Macroaneurysms Associated With Retinal Branch Vein Occlusion Scott W. Cousins, M.D., Harry W. Flynn, [r., M.D., and John G. Clarkson, M.D. We reviewed the clinical findings, color photographs, and fluorescein angiograms of 147 patients (154 involved eyes) who participated in the Branch Vein Occlusion Study through the Bascom Palmer Eye Institute for the presence of macroaneurysms within or adjacent to the zone of branch vein occlusion. Four distinct categories of macroaneurysms included arterial, venous, capillary, and collateral-associated types. We identified 59 lesions among 25 of the 154 (16.2%) affected eyes. Intraretinal lipid exudation and hemorrhage were commonly observed around the lesions, but these complications usually did not affect visual acuity. Macroaneurysms were associated with capillary nonperfusion on fluorescein angiography. Retinal neovascularization occurred in eight of 25 eyes (32%) with macroaneurysms. DISTINCTIVE MICROVASCULAR CHANGES develop within the retinal vasculature after retinal vein occlusion. Microaneurysms, telangiectatic and nonperfused capillaries, and dilated collateral vessels are the common sequelae of branch and central vein occlusion.' Although the clinical features and course of arterial macroaneurysms have been well established.V it is not well known that similar macrovascular changes can be associated with retinal vein occlusions.' A review of published reports showed five welldocumented case reports of these Iesions.v" and one additional Japanese series appeared."
Accepted for publication Feb. 26, 1990. From the Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, Florida. This study was supported in part by the Public Health Service Research grant No. EY06749 and grant No. EY02180 from the Department of Health and Human Services, National Institutes of Health, National Eye Institute, Bethesda, Maryland. Dr. Cousins was a Heed Foundation Fellow (1986-1987). Reprint requests to Scott W. Cousins, M.D., Bascom Palmer Eye Institute, P.O. Box 016880, Miami, FL 33101.
©AMERICAN JOURNAL OF OPHTHALMOLOGY
109:567-570,
Nevertheless, it was our clinical impression that macroaneurysms occurred more commonly in venous occlusive disease than these few reports suggested.
Patients and Methods We reviewed the clinical, photographic, and angiographic records of 147 patients (154 involved eyes), who enrolled in the Branch Vein Occlusion Study":" through the Bascom Palmer Eye Institute, for the occurrence of macroaneurysms. One of us (J.G.c.) had been the principal investigator for the Branch Vein Occlusion Study and had examined each patient during the follow-up period (1977-1985). We defined a macroaneurysm on the basis of clinical features, color photographs, and fluorescein angiographic criteria. Clinically, a macroaneurysm was identified from color fundus slides as a red saccular or fusiform structure measuring 100 urn or larger in size. The size was estimated by comparing the macroaneurysm to the diameter of the branch arteries (100 um) from uninvolved quadrants as they crossed the optic disk. Angiographically, macroaneurysms showed discrete filling of the lesion with dye (Fig. 1) and showed persistent discrete hyperfluorescence for 180 seconds (Fig. 2). The angiographic criteria were based on our impression that microaneurysms and nonaneurysmal structures were not usually hyperfluorescent after 180 seconds, whereas neovascularization usually showed significant leakage by that time. The size criteria were selected on the basis of the definition by Wise, Dollery, and Henkind," who defined microaneurysms as structures smaller than 100 urn. In this series, macroaneurysms were classified by size: small, 100 to 149 ~m; medium, 150 to 249 um; and large, 250 urn or larger (Table 1). Four distinct types of macroaneurysms occurred within or adjacent to the zone of branch MAY,
1990
567
568
May, 1990
AMERICAN JOURNAL OF OPHTHALMOLOGY
Fig. 1 (Cousins, Flynn, and Clarkson). Fluorescein angiogram during the early venous filling phase shows early filling of a large macroaneurysm associated with a collateral vessel (arrow).
vein occlusion. Arterial macroaneurysms were identified as saccular or fusiform dilations along major branch arteries or arterioles within the area of branch vein occlusion, with filling from the arterial circulation. Arterial macroaneurysms were indistinguishable from typical idiopathic arterial macroaneurysms except that they were located adjacent to or within the zone of the branch vein occlusion. Venous macroaneurysms were similar-appearing lesions associated with major branch veins or venules, but they exhibited a delay in filling because they filled from the venous circulation. Capillary macroaneurysms were not associated directly with a larger vessel but, on fluorescein angiography, they filled from the adjacent capillary bed and showed no clearly identifiable feeder vessels (Fig. 3). The fourth type of macroaneurysm was the collateral-associated form (Figs. 1, 2, and 4). This last category was associated
Fig. 2 (Cousins, Flynn, and Clarkson). The late phase of the fluorescein angiogram shown in Figure 1 shows persistent filling and dye leakage from the macroaneurysm and the telangiectatic parafoveal capillaries. The macroaneurysm drains into another collateral vessel (arrow) that was not easily identified in the earlier phase of the angiogram.
with clearly identifiable dilated collateral vessels on fluorescein angiography.
Results
The records of 24 patients (25 eyes) who had macroaneurysms were studied. In these 25 eyes, 59 macroaneurysms were identified.
TABLE 1 TYPES OF MACROANEURYSMS IN 25 EYES
TYPE
NO. OF LARGE MEDIUM SMALL LESIONS (;,,250 I'M) (150-249 I'M) (100-149 I'M)
Arterial
12
2
Venous
9
2
22 16
4 5
9 6 12 7
59
13
34
Capillary Collateral Total No. of lesions
1 1
6 4 12
Fig. 3 (Cousins, Flynn, and Clarkson). The fluorescein angiogram shows a medium-sized capillary macroaneurysm (arrow).
Vol. 109, No.5
569
Macroaneurysms and Branch Vein Occlusion
TABLE 2 DISTRIBUTION OF LESIONS IN 25 EYES TYPE
o
Capillary
7
Total No. of eyes
Twelve macroaneurysms (20.3%) were small, 34 (57.6%) were medium-sized, and 13 (22.0%) were large. Of the 59 lesions, 12 were arterial, nine were venous, 22 were capillary, and 16 were collateral-associated (Table I). These lesions, however, were distributed among only 25 of the 154 study eyes (Table 2). Ten eyes had a mixture of aneurysm types, and nearly all possible combinations occurred. Ten eyes had only one lesion, five eyes had two lesions, and ten eyes had three or more lesions. We also analyzed the course of the macroaneurysms. The mean time of macroaneurysm identification was 23.3 months after entry into the Branch Vein Occlusion Study (range, zero to 72 months). Eleven of 25 eyes (44.0%) had macroaneurysms that involuted within a mean interval of approximately 14.0 months after identification. Additional macroaneurysms developed in 11 of 25 (44.0%) eyes, with an average interval of 40.1 months from the time of detection of the first macroaneurysm to the detection of additional macroaneurysms. The average age of our patients was 70.6 years. Thirteen (54.2%) patients were women and 11 (45.8%) were men. Of the 25 eyes, 14 received retinal photocoagulation. Six eyes (24.0%) developed clinically significant cystoid macular edema. Fifteen of 24 patients (62.5%) were receiving treatment for hypertension or coronary artery disease. Seven of the 24 patients (29.2%) developed bilateral branch vein occlusions, but only one patient developed bilateral macroaneurysms.
2
Arterial Venous Collateral Mixed types
Fig. 4 (Cousins, Flynn, and Clarkson). The fluorescein angiogram shows an isolated collateral-associated macroaneurysm (arrow) in a zone of severe capillary nonperfusion.
NO. OF EYES
6 10
25
The macroaneurysms were associated with a relatively high rate of surrounding intraretinal lipid (ten of 25 eyes [40.0% D or hemorrhage (nine of 25 eyes [36.0% D. Lipid exudation involving the macula, either associated with the macroaneurysm or with other microvascular changes caused by the branch vein occlusion, occurred in three of 25 eyes (12.0%). Similarly, intraretinal hemorrhage involved the macula only in one eye. In 13 of 25 eyes (52.0%), at least one macroaneurysm was located within the posterior pole. In eyes with multiple lesions, most of the additional macroaneurysms were located peripherally. Most posteriorly located macroaneurysms, however, were far enough away from the macula that lipid exudation or hemorrhage usually did not affect the fovea. On fluorescein angiography, macroaneurysms were generally associated with capillary nonperfusion adjacent to or near the lesion. Of the 25 eyes with macroaneurysms, 21 (84.0%) were in the "severe nonperfusion" category by the Branch Vein Occlusion Study criteria (at least five disk areas of capillary nonperfusion)." Eight of 25 eyes (32.0%) with macroaneurysms developed neovascularization at some time during the course of the Branch Vein Occlusion Study. However, 14 of the eyes in this study received photocoagulation, which might have altered the course of these eyes by lowering the rate of development of neovascularization. A subgroup of seven eyes had one or more macroaneurysms and a zone of severe nonperfusion in an untreated eye. Within this subgroup, four of seven eyes (57.1 %) had or eventually developed neovascularization. An appropriate control group was not available for our study population, which prevented a formal statistical comparison.
570
May, 1990
AMERICAN JOURNAL OF OPHTHALMOLOGY
Discussion
Macroaneurysms of four distinct types occurred in 25 of 154 (16.2%) eyes studied. Although associated vascular leakage was observed in many of these lesions, most macroaneurysms did not affect vision because most of the lesions were not located in the macula. Such macroaneurysms are easily overlooked clinically because of their noncentral location, the coexistence of other diffuse microvascular changes, and commonly associated diffuse intraretinal hemorrhages. Although a control group was not available, we compared the characteristics of the eyes we studied to those of the entire Branch Vein Occlusion Study population. The age and sex distribution appeared to be similar. The average age of our patients was 76.0 years. The average age of the patients in the Branch Vein Occlusion Study was not stated, but 90% of the patients were between 50 and 80 years of age. 10,11 Thirteen of 24 (54.2%) of our patients were women, and 146 of the 319 (45.8%) of patients in Group I (three to eight months since onset) of the Branch Vein Occlusion Study and 42 of 82 (51.2%) of those in Group II (similar to Group I, but with retinal neovascularization) were women. Fifteen of 24 patients (62.5%) with macroaneurysms were receiving treatment for hypertension. In the Branch Vein Occlusion Study, 156 of 319 (48.9%) of Group I patients were treated for hypertension, and 15 (5%) were also being treated for diabetes mellitus. We compared the clinical course of eyes that had macroaneurysms with the course of eyes in the Branch Vein Occlusion Study. Six of 25 eyes (24.0%) developed clinically significant cystoid macular edema in our patient group, and 139 of 502 eyes (27.7%) developed it in the entire study. Seven of the 24 patients (29.2%) developed bilateral vein occlusions in this study as compared to only 9% of patients in the Branch Vein Occlusion Study. In our series, only one patient with bilateral vein occlusions developed a macroaneurysm in each eye. Eight of 25 eyes (32.0%) in our group developed neovascularization. In Group I of the Branch Vein Occlusion Study, 19 of 160 treated eyes (11.9%) and 35 of 159 untreated eyes (22.0%) developed new vessels. However, 27 of 87 untreated eyes (31.0%) with severe (at least five disk areas) capillary nonperfusion eventually developed new vessels. Thus, eyes with macroaneurysms had the same frequency of neovas-
cularization as those with severe nonperfusion in the Branch Vein Occlusion Study. The clinical significance of a macroaneurysm in an eye with a branch vein occlusion is its association with ischemic retina as demonstrated by capillary nonperfusion on fluorescein angiography. Twenty-one of 25 eyes (84.0%) with macroaneurysms had severe (at least five disk areas) capillary nonperfusion. Eight of the 25 eyes (32.0%) developed neovascularization. In the small subgroup of untreated patients who had severe capillary nonperfusion and a macroaneurysm, however, four of seven (57.1 %) of the eyes eventually developed neovascularization. References 1. Archer, D. B., Ernest, J. T., and Newell, F. W.: Classification of branch vein occlusion. Trans. Am. Acad. Ophthalmol. Otolaryngol. 78:148, 1973. 2. Robertson, D. M.: Macroaneurysm of the retinal arteries. Trans. Am. Acad. Ophthalmol. Otolaryngol. 77:55, 1973. 3. Gass, J. D. M.: Stereoscopic Atlas of Macular Diseases. Diagnosis and treatment. St. Louis, C. V. Mosby, 1987, pp. 422-435. 4. Becker, B., and Post, 1. T.: Retinal vein occlusion. Clinical and experimental observations. Am. J. Ophthalmol. 34:677, 1951. 5. Lavin, M. J., Marsh, R. J., Peart,S., and Rehman, A.: Retinal arterial macroaneurysms. A retrospective review of 40 patients. Br. J. Ophthalmol. 71:817, 1987. 6. Margargal, 1. E., Augsburger, J. J., Hyman, D., and Townsend, R.: Venous macroaneurysm following branch retinal vein obstruction. Ann. Ophthalmol. 12:685, 1980. 7. Schulman, J., [arnpol. 1. M., and Goldberg, M. F.: Large capillary aneurysms secondary to retinal vein obstruction. Br. J. Ophthalmol. 65:36, 1981. 8. Sandborn, G. E., and Margargal, 1. E.: Venous macroaneurysms associated with branch retinal vein obstruction. Ann. Ophthalmol. 16:464, 1984. 9. Takeda, M., and Kimura,S.: Large capillary aneurysms secondary to retina branch vein occlusion. Jpn. J. Ophthalmol. 36:315, 1982. 10. The Branch Vein Occlusion Study Group: Argon laser photocoagulation for macular edema in branch vein occlusion. Am. J. Ophthalmol. 98:271, 1984. 11. - - : Argon laser scatter photocoagulation for prevention of neovascularization and vitreous hemorrhage in branch vein occlusion. Arch. Ophthalmol. 104:34, 1986. 12. Wise, G., Dollery, c.. and Henkind, P.: The Retinal Circulation. New York, Harper and Row, 1971, pp. 278-284.
Accepted for publication Feb. 26, 1990. From the Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, Florida. This study was supported in part by the Public Health Service Research grant No. EY06749 and grant No. EY02180 from the Department of Health and Human Services, National Institutes of Health, National Eye Institute, Bethesda, Maryland. Dr. Cousins was a Heed Foundation Fellow (1986-1987). Reprint requests to Scott W. Cousins, M.D., Bascom Palmer Eye Institute, P.O. Box 016880, Miami, FL 33101.
©AMERICAN JOURNAL OF OPHTHALMOLOGY
109:567-570,
Nevertheless, it was our clinical impression that macroaneurysms occurred more commonly in venous occlusive disease than these few reports suggested.
Patients and Methods We reviewed the clinical, photographic, and angiographic records of 147 patients (154 involved eyes), who enrolled in the Branch Vein Occlusion Study":" through the Bascom Palmer Eye Institute, for the occurrence of macroaneurysms. One of us (J.G.c.) had been the principal investigator for the Branch Vein Occlusion Study and had examined each patient during the follow-up period (1977-1985). We defined a macroaneurysm on the basis of clinical features, color photographs, and fluorescein angiographic criteria. Clinically, a macroaneurysm was identified from color fundus slides as a red saccular or fusiform structure measuring 100 urn or larger in size. The size was estimated by comparing the macroaneurysm to the diameter of the branch arteries (100 um) from uninvolved quadrants as they crossed the optic disk. Angiographically, macroaneurysms showed discrete filling of the lesion with dye (Fig. 1) and showed persistent discrete hyperfluorescence for 180 seconds (Fig. 2). The angiographic criteria were based on our impression that microaneurysms and nonaneurysmal structures were not usually hyperfluorescent after 180 seconds, whereas neovascularization usually showed significant leakage by that time. The size criteria were selected on the basis of the definition by Wise, Dollery, and Henkind," who defined microaneurysms as structures smaller than 100 urn. In this series, macroaneurysms were classified by size: small, 100 to 149 ~m; medium, 150 to 249 um; and large, 250 urn or larger (Table 1). Four distinct types of macroaneurysms occurred within or adjacent to the zone of branch MAY,
1990
567
568
May, 1990
AMERICAN JOURNAL OF OPHTHALMOLOGY
Fig. 1 (Cousins, Flynn, and Clarkson). Fluorescein angiogram during the early venous filling phase shows early filling of a large macroaneurysm associated with a collateral vessel (arrow).
vein occlusion. Arterial macroaneurysms were identified as saccular or fusiform dilations along major branch arteries or arterioles within the area of branch vein occlusion, with filling from the arterial circulation. Arterial macroaneurysms were indistinguishable from typical idiopathic arterial macroaneurysms except that they were located adjacent to or within the zone of the branch vein occlusion. Venous macroaneurysms were similar-appearing lesions associated with major branch veins or venules, but they exhibited a delay in filling because they filled from the venous circulation. Capillary macroaneurysms were not associated directly with a larger vessel but, on fluorescein angiography, they filled from the adjacent capillary bed and showed no clearly identifiable feeder vessels (Fig. 3). The fourth type of macroaneurysm was the collateral-associated form (Figs. 1, 2, and 4). This last category was associated
Fig. 2 (Cousins, Flynn, and Clarkson). The late phase of the fluorescein angiogram shown in Figure 1 shows persistent filling and dye leakage from the macroaneurysm and the telangiectatic parafoveal capillaries. The macroaneurysm drains into another collateral vessel (arrow) that was not easily identified in the earlier phase of the angiogram.
with clearly identifiable dilated collateral vessels on fluorescein angiography.
Results
The records of 24 patients (25 eyes) who had macroaneurysms were studied. In these 25 eyes, 59 macroaneurysms were identified.
TABLE 1 TYPES OF MACROANEURYSMS IN 25 EYES
TYPE
NO. OF LARGE MEDIUM SMALL LESIONS (;,,250 I'M) (150-249 I'M) (100-149 I'M)
Arterial
12
2
Venous
9
2
22 16
4 5
9 6 12 7
59
13
34
Capillary Collateral Total No. of lesions
1 1
6 4 12
Fig. 3 (Cousins, Flynn, and Clarkson). The fluorescein angiogram shows a medium-sized capillary macroaneurysm (arrow).
Vol. 109, No.5
569
Macroaneurysms and Branch Vein Occlusion
TABLE 2 DISTRIBUTION OF LESIONS IN 25 EYES TYPE
o
Capillary
7
Total No. of eyes
Twelve macroaneurysms (20.3%) were small, 34 (57.6%) were medium-sized, and 13 (22.0%) were large. Of the 59 lesions, 12 were arterial, nine were venous, 22 were capillary, and 16 were collateral-associated (Table I). These lesions, however, were distributed among only 25 of the 154 study eyes (Table 2). Ten eyes had a mixture of aneurysm types, and nearly all possible combinations occurred. Ten eyes had only one lesion, five eyes had two lesions, and ten eyes had three or more lesions. We also analyzed the course of the macroaneurysms. The mean time of macroaneurysm identification was 23.3 months after entry into the Branch Vein Occlusion Study (range, zero to 72 months). Eleven of 25 eyes (44.0%) had macroaneurysms that involuted within a mean interval of approximately 14.0 months after identification. Additional macroaneurysms developed in 11 of 25 (44.0%) eyes, with an average interval of 40.1 months from the time of detection of the first macroaneurysm to the detection of additional macroaneurysms. The average age of our patients was 70.6 years. Thirteen (54.2%) patients were women and 11 (45.8%) were men. Of the 25 eyes, 14 received retinal photocoagulation. Six eyes (24.0%) developed clinically significant cystoid macular edema. Fifteen of 24 patients (62.5%) were receiving treatment for hypertension or coronary artery disease. Seven of the 24 patients (29.2%) developed bilateral branch vein occlusions, but only one patient developed bilateral macroaneurysms.
2
Arterial Venous Collateral Mixed types
Fig. 4 (Cousins, Flynn, and Clarkson). The fluorescein angiogram shows an isolated collateral-associated macroaneurysm (arrow) in a zone of severe capillary nonperfusion.
NO. OF EYES
6 10
25
The macroaneurysms were associated with a relatively high rate of surrounding intraretinal lipid (ten of 25 eyes [40.0% D or hemorrhage (nine of 25 eyes [36.0% D. Lipid exudation involving the macula, either associated with the macroaneurysm or with other microvascular changes caused by the branch vein occlusion, occurred in three of 25 eyes (12.0%). Similarly, intraretinal hemorrhage involved the macula only in one eye. In 13 of 25 eyes (52.0%), at least one macroaneurysm was located within the posterior pole. In eyes with multiple lesions, most of the additional macroaneurysms were located peripherally. Most posteriorly located macroaneurysms, however, were far enough away from the macula that lipid exudation or hemorrhage usually did not affect the fovea. On fluorescein angiography, macroaneurysms were generally associated with capillary nonperfusion adjacent to or near the lesion. Of the 25 eyes with macroaneurysms, 21 (84.0%) were in the "severe nonperfusion" category by the Branch Vein Occlusion Study criteria (at least five disk areas of capillary nonperfusion)." Eight of 25 eyes (32.0%) with macroaneurysms developed neovascularization at some time during the course of the Branch Vein Occlusion Study. However, 14 of the eyes in this study received photocoagulation, which might have altered the course of these eyes by lowering the rate of development of neovascularization. A subgroup of seven eyes had one or more macroaneurysms and a zone of severe nonperfusion in an untreated eye. Within this subgroup, four of seven eyes (57.1 %) had or eventually developed neovascularization. An appropriate control group was not available for our study population, which prevented a formal statistical comparison.
570
May, 1990
AMERICAN JOURNAL OF OPHTHALMOLOGY
Discussion
Macroaneurysms of four distinct types occurred in 25 of 154 (16.2%) eyes studied. Although associated vascular leakage was observed in many of these lesions, most macroaneurysms did not affect vision because most of the lesions were not located in the macula. Such macroaneurysms are easily overlooked clinically because of their noncentral location, the coexistence of other diffuse microvascular changes, and commonly associated diffuse intraretinal hemorrhages. Although a control group was not available, we compared the characteristics of the eyes we studied to those of the entire Branch Vein Occlusion Study population. The age and sex distribution appeared to be similar. The average age of our patients was 76.0 years. The average age of the patients in the Branch Vein Occlusion Study was not stated, but 90% of the patients were between 50 and 80 years of age. 10,11 Thirteen of 24 (54.2%) of our patients were women, and 146 of the 319 (45.8%) of patients in Group I (three to eight months since onset) of the Branch Vein Occlusion Study and 42 of 82 (51.2%) of those in Group II (similar to Group I, but with retinal neovascularization) were women. Fifteen of 24 patients (62.5%) with macroaneurysms were receiving treatment for hypertension. In the Branch Vein Occlusion Study, 156 of 319 (48.9%) of Group I patients were treated for hypertension, and 15 (5%) were also being treated for diabetes mellitus. We compared the clinical course of eyes that had macroaneurysms with the course of eyes in the Branch Vein Occlusion Study. Six of 25 eyes (24.0%) developed clinically significant cystoid macular edema in our patient group, and 139 of 502 eyes (27.7%) developed it in the entire study. Seven of the 24 patients (29.2%) developed bilateral vein occlusions in this study as compared to only 9% of patients in the Branch Vein Occlusion Study. In our series, only one patient with bilateral vein occlusions developed a macroaneurysm in each eye. Eight of 25 eyes (32.0%) in our group developed neovascularization. In Group I of the Branch Vein Occlusion Study, 19 of 160 treated eyes (11.9%) and 35 of 159 untreated eyes (22.0%) developed new vessels. However, 27 of 87 untreated eyes (31.0%) with severe (at least five disk areas) capillary nonperfusion eventually developed new vessels. Thus, eyes with macroaneurysms had the same frequency of neovas-
cularization as those with severe nonperfusion in the Branch Vein Occlusion Study. The clinical significance of a macroaneurysm in an eye with a branch vein occlusion is its association with ischemic retina as demonstrated by capillary nonperfusion on fluorescein angiography. Twenty-one of 25 eyes (84.0%) with macroaneurysms had severe (at least five disk areas) capillary nonperfusion. Eight of the 25 eyes (32.0%) developed neovascularization. In the small subgroup of untreated patients who had severe capillary nonperfusion and a macroaneurysm, however, four of seven (57.1 %) of the eyes eventually developed neovascularization. References 1. Archer, D. B., Ernest, J. T., and Newell, F. W.: Classification of branch vein occlusion. Trans. Am. Acad. Ophthalmol. Otolaryngol. 78:148, 1973. 2. Robertson, D. M.: Macroaneurysm of the retinal arteries. Trans. Am. Acad. Ophthalmol. Otolaryngol. 77:55, 1973. 3. Gass, J. D. M.: Stereoscopic Atlas of Macular Diseases. Diagnosis and treatment. St. Louis, C. V. Mosby, 1987, pp. 422-435. 4. Becker, B., and Post, 1. T.: Retinal vein occlusion. Clinical and experimental observations. Am. J. Ophthalmol. 34:677, 1951. 5. Lavin, M. J., Marsh, R. J., Peart,S., and Rehman, A.: Retinal arterial macroaneurysms. A retrospective review of 40 patients. Br. J. Ophthalmol. 71:817, 1987. 6. Margargal, 1. E., Augsburger, J. J., Hyman, D., and Townsend, R.: Venous macroaneurysm following branch retinal vein obstruction. Ann. Ophthalmol. 12:685, 1980. 7. Schulman, J., [arnpol. 1. M., and Goldberg, M. F.: Large capillary aneurysms secondary to retinal vein obstruction. Br. J. Ophthalmol. 65:36, 1981. 8. Sandborn, G. E., and Margargal, 1. E.: Venous macroaneurysms associated with branch retinal vein obstruction. Ann. Ophthalmol. 16:464, 1984. 9. Takeda, M., and Kimura,S.: Large capillary aneurysms secondary to retina branch vein occlusion. Jpn. J. Ophthalmol. 36:315, 1982. 10. The Branch Vein Occlusion Study Group: Argon laser photocoagulation for macular edema in branch vein occlusion. Am. J. Ophthalmol. 98:271, 1984. 11. - - : Argon laser scatter photocoagulation for prevention of neovascularization and vitreous hemorrhage in branch vein occlusion. Arch. Ophthalmol. 104:34, 1986. 12. Wise, G., Dollery, c.. and Henkind, P.: The Retinal Circulation. New York, Harper and Row, 1971, pp. 278-284.
