Natural course of nonaphakic cystoid macular edema

SURVEY OF OPHTHALMOLOGY

VOLUME 28. SUPPLEMENT.

MAY 1984

Natural Course of Nonaphakic Cystoid Macular Edema GABRIEL COSCAS, M.D., AND ALAIN GAUDRIC,

M.D.

Eye University Clinic of Creteil, Paris, France

Abstract.

We evaluated the longterm natural history of nonaphakic cystoid macular edema (CME) in a retrospective study of 130 out of 557 CME cases recorded in the past ten years. A listing ofcauses was provided and the cases divided into two groups: those with perifoveal leakage and those with deep subretinal leakage. In cases ofdiabetic retinopathy, 60 patients who were followed up for more than three years had noncystoid or cystoid macular edema. The occurrence and persistence ofa large central fovea1 cyst usually resulted in a severe decrease in visual acuity. Hard exudates, present in 60% of cases, seemed to influence visual prognosis when they were inside the fovea1 avascular zone. In cases ofvenous occlusion, chronic CME increased the risk of a central cyst and was the major cause of a macular scar. In cases of uveitis and vasculitis, the restoration of macular capillary wall competence was possible when inflammation decreased. Disturbances in the macular pigment epithelium were also shown to produce poor visual acuity. (Surv Opbthalmol 28(Suppl):47 l-484, 1984)

Key words. fluorescein

diabetes angiography

l

cystoid macular edema noncystoid macular edema l

l

w

central retinal vein occlusion retinal branch vein occlusion l

l

Subjects and Methods

hereas the natural history of Irvine-Gass syndrome is well known, less attention has been paid to the natural history and prognosis of cystoid macular edema (CME) resulting from other conditions. Numerous disease states produce CME. In some cases, macular edema is the major cause of visual loss, especially in retinal vascular diseases, while in other cases it is only an incidental finding.‘7,‘H In this retrospective study, we collected data to record the different etiologies of CME and to define the features of CME related to the various conditions. Diagnosis of CME was based on the following fluorescein angiographic findings: presence at the tenth minute after injection of dye pooling in well delineated cystoid spaces, occupying part or all of the macular area. From our chart review, we analyzed 130 cases of CME in retinal vascular and inflammatory diseases followed for more than three years in an attempt to answer the following questions. 1. Which cases of CME spontaneously resolve? 2. How long is persistent CME consistent with good or fair visual acuity? 3. Does visual function always increase with the resolution of CME? 4. Which signs indicate poor visual prognosis? 5. What is the end point of the natural course of CME?

The patients raphy Clinic CME,

cases were selected from the files of 11,000 who had undergone fluorescein angiogin the past ten years at the Eye University of Creteil (Paris XII). Among 557 cases of only 130 were available for a longterm study.

DIABETES Macular edema in 60 eyes (38 patients) was followed up for a period ranging from three to nine years (mean, five years). In 30 cases, edema was noncystoid (NCME) at the beginning of the followup, but became cystoid (CME) within three years. In 42 cases, panretinal photocoagulation was performed either before or during the follow-up. All cases that underwent photocoagulation to the posterior pole were excluded. RETINAL

VEIN OCCLUSION

A total of38 vein occlusion

eyes (37 patients) with central retinal (CRVO) was examined within three

months after the onset of symptoms. The follow-up period ranged from three to seven years (mean, four years).

In all cases,

Twenty

eyes

(20

vein occlusion were onset of symptoms.

CME

lasted

patients) seen They

at least with

within were

branch

three months of followed up from

three to five years (average, 3.5 years). CME lasted at least six months. 471

six months.

retinal

In all cases,

472

Surv Ophthalmol

28 (Suppl)

TABLE Causes

May

COSCAS AND GAUDRIC

1984

1

of CystoidMacular Edema, Based on Stud_yof 557 of 11,000 Patients

Over a Ten-Year

Period

Number of Patients

Cause I. Leakage of perifoveal retinal capillaries A. Post-ocular surgery 1. Irvine-Gass syndrome 2. Retinal detachment surgery 3. Pars plana vitrectomy 4. Hypotony 5. Photocoagulation around a localized retinal detachment B. Retinal vascular disorders 1. Diabetic retinopathy 2. Branch retinal vein occlusion 3. Central retinal vein occlusion 4. Retinal telangiectasia (Coats’ disease, macular telangiectasia; dominant CME6.20 5. Hypertensive retinopathy 6. Carotid-cavernous listula C. Intraocular inflammation 1. Pars planitis, choroiditis 2. Birdshot retinochoroidopathy 3. Retinal vasculitis (Eales’, Behcet’s, sarcoidosis, tuberculosis) D. Retinal degeneration 1. Retinitis pigmentosa 2. Surface wrinkling retinopathy E. Chronic optic disc edema F. Drugs 1. Epinephrine in aphakia 2. Perhexilline (Pexid@) II. Subretinal leakage A. Subretinal neovascular membrane B. Choroidal tumors I. Malignant melanoma 2. Hemangioma C. Diffuse retinal pigment epitheliopathy D. Pseudo-vitelliform dystrophy E. After severe blunt injury

50

1 1 1 2

220 83 104 8

23 5 13

3 2 1 3 1 28 1 2 1 1 - 1 557

UVEITIS Ten cases (12 eyes) with posterior uveitis (vitritis, 2 eyes; vasculitis, 4 eyes; birdshot retinochoroidopathy, 4 eyes; pars planitis, 2 eyes) developed CME lasting for more than six months. They were followed for more than three years. Cases with cloudy media, vitreous hemorrhages, and tractional detachment were excluded from the study.

Results CAUSES OF CYSTOID

MACULAR

EDEMA

The clinical conditions in which CME has been found are various and have been published.4,‘4,30 Our cases fell into two main etiological groups: CME with perifoveal capillary leakage and CME

Fig. 1. Hypertensive

retinopathy

with sectorial

CME.

with deep subretinal leakage (Table 1). Cases of CME without fluorescein pooling were excluded from this study. CME presents different features according to its etiologic condition. Characteristics of CME in Irvine-Gass syndrome, diabetes, retinal vein occlusion, and uveitis are well documented.‘0J2~‘7J8~2’ We wish to emphasize the features of some less frequently seen etiologies. Cystoid Macular Edema With Perifoveal Capillary Leakage We observed two cases of CME after laser treatment for localized peripheral rhegmatogenous retinal detachment. Although the retinal detachment remained limited a macular pucker occurred some

NATURAL

COURSE

OF NONAPHAKIC

CYSTOID

MACULAR

EDEMA

473

Fig. 2. Coats’

disease. Top: Before treatment, temporomacular vascular ectasia. Bottom: After treatment, persistence of sectorial CME due to macular telangiectasia.

Retinitis pigmentosa demonstrates cystoid edema in 15% ofpatients with capillary leakage around the disc and the temporal vessels.* Of 30 patients with retinitis pigmentosa, we observed three cases of CME; in two cases the fluorescein pattern was typical, with widespread capillary leakage, and in one case capillary leakage was localized in the fovea1 area (Figs. 3 and 4). One case of optic disc drusen with longstanding papilledema had chronic CME. In one case treated with perhexilline (Pexid@) for heart disease, we observed as a side effect a very large area of CME associated with papilledema and polyneuritis” . - . (Fig. 5).

weeks later, probably related to the non-closure of the retinal tear. In recent reviews of systemic hypertension, the occurrence of CME has not been reported.2.‘” In fact, of 150 cases of hypertensive retinopathy, we observed only one case of CME that was related to ischemic perifoveal damage (Fig. 1). It is well known that isolated macular telangiectasia produces temporal or total CME.” Macular telangiectasia has been frequently demonstrated in Coats’ disease2” with CME visible only on angiography. In our study, CME was the main cause of persistent visual loss, even after successful treatment of peripheral ectasia and resorption of exudates and subretinal fluid (Fig. 2).

Cystoid Macular Edema With Subretinal Leakage Macular subretinal new vessels are often associated with biomicroscopically visible CME. However, its angiographic pattern is difficult to identify because of superimposed deep subretinal leakage (Fig. 6). Choroidal tumors (e.g., melanoma and hemangioma) may induce CME when they develop close to the macula. In some cases, however, peripheral melanomas have been associated with CME even when the tumor and the serous detachment did not extend to the macula. ‘*I9We observed a similar case with a melanoma nasal to the disc (Fig. 7). In idiopathic central serous choroidopathy, CME is very uncommon but was observed by us in one

474

Surv Ophthalmol

28 (Suppl)

TABLE Natural

May

NCME*

CME With Central cyst

30 Cases

27 Cases

18 Cases

14

6

1

8

11

1

6

6

10

0 0 2 30

2 2 0 27

2 2 2 18

*NCME: Noncystoid macular edema. **CME: Cystoid macular edema. tRPE: Retinal pigment epithelium. $Total is more than 60 because some cases have been entered in two successive groups when the follow-up in each group was superior to 4 years.

case of diffuse retinal pigment epitheliopathy (PRPE) (Fig. 8). In all of these cases, it was difficult to ascertain whether the filling of cystoid spaces occurred from the subretinal space or the retinal capillaries, or both.

The longterm three conditions: and uveitis.

Fig.

4. Retinitis

follow-up diabetes,

pigmentosa

AND

GAUDRIC

with

peripapillary

Edema

CME** Without Central cyst

Natural

COSCAS

2

Course of 60 Cases of Diabetic Macular (Follow-up 3 to 5 Years)

NCME CME Without Central Cyst CME With Central Cyst Epiretinal Membrane RPEt Disturbance Lamellar Hole Total:

1984

History

3. Retinitis and CME.

pigmentosa

CME.

leakage

DIABETES

In diabetic retinopathy, maculopathy usually progessed very slowly from mild to severe forms (Fig. 9 and Table 2). Three groups were distinguished and are described below. Group I: Noncystoid

of CME was studied in retinal vein occlusion,

with isolated

Fig.

Macular Edema

A total of 30 eyes were seen initially at this stage. The perifoveal leakage was fuzzy, without sharply demarcated cystoid hyperfluorescent spaces (Fig.

line treatment.

Fig. 6. Macular

subretinal

new vessels with CME.

10) even at late phases in the angiogram (10 and 30 minutes). The mean visual acuity was 0.85 at presentation (ranging from 1.O to 0.4). After five years of follow-up, visual acuity was still 0.6, but it decreased to 0.45 after six years. At the end of the follow-up period (mean, five years) 16 of 30 eyes became cystoid, including six with a central cyst and two with a lamellar hole, and 14 remained noncystoid, even if there was transient CME (four cases). Group II: Cystoid Macular Edema; Without a Central Cyst in the Fovea1 Avascular Zone (Fig. 11) This

group

initially

included

27 eyes

(previous

Fig. 7. CME associated

with nasally

developed

choroidal

melanoma.

Fig. 8. after

Diffuse retinal pigment epitheliopathy.

injection.

Left: Composite aqiogram

at an early phase. Right: CME visible at 30 min

Surv Ophthalmol

476

28 (Suppl) May 1984

+ Pattern

of

macular

Non cyrfoid

I 12

34

16

edema

macular

COSCAS

AND GAUDRIC

at presentation :

edema

(NCME)

: 30

, YEARS

Fig. 9. Natural course of visual acuity of diabetic macular

edema.

duration of CME was not always known). Mean visual acuity was 0.65 (ranging from 1 .O to 0.2). After four years of follow-up, it remained at 0.50. At the end of the follow-up period (mean, four years), 6 of the 27 had a large central cyst in the fovea1 avascular zone, 11 cases remained unchanged, and in 6 cases, CME became noncystoid. Epiretinal membrane occurred in two cases, and central pigmentary disturbances developed in two others. Group III: Cystoid Macular Edema With a Large Central Cyst in the Fovea1 Avascular Zone (Fig. 12) There

were

Fig. Il. Diabetic without central

18 eyes in this group

maculopathycyst.

Cystoid

at the initial

macular

edema

Fi_g. IO. Diabetic edema (NCME).

maculopathy

-

noncystoid

macular

examination. Mean visual acuity was 0.35 (ranging from 1.0 to 0.1). It decreased to 0.30 after three years of follow-up and to 0.10 after five years. After a mean follow-up of four years, CME became noncystoid in one case; the large central cyst disappeared in one case; the central cyst persisted in 14 cases (10 cases were unchanged, 2 had fovea1 pigment epithelial disturbances and 2 had an epiretinal membrane); and a lamellar hole developed in 2 cases.

Fig. 12. Diabetic maculopathy with central cyst.

-

cystoid

macular

edema

NATURAL

COURSE

Fig. 13. Diabetic

The been

OF NONAPHAKIC

maculopathy

role of hard considered

they were

extrafoveolar

exudates

in diabetic

in relation

to visual

They have been observed zone,

-

located

outside

they did not seem

CYSTOID

in 60%

MACULAR

exudates.

CME

has

prognosis.

of the cases. When

of the fovea1 avascular to influence

visual

acuity

Diabetic

Fig. II. exudates.

capillary creased

permeability

angiography

tous first group,

Patients

RETINAL suffering

VEIN OCCLUSION

from central

retinal

Fig. 1.5. Central vein occlusion (edematous months later (visual acuity: 1.0).

-

Although

centrofoveolar

according

and leakage.

dilated

with

or the capillary

sus 0.5)

CENTRAL

perfusion

bed is mainly

large nonperfused

14).

maculopathy

sion can be divided into two groups status of capillary

(Fig. 13). When they were inside the fovea1 avascular zone, visual acuity decreased markedly (0.3 ver(Fig.

477

EDEMA

and

leakage

to the

Either

shows

the

an in-

on fluorescein

bed is ischemic

with

territories.3

CME

occurred

mainly

it may be observed

in the edemain the ischemic

group as well. Nevertheless, in this last group visual prognosis remained very poor because of capillary

vein occlu-

nonperfusion

and

was not directly

related

type). Left: CME

6 months after onset. Kight: Disappearance

to the

of CME

18

Fig. 16. Central retinal vein occlusion (edematous type). Persistent CME 18 months later (visual acuity: 0.05).

Fig. 17. Branch retinal vein occlusion. CME 12 months later (visual acuity:

Top: Sectorial 0.6).

CME

IAt:

Noncystoid

6 months

macular

after onset

edema

3 months

(visual acuity:

after onset.

0.6). Bottom:

Right:

Persistent

NATURAL COURSE OF NONAPHAKIC

presence or absence of CME.3’ We studied 38 eyes (37 patients) with edematous central retinal vein occlusion. As far as we knew the exact date ofvenous occlusion, CME occurred within the first two months of the disease. In fact, it was often difficult to detect early stages of CME because of retinal hemorrhage in the macular area. The mean visual acuity at presentation was 0.33, ranging from counting fingers to 1.0. CME lasted more than six months and was followed for three to seven years. To assess the role of CME duration, we distinguished two groups according to visual acuity at the end of the follow-up period (Table 3). Group I: Final Visual Acuity Equal to or Better Than 0.5 The mean visual acuity in this group of 11 patients was 0.34. At the end of the follow-up period, the CME had resolved in nine patients (Fig. 15). The two remaining patients had CME lasting more than three years with surprisingly good vision (0.6 and 0.8). A central cyst was present at one stage of evolution in live cases, but we have not been able to detect that the presence or absence of this central cyst made any difference in final visual acuity. Opticociliary circulation developed in seven (63%) patients. Group II: Final Visual Acuity Worse Than 0.5 The mean initial visual eyes was 0.2. Poor final

Fig.

acuity visual

in this group of 27 acuity in 18 eyes

18. Retinal vasculitis (periphlebitis). (visual acuity: 0.3). Right: Disappearance

479

CYSTOID MACULAR EDEMA TABLE

3

Duration of Cystoid Macular Edema and Final Visual Acuity in 38 Cases of Central Retinal Vein Occlusion (Follow-up, 3-7 Years)

Final VA Initial VA (m) Mean CME Duration Presence of Central Cyst Opticociliary Collaterals

2 0.5

< 0.5

I1 Cases

27 Cases

0.35 8.4 months 45% 63%

0.20 29 months 85% 22%

(66%) was related to chronic CME (mean duration, 29 months). A central cyst was present at one stage of evolution in 23 eyes (85%) (Fig. 16). CME resolved in only nine eyes after at least 18 months duration. In association with CME, macular ischemia developed later in the course in two cases. At the end of the follow-up, macular scars occurred in seven cases, macular pucker in two cases, and lamellar hole in two cases. Opticociliary circulation developed in only six eyes (22%). RETINAL BRANCH VEIN OCCLUSION (FIG. 17) CME is the main cause of longterm poor vision in retinal branch vein occlusion. In a previous study,” we noticed that CME was present in 36% of cases lasting more than one year. In 20 cases of retinal branch vein occlusion, mean duration of CME was 15 months with a minimum of six months. At the end of the follow-up period, CME resolved in six eyes. The mean final visual acuity was 0.45 (ranging

Left: CME with inflammatory of CME 2 years later (visual

obstruction of superotemporal acuity: 0.8).

vein branches

Surv Ophthalmol

28 (Suppl) May 1984

COSCAS

AND GAUDRIC

Top: Non cystoid macular edema (visual acuity: 0.6). Bottom: Widespread cystoid edema 3 years later (visual acuity: 0.1).

Fig. 19. Birdshot retinochoroidopathy.

from 0.9 to 0.05). In three cases, CME resolved with a good final visual acuity (better than 0.5). This relatively good outcome after more than one year’s duration may be explained by the sectorial pattern of CME

UVEITIS

in retinal

branch

vein occlusion.

AND VASCULITIS

In this group,

12 eyes (11 patients)

had a mean

visual acuity of 0.5 (ranging from 0.05 to 1 .O) when CME was discovered. In all cases, CME lasted at least six months (mean, 21 months). In nine eyes, CME had resolved at the end of follow-up, and

visual acuity

had increased

in six cases.

In retinal

CME may be due to inflammatory vasculitis, venous occlusion. It can resolve (as it did in two of our two cases) if the circulation is reestablished (Fig. 18). In vitritis and pars planitis with diffuse edema, CME was associated with dilatation,

retinal abnor-

mal permeability, and leakage from the macular capillary bed. Macular edema may diminish when inflammation regresses. In some cases, CME gradually increased leading to severe visual impairment (Fig. 19).

NATURAL

COURSE

OF NONAPHAKIC

CYSTOID

MACULAR

481

EDEMA

Fig. 20. Longstanding

CME in central retinal vein occlusion. Top: Persistence of CME 1 year after onset. Bottom: Scarring of centrofoveolar retinal pigment epithelium 1 year later (visual acuity: 0.05).

Discussion The course of nonaphakic CME is related etioiogy. In some cases it is only an incidental ing in the evolution of an underlying disease

to its find(cho-

roidal tumor, subretinal new vessels, retinitis pigmentosa, retinal detachment). More frequently,

CME is due to the involvement of macular capillaries by an obstructive or inflammatory vascular retinal disease (diabetes, venous occlusion, uveitis). It then becomes a major cause of low vision. The Iongterm prognosis of CME is then dependent on the recovery or compensation these cases, the prognosis

of the causal disease. In is related to the possible

482

Surv Ophthalmol

restoration of the blood-retinal barrier of the macular capillaries. Restoration of the macular capillary wall competence is possible in retinal vein occlusion when intraluminal venous pressure decreases, but in other cases CME persists in spite of venous caliber normalization. Improvement ofCME is also possible in uveitis (when inflammation declines), and in vasculitis (if the occluded vessels become patent again). Even in diabetic retinopathy, spontaneous resolution or a decrease of CME can be observed. This possibility is rare, for usually the evolution undergoes a progessive but relatively slow impairment.‘,“j The occurrence of CME after panretinal photocoagulation is a peculiar situation. It may regress the capability after some months, I6 demonstrating of the blood-retinal barrier to improve even in diabetic patients. When CME disappears or converts to a noncystoid macular edema, visual acuity can improve, even if CME has persisted for several months or years. This functional improvement is well known in the Irvine-Gass syndrome.13,2g We have also noted it in other causes of CME, even if the anatomic restoration is less frequent than in aphakic CME. In fact, the persistence or increase of CME in these diseases is the major cause of poor visual prognosis. Even when CME persists, the deterioration of visual acuity appears to be quite slow. Modifica-

Fig. 22. Exudates associated CME maculopathy.

in central cyst of with

COSCAS

28 (Suppl) May 1984

diabetic

AND GAUDRIC

Fig. 21. Hemorrhages

in central cysts of cystoid macular edema associated with retinal branch vein occlusion.

tions in the pattern of cystoid spaces are frequent. The occurrence and persistence of a large central fovea1 cyst usually results in a severe decrease of visual acuity. There are, however, some unusual cases in which the disappearance of a central cyst is followed by a functional improvement. In central vein occlusion, a short period of CME might be related to the development of opticociliary collateral circulation, al-

NATURAL

COURSE OF NONAPHAKIC

CYSTOID MACULAR

483

EDEMA

Fig. 23. Early fluorescence

of central cyst. Foveolar pigment defect is more visible with blue light. A: Green light. B: Blue light. C: Red light. D: Angiogram, early phase. E: Angiogram, 10 min.

though

it is inconstant.

tence of a central

In some other

cyst is compatible

cases,

persis-

with good vision

capillaries. circinate

for many months or years. Chronic CME increases the risk of a central cyst and is the major cause of

always

development

impaired

of macular

ment epithelium

scar at the level of the pig-

(Fig. 20)) or more rarely of a macu-

Photocoagulation rings eliminates

affect

macular

perifoveal

conditions,

CME

is not the sole cause

of macular damage. During retinal vein occlusion, hemorrhages can be located in the cystoid spaces (Fig. 2 1) . Moreover, deep retinal hemorrhages can be responsible

not only for the initial visual loss but

also for pigmentary fovea1 scarring. In diabetic retinopathy, exudates

capillary

CME perfusion arcade

can

be associated

and a breakdown

nosis is associated

with of the

with an enlargement

zone. An extremely with impaired

of

poor visual progperfusion

macular capillaries and CME.” Disturbances in the macular xanthophyll,

of the visible

in blue light, may be responsible for the early hyperfluorescence of a central cyst (Fig. 23). A macular hole could result from a breakdown

in the fovea are

of the

but does not

the CME.“.”

In all conditions,

the avascular

lar pucker. In several

to the center

the exudates

thinned

wall of the central

of the internal,

cyst. An epiretinal

mem-

associated with CME in 60% ofcases. Visual acuity is profoundly decreased when a plaque of exudates accumulates in the center of the fovea.‘” Perifoveal

brane These

can also be associated with a lamellar hole. lesions occur only after a long evolution of

CME

of at least five years.

exudates

The ultimate end point is irreversible ation of the retinal macular tissue.

and lipids can fill some cystoid spaces (Fig. 22). Th eir influence on the natural course ofCME is, however, unclear. They commonly originate from intraretinal microvascular abnormalities and from large microaneurysms distant from the foveola, in contrast to the CME that is maintained

can be stellar

by leakage

from the damaged

perifoveal

degener-

References 1. Blankenship

GW: Diabetic macular edema and argon laser photocoagulation. Ophthalmologv 86:69-75. 1979 2. Chaine GJ, Kohner EM: Hypertensive retinopathy, in Sleight P

484

7. 8. 9. 10.

11. 12.

13. 14. 15.

16.

17.

18.

Surv Ophthalmol

28 (Suppl) May 1984

(ed): Hypertension. London, Buttersworth, 1982, pp 92-l 16 Coscas G, Dhermy P: Occias~oons Veinneuses Re’tiniennes, Paris, Masson, 1978, p 471 Coscas G: Maculopathies oedemateuses. Bull Sot Ophthalmol Fr 1:260, 1972 Debbasch S: Oedeme papillaire iatrogtne. Clin Ophthalmol 4:109-l 16, 1977 Deutman AF, Pinckers AJLG, Aan de Kerk AL: Dominantly inherited cystoid macular edema. AmJ Ophthalmol82:54&548, 1976 Fine BS, Brucker AJ: Macular edema and cystoid macular edema. Am J Ophthalmol92:46t%481, 1981 Fishman GA, Maggiano.IM, Fishman M: Fovea1 lesions seen in retinitis pigmentosa. Arch Ophthalmol 9.5:1993-1996, 1977 Fishman GA. Trimble S. Rabb MF. Fishman M: Pseudovitelliform macular degeneration. Arch &hthalmol 95:73-76, 1977 Gass JDM, Norton EWD: Cystoid macular edema and papilledema following cataract extraction. A fluorescein funduscopic and angiographic study. Arch Ophthalmol 76:64&661, 1966 Gass JDM, Oyakawa RT: Idiopathic juxtafoveolar retinal telangiectasis. Arch Ophthalmol 100:769-780, 1982 Gutman FA: Macular edema in branch retinal vein occlusion: prognosis and management. Trans Am Acad Ophthalmol Otolaryngo! 83:48&495, 1977 Hitchings RA: Aphakic macular edema. A two-year follow-up study. Br J Ophthalmol61:628-630, 1977 Irvine AR: Cystoid maculopathy. Sum Ophthalmol21:1-17, 1976 Kirkendall WM: Retinal changes of hypertension, in Mausolf FA (ed): The Eye in Systemic Disease. St Louis, CV Mosby, 1975, p 212-222 Little H: Complications of argon laser photocoagulation in the treatment of diabetic retinopathy, in Little H, Jack RL (eds): Diabetic Retinopathy. Pathogenesis and Treatment. Thieme Stratton. New York 1983, pp 285-296 Maumenee AE: Clinical entities in “uveitis”: An approach to the study of intraocular inflammation. Am J Ophthalmol69: l-27, 1970 McMeel JW, Trempe CL, Franks EB: Diabetic maculopathy. Trans Am Acad Ophthalmol Otolaryngol 83:47&487, 1977

COSCAS

AND GAUDRIC

19. Newsom WA, Hood CI, Horwitz JA: Cystoid macular edema: histopathologic and angiographic correlation. A clinicopathologic case report. Tram Am Acad Ophthalmoi Otolaryngol76: 100% 1009, 1972 20. Notting JGA, Pinckers AJLG: Dominant cystoid macular dystrophy. Am J Ophthalmol83:234-241, 1977 21. Patz A, Schatz H, Berkow JW, et al: Macular edema, an overlooked complication of diabetic retinopathy. Tranr Am Acud Ophthalmol Otolaryngol 77:34-42, 1973 22. Reeser F, Fleishman J, Williams FA, Goldman A: Efficacy of argon laser photocoagulation treatment of circinate diabetic retinopathy. Am J Ophthalmol 92:762-767, 1981 23. Sigurdsson R, Begg IS: Organised macular plaques in exudative diabetic maculopathy. Br J Ophthnlmol 64:392-397, 1980 24. Spalter HF: Photocoagulation of circinate maculopathy in diabetic retinopathy. Am J Ophthalmol 71:242-250, 1971 25. Spitznas M, Joussen F, Wessing A, Meyer-Schwickerath G: Coats’ disease. Albrecht uon Graefes Arch Klin Exp Ophthalmol 195:241-250, 1975 26. Townsend C, Bailey J, Kohner EM: Xenon arc photocoagulation for the treatment ofdiabetic maculopathy. Interim report of a multicentre controlled study. Br J Ophthalmol64:385-39 1, 1980 27. Tso MOM: Pathology of cystoid macular edema. Ophthalmol @:902-915, 1982 28. Tso MOM: Pathological study of cystoid macular edema. Tram Obhthnlmol Sot UK 100:408-413. 1980 29. Yannuzzi LA, Klein RM, Wallyn RH, et al: Inelfectiveness of indomethacin in the treatment of chronic cystoid macular edema. Am J Ophthalmol 84:517-519, 1977 30. Yanoff M, Fine BS: Ocular Putholou. Philadelphia, Harper and Row, 1982, p 916 31. Zegarra H, Gutman FA, Conforto J: The natural course of central retinal vein occlusion. Ophthalmology &:1931-1939, 1979

Reprint requests should be addressed to Gabriel Coscas, M.D., Eye University Clinic ofCretei1 (Paris XII), 40, Avenue de Verdun, 94010 CRETEIL - FRANCE.