Journal of Neuro-Ophthalmology, September 1993, Volume 13, Issue 3

Carotid artery disease in vascular ocular syndromes.

We prospectively investigated 83 consecutive patients with vascular ocular syndromes: 19 suffered from amaurosis fugax attacks, 23 had occlusions of the central retinal artery or a branch retinal artery occlusion, 26 had a central retinal vein occlusion or a branch retinal vein occlusion, and another 15 exhibited an anterior ischemic optic neuropathy. In 5 patients bilateral symptoms occurred; thus a total of 88 eyes were affected. All patients underwent a neurological examination and ultrasound investigations of the carotid arteries, including continuous wave (cw)-Doppler-sonography and duplex ultrasound. Stenosis of more than 50% diameter reduction and occlusion of the internal carotid artery ipsilateral to the symptomatic eye were significantly more frequent in amaurosis fugax attacks and central or branch retinal artery occlusion than in central or branch retinal vein occlusion or anterior ischemic optic neuropathy (p < .025). Additionally, the analysis of plaque surface and echogenicity of the plaques on the affected side with a high-resolution duplex scan uncovered that ulcerated plaque surfaces and plaques with a heterogeneous echogenicity were found significantly more frequent in the internal carotid arteries of patients with amaurosis fugax attacks and central or branch retinal artery occlusions than in patients with anterior ischemic optic neuropathy (p < .04) or central and branch retinal vein occlusion (p < .025). We conclude that amaurosis fugax attacks and central retinal artery or branch retinal artery occlusions are due to arterio-arterial embolization from ulcerated and heterogeneous carotid artery plaques.(ABSTRACT TRUNCATED AT 250 WORDS)

Journal of Clinical Neuro-ophthalmology 13(3): 175-180, 1993. Carotid Artery Disease in Vascular Ocular Syndromes M. Muller, M.D., K. Wessel M.D., E. Mehdorn, M.D., D. Kampf, M.D., and C. M. Kessler, M.D. © 1993 Raven Press, Ltd., New York We prospectively investigated 83 consecutive patients with vascular ocular syndromes: 19 suffered from amau­rosis fugax attacks, 23 had occlusions of the central ret­inal artery or a branch retinal artery occlusion, 26 had a central retinal vein occlusion or a branch retinal vein occlusion, and another 15 exhibited an anterior ischemic optic neuropathy. In 5 patients bilateral symptoms oc­curred; thus a total of 88 eyes were affected. All patients underwent a neurological examination and ultrasound investigations of the carotid arteries, including continu­ous wave (cw)-Dopplersonography and duplex ultra­sound. Stenosis of more than 50% diameter reduction and occlusion of the internal carotid artery ipsilateral to the symptomatic eye were significantly more frequent in amaurosis fugax attacks and central or branch retinal artery occlusion than in central or branch retinal vein occlusion or anterior ischemic optic neuropathy (p < .025). Additionally, the analysis of plaque surface and echogenicity of the plaques on the affected side with a high-resolution duplex scan uncovered that ulcerated plaque surfaces and plaques with a heterogeneous echogenicity were found significantly more frequent in the internal carotid arteries of patients with amaurosis fugax attacks and central or branch retinal artery occlu­sions than in patients with anterior ischemic optic neu­ropathy (p < .04) or central and branch retinal vein oc­clusion (p < .025). We conclude that amaurosis fugax attacks and central retinal artery or branch retinal artery occlusions are due to arterioarterial embolization from ulcerated and heterogeneous carotid artery plaques. Be­cause of the proven sensitivity of Duplex ultrasound and cw-Dopplersonography to diagnose atherosclerotic ca­rotid artery disease, these techniques should be used in order to evaluate the patients with retinal arterial isch­emia. Key Words: Vascular ocular syndromes-Carotid ath­erosclerosis- Carotid ultrasound examination. From the Departments of Neurology and Ophthalmology (E.M.), Medical University of Lubeck, Germany. Address correspondence and reprint requests to Dr. C. M. Kessler, Department of Neurology, Medical University of Lubeck, Ratzeburger Allee 160, Lubeck 1, Germany. 175 In 1952, Fisher (1) was the first who described patients with transient monocular blindness and subsequent contralateral hemiparesis. Since then, vascular eye syndromes as heraldic symptoms of cerebral ischemia have raised increasing interest (2-8). Whereas amaurosis fugax attacks with tran­sient monocular blindness have been recognized to be due to cholesterol or platelet emboli (7,9-11) mostly originating from atherosclerotic lesions of the carotid bifurcation, the pathogenesis of other ischemic ocular syndromes with sudden onset, as, for example, central retinal artery occlusion or branch retinal artery occlusion, is still uncertain. The discussion of the cause of the latter ischemic eye diseases centers on embolization either from carotid arteries or from the heart, local thrombosis, or hemodynamic mechanisms due to low perfu­sion (3,4,9,12,13). Other vascular eye diseases, such as the anterior ischemic optic neuropathy, the central retinal vein occlusion, or the branch retinal vein occlusion, are also associated with vascular risk factors, but the etiology of these circulatory eye diseases is still not certain (1~20). To evaluate the pathogenetic role of extracranial carotid atherosclerotic stenoses in various vascular ocular syndromes, we prospectively examined the carotid arteries of 83 patients with vascular eye syndromes with continuous wave Doppler and duplex ultrasound. MATERIAL AND METHODS Patients We studied 83 consecutive patients (46 male, 37 female, mean age 67 ± 15 years) suffering from vascular ocular diseases. Patients with a history of migraine, vasculitis, chronic inflammatory disease of the central nervous system, or glaucoma were excluded from the study. In the 83 patients evalu­ated, 88 eyes were affected. 176 M. MULLER ET AL. Nineteen patients experienced amaurosis fugax attacks in 20 eyes (once bilaterally at different times). The frequency of amaurosis fugax attacks occurring in each eye was once in 12 eyes, twice in 2 eyes, three, four, five, and seven times in 1 eye, and more than 10 times in 2 eyes. Amaurosis fugax attacks were diagnosed according to the definition of The Amaurosis Fugax Study Group (21), when patients described a diminished or absent vision in one eye that progressed over a few seconds and usually lasted a few seconds or a few minutes. Central retinal artery occlusion had occurred in 12 patients and branch retinal artery occlusion in 11 patients. Central retinal artery occlusion was diag­nosed when a sudden painless loss of vision oc­curred in one eye with the ophthalmoscopically typical findings of a pale retina, the narrowing of the retinal arteries, a swollen optic disc, and a mac­ular cherry red spot; branch retinal artery occlu­sion was diagnosed by the demonstration of a pale ischemic territory of one branch retinal artery (4,9, 13). Fifteen patients suffered from anterior ischemic optic neuropathy in 17 eyes (2 bilaterally). An an­terior ischemic optic neuropathy was diagnosed in patients with an acute nonprogressive loss of vi­sion with fundoscopically proven pallid disc swell­ing, with or without hemorrhages (14,15). Eleven patients had central retinal vein occlu­sions, another 15 patients branch retinal vein oc­clusions in 17 eyes (2 bilaterally). These diagnoses were based on severe visual impairment accompa­nied by venous engorgement and intraretinal and peripapillary hemorrhages (18-20). All patients underwent clinical examinations by both a neurologist and an ophthalmologist. Cere­brovascular risk factors and preexisting cerebro­vascular or cardiac diseases (previous myocardial infarction, angina pectoris, and atrial fibrillation) were registered. The ophthalmological examina­tion included the measurements of the visual acu­ity, the visual field using the Goldmann scheme, the intraocular pressure, and an ophthalmoscopic examination. The neurological examination was focused on transient or persistent hemispheric syndromes ipsilateral to the affected eyes and was classified as transient ischemic attack, minor stroke without a permanent functional neurological defi­cit, or major stroke with a permanent handicap­ping neurological deficit. Carotid Artery Ultrasound Examination The carotid arteries of all patients were exam­ined by both the continuous wave Doppler ultra- JCll/1 Neuro-ophtllalmol, Vol. 13, No.3, 1993 sound (Velocimetre 0.800, Delalande, Paris), us­ing a 4-MHz probe, and the duplex ultrasound (Sonotron 300, Sonotron, Cologne). Continuous wave Doppler recognizes extracranial stenoses if the narrowing of the artery diameter exceeds 50% (22). Pathologic findings in the internal carotid ar­tery were classified into normal, stenotic (reduc­tion of 50% to 99% of the vessel's diameter), and occlusion of the internal carotid artery. The duplex ultrasound examination was performed with a 7.5­MHz probe by visualizing the longitudinal axis as well as the cross section of the carotid bifurcation and the internal carotid artery. The internal carotid artery was distinguished from the external carotid artery by the typical pulsed wave Doppler signal from both arteries and the changing of the flow pattern in the external carotid artery induced by rhythmic compression of the superior temporal ar­tery (22). Nonstenotic atherosclerotic plaques and stenotic atherosclerotic lesions in the internal ca­rotid artery and the common carotid artery were classified as follows (23): 1. The plaque surface was differentiated into smooth or ulcerated. A smooth plaque surface was diagnosed if no disconnection of the covering layer was obvious, plaques with an interrupted surface were classified as ulcerated. Calcified plaques with echo shadowing where classified as undetermined regarding their surface. 2. The plaque structure was classified as homo­geneous if it consisted mostly of a medium or high level of echogenicity, and as heterogeneous if it appeared with low, medium, or high echointensi­ties. A classification of plaques according to these crite­ria correlates well to the histological finding (24­27) of mainly fibrotic, stable plaques with a smooth plaque surface and a homogeneous echogenicity (Fig. 1) and unstable, complicated plaques with an ulcerated plaque surface and a heterogeneous echogenicity (Fig. 2). Stastical Analysis The statistical analysis was performed with stu­dent's t-test and Fisher's exact test. RESULTS There was no significant age difference between the different diagnostic groups (Table 1). Hyper­tension was the major vascular risk factor in all ~iagnos~ic groups; it was found in 42% of all pa­tients WIthout prevalence for a special group (Ta- CAROTID DISEASE IN VASCULAR OCULAR SYNDROME 177 FIG. 1. Duplex ultrasound image of the carotid bifurcation showing a stable plaque (arrow) with smooth surface and a homogeneous echoge­nicity. ble 1). Seventeen (20%) of the 83 patients suffered from diabetes mellitus, 16 (19%) had hyperlipid­emia; again, there was no prevalence for one diag­nostic group. Eighteen (21 %) of the 83 patients had ischemic heart disease (previous myocardial infarction and angina pectoris) and 9 patients (9.5%) had atrial fibrillation, both conditions well known for their risk of cardiogenic embolism (28). None of the pa­tients previously experienced an ipsilateral major hemispheric stroke; hemispheric transient isch­emic attacks or minor strokes ipsilateral to the af­fected eyes were found in 14 patients: in 5 of the 20 cases with amaurosis fugax attacks (25%), in 5 of the 23 cases with central or branch retinal artery occlusion (21 %), in 2 of the 17 cases with anterior ischemic optic neuropathy (11 %), and in 3 of the 28 cases with central or branch retinal vein occlusion (9.5% ). Table 2 shows the results of the continuous wave Doppler and the duplex ultrasound exami­nations: we identified 57 pathological ultrasound findings ipsilateral to the affected eyes: 40 low­degree stenoses «50%), and 13 high-degree (>50%) stenoses. Further, we diagnosed 4 occlu­sions of the ipsilateral internal carotid artery. Most ipsilateral atherosclerotic lesions were found in the 19 patients with the 20 amaurosis fugax attacks; in 16 (80%) of these 20 internal ca­rotid arteries we found 9 high-degree stenoses and 7 low-degree stenoses. Also, patients with retinal infarctions had a high incidence of atherosclerotic carotid artery disease (18 of 23: 77%) with a dom­inance of low-degree stenoses; in this group we found 11 low-degree stenoses, 3 high-degree stenoses, and all 4 of the internal carotid artery occlusions. Nevertheless, there was no statistical difference between the patients with amaurosis fuga x attacks and the patients with central or branch retinal artery occlusion with regard to the incidence of atherosclerotic internal carotid artery lesions. The incidence of ipsilateral internal carotid artery stenoses and occlusions was significantly lower in patients with anterior ischemic optic neu­ropathy or central and branch retinal vein occlu­sion as compared to the patients with amaurosis fugax attacks or with central or branch retinal ar­tery occlusion (p = .01). High-degree stenoses or occlusions of the ipsilateral internal carotid artery were also significantly lower in patients with an- FIG. 2. Duplex ultrasound of a heterogeneous ulcer­ated plaque at the internal carotid artery origin. bulb, carotid bulbus; ICA, internal carotid artery; CCA, com­mon carotid artery. / Clill Neuro-ophthallllol, Vol. 13, No.3, 1993 178 M. MULLER ET AL. TABLE 1. Mean age, sex, and vascular risk factors in patients with vascular ocular syndromes Mean age Males Females Hypertension Diabetes mellitus Hyperlipidemia Atrial fibrillation Ischemic heart disease Amaurosis fugax (n = 19) 64:!: 9 13 6 6 2 4 25 Central/branch retinal artery occlusion (n = 23) 71 :!: 9 10 13 12 75 2 5 Anterior ischemic optic neuropathy (n = 15) 66 :t 16 8 7 6 4 2 13 Central/branch retinal vein occlusion (n = 26) 69 :t 17 15 11 12 4 5 45 DISCUSSION Amaurosis fugax attacks are transient retinal ischemic conditions, whereas central or branch ret-terior ischemic optic neuropathy or central or branch retinal vein occlusion as compared to the patients with amaurosis fugax attacks (p = .001) or with central or branch retinal artery occlusion (p = .03). The analysis of the plaque morphology (Table 3) showed a high prevalence of ulcerated and hetero­geneous plaques in patients with amaurosis fugax attacks and central or branch retinal artery occlu­sion as compared to the patients with anterior isch­emic optic neuropathy (p < .005 for ulcerated plaque surface, p < .04 for heterogeneous echoge­nicity) and central or branch retinal artery occlu­sion (p < .025 for ulcerated plaque surface; p < .01 for heterogeneous echogenicity). In amaurosis fugax attacks and central or branch retinal artery occlusion plaques mostly represented the unstable type with ulcerated surface and heterogeneous echogenicity. Of the 19 patients with amaurosis fugax attacks, only 3 showed normal carotid arteries; 2 of them had ischemic heart disease as a possible source of embolism. Ischemic heart disease and atrial fibril­lation concomitantly with normal carotid arteries were found in 4 of the 23 patients with central or branch retinal artery occlusions and in 2 of the 15 patients with anterior ischemic optic neuropathy. inal artery occlusions represent an irreversible ret­inal infarct. Both conditions are suspected to have the same pathogenesis. Several angiographic stud­ies demonstrated a high incidence of ipsilateral ca­rotid atherosclerosis in patients with amaurosis fugax attacks and central or branch retinal artery occlusion (2-5,7-10,12,29-33). The frequent obser­vation of passing platelet aggregates or cholesterol emboli (9-11,33) has led to the assumption that retinal ischemia is most often due to arterioarterial embolization derived from ipsilateral carotid ath­erosclerotic lesions. In the present study we found a high correlation between ultrasonographically detectable internal carotid artery atherosclerosis and amaurosis fugax and central or branch retinal artery occlusion. In addition we found a signifi­cantly higher number of plaque ulcerations as well as of plaques with a heterogeneous echogenicity in these groups of patients as compared to the other vascular eye diseases, such as anterior ischemic optic neuropathy and central or branch retinal vein occlusion. The comparison between ultrasound morphol­ogy and histological findings proved that the B-mode ultrasound technique is highly sensitive for the assessment of carotid plaque ulcerations (24,25), which are closely associated with ipsilat­eral cerebral hemispheric ischemia due to embo­lization (26,34,35). Apart from the plaque surface, B-mode ultrasound also characterizes the structure of carotid plaques. The plaques with a homoge- TABLE 2. Ultrasound findings (continuous wave Doppler, duplex ultrasound) of the ipsilateral internal carotid artery in 83 patients with vascular ocular syndrome Ultrasound finding Stenosis < 50% Stenosis> 50% Occlusion Total Amaurosis fugax (n = 20) 7 9' 16" Central/branch retinal artery occlusion (n = 23) 11 3 4 18" Anterior ischemic optic neuropathy (n = 17) 6 6 Central/branch retinal vein occlusion (n = 28) 16 1 17 , P "" .025: for Amaurosis fugax versus anterior ischemic optic neuropathy and central/branch retinal vein occlusion. " p = .01: for amaurosIS fugax and central/branch retinal artery occlusion versus anterior ischemic optic neuropathy. I (/ill Neuro-ophtiullmol. Vol. 13. No.3. 1993 CAROTID DISEASE IN VASCULAR OCULAR SYNDROME TABLE 3. Ultrasound characteristics of ipsilateral internal carotid artery plaques 179 Central/branch retinal Anterior ischemic Central/branch retinal Amaurosis fugax artery occlusion optic neuropathy vein occlusion Plaquemorphology (n = 16)" (n = 14)" (n = 6) (n = 17)" Surface Smooth 4 4 6 12 Ulcerated 10' r 3 Echogenicity Homogeneous 5 5 5 14 Heterogeneous 11" 9" 1 3 " In 2 cases of amaurosis fugax attacks, in 3 of central/branch retinal artery occlusions, and in 2 of central/branch retinal vein occlUSions the plaque surface could not be determined due to echoshadowing. • p "" .025; "p "" .04: for amaurosis fugax and central/branch retinal artery occlusion versus anterior ischemic optic neuropathy and central/branch retinal vein occlusion. neous echogenicity are histologically shown to consist of homogeneous fibrotic tissue, whereas the ultrasound appearance of a heterogeneous plaque corresponds to plaque hemorrhage that is associated with a mixed histology, including cellu­lar infiltrates, cholesterol detritus, or necrotic ma­terial (24-26). In a previous study (36), we found that most of the heterogeneous plaques were also ulcerated and that plaque hemorrhage developed mainly by blood inflow from the lumen through the damaged plaque surface. It is thus necessary to distinguish between stable plaques with a homo­geneous plaque structure and a smooth surface and unstable plaques characterized by a mixed plaque structure and an ulcerated plaque surface. Nearly all carotid plaques in anterior ischemic op­tic neuropathy and central or branch retinal vein occlusion were of the stable type, which suggests there is no pathophysiological association between the carotid artery plaques and these eye diseases. Additionally, most hemispheric transient ischemic attacks and minor strokes occurred distally to the unstable carotid plaques of the patients with amaurosis fugax attacks and central or branch ret­inal artery occlusions, thus proving their high risk of subsequent embolization. In the group of pa­tients with central or branch retinal vein occlusion only one ipsilateral internal carotid artery stenosis of more than 50% was found. The remaining 12 internal carotid artery stenoses of more than 50% were related to the retinal ischemic syndromes amaurosis fugax, and central or branch retinal ar­teryocclusion. We found in all patients with internal carotid artery occlusions ipsilateral either central or branch retinal artery occlusions. On the other hand, there was also a large number of low-degree stenoses in this patients group. This finding proves that ulcer­ated and heterogeneous low-degree carotid stenoses posses an increased risk of embolization. The observation that embolization from ulcerated low-degree stenosis is mainly platelet dependent (37) is confirmed by experimental data showing the development of platelet thrombosis following platelet activation due to high shear stress and contact with the subendothelial structure in ulcer­ated areas (38). Thus, patients with amaurosis fugax and central or branch retinal artery occlusion should be routinely investigated by ultrasound ex­amination of the carotid artery in order to select patients not only at a high risk of recurrent ocular ischemia but also at risk of developing cerebral in­farction, so that an appropriate treatment, either carotid surgery or antiplatelet medication, can be indicated. Ultrasound has the advantage of being noninvasive, without any side effects, whereas in­tra- arterial angiography can be associated with complications (39). The use of MR angiography (40) still remains limited because of the uncertainty of estimating the degree of stenosis as well as plaque surface characteristics. The frequency of retinal ischemia due to cardio­genic embolism is unknown (28,30,33). In our study, we found the same incidence of atrial fibril­lation or ischemic heart disease in all examined groups of patients, which not only reflects the high average age of all patients but also the incidence of vascular risk factors within these patients (8­10,12,15,19,20,29,33). As shown in a previous paper (41), delayed vi­sual evoked potentials are a common finding in patients with anterior ischemic optic neuropathy (AlaN). This fact reinforces the hypothesis that the anterior ischemic optic neuropathy represents the microangiopathic type of eye ischemia with oc­clusion of the small vessels supplying the optic nerve, often resulting in demyelinization (14,15). As AlaN probably represents a small vessel dis­ease, its significant correlation with cerebrovascu­lar and cardiovascular events (41,42) emphasizes that AlaN is an important marker of a systemic vascular disease. , Clin Neuro-ophthalmol. Vol. 13. No.3. 1993 180 M. MULLER ET AL. REFERENCES 1. Fisher M. Transient monocular blindness associated with hemiplegia. Arch Ophthalmol 1952;47:167-203. 2. Marshall I. Meadows S. The natural history of amaurosis fugax. Brain 1968;91:419-33. 3. Lorentzen SE. Occlusion of the central retinal artery: a fol­low up. Acta Ophthalmol 1969;47:690-703. 4. Kollartis CR, Lubow M, Hissong SL. Retinal stroke: I. In­cidence of carotid atheroma. JAMA 1972;222:1273-5. 5. Hooshmand H, Vines FS, Lee HM, Grindal A. Amaurosis fugax: diagnostic and therapeutic aspects. Stroke 1975;5: 643-7. 6. Pessin MN, Duncan GW, Mohr JP, Poskanzer DC. Clinical and angiographic features of carotid transient ischemic at­tacks. N Engl J Med 1977;296:358-62. 7. Adams PH, Putman SF, Corbett J], Sires BP, Thompson HS. Amaurosis fugax: the results of arteriographie in 59 patients. Stroke 1983;17:742-4. 8. Toole C]M, Ross Russell RW. Mortality and stroke after amaurosis fugax. J Neurol Neurosurg Psychiatry 1985;48: 902-5. 9. Hollenhorst RW. Vascular status of patients who have cho­lesterol emboli in the retina. Am J OphthalmoI1966;61:1159­64. 10. Pfaffenbach DO, Hollenhorst RW. Morbidity and survivor­ship of patients with embolic cholesterol crystals in the oc­ular fundus. Am J Ophthalmol 1973;75:66-72. 11. Kessler Ch, Reuther R, Rosch M. Dual isotope carotid scin­tigraphy in patients with amaurosis fugax attacks. Eur Arch Psychiatr Neurol Sci 1984;234:106-11. 12. Savino PI. Glaser JS, Cassady J. Retinal stroke: is the patient at risk? Arch Ophthalmol 1977;95:1185-9. 13. von Graefe A. Uber Embolie der Arteria centralis retinae als Ursache plotzlicher Erblindung. Albrecht von Graefes Arch OphthalmoI1859;5:136-40. 14. Hayreh SS. Anterior ischaemic optic neuropathy. Arch NelI­roI1981; 38:675-8. 15. Boghen DR, Glaser JS. Ischaemic optic neuropathy. The clinical profile and natural history. Brain 1975;98:689-708. 16. Gmeiner H-L Kompf 0, Ruprecht KW. Ischamisch bed­ingte Augenerkrankungen und extrakranielle Carotiss­tenosen/- verschliisse. In: Fischer P-A, Baas H, Enzens­berger W, eds. Verhandlungen der Deutschen Gesellschaft fur Neurologie. Bd. 5. Springer-Verlag, Berlin, 1989:1044-7. 17. Bogousslavsky I. Regli F, Zografos L, Uske A. Optico­cerebral syndrome: simultaneous hemodynamic infarction of optic nerve and brain. Neurology 1987;37:263-8. 18. Weinberg 0, Dodwell DG, Fern SA. Anatomy of arteriove­nous crossings in branch retinal vein occlusion. Am J Oph­thalmol 1990;109:298-302. 19. Zegarra H, Gutman FA, Zakov N, Carim M. Partial occlu­sion of the central retinal vein. Am J Ophthalmol 1983;96: 330-7. 20. Peternel P, Keber 0, Videcnik V. Carotid arteries in central retinal vessel occlusion as assessed by Doppler ultrasound. Br J Ophthalmol 1989;73:880-3. 21. The Amaurosis Fugax Study Group. Current management of amaurosis fugax. Stroke 1990;21:201-8. 22. von Reutern GM. Biidingen HJ. U1traschalldiagnostik der hirnversorgenden Arterien. Thieme, Stuttgart, 1989:47-209. 23. von Maravic M, Kessler Ch, Wessel K, Kompf D. Pathoge­netische Differenzierung von zerebrovaskularen Erkran- I Lilli NCllro-"I"llhalmol, Vol. 13, No.3, 1993 kungen. In: Fischer P-A, Baas H, Enzensberger W, eds. Verhandl Dtsch Gesell Neurol 1989;5:1009-12. 24. O'Donnell TF, Erdoes L, Mackey WC, McCullough J, Shep­ard A, Heggerick P, et al. Correlation of B-mode ultrasound imaging and arteriography with pathologic findings at ca­rotid endarterectomy. Arch Surg 1985;120:443-9. 25. Rubin JR, Bondi JA, Rhodes RS. Duplex scanning versus conventional arteriography for the evaluation of carotid ar­tery plaque morphology. Surgery 1987;102:749-55. 26. Reilly LM, Lusby RJ, Hughes L, Ferrell LD, Stoney RJ, Ehrenfeld WK. Carotid plaque histology using real-time ul­trasonography. Am J Surg 1983;146:188-93. 27. Cerebral Embolism Task Force. Cardiogenic brain embo­lism. Arch Neurol 1989;46:727-43. 28. Hurwitz BJ, Heyman A, Wilkinson WE, Haynes CS, Utley CM. Comparison of amaurosis fugax and transient cerebral ischemia: a prospective clinical and arteriographic study. Ann Neurol 1985;18:698-704. 29. Ross Russell RW. Natural history of amaurosis fugax. In: Bernstein EB, ed. Amaurosis fugax. Springer-Verlag, New York, 1988:174-82. 30. Harrison MJG. Angiography in amaurosis fugax. In: Bern­stein EB, ed. Amaurosis fugax. Springer-Verlag, New York, 1988:227-35. 31. Aasen J, Kerty E, Russell 0, Bakke SJ, Nyberg-Hansen R. Amaurosis fugax: clinical, doppler and angiographic find­ings. Acta Neurol Scand 1988;77:450-5. 32. Wilson LA, Warlow CP, Ross Russell RW. Cardiovascular disease in patients with retinal arterial occlusion. Lancet 1979;1:292-4. 33. Johnson JM, Kennelly MM, Decesare 0, Morgan S, Spar­row A. Natural history of asymptomatic carotid plaque. Arch Surg 1985;120:1010-12. 34. Sterpetti AV, Schultz RD, Feldhaus RJ, Davenport KL, Ri­chardson M, Farina C, et al. Ultrasonographic features of carotid plaque and the risk of subsequent neurological def­icits. Surgery 1988;104:652--60. 35. von Maravic C, Kessler Ch, von Maravic M, Hohlbach G, Kompf D. Clinical relevance of intraplaque hemorrhage in the internal carotid artery. Eur J Surg 1991;157:185-8. 36. McBrien OJ, Bradley RD, Ashton N. The nature of retinal emboli in stenosis of the internal carotid artery. Lancet 1963; 1:697-9. 37. Soloway HB, Aronson SM. Atheromatous emboli to central nervous system: report of 16 cases. Arch Neurol 1964;11:657­67. 38. Thompson PO, Mastaglia FL, Carrol WM. Anterior isch­aemic optic neuropathy: a correlative clinical and visual evoked potential study of 18 patients. J Neurol Neurosurg PSl/chiatYl/ 1986;49:128-35. 39. Dian JE: Gates PC, Fox AJ. Clinical events follOWing neu­roangiography: a prospective study. Stroke 1987;18:997­1004. 40. Edelmann RR, Mattie EM, Atkinson OJ. MR angiography. AJR 1990;154:937-46. 41. Guyer DR, Miller NR, Auer CL, Fine SL. The risk of cere­brovascular and cardiovascular disease in patients with an­terior ischemic optic neuropathy. Arch Ophthalmol 1985;103: 1136-42. 42. Sawle GV, James CB: Ross Russell RW. The natural history of non-artennc antenor Ischemic optic neuropathy. J Neurol Neurosurg Psychiatry 1990;53:830-3.