TW. The epidemiology of haemostatic and other variables in coronary artery disease. In: Verstraete M, Vermylen J, Lijnen R, Arnout J, eds. Thrombosis and Haemostasis. Leuven: University Press, 1987; 37-60 (Thrombosis Xlth Haemostasis).
55
Abstract: Increased intraocular pressure is accepted as a primary etiologic factor for the atrophy of the optic nerve head and visual field defects of hightension glaucoma. Other factors must be present to explain these findings in low-tension glaucoma. One of the current theories is that low-tension glaucoma is the result of decreased optic nerve perfusion on the basis of vascular disease or other factors such as altered blood viscosity. This study compared the noninvasive vascular profiles, coagulation tests, and rheological profiles of 46 consecutive cases of low-tension glaucoma with 69 similarly unselected cases of high-tension glaucoma and 47 age-matched controls. Despite the multifactorial approach and the use of previously validated objective tests, no significant group differences were detected with any of the above investigations. If vascular disease is important in the etiology of low-tension glaucoma, then it must be localized or vasospastic since this study does not support the concept of a generalized vascular etiology, either of an atheromatous or hyperviscous nature, for the genesis of low-tension glaucoma. Ophthalmology 1990; 97:49-55
Glaucoma is a condition characterized by excavation and atrophy of the optic nerve head and characteristic visual field defects. The condition was first recognized in the last century and, until recently, was thought to be almost invariably associated with increased intraocular pressure (lOP). It was hypothesized that increased lOP either decreased optic nerve perfusion, giving rise to the degenerative optic nerve changes, or produced mechanical pressure damage to the axons of the nerve head. 1,2 The development of applanation tonometry confirmed earlier suggestions that not all glaucomatous optic atrophy was associated with increased lOPs. This raised the possibility that either there were two separate diseases (i.e., hightension glaucoma and low-tension glaucoma) with entirely Originally received: June 5, 1989. Revision accepted: July 24, 1989. From the Departments of Pathology, Vascular Surgery, and Ophthalmology, University of British Columbia, Vancouver. Supported by a grant from the B. C. Health Care Research Foundation. Reprint requests to Cedric J. Carter, MB, Department of Laboratory Medicine, 2211 Westbrook Mall, Vancouver, Canada V6T 2B5.
different etiologies that happened to have the same pathologic appearances or that increased lOP was one of several risk factors for an underlying pathologic process. 3 This controversy remains unresolved. Some epidemiologic studies of low-tension glaucoma suggest that patients with low-tension glaucoma have an excess of ischemic cerebral and cardiovascular disease and that this may explain the ischemic optic nerve degeneration. 2 Further epidemiologic evidence for an ischemic etiology includes the age dependency oflow-tension glaucoma. Some series also have shown statistical associations with the presence of diabetes and vascular hypotension. An earlier study of 29 confirmed cases of low-tension glaucoma showed that these patients, relative to agematched controls, had positive histories of major vascular crises, including myocardial infarction and perioperative hypotension. 3 In addition, these patients appeared to have increased platelet adhesion and increased prestress euglobulin lysis times. The latter is a measure of fibrinolytic activity and is believed to be a reflection of free tissue plasminogen activator levels. This relatively small study did not include a control group of patients with hightension glaucoma.
49
OPHTHALMOLOGY
•
JANUARY 1990
Subsequent investigators looked at other factors known to be associated with more generalized arterial disease. Whole blood and plasma viscosity, which have been demonstrated to be abnormal in patients with myocardial infarction, angina, and intermittent claudication, were studied in glaucoma patients. 4 A large European study compared 83 patients diagnosed as having low-tension glaucoma with 23 patients with high-tension glaucoma and 50 control patients. s Higher whole blood and plasma viscosity values were shown in the patients with low-tension glaucoma relative to controls. In addition, plasma viscosity was higher in low-tension glaucoma patients relative to high-tension glaucoma cases. It is not clear how patients were selected for this study since there were many more low-tension glaucoma patients than high-tension glaucoma patients. If this were a true random selection of patients, the latter should have been far more common. Another unexplained feature of this study was the 25% incidence of ischemic vascular disease in the 50 control patients compared to the absence of ischemic vascular disease in the 23 high-tension glaucoma patients. This suggests that high-tension glaucoma in some way protects against vascular disease. From a biologic point of view, this is unlikely. Despite the methodologic limitations in some of the above studies, the overall conclusions are in keeping with a vascular contribution to the development of low-tension glaucoma. It is now possible to take other approaches to the documentation and demonstration of vascular disease. Since the earlier reports, major developments have occurred in noninvasive vascular diagnostic techniques. Measurements that can be made include segmental Doppler arterial pressures, strain gauge plethysmography for smaller vessels and digits, and duplex ultrasound examination of larger arteries. 6 None of these techniques has been applied systematically to patients with glaucoma. An alternative approach to the demonstration of generalized vascular disease is to look for the presence of various blood markers that have been associated with acute arterial events and more generalized atherosclerosis. In population surveys, elevated levels of fibrinogen, factor VII, and factor VIII were associated with an increased incidence of ischemic heart disease. 7 Another blood finding associated with generalized vascular disease is an increase in the euglobulin lysis time (ELT). In its current form, the time taken for a euglobulin fraction plasma clot to lyse is determined before the application of ischemic stress to the upper limbs. The normal population shows a shortening of the lysis time following ischemic stress to the forearm. Patients with a predisposition to, or who have developed vascular disease often show a relative decrease in the poststress shortening of the ELT. 8.9 Other investigators looked at markers of platelet activation. Platelet deposition is thought to be a fundamental event in the development of generalized atheroma. to In addition, in cerebrovascular disease, platelet microemboli have been demonstrated to cause a variety of ophthalmologic and cerebrovascular conditions. II Platelet microemboli are not easily demonstrated; however, it is possible to demonstrate an increase in markers of platelet 50
•
VOLUME 97
•
NUMBER 1
release, such as raised levels of beta thromboglobulin (BTG) in generalized vascular disease and in various ophthalmologic conditions including retinal vein occlusions. 12- 14 Another method that has been used in the investigation of suspected platelet-mediated vascular disease is to look for spontaneous platelet aggregation. This phenomenon has been reported in a variety of vascular diseases, including primary thrombocythemia, transient ischemic attacks, stroke, acute myocardial infarction, and peripheral vascular disease. ls . 16 When this technique was applied to a large group of patients with low-tension glaucoma and classic high-tension glaucoma, an association between spontaneous platelet aggregation and low-tension glaucoma was observed. 17 Defining the etiology of low-tension glaucoma is of therapeutic importance. If a clear relationship exists between vascular disease and low-tension glaucoma, then a case can be made for clinical trials of medications that have been shown to be effective in other forms of cerebrovascular disease (e.g., the use of an antiplatelet agent such as acetylsalicyclic acid).18 This study assembled a consecutive group of patients with high-tension and low-tension glaucoma together with an age- and sex-matched cohort with no evidence ofglaucoma. All patients and subjects underwent a complete noninvasive vascular assessment using segmental Doppler pressures and strain gauge plethysmography. Whole blood and plasma viscosities, antithrombin III, factor VII, factor VIII, and fibrinogen levels were measured. The vascular fibrinolytic status was examined by pre- and poststress ELT. Evidence for platelet activation was examined by measuring 24-hour urinary BTG excretion. Through the use of a variety of diagnostic modalities, including peripheral vascular testing, coagulation profiles, and rheological measurements, we investigated whether it was possible to show vascular differences between the glaucomatous and nonglaucomatous population.
MATERIALS AND METHODS PATIENTS AND STUDY SUBJECfS
One hundred sixty-two subjects were recruited. Sixtynine patients had been diagnosed as having primary openangle, high-tension glaucoma, with characteristic field defects, funduscopic appearances, and lOPs in excess of 21 mmHg. Forty-six were low-tension glaucoma patients as defined by field defects of classic glaucoma type, open anterior chamber angles, no history of uveitis or other ocular disease, and lOPs consistently 21 mmHg or less, including diurnal testing. Both groups of glaucoma patients were consecutive patients presenting for evaluation at the University of British Columbia glaucoma clinics. The 47 control subjects came from two sources. The majority were spouses of glaucoma patients. These subjects were chosen because there was no question of consanguinity and they would have been exposed to similar environmental factors. The remaining controls were patients admitted for elective cataract surgery in whom glaucoma
CARTER et al
•
ETIOLOGY OF LOW-TENSION GLAUCOMA
had been excluded. Because the underlying etiology for glaucoma is not known, no attempt was made to match control subjects with the patients for factors other than smoking and age. This avoided the problem of overmatching. 19 Informed consent was obtained for this research project from all patients and control subjects in accordance with the ethics approval granted by the University of British Columbia's Ethics Approval Committee.
surements were taken, and brachial to ankle pressure rations were calculated. These measurements were taken before and after exercise. The exercise regimen consisted of treadmill walking using a 10-minute Bruce protocol or until ischemic pain intervened. Small arterial vessels were evaluated by mercury strain gauge plethysmography pressure measurements of the fingers and toes using a Medsonic SPG 16 plethysmograph. 24
OPHTHALMOLOGIC TESTING
All control subjects had ocular evaluations that excluded the presence of high-tension or low-tension glaucoma. The glaucoma patients had diurnal tension studies and perimetry repeated. All laboratory and vascular investigations apart from the initial ophthalmologic documentation were performed without knowledge of the patient's disease category. Blind testing was done to avoid observer bias. 20 COAGULATION AND HEMATOLOGY TESTING
All subjects had a full peripheral blood count performed with various coagulation assays. Coagulant factor VII levels were done using a prothrombin time assay (Simplastin Automated, Organon Teknika, Durham, NC) and factor VII deficient plasma (Pacific Hemostasis, Ventura, CA). Coagulant factor VIII levels were determined by a onestage activated partial thromboplastin time (APTT) -based assay (Automated APTT, Organon Teknika, Durham, NC) and factor VIII deficient plasma (Pacific Hemostasis, Ventura, CA). Fibrinogen levels were determined using a thrombin clotting assay (American Dade, AHS, Puerto Rico). Antithrombin levels (ATIII) were done spectrophotometrically (Dupont ACA III, Dupont, Wilmington, DE). Euglobulin lysis times were performed both pre- and poststress. 21 The stress consisted of inflating a sphygmomanometer cuff on the upper arm to 90 mmHg and maintaining this pressure for 5 minutes. All blood samples for the coagulation assays were put into a standard buffered 3.8% sodium citrate anticoagulant (Becton Dickinson, Rutherford, NJ). A 24-hour urine specimen was taken and a BTG assay (Amersham, Arlington Heights, IL) performed using a modification of the standard plasma assay. 22 BIOCHEMICAL TESTING
Total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol and triglyceride levels were tested using spectrophotometric assays (Kodak Ektachem, Rochester, NY). Additional investigations included hemoglobin Ale levels to document biochemical evidence of glucose intolerance (Bio-Rad, Hercules, CA). VASCULAR TESTING
All subjects underwent noninvasive vascular testing. To assess the status of the larger arteries, segmental arterial pressures were determined using occlusive cuffs at two levels on the lower limb and one level on the upper limb using a Medsonic bi-directional DopplerY Direct mea-
RHEOLOGY TESTING
Plasma viscosity was determined using a CannonManning semi-micro type viscometer. Whole blood viscosity was examined using a Contraves LS-2 Couette viscometer with a concentric cylinder measuring system that had a stationary inner cylinder and a rotating outer cylinder. Measurements were done at 3rC, with nominal shear rates ranging from 0.031 S-l to 124.0 S-l and hematocrits ranging from 0.25% to 0.5% including the patient's own venous hematocrit. These tests produced a series of viscosity readings that were incorporated into an analytical procedure that generated a constitutive function that best described the rheological behavior of whole blood. 25 This function describes the shear rate and hematocrit dependence of viscosity in terms of four numerical parameters (Xt, X 2 , X 3 , and X 4 ) whose values completely describe the flow behavior of the sample within the operating limits used. These four parameters are independent of shear rate and hematocrit and can be used to numerically describe the rheological behavior of any given blood sample. The values of the four parameters determined for any sample are then compared statistically with various other individuals and clinical groupings to provide an objective comparison of the flow behavior of blood. STATISTICAL ANALYSIS
Analysis of quantitative data was made by comparison of variances. Because there were three patient groups, oneway analysis of variance was performed with the NewmanKeuls correction for comparing group means. 26 Analysis for the presence of vascular disease was made on the basis of absolute arterial readings and on the criteria of patients being designated as normal or abnormal according to previously determined threshold values of normalcy. The latter results were analyzed on the basis of 2 X K contingency table using the chi-square statistic. 27
RESULTS The ages of the comparative groups, male to female ratios, and the percentage of smokers are shown in Table 1. The three groups appear to be comparable with respect to mean age, age distribution, and smoking habits. Males appear to be slightly overrepresented in the high-tension glaucoma group. Since these were consecutive and unselected patients, there are no obvious explanations for this observation. 51
OPHTHALMOLOGY
•
JANUARY 1990
LTG
Controls
Age (yrs) Range
66 37-86
64 42-84
66 66-85
Male to female ratio
42:27
16:30
20:27
33
25
35
Percentage of smokers HTG
=
high-tension glaucoma; LTG
=
VOLUME 97
•
NUMBER 1
cholesterol ratios, and hemoglobin Ale levels are summarized in Table 3. No statistically significant differences were present.
Table 1. Patient Characteristics HTG
•
VASCULAR TESTS
Segmental arterial pressures. Segmental arterial pressures are expressed as ankle to brachial ratios. The group means are shown in Table 4. An alternative analysis of these data is to determine the proportion of patients with ratios below 1.0 (Table 4). No statistically significant differences were demonstrated among the three groups using either analytical approach. Strain gauge plethysmography. The results of the strain gauge pressures in the fingers and toes relative to brachial and tibial pressures are shown in Table 5. The first section of the table shows the mean pressure drops occurring in the brachial and digital arteries and tibial and toe arteries, respectively. The second section of the table shows the number of subjects with brachial-digital and tibial-toe arterial pressure decreases of greater than 40 mmHg and 70 mmHg, respectively. No statistically significant differences were detected after analysis by analysis of variance in the first section and by the chi-square statistic in the second section.
low-tension glaucoma.
COAGULATION STUDIES
Coagulation studies are summarized in Table 2. The various coagulation tests for each experimental group were compared using one-way analysis of variance with a Newman-Keuls multiple comparisons test. No statistically significant differences among the mean values were demonstrated. BIOCHEMICAL TESTING
Values obtained for cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, high-density lipoprotein cholesterol to total
Table 2. Coagulation Studies Mean ± 1 SO
LTG
HTG 1.148 1.110 3.075 2.87 418.7 282.8 36.4 1.04
Factor VII (U/ml) Factor Vlll:c (U/ml) Fibrinogen (g/I) BTG at 24 hrs Prestress EL T (mins) Poststress EL T (mins) Percent decrease' AT III (U/ml) SO = standard deviation; HTG III; EL T = euglobulin lysis time.
=
, Calculated with the following equation:
(pr~post) X re
1.158 ± 0.219 1.084 ± 0.385 3.126 ± 0.671 2.79 ± 0.27 441.8 ± 217.8 308.6 ± 243.5 35.8 ± 24.6 1.07 ± 0.13
± 0.219 ± 0.385 ± 0.671 ± 0.29 ± 23.5 ± 219.2 ± 24.6 ± 0.15
high-tension glaucoma; LTG
Controls
low-tension glaucoma; BTG
=
1.139 1.435 3.318 2.62 540.5 389.8 30.6 1.08
± 0.358 ± 0.431 ± 0.731 ± 0.25 ± 201.5 ± 242.7 ± 27.0 ± 0.12
beta thromboglobulin; AT III
=
=
antithrombin
100.
Table 3. Biochemical Tests Mean ± SO
LTG
HTG Total cholesterol (mmol/I) HOL cholesterol (mmol/I) LOL cholesterol (mmol/I) Triglycerides (mmol/I) HOL cholesterol/total cholesterol Hemoglobin Ale SO = standard deviation; HTG LOL = low-density lipoprotein.
52
=
5.80 1.86 3.84 1.67 0.21 0.053
± ± ± ± ± ±
high-tension glaucoma; LTG
0.88 0.31 0.90 1.07 0.07 0.005 =
5.89 1.202 3.39 1.90 0.21 0.060
± ± ± ± ± ±
low-tension glaucoma; HOL
Controls
1.29 0.42 1.51 1.35 0.08 0.037 =
5.64 1.31 3.55 1.72 0.24 0.054
high-density lipoprotein;
± ± ± ± ± ±
1.28 0.33 1.02 0.94 0.07 0.009
CARTER et al
•
ETIOLOGY OF LOW·TENSION GLAUCOMA
Table 4. Arterial Testing by Segmental Doppler
LTG
HTG
Ankle to brachial ratios (mean ±1 SD) Subjects with ratio 40 mmHg) Ankle/toe (>70 mmHg)
%8 2%8
%8 2%8
211a6
%6 1%6
1%1
11a6
%1 1°/31
71a1
HTG = high-tension glaucoma; LTG = low-tension glaucoma; R = right; L = left; SD = standard deviation. Table 6. Viscosity Testing-Constitutive Function
Parameter (mean ± 1 SD)
X1 X2 X3
X4
Plasma viscosity (mean + 1 SD)
HTG (n = 69)
LTG (n
2.356 ± 1.4 5.1 ± 1.1 0.53 ± 0.05 0.003 1.27 ± 0.6
2.16 ± 1.5 4.6 ± 1.3 0.51 ± 0.12 -0.003 1.29 ± 0.9
=
45)
Controls (n
=
47)
2.21 ± 1.5 5.22 ± 1.0 0.53 ± 0.06 0.001 1.27 ± 0.8
HTG = high-tension glaucoma; LTG = low-tension glaucoma; SD = standard deviation. RHEOLOGY
The comparative values of the four discriminant parameters derived from whole blood viscometry and the plasma viscosity values are shown in Table 6. No significant differences were demonstrated among the group means. For comparison with other studies that have not used the constitutive function approach to the characterization of blood flow properties, the whole blood viscosity is shown at the nominal shear rates of 19.7 and 124 S-1 at native hematocrit (Table 7). It should be noted that because blood is a non-Newtonian fluid, these are the nominal shear rates as determined by the instrument setting rather than the actual shear rates occurring in the test system. Nominal shear rates have been used by previous investigators.
DISCUSSION The exact pathogenesis of the glaucomas remains unanswered. It is clear that an elevated lOP is an important
risk factor that also has a causal relationship to glaucomatous damage. There are, however, many optic nerves that do not appear to be damaged by prolonged increased lOP. In other cases, marked and progressive damage occurs despite a statistically normal or even low lOP. Other factors must be present, and a comparison oflow-tension glaucoma with high-tension glaucoma is an obvious area of study to demonstrate such factors. Vascular etiologies have been implicated in the development of the glaucomas. The vascular factors could be organic or vasospastic or combinations of these. In other organic vascular pathologies, noninvasive diagnostic techniques and various blood markers, including elevated fibrinogen, factor VII, factor VIII, euglobulin lysis times, and markers of platelet activation have shown abnormalities. Iflow-tension glaucoma has an organic vascular basis like transient ischemic attack, myocardial infarction, stroke, and peripheral ~~s cular disease, one would expect diagnostic abnormahties ip patients with low-tension glaucoma. Because there is no direct method for definitively measuring in vivo the blood supply provided by ~he ophthalmic artery and its distal branches, documentatIOn 53
OPHTHALMOLOGY
•
JANUARY 1990
Table 7. Viscosity Testing Mean ± SO Shear Rate
HTG (n = 69)
LTG (n = 46)
Controls (n = 63)
Peak 19.7 S·l Steady state 124 S-l
10.09 ± 1.8 8.85 ± 1.6 5.87 ± 0.7
9.78 ± 2.0 8.47 ± 1.8 5.73 ± 0.9
9.34 ± 1.8 8.21 ± 1.6 5.59 ± 0.7
SO = standard deviation; HTG LTG = low-tension glaucoma.
=
high-tension glaucoma;
of vascular factors in low-tension glaucoma must be by indirect investigations such as we have undertaken. Compared with previous publications, this is one of the more extensive studies. Care was taken to avoid bias in selection of patients and controls. An obligate consequence of hospital or specific referral center-based studies is the generation of false associations. For example, if one condition initiates a hospital visit, then nonspecific laboratory and investigational findings will detect other, but not necessarily directly related, conditions. The selection of appropriate control groups helps prevent this source of bias. Two control groups were selected for this study. The first control group comprised nonglaucomatous subjects who were selected to allow an age match. The second control group consisted of carefully documented consecutive cases of high-tension glaucoma. By chance, they happened to be of similar age to the low-tension glaucoma cohort. Selection to permit a sex match was not possible since in the glaucoma patients there was a different sex distribution between the high- and low-tension glaucoma cases. As such, with the exception of sex, the experimental groups show no evidence of major imbalance or any obvious selection artifacts. The underlying rationale of this study was that if there was a major vascular component to low-tension glaucoma, one should be able to demonstrate this by techniques that had already been used to implicate a vascular etiology in low-tension glaucoma or other defined vascular pathologies. Because increased lOP could be an additional risk factor for a primary glaucomatous predisposition, as exemplified by low-tension glaucoma, a high-tension glaucoma group was an important comparison group. If the above hypothesis was correct, then the high-tension glaucoma test results might be expected to fall between the low-tension glaucoma and control results. The similarity of the vascular laboratory results between the comparative groups excludes major differences in overall vascular pathology but cannot exclude the presence of some highly localized process. Earlier studies relied on clinical history to determine vascular status. There could be a problem with the relative sensitivity of the objective vascular tests used in this study, but these same tests have been validated in most of the cardiovascular conditions previously described as being associated with low-tension glaucoma.6.23.24 The coagulation studies were not informative. Due to problems of test reproducibility, we were not able to repeat 'i4
•
VOLUME 97
•
NUMBER 1
the earlier studies that examined spontaneous platelet aggregation and platelet adhesion, but we were able to measure clotting factors VII and VIII and fibrinogen. Recent studies have confirmed that raised levels offactor VII and fibrinogen are present in people with atherosclerosis.z s The magnitude of these differences are in the region of 10% and should have been detectable in a study of this size. Similarly, no significant group differences were observed in 24-hour urinary BTG excretion. All these negative findings are in keeping with the results of the direct vascular testing. In view of the close association between lipids and vascular disease, a full lipid screen was performed, but, again, no group differences were demonstrated. The final series of investigations were the studies on plasma and whole blood viscosity. Two previous studies implicated increased blood viscosity in the genesis oflowtension glaucoma. We were unable to confirm these findings. It should be noted that our method of analyzing whole blood viscosity is fundamentally different from that of previous investigators, and there is no data on the comparative sensitivities of various rheological techniques for the investigation of vascular disease. However, data obtained using a more traditional approach (Table 7) did not show any distinction among the groups. Although differences exist between this study and earlier studies in terms of specific methods, the study was balanced and had appropriate controls and adequate statistical power. The lack of any positive findings for incrp :4sed atherosclerosis or abnormal rheology is important because if there is any vascular basis for low-tension glaucoma, then it is mediated through other factors such as vasospasticity. In summary, this study failed to demonstrate group differences for the markers of atherosclerotic vascular disease among patients with high-tension and low-tension glaucoma and nonglaucomatous controls. Similarly, no rheologic differences were detected. Using the above methods and statistical analysis, this study does not support an organic vascular pathology for low-tension glaucoma.
REFERENCES 1. Grehn F, Prost M. Function of retinal nerve fibers depends on perfusion pressure: neurophysiologic investigations during acute intraocular pressure elevation. Invest Ophthalmol Vis Sci 1983; 24:347-53. 2. Spaeth GL. Fluorescein angiography: its contributions towards understanding the mechanisms of visual loss in glaucoma. Trans Am Ophthalmol Soc 1975; 73:491-553. 3. Orance SM, Sweeney VP, Morgan RW, Feldman F. Studies of factors involved in the production of low tension glaucoma. Arch Ophthalmol 1973; 89:457-65. 4. Trope GE, Salinas RG, Glynn M. Blood viscosity in primary openangle glaucoma. Can J Ophthalmol1987; 22:202-4. 5. Klaver JHJ, Greve EL, Goslinga H, et al. Blood and plasma viscosity measurements in patients with glaucoma. Br J Ophthalmol 1985; 69: 765-70. 6. Barnes RW. Noninvasive diagnostic techniques in peripheral vascular disease. Am Heart J 1979; 97:241-58. 7. Meade TW, North WRS, Chakrabarti R, et al. Haemostatic function and cardiovascular death: early results of a prospective study. Lancet 1980; 1: 1050-4.
CARTER et al
•
ETIOLOGY OF LOW-TENSION GLAUCOMA
8. Balkau-Ulutsin S. Fibrinolytic system in atherosclerosis. Semin Thromb Hemost 1986; 12:91-101 . 9 . Hamsten A, De Faire U, Walldius G, et al. Plasminogen activator inhibitor in plasma: risk factor for recurrent myocardial infarcation. Lancet 1987; 2:3-9. 10. Ross R. George Lyman Duff Memorial Lecture. Atheroscerosis: a problem of the biology of arterial wall cells and their interactions with blood components. Arteriosclerosis 1981 ; 1:293-311. 11 . Russell RWR . Observations on the retinal blood-vessels in monocular blindness. Lancet 1961 ; 2:1 422-8. 12. Cella G, Zahavi J, de Haas HA, Kakkar W ./Hhromboglobulin, platelet production time and platelet function in vascular disease. Br J Haematol 1979; 43:127-36. 13. Dodson PM, Westwick J, Marks G, et al. /3-thromboglobulin and platelet factor 4 levels in retinal vein occlusion. Br J Ophthalmol 1983; 67: 1430-46. 14. Borsey DQ, Prowse CV, Gray RS, et al. Platelet and coagulation factors in proliferative diabetic retinopathy. J Clin Patho11984; 37:659-64. 15. Lowe GOO, Reavey MM, Johnston RV , et al. Increased platelet aggregates in vascular and non-vascular illness: correlation with plasma fibrinogen and effect of ancrod . Thromb Res 1979; 14:377-86. 16. Wu KK , Hoak JC. Spontaneous platelet aggregation in arterial insufficiency: mechanisms and implications. Thrombos Haemost 1976; 35:702-11 . 17. Hoyng PFJ, Greve EL, Frederikse K , et aI. Platelet aggregation and glaucoma. Doc Ophthalmol1985; 61 :167- 73.
18. The Canadian Cooperative Stroke Study Group. A randomized trial of aspirin and sulphinpyrazone in threatened stroke. N Engl J Med 1978; 299:53-9. 19. Miettinen OS. Matching and design efficiency in retrospective studies. Am J Epidemiol 1970; 91: 111-22. 20. Sackett DL. Bias in analytic research. J Chron Dis 1979; 32:51-63. 21. latridis SG, Ferguson JH. Active Hageman factor: a plasma lysokinase of the human fibrinolytic system. J Clin Invest 1962; 41 :1277-87. 22. Dawes J, Smith RC, Pepper OS. The release, distribution and clearance of human /3-thromboglobulin and platelet factor 4. Thromb Res 1978; 12:851-61. 23. Knox RA, Strandness DE. Ultrasound techniques for evaluation of lower extremity arterial occlusion . Semin Ultrasound 1981 ; 4:264-75. 24. Sumner DS. Mercury strain-gauge plethysmography. In: Bemstein EF, ed. Non-invasive Diagnostic Techniques in Vascular Disease, 2nd ed. SI. Louis: CV Mosby, 1982; 117-35. 25. Easthope PL, Brooks DE. A comparison of rheological constitutive functions for whole human blood. Biorheology 1980; 17:235-47. 26. Armitage P. Statistical Methods in Medical Research. Oxford: Blackwell, 1977; 189-268. 27. Colton T. Statistics in Medicine. Boston: Little, Brown and Company, 1974; 151-88. 28. Meade TW. The epidemiology of haemostatic and other variables in coronary artery disease. In: Verstraete M, Vermylen J, Lijnen R, Arnout J, eds. Thrombosis and Haemostasis. Leuven: University Press, 1987; 37-60 (Thrombosis Xlth Haemostasis).
55
