- Email: [email protected]
Diabetic Macular Edema
Diabetic Macular Edema A Review GEORGE H. BRESNICK, MD
Abstract: Diabetic macular edema can be classified into a focal variety, characterized by focal leakage from microaneurysms, often with accumulation of extravascular lipoprotein in a circinate pattern around the focal leakage, and a diffuse variety, with diffuse leakage from retinal vessels often accompanied by cystoid macular changes. Laser photocoagulation is directed at microaneurysms for focal leakage and is applied in a grid pattern for diffuse leakage. Several prospective randomized clinical trials have shown that laser-treated eyes fare better than untreated eyes: there is a higher rate of modest visual improvement and a lower rate of visual deterioration in eyes treated with laser photocoagulation. In patients with diabetic macular edema, especially the diffuse variety, systemic factors also may playa pathogenic role. Fluid retention and hypertension due to cardiovascular and renal disease exacerbate retinal capillary leakage. Correction of systemic abnormalities (reduced blood pressure, diuresis) may reduce macular edema and should be included as part of the total management of patients with diabetic macular edema. [Key words: diabetes mellitus, diabetic retinopathy, laser photocoagulation, macular edema, retinal pigment epithelium, systemic hypertension.] Ophthalmology 93:989-997, 1986
Edema of the macula in diabetic patients is a frequent component of diabetic retinopathy, as well as a common cause of visual impairment in this disorder. Edema within one-disc diameter of the center of the macula is found in about 9.0% of the diabetic population, approximately 40% of whom have central macular involvement, according to population-based statistics. I As the severity of overall retinopathy increases, the proportion of eyes with macular edema also increases: 3% in eyes with mild nonproliferative diabetic retinopathy (NPDR), 38% with moderate to severe NPDR, and 71 % with proliferative diabetic retinopathy (PDR).* From the Department of Ophthalmology, University of Wisconsin, Madison. Presented in part at the Annual Meeting of the American Academy of Ophthalmology, San Francisco, California, September 29 to October 3, 1985. Reprint requests to George H. Bresnick, MD, Department of Ophthalmology, University of Wisconsin, 600 Highland Avenue, Madison, WI 53792. • Computed from data reported for younger-onset diabetic patients with 10 plus years of diabetes and older-onset diabetic patients with 0-14 and 15 plus years of diabetes. 1
Macular edema occurs earlier after discovery of diabetes in older-onset diabetic patients (diagnosed after age 30) compared with younger-onset diabetic patients (diagnosed before age 30): it is found in only 0.5% of younger-onset diabetic patients with up to ten years of diabetes, whereas in older-onset diabetic patients edema is present in about 3% of non-insulin-taking and 8% of insulin-taking patients with up to three years of diabetes, and about 5% of noninsulin-taking and 10% of insulin-taking patients with at least ten years of diabetes. I Among patients with moderate to severe NPDR, a large proportion (63%) of older-onset diabetic patients (duration 15+ years) have macular edema compared with a much smaller proportion (20%) of younger-onset diabetic patients (duration 10+ years); in the presence ofPDR, however, the proportion with edema in both groups are about the same (70 and 74%, respectively). Because there are many more older-onset diabetic patients in the general population, one can expect to see a greater total number of older-onset compared with younger-onset diabetic patients with macular edema. In the presence of macular edema, older-onset patients are also more likely to have visual impairment; 50% of older-onset diabetic patients 989
OPHTHALMOLOGY
•
JULY 1986
Table 1. Classification of Diabetic Macular Edema I. Focal leakage A. Focal hard exudate rings B. Multifocal edema C. Perifoveolar edema and exudate II. Diffuse leakage A. Diffuse edema (cystoid) B. Systemic factors 1. Fluid retention (cardiac, renal) 2. Severe hyperten~ion 3. Pregnancy C. Cystoid macular edema followin·g panretinal photocoagulation.
•
VOLUME 93
•
NUMBER 7
with macular edema have worse than 20/40 vision, compared with only 20% of younger-onset diabetic patients. 2
METHODS This review is based primarily on the published results from randomized clinical trials of photocoagulation for diabetic macular edema. Clinical material from the University of Wisconsin Retina Service is used for illustrative examples.
Fig I. F al ma ular edema. 9-}ear-old \\hlte woman. diabctl 10 cars. Top/ell. nght C) macular edema \\lth center Inv Ivement. multiple mlcroaneury m scattered around po terior pole. and hard exudate de . lied bcl\\ecn the dl and macula. 20/ O. l op ('1'1/1('(. nuore (In anglogr.tm. right e~e. how lu tel", of leak) mlcroaneury~m\. l i)/I TlKIII . right e)c: IX monthl later. untreatcd. c\udatc has c\tended Into the center of the macula. redu ed to 20/100. ('('I/Ief /t'II. leli eye: macular edema \\lIh center In\o!\cmel1t. ultlple mlcroancury m . hard c\udate. and cOllon-\\ool ,pots arc present. 20/30. Cel/lef. nu n:!>Ccln angt gr.tm. len c.. latc pha!>C h \\\ clu~tcr.. of lea!.) microalll.:ury\m and area fcapillary nonpcrfu Ion. e. p<:clall) tcmporalto the macula. ('el/l('f flghl. left eye. focal argon la!>Cr treatment of mlcroancury m and pametlnal photocoagulation . /Jill/lilli/eli. left eye: one year follo\\lng initlalla r treatment and i\ month\ follo\\lng a 'cond focal treatment of mlcroaneury\m . a ular edema has rc hoo c\cept for mall amount of cxudate nasal to foveola . stable at 201 O.
990
BRESNICK
•
DIABETIC MACULAR EDEMA
Fig 2. Diffuse macular edema, 25-year-old white woman, insulin-taking diabetic patient for 14 years. Top left. left eye; cystoid macular edema with center involvement. There is diffuse edema throughout posterior pole; note absence of hard exudate. VA 20/30. Top right. fluorescein angiogram, left eye; widespread capillary dilation and leakage in the early arteriovenous phase. Bottom left. right eye; cystoid macular edema with center involvement. Note multiple dark, blot hemorrhages temporally. VA 20/30. Bottom right. fluorescein angiogram, left eye; late pooling of dye in cystoid spaces and diffuse staining of entire posterior retina.
RESULTS CLINICAL CHARACfERISTICS
Diabetic macular edema can be subdivided into focal and diffuse categories, although many eyes show features of both (Table 1). Focal macular edema. Clusters of microaneurysms usually are found within areas offocal retinal edema. Fluorescein angiography clearly demonstrates that microaneurysms on the retinal capillaries are a major source of leakage (Fig 1). Hard exudate rings may surround foci of retinal edema; the "exudates" are deposits of lipoprotein in the outer retinal layers or beneath the retina in severe cases. 3 Sometimes circinate rings develop around cottonwool spots; this is not surprising, since leaky microaneurysms and dilated capillaries often form in the capillary bed surrounding the occluded terminal arteriole that produces the cotton-wool spot. 4 Focal leakage from microaneurysms causing focal edema is important to recognize, because it is amenable to focal laser treatment (see below). Diffuse macular edema. Diffuse macular edema results
from a generalized breakdown of the inner blood-retinal barrier (BRB), in which, not only microaneurysms, but also retinal capillaries and even arterioles leak diffusely5.6 (Fig 2). The capillary bed is diffusely dilated and may show widened intercapillary spaces in fluorescein angiograms, corresponding histologically to dilated hypercellular capillaries and numerous acellular occluded capillaries. 4 Whether the capillary closure or the dilation is the primary pathogenic event is unclear. Hard exudate is usually absent in eyes with diffuse edema, suggesting that the generalized defect in the inner BRB is insufficient to allow larger lipoprotein molecules to pass into the extravascular space. Since the retinal pigment epithelium (RPE), forming the outer BRB, contributes to normal fluid dynamics in the retinal extracellular and subretinal space, it has been suggested that RPE dysfunction also might contribute to the formation or persistence of diffuse edema in diabetes. 6 Proof of this is lacking in human diabetic patients, but in experimental streptozotocin diabetes in animals a breakdown of the outer BRB has been demonstrated by tracer studies with electron microscopy.7 Because diffuse macular edema is accompanied by a generalized breakdown 991
OPHTHALMOLOGY
•
JULY 1986
•
VOLUME 93
•
NUMBER 7
Table 2. Study Designs of Randomized Trials of Photocoagulation for Diabetic Macular Edema Entrance Criteria Patz et al 18
Fluorescein Angiography
Photocoagulation Technique
Follow-up Treatment
Argon laser
No PDR 2 eyes symmetric VA:5 20/40
Not required
Multicentre Trial British Diabetes Association 19.20
NPDR and mild PDR 2 eyes symmetric VA:5 6/9 with macular edema > 6/9 with hard exudate "encroaching on fovea"
Not required
Xenon photocoagulation Technique not specified
Not specified
Blankenship8
No PDR 2 eyes symmetric VA> 20/100
Required Identify focal and diffuse leakage
Argon laser Focal leak: 1OO-~ spots Diffuse leak: "grid" with 100-~ spots
Fluorescein if acuity worse and treatment if leakage
Olk11
No PDR 2 eyes symmetric VA < 20/32, > 20/200
Required Identify focal and diffuse leakage
Argon laser Focal leak: 50-100-~ spots Diffuse leak: "grid" with 100-200-~ spots
Fluorescein every four months and treatment if any leakage
ETDRS 21
No PDR At least one eye macular edema VA 2 20/200
Required Argon laser Identify focal and Focal leak: 50-200-~ spot diffuse leakage, Diffuse leak and capillary capillary drop-out nonperfusion: "grid" with 50-200-~ spots
PDR = proliferative diabetic retinopathy; NPDR = nonproliferative diabetic retinopathy; VA = visual acuity; ETDRS Retinopathy Study.
of the BRB, photocoagulation cannot be directed at focal leaks; for this reason, a scatter-type grid treatment has been developed. 8- 11 In addition, correction of systemic abnormalities such as hypertension and severe fluid retention sometimes ameliorates the edema. 5,6,'2 Macular ischemia. If edema were the only factor in diabetic macular disease, the visual prognosis would probably be better than it is; retinal ischemia, however, is often a major complicating feature. Some degree of capillary nonperfusion is probably always present in the edematous diabetic macula, either focally in relation to microaneurysms and cotton-wool spots or more diffusely in the presence of generalized capillary dilation. 4 Enlargement of the foveal avascular zone (FAZ) due to capillary closure is found in many diabetic eyes, with or without macular edema, but it is usually of only minor consequence to vision. 13 More serious is extensive capillary and arteriolar closure, which, in combination with edema, has a poor visual prognosis.1 4 ,18 Macular edema following panretinal photocoagulation (PRP). Some eyes that receive PRP for PDR have macular edema develop for the first time or show a worsening of edema after photocoagulation treatment. 16.17 The edema is usually of the diffuse variety. In a study of 175 eyes treated with PRP for PDR, 75 (43%) eyes had increased macular edema develop six to ten weeks after treatment. 17 The increased edema resolved in 28 eyes but persisted in 47 eyes at the time of the last follow-up visit. (Median
992
Not specified
spot Treat areas of edema; focal versus diffuse Not specified Fluorescein not routine 50-100-~
=
Fluorescein and treatment if "clinically significant macular edema" present Early Treatment Diabetic
follow-up for all eyes was 15 months.) Of the 47 eyes with a persistent post-laser increase in macular edema, 16 lost two or more lines of vision by the last follow-up visit, and in 14 ofthese the vision loss was attributable to macular edema. Thus, out of 175 eyes, 14 (8%) had a persistent post-operative increase in macular edema and vision loss develop. (Median follow-up for this subgroup was 14 months.) (Since this was not a randomized trial, the natural occurrence of macular edema in untreated eyes could notbe compared with the post-treatment incidence.) TREATMENT TECHNIQUES
The techniques to be described represent a compilation of techniques described in previously published randomized clinical trials.8.11.18-21 The specific treatment used in each trial is summarized in Table 2. The treatment approach differs, depending on whether focal or diffuse edema is present. Focal macular edema. The goal of treatment is to coagulate the walls ofthe leaky microaneurysms (Figs 1,3). The argon laser is preferred, because absorption of the blue-green or green wavelengths by hemoglobin is desired. Since the blue wavelengths are also absorbed by macular xanthophyll, a green-only argon laser 2 or a blue-absorbing contact lens23 have been advocated to minimize sensory retinal damage. All microaneurysms in areas of edematous macula are treated, except those closer than 300 f.L to the
BRESNICK • DIABETIC MACULAR EDEMA
Diabet ic Macul ar Edema Treatable Lesions
Diabet ic Macul ar Edema Treatable Lesions Fig 3. Left, diagram to show focal laser treatment of microaneu rysms for focal edema; a 100-200-1' spot whitens the background and a 50-75-1' spot whitens or darkens the microaneurysm. Right, diagram to show grid treatment of diffuse macular edema; 100-200-1' spots are applied to areas of diffuse leakage (and capillary nonperfusion), up to 500 I' of center of macula.
•
e.
•:
00
Focal Edema .................. ..Treat Focan:es ions
/
Diffuse Edema ..................... Laser Grid
Fig 4. Perifoveal exudate, 43year-old white man, insulintaking diabetic patient for 22 years. Top left, left eye; hard exudate deposited adjacent to the foveola; no macular edema in center. VA 20/20. Top right, fluorescein angiogram, left eye; microaneurysms distributed mainly in the perifoveolar capillary bed. Bottom left, left eye; focal argon laser treatment of microaneurysms, mostly perifoveolar. Bottom right, left eye; 12 months following treatment. Exudate has resolved. VA 20/15.
foveal center. Treatm ent in the first session may be restricted to microaneurysms farther than 500 JL from the center, and, if the edema doesn't clear, those closer than
500 JL and those persisting elsewhere may be treated at a second session two to three months later. The exudate often takes four to six months to resorb completely, since
993
OPHTHALMOLOGY •
JULY 1986 •
VOLUME 93 •
NUMBER 7
Fig 5. Diffuse macular edema, 52-year-old white man, insulin-taking diabetic patient for one year. Proliferative retinopathy with cystoid macular edema treated in both eyes with pametinal photocoagulation, December 1982. Developed rapidly progressive renal failure, massive proteinuria, pleural and pericardial effusion, April 1983. BP 224/100; cystoid macular edema worse in both eyes. Top leji, fluorescein angiogram, right eye, diffuse macular leakage and cystoid change. April 1983. VA 7/ 200. Note temporal ischemia and disc new vessels. Top right, fluorescein angiogram, left eye, diffuse macular leakage, similar to right eye. April 1983. VA 11/200. Bottom leji. patient lost 9 kg on fluid restriction, antihypertensives, and diuretics. Cystoid macular edema resolved. BP 152/84. Fluorescein angiogram, right eye, only mild leakage in macula; no edema. Note that with reduction in diffuse leakage, more microaneurysms are visible. (Compare Fig 5, lOP leji) July 1983. V A 20/200. Bottom right, fluorescein angiogram, left eye, mild leakage in macula; no edema. More microaneurysms visible with reduction of diffuse leakage. July 1983. VA 20170.
phagocytosis by macrophages, probably the major mechanism of lipoprotein removal, is a slower process than fluid resorption. Focal edema and exudate develops in some patients primarily in the perifoveolar region, with microaneurysms and focal capillary drop-out in the perifoveolar capillary bed (Fig 4). Treatment of such cases is challenging because of the proximity of the microaneurysms to the foveola, but with retrobulbar anesthesia, direct treatment of the microaneurysms, and care to minimize RPE and photoreceptor damage resolution can be obtained (Fig 4). Mechanism of focal treatment. Although it is generally assumed that this type of focal treatment works by directly coagulating microaneurysms, it is also possible that laser effects on the RPE that inevitably accompany this treatment may play a role in the resolution of focal edema and circinate rings (see "Grid Treatment" below). Diffuse macular edema. Laser spots applied in a grid pattern have been used to treat diffuse macular edema. 8- 1I .21 A pattern of 50-200-~ spots oflight to me994
dium intensity is placed throughout the areas of macular edema, up to 500 ~ of the center of the fovea; nonedematous retina is not treated. Some also advocate additional grid treatment of nonperfused retina, usually temporal to the macula21 (Fig 3). Mechanism of grid treatment. The mechanism of action of the grid treatment is unknown. The RPE and outer retina are acutely damaged by coagulation, but there is experimental evidence that the RPE regenerates and the outer BRB is restored.2 4 Hypotheses to explain the mechanism of action that take this into account include the following: RPE debridement to remove abnormal cells, and removal of some of the photoreceptors to reduce the metabolic needs of the outer retina and to allow the choriocapillaries to deliver oxygen to ischemic inner retina. 6,25 An interesting indirect effect on the retinal vasculature of laser damage to the RPE and outer retina has been found experimentally.26 Autoradiography (thymidine uptake) of retinal vessels overlying mild retinal lesions produced by the krypton laser (which coagulates the RPE and outer
BRESNICK •
DIABETIC MACULAR EDEMA
Table 3. Changes in Visual Acuity Two Years After Photocoagulation Treatment for Diabetic Macular Edema: Five Randomized Clinical Trials Vision Improved by 2 or More Lines Patz et al 18 Treated eyes Control eyes
(%)
Vision Unchanged ±1 Line (%)
Vision Worse by 2 or More Lines (%)
Total No. Eyes
17 (27) 6 (10)
42 (66) 17 (27)
4 (10) 40 (63)
63 63
49 (60) 35 (43)
32 (40) 46 (57)
81 81
Multicentre Trial British Diabetes Association 19.20 * Treated eyes Control eyes Blankenship8 Treated eyes Control eyes
5 (17) 1 (3)
18 (60) 16(53)
7 (23) 13 (43)
30 30
Olk ll Treated eyes Control eyes
19 (45) 3 (8)
19 (45) 18 (49)
4 (10) 16 (43)
42 37
Vision Improved by More Than 1 Line (start < 20/40)
Vision Unchanged (%)
Vision Worse by 3 or More Lines (%)
Total Eyes
65 (16) 88 (11)
321 (77) 595 (73)
30 (7) 129 (16)
416 812
ETDRSfl Treated eyes Control eyes
* Eyes reported as worse if visual acuity worse by one line or more; all other eyes pooled as same or better.
t Two-year data for comparison with other randomized trials provided by the courtesy of the Early Treatment Diabetic Retinopathy Study (ETDRS)
Research Group.
retina but doesn't directly affect the retinal vasculature) showed induced proliferation of endothelial cells in the retinal capillaries and venules. On this basis, the grid treatment might work by stimulating an endothelial repair process and perhaps restore thereby the inner BRB. It is also important to note that diffuse macular edema may spontaneously wax and wane or may respond to correction of systemic abnormalities. For example, resolution of diffuse macular edema has been described following diuresis, dialysis, or kidney transplantation in some diabetic patients with systemic fluid retention. 6 ,15,27 (Fig 5). Similarly, diffuse macular edema in pregnant diabetic women may resolve after parturition. 5 Because of these vagaries, evaluation of new treatments for diffuse edema requires adequate control groups and attention to the patient's systemic status. RESULTS OF CLINICAL TRIALS
There is now a total of five published randomized clinical trials of photocoagulation for diabetic macular edema that can be used as guidelines for patient management. 8,II,18-21 Although the treatment protocols and case selections of each differ, the conclusions are similar: treatment tends to stabilize vision in most eyes by slowing further deterioration, but relatively few eyes improve in vision. Except in the British Multicenter Trial, only eyes with NPDR were entered (Table 2).19,20 In the British
Trial, only mild PDR was allowed, and the visual acuity results are somewhat confounded by inclusion of eyes in both treated and untreated categories that lost substantial vision from vitreous hemorrhage. The type of treatment used in each trial is listed in Table 2. All but one trial employed the argon laser; the British Multicenter Trial used xenon photocoagulation. 19,20 Among the argon laser trials, treatment with a small spot (50-100 J,L) was applied to areas of focal leakage or microaneurysms, and in the three most recent trials a grid pattern (1 00-200-J,L spots) was used in areas of diffuse leakage. 8 ,II,18,21 In addition, the Early Treatment Diabetic Retinopathy Study (ETDRS) used a grid pattern in areas of capillary nonperfusion when diffuse edema was present. 21 It should be emphasized for patient management that many eyes received multiple treatment sessions in these trials. The compulsiveness with which the need for retreatment was assessed and then carried out, however, varied considerably from trial to trial. In the trial by Olk, II fluorescein angiography was repeated at each four-month follow-up visit, and additional treatment was applied for persistent leakage. In the ETDRS, eyes with persistent "clinically significant macular edema" (see below for definition) also had mandatory follow-up angiography, and additional treatment was required if treatable lesions were present. 21 In the British Trial, the initial and follow-up treatment protocol was not prescribed in a standard fash995
OPHTHALMOLOGY
•
JULY 1986 •
ion. In the early trial of Patz et al, angiography was not required and the treatment protocol was not standardized; in the trial of Blankenship, a follow-up routine also was not specified. 8, 18-20 The results of each of the randomized trials are summarized in Table 3 in the form of two-year visual acuity data. While the reporting techniques vary somewhat, the results are really quite comparable. A statistically significant reduction in vision loss could be attributed to treatment except in the smallest trial, where only a trend in this direction was found. 8 Most treated eyes showed no change in vision, and, except for one trial, relatively few treated eyes improved in vision. II The study by Olk, which selected eyes with predominantly diffuse edema, showed a surprisingly high rate of improvement (45%) in treated compared with untreated eyes (8%). II In general, however, the most important treatment effect was a reduction in the proportion of eyes that showed a further loss of vision over the two-year follow-up periods. Because of the large scope of the ETDRS and the recent publication of the study's photocoagulation results for macula edema, a detailed discussion of this trial will be presented. The first ETDRS report described the results of focal laser treatment in the subgroup of eyes with mild to moderate NPDR and macular edema.21 Eyes were randomized to either immediate photocoagulation or deferral of photocoagulation. (The immediate treatment group was further randomized to either immediate focal treatment for macular edema or immediate panretinal photocoagulation followed by focal treatment if the macular edema did not improve.) The report compares only the results of immediate focal treatment of macular edema (754 eyes) with the results of deferral of treatment unless DRS-High Risk Characteristics develop (1490 eyes)?8 Either the argon blue-green or green-only laser was used; both focal treatment of microaneurysms and grid treatment of areas of diffuse leakage and nonperfusion were applied (as described above). The immediate treatment eyes were less likely than the deferred treatment eyes to have significant visual acuity loss, defined as a loss of three or more lines on the ETDRS visual acuity chart, equivalent to a doubling of the visual angle. At three years' follow-up, 12% of the immediate treatment eyes and 24% of the deferred treatment eyes had significant visual acuity loss. While the major treatment effect was to prevent visual acuity loss, a modest improvement in visual acuity also occurred more frequently in treated eyes. Among eyes with less than 20/40 vision before treatment, approximately 20% of eyes assigned to deferral of treatment and 40% of eyes assigned to immediate treatment had gained more than one line of vision after two years of follow-up; improvement of three or more lines was uncommon in both groups. When eyes were subdivided by the severity of macular edema at baseline, the treatment effect (reduction of significant vision loss) was much more pronounced among eyes with clinically significant macular edema than among eyes with less severe edema. The ETDRS defined "clinically significant macular edema" as anyone of the following criteria: (I) thickening of the retina located
996
VOLUME 93
•
NUMBER 7
=:;;500 II- from the center of the macula; (2) hard exudates (with thickening of the adjacent retina) located =:;;500 II-
from the center of the macula; (3) a zone of retina thickening I disc area or larger in size, located =:;; I disc diameter from the center of the macula. In eyes with clinically significant macular edema at baseline the treatment effect was statistically significant by the eight-month follow-up visit and remained significant thereafter. Eyes with less severe edema at baseline did not show a significant treatment effect until after almost three years of follow-up; the proportion of eyes with vision loss before that time was relatively low in both treatment groups. In view of these results, the ETDRS recommended immediate focal argon laser treatment of eyes with clinically significant macular edema; before the development of clinically significant macular edema the risk of vision loss is low, and there is no evidence that treatment at this earlier stage is of added benefit. It should also be emphasized that a beneficial treatment effect was demonstrated for eyes with all levels of starting visual acuity; eyes with very good starting acuity were more likely to maintain that level if treated, and eyes with starting acuity less than 20/40 were more likely to show moderate improvement of acuity as well as more likely to maintain vision if treated. Thus, strong arguments can be made to treat eyes with clinically significant macular edema and good visual acuity to prevent future loss of vision, as well as to treat eyes with reduced acuity to prevent further loss.
DISCUSSION The randomized clinical trials for diabetic macular edema consistently demonstrate a beneficial treatment effect of photocoagulation therapy. Treatment reduces the rate of vision loss, and in a minority of eyes treatment may cause a modest improvement of vision. On the basis of these results, eyes with clinically significant macular edema, as defined in the ETDRS, and NPOR should receive prompt laser treatment. The results of the ETDRS also indicate that eyes with macular edema that is not Clinically significant can be watched and treated if the edema progresses to the clinically significant category. The ETDRS21 and other recent trials8, II have achieved a beneficial effect with the combined use of focal spot treatment of focal leaks and grid treatment for diffuse leakage (and for areas of capillary drop-out in the ETORS). Repeat fluorescein angiography and follow-up treatments for persistent clinically significant macular edema may be needed to achieve results comparable to those in the more recent trials. I 1,21 Since eyes with POR were not included in the randomized trials reported (except in the British Multicenter Trial), the results of treatment of macular edema with and without POR cannot be compared. There is some evidence that PRP for PDR can initiate or exacerbate macular edema,16,17 but there are limited data available to determine the efficacy of focal laser treatment of macular edema when it is combined with PRP for POR. The
BRESNICK •
DIABETIC MACULAR EDEMA
ETDRS has randomized eyes with severe NPDR or mild PDR to immediate laser treatment or deferral of treatment unless DRS-High Risk Characteristics develop. The eyes assigned to immediate laser treatment have PRP as well as focal laser treatment for macular edema if indicated. The results of this part of the ETDRS when available should provide useful data on the effectiveness of laser treatment of macular edema in the presence of more extensive treatment (PRP) for severe NPDR and mild PDR.
REFERENCES 1. Klein R, Klein BEK, Moss SE, et al. The Wisconsin Epidemiologic Study of Diabetic Retinopathy. IV. Diabetic macular edema. Ophthalmology 1984; 91 :1464-74. 2. Klein R, Klein BEK, Moss SE. Visual impairment in diabetes. Ophthalmology 1984; 91:1-8. 3. Bresnick GH. Diabetic retinopathy. In: Peyman GA, Sanders DR, Goldberg MF, eds. Principles and Practice of Ophthalmology. Philadelphia: WB Saunders, 1980; 1237-41. 4. Bresnick GH, Engerman R, Davis MD, et aI. Patterns of ischemia in diabetic retinopathy. Trans Am Acad Ophthalrnol Otolaryngol 1976; 81 :694-709. 5. Kearns M, Hamilton AM, Kohner EM. Excessive permeability in diabetic maculopathy. Br J Ophthalmol1979; 63:489-97. 6. Bresnick GH. Diabetic maculopathy; a critical review highlighting diffuse macular edema. Ophthalmology 1983; 90:1301-17. 7. Tso MOM, Cunha-Vaz JGF, Shih CY, Jones CWo A clinicopathologic study of blood-retinal barrier in experimental diabetes. ARVO Abstracts. Invest Ophthalmol Vis Sci 1979; 18(Suppl):169. 8. Blankenship GW. Diabetic macular edema and argon laser photocoagulation: a prospective randomized study. Ophthalmology 1979; 86:69-78. 9. Whitelocke RAF, Kearns M, Black RK, Hamilton AM. The diabetic maculopathies. Trans Ophthalmol Soc UK 1979; 99:314-20. 10. McDonald HR, Schatz H. Grid photocoagulation for diffuse macular edema. Retina 1985; 5:65-72. 11. Olk RJ. Modified grid argon (blue-green) laser photocoagulation for diffuse diabetic macular edema. Ophthalmology 1986; 93:938-48. 12. Ferris FL III, Patz A. Macular edema. A complication of diabetic retinopathy. Surv Ophthalmol1984; 28:452-61.
13. Bresnick GH, Condit R, Syrjala S, et al. Abnormalities of the foveal avascular zone in diabetic retinopathy. Arch Ophthalmol1984; 102: 1286-93. 14. Ticho U, Patz A. The role of capillary perfusion in the management of diabetic macular edema. Am J Ophthalmol1973; 76:880-6. 15. Bresnick GH, de Venecia G, Myers FL, et al. Retinal ischemia in diabetic retinopathy. Arch Ophthalmol1975; 93:1300-10. 16. Meyers SM. Macular edema after scatter laser photocoagulation for proliferative diabetic retinopathy. Am J Ophthalmol1980; 90:210-6. 17. McDonald HR, Schatz H. Macular edema following panretinal photocoagulation. Retina 1985; 5:5-10. 18. Patz A, Schatz H, Berkow JW, et al. Macular edema-an overlooked complication of diabetic retinopathy. Trans Am Acad Ophthalmol 010laryngol1973; 77:0P34-42. 19. Multicentered Controlled Study. Photocoagulation in treatment of diabetic maculopathy. Lancet 1975; 2:1110-3. 20. Townsend C, Bailey J, Kohner E. Xenon arc photocoagulation for the treatment of diabetic maculopathy; interim report of a multicentre controlled clinical study. Br J OphthalmoI1980; 64:385-91. 21. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular edema; Early Treatment Diabetic Retinopathy Study report number 1. Arch OphthalmoI1985; 103: 1796806. 22. Trempe CL, Mainster MA, Pomerantzeff 0, et al. Macular photocoagulation; optimal wavelength selection. Ophthalmology 1982; 89: 721-8. 23. Eisner A, Chenoweth RG. A low-cost contact lens for green-only argonlaser photocoagulation. Retina 1984; 4:126-8. 24. Wallow IH. Repair of the pigment epithelial barrier following photocoagulation. Arch Ophthalmol1984; 102:126-35. 25. Weiter JJ, Zuckerman R. The influence of the photoreceptor-RPE complex on the inner retina; an explanation for the beneficial effects of photocoagulation. Ophthalmology 1980; 87:1133-9. 26. MarShall J, Clover G, Rothery S. Some new findings on retinal irradiation by krypton and argon lasers. Doc Ophthalmol Proc Ser 1984; 36:2137. 27. Aiello LM, Rand LI, Briones JC, et al. Nonocular clinical risk factors in the progression of diabetic retinopathy, In: Little HL, Jack RL, Patz A, Forsham PH, eds. Diabetic Retinopathy. New York: Thieme-Stratton, 1983; 21-32. 28. Diabetic Retinopathy Study Research Group. Photocoagulation treatment of diabetic retinopathy; clinical application of diabetic retinopathy study (DRS) findings, DRS report number 8. Ophthalmology 1981; 88:583-600.
997
Abstract: Diabetic macular edema can be classified into a focal variety, characterized by focal leakage from microaneurysms, often with accumulation of extravascular lipoprotein in a circinate pattern around the focal leakage, and a diffuse variety, with diffuse leakage from retinal vessels often accompanied by cystoid macular changes. Laser photocoagulation is directed at microaneurysms for focal leakage and is applied in a grid pattern for diffuse leakage. Several prospective randomized clinical trials have shown that laser-treated eyes fare better than untreated eyes: there is a higher rate of modest visual improvement and a lower rate of visual deterioration in eyes treated with laser photocoagulation. In patients with diabetic macular edema, especially the diffuse variety, systemic factors also may playa pathogenic role. Fluid retention and hypertension due to cardiovascular and renal disease exacerbate retinal capillary leakage. Correction of systemic abnormalities (reduced blood pressure, diuresis) may reduce macular edema and should be included as part of the total management of patients with diabetic macular edema. [Key words: diabetes mellitus, diabetic retinopathy, laser photocoagulation, macular edema, retinal pigment epithelium, systemic hypertension.] Ophthalmology 93:989-997, 1986
Edema of the macula in diabetic patients is a frequent component of diabetic retinopathy, as well as a common cause of visual impairment in this disorder. Edema within one-disc diameter of the center of the macula is found in about 9.0% of the diabetic population, approximately 40% of whom have central macular involvement, according to population-based statistics. I As the severity of overall retinopathy increases, the proportion of eyes with macular edema also increases: 3% in eyes with mild nonproliferative diabetic retinopathy (NPDR), 38% with moderate to severe NPDR, and 71 % with proliferative diabetic retinopathy (PDR).* From the Department of Ophthalmology, University of Wisconsin, Madison. Presented in part at the Annual Meeting of the American Academy of Ophthalmology, San Francisco, California, September 29 to October 3, 1985. Reprint requests to George H. Bresnick, MD, Department of Ophthalmology, University of Wisconsin, 600 Highland Avenue, Madison, WI 53792. • Computed from data reported for younger-onset diabetic patients with 10 plus years of diabetes and older-onset diabetic patients with 0-14 and 15 plus years of diabetes. 1
Macular edema occurs earlier after discovery of diabetes in older-onset diabetic patients (diagnosed after age 30) compared with younger-onset diabetic patients (diagnosed before age 30): it is found in only 0.5% of younger-onset diabetic patients with up to ten years of diabetes, whereas in older-onset diabetic patients edema is present in about 3% of non-insulin-taking and 8% of insulin-taking patients with up to three years of diabetes, and about 5% of noninsulin-taking and 10% of insulin-taking patients with at least ten years of diabetes. I Among patients with moderate to severe NPDR, a large proportion (63%) of older-onset diabetic patients (duration 15+ years) have macular edema compared with a much smaller proportion (20%) of younger-onset diabetic patients (duration 10+ years); in the presence ofPDR, however, the proportion with edema in both groups are about the same (70 and 74%, respectively). Because there are many more older-onset diabetic patients in the general population, one can expect to see a greater total number of older-onset compared with younger-onset diabetic patients with macular edema. In the presence of macular edema, older-onset patients are also more likely to have visual impairment; 50% of older-onset diabetic patients 989
OPHTHALMOLOGY
•
JULY 1986
Table 1. Classification of Diabetic Macular Edema I. Focal leakage A. Focal hard exudate rings B. Multifocal edema C. Perifoveolar edema and exudate II. Diffuse leakage A. Diffuse edema (cystoid) B. Systemic factors 1. Fluid retention (cardiac, renal) 2. Severe hyperten~ion 3. Pregnancy C. Cystoid macular edema followin·g panretinal photocoagulation.
•
VOLUME 93
•
NUMBER 7
with macular edema have worse than 20/40 vision, compared with only 20% of younger-onset diabetic patients. 2
METHODS This review is based primarily on the published results from randomized clinical trials of photocoagulation for diabetic macular edema. Clinical material from the University of Wisconsin Retina Service is used for illustrative examples.
Fig I. F al ma ular edema. 9-}ear-old \\hlte woman. diabctl 10 cars. Top/ell. nght C) macular edema \\lth center Inv Ivement. multiple mlcroaneury m scattered around po terior pole. and hard exudate de . lied bcl\\ecn the dl and macula. 20/ O. l op ('1'1/1('(. nuore (In anglogr.tm. right e~e. how lu tel", of leak) mlcroaneury~m\. l i)/I TlKIII . right e)c: IX monthl later. untreatcd. c\udatc has c\tended Into the center of the macula. redu ed to 20/100. ('('I/Ief /t'II. leli eye: macular edema \\lIh center In\o!\cmel1t. ultlple mlcroancury m . hard c\udate. and cOllon-\\ool ,pots arc present. 20/30. Cel/lef. nu n:!>Ccln angt gr.tm. len c.. latc pha!>C h \\\ clu~tcr.. of lea!.) microalll.:ury\m and area fcapillary nonpcrfu Ion. e. p<:clall) tcmporalto the macula. ('el/l('f flghl. left eye. focal argon la!>Cr treatment of mlcroancury m and pametlnal photocoagulation . /Jill/lilli/eli. left eye: one year follo\\lng initlalla r treatment and i\ month\ follo\\lng a 'cond focal treatment of mlcroaneury\m . a ular edema has rc hoo c\cept for mall amount of cxudate nasal to foveola . stable at 201 O.
990
BRESNICK
•
DIABETIC MACULAR EDEMA
Fig 2. Diffuse macular edema, 25-year-old white woman, insulin-taking diabetic patient for 14 years. Top left. left eye; cystoid macular edema with center involvement. There is diffuse edema throughout posterior pole; note absence of hard exudate. VA 20/30. Top right. fluorescein angiogram, left eye; widespread capillary dilation and leakage in the early arteriovenous phase. Bottom left. right eye; cystoid macular edema with center involvement. Note multiple dark, blot hemorrhages temporally. VA 20/30. Bottom right. fluorescein angiogram, left eye; late pooling of dye in cystoid spaces and diffuse staining of entire posterior retina.
RESULTS CLINICAL CHARACfERISTICS
Diabetic macular edema can be subdivided into focal and diffuse categories, although many eyes show features of both (Table 1). Focal macular edema. Clusters of microaneurysms usually are found within areas offocal retinal edema. Fluorescein angiography clearly demonstrates that microaneurysms on the retinal capillaries are a major source of leakage (Fig 1). Hard exudate rings may surround foci of retinal edema; the "exudates" are deposits of lipoprotein in the outer retinal layers or beneath the retina in severe cases. 3 Sometimes circinate rings develop around cottonwool spots; this is not surprising, since leaky microaneurysms and dilated capillaries often form in the capillary bed surrounding the occluded terminal arteriole that produces the cotton-wool spot. 4 Focal leakage from microaneurysms causing focal edema is important to recognize, because it is amenable to focal laser treatment (see below). Diffuse macular edema. Diffuse macular edema results
from a generalized breakdown of the inner blood-retinal barrier (BRB), in which, not only microaneurysms, but also retinal capillaries and even arterioles leak diffusely5.6 (Fig 2). The capillary bed is diffusely dilated and may show widened intercapillary spaces in fluorescein angiograms, corresponding histologically to dilated hypercellular capillaries and numerous acellular occluded capillaries. 4 Whether the capillary closure or the dilation is the primary pathogenic event is unclear. Hard exudate is usually absent in eyes with diffuse edema, suggesting that the generalized defect in the inner BRB is insufficient to allow larger lipoprotein molecules to pass into the extravascular space. Since the retinal pigment epithelium (RPE), forming the outer BRB, contributes to normal fluid dynamics in the retinal extracellular and subretinal space, it has been suggested that RPE dysfunction also might contribute to the formation or persistence of diffuse edema in diabetes. 6 Proof of this is lacking in human diabetic patients, but in experimental streptozotocin diabetes in animals a breakdown of the outer BRB has been demonstrated by tracer studies with electron microscopy.7 Because diffuse macular edema is accompanied by a generalized breakdown 991
OPHTHALMOLOGY
•
JULY 1986
•
VOLUME 93
•
NUMBER 7
Table 2. Study Designs of Randomized Trials of Photocoagulation for Diabetic Macular Edema Entrance Criteria Patz et al 18
Fluorescein Angiography
Photocoagulation Technique
Follow-up Treatment
Argon laser
No PDR 2 eyes symmetric VA:5 20/40
Not required
Multicentre Trial British Diabetes Association 19.20
NPDR and mild PDR 2 eyes symmetric VA:5 6/9 with macular edema > 6/9 with hard exudate "encroaching on fovea"
Not required
Xenon photocoagulation Technique not specified
Not specified
Blankenship8
No PDR 2 eyes symmetric VA> 20/100
Required Identify focal and diffuse leakage
Argon laser Focal leak: 1OO-~ spots Diffuse leak: "grid" with 100-~ spots
Fluorescein if acuity worse and treatment if leakage
Olk11
No PDR 2 eyes symmetric VA < 20/32, > 20/200
Required Identify focal and diffuse leakage
Argon laser Focal leak: 50-100-~ spots Diffuse leak: "grid" with 100-200-~ spots
Fluorescein every four months and treatment if any leakage
ETDRS 21
No PDR At least one eye macular edema VA 2 20/200
Required Argon laser Identify focal and Focal leak: 50-200-~ spot diffuse leakage, Diffuse leak and capillary capillary drop-out nonperfusion: "grid" with 50-200-~ spots
PDR = proliferative diabetic retinopathy; NPDR = nonproliferative diabetic retinopathy; VA = visual acuity; ETDRS Retinopathy Study.
of the BRB, photocoagulation cannot be directed at focal leaks; for this reason, a scatter-type grid treatment has been developed. 8- 11 In addition, correction of systemic abnormalities such as hypertension and severe fluid retention sometimes ameliorates the edema. 5,6,'2 Macular ischemia. If edema were the only factor in diabetic macular disease, the visual prognosis would probably be better than it is; retinal ischemia, however, is often a major complicating feature. Some degree of capillary nonperfusion is probably always present in the edematous diabetic macula, either focally in relation to microaneurysms and cotton-wool spots or more diffusely in the presence of generalized capillary dilation. 4 Enlargement of the foveal avascular zone (FAZ) due to capillary closure is found in many diabetic eyes, with or without macular edema, but it is usually of only minor consequence to vision. 13 More serious is extensive capillary and arteriolar closure, which, in combination with edema, has a poor visual prognosis.1 4 ,18 Macular edema following panretinal photocoagulation (PRP). Some eyes that receive PRP for PDR have macular edema develop for the first time or show a worsening of edema after photocoagulation treatment. 16.17 The edema is usually of the diffuse variety. In a study of 175 eyes treated with PRP for PDR, 75 (43%) eyes had increased macular edema develop six to ten weeks after treatment. 17 The increased edema resolved in 28 eyes but persisted in 47 eyes at the time of the last follow-up visit. (Median
992
Not specified
spot Treat areas of edema; focal versus diffuse Not specified Fluorescein not routine 50-100-~
=
Fluorescein and treatment if "clinically significant macular edema" present Early Treatment Diabetic
follow-up for all eyes was 15 months.) Of the 47 eyes with a persistent post-laser increase in macular edema, 16 lost two or more lines of vision by the last follow-up visit, and in 14 ofthese the vision loss was attributable to macular edema. Thus, out of 175 eyes, 14 (8%) had a persistent post-operative increase in macular edema and vision loss develop. (Median follow-up for this subgroup was 14 months.) (Since this was not a randomized trial, the natural occurrence of macular edema in untreated eyes could notbe compared with the post-treatment incidence.) TREATMENT TECHNIQUES
The techniques to be described represent a compilation of techniques described in previously published randomized clinical trials.8.11.18-21 The specific treatment used in each trial is summarized in Table 2. The treatment approach differs, depending on whether focal or diffuse edema is present. Focal macular edema. The goal of treatment is to coagulate the walls ofthe leaky microaneurysms (Figs 1,3). The argon laser is preferred, because absorption of the blue-green or green wavelengths by hemoglobin is desired. Since the blue wavelengths are also absorbed by macular xanthophyll, a green-only argon laser 2 or a blue-absorbing contact lens23 have been advocated to minimize sensory retinal damage. All microaneurysms in areas of edematous macula are treated, except those closer than 300 f.L to the
BRESNICK • DIABETIC MACULAR EDEMA
Diabet ic Macul ar Edema Treatable Lesions
Diabet ic Macul ar Edema Treatable Lesions Fig 3. Left, diagram to show focal laser treatment of microaneu rysms for focal edema; a 100-200-1' spot whitens the background and a 50-75-1' spot whitens or darkens the microaneurysm. Right, diagram to show grid treatment of diffuse macular edema; 100-200-1' spots are applied to areas of diffuse leakage (and capillary nonperfusion), up to 500 I' of center of macula.
•
e.
•:
00
Focal Edema .................. ..Treat Focan:es ions
/
Diffuse Edema ..................... Laser Grid
Fig 4. Perifoveal exudate, 43year-old white man, insulintaking diabetic patient for 22 years. Top left, left eye; hard exudate deposited adjacent to the foveola; no macular edema in center. VA 20/20. Top right, fluorescein angiogram, left eye; microaneurysms distributed mainly in the perifoveolar capillary bed. Bottom left, left eye; focal argon laser treatment of microaneurysms, mostly perifoveolar. Bottom right, left eye; 12 months following treatment. Exudate has resolved. VA 20/15.
foveal center. Treatm ent in the first session may be restricted to microaneurysms farther than 500 JL from the center, and, if the edema doesn't clear, those closer than
500 JL and those persisting elsewhere may be treated at a second session two to three months later. The exudate often takes four to six months to resorb completely, since
993
OPHTHALMOLOGY •
JULY 1986 •
VOLUME 93 •
NUMBER 7
Fig 5. Diffuse macular edema, 52-year-old white man, insulin-taking diabetic patient for one year. Proliferative retinopathy with cystoid macular edema treated in both eyes with pametinal photocoagulation, December 1982. Developed rapidly progressive renal failure, massive proteinuria, pleural and pericardial effusion, April 1983. BP 224/100; cystoid macular edema worse in both eyes. Top leji, fluorescein angiogram, right eye, diffuse macular leakage and cystoid change. April 1983. VA 7/ 200. Note temporal ischemia and disc new vessels. Top right, fluorescein angiogram, left eye, diffuse macular leakage, similar to right eye. April 1983. VA 11/200. Bottom leji. patient lost 9 kg on fluid restriction, antihypertensives, and diuretics. Cystoid macular edema resolved. BP 152/84. Fluorescein angiogram, right eye, only mild leakage in macula; no edema. Note that with reduction in diffuse leakage, more microaneurysms are visible. (Compare Fig 5, lOP leji) July 1983. V A 20/200. Bottom right, fluorescein angiogram, left eye, mild leakage in macula; no edema. More microaneurysms visible with reduction of diffuse leakage. July 1983. VA 20170.
phagocytosis by macrophages, probably the major mechanism of lipoprotein removal, is a slower process than fluid resorption. Focal edema and exudate develops in some patients primarily in the perifoveolar region, with microaneurysms and focal capillary drop-out in the perifoveolar capillary bed (Fig 4). Treatment of such cases is challenging because of the proximity of the microaneurysms to the foveola, but with retrobulbar anesthesia, direct treatment of the microaneurysms, and care to minimize RPE and photoreceptor damage resolution can be obtained (Fig 4). Mechanism of focal treatment. Although it is generally assumed that this type of focal treatment works by directly coagulating microaneurysms, it is also possible that laser effects on the RPE that inevitably accompany this treatment may play a role in the resolution of focal edema and circinate rings (see "Grid Treatment" below). Diffuse macular edema. Laser spots applied in a grid pattern have been used to treat diffuse macular edema. 8- 1I .21 A pattern of 50-200-~ spots oflight to me994
dium intensity is placed throughout the areas of macular edema, up to 500 ~ of the center of the fovea; nonedematous retina is not treated. Some also advocate additional grid treatment of nonperfused retina, usually temporal to the macula21 (Fig 3). Mechanism of grid treatment. The mechanism of action of the grid treatment is unknown. The RPE and outer retina are acutely damaged by coagulation, but there is experimental evidence that the RPE regenerates and the outer BRB is restored.2 4 Hypotheses to explain the mechanism of action that take this into account include the following: RPE debridement to remove abnormal cells, and removal of some of the photoreceptors to reduce the metabolic needs of the outer retina and to allow the choriocapillaries to deliver oxygen to ischemic inner retina. 6,25 An interesting indirect effect on the retinal vasculature of laser damage to the RPE and outer retina has been found experimentally.26 Autoradiography (thymidine uptake) of retinal vessels overlying mild retinal lesions produced by the krypton laser (which coagulates the RPE and outer
BRESNICK •
DIABETIC MACULAR EDEMA
Table 3. Changes in Visual Acuity Two Years After Photocoagulation Treatment for Diabetic Macular Edema: Five Randomized Clinical Trials Vision Improved by 2 or More Lines Patz et al 18 Treated eyes Control eyes
(%)
Vision Unchanged ±1 Line (%)
Vision Worse by 2 or More Lines (%)
Total No. Eyes
17 (27) 6 (10)
42 (66) 17 (27)
4 (10) 40 (63)
63 63
49 (60) 35 (43)
32 (40) 46 (57)
81 81
Multicentre Trial British Diabetes Association 19.20 * Treated eyes Control eyes Blankenship8 Treated eyes Control eyes
5 (17) 1 (3)
18 (60) 16(53)
7 (23) 13 (43)
30 30
Olk ll Treated eyes Control eyes
19 (45) 3 (8)
19 (45) 18 (49)
4 (10) 16 (43)
42 37
Vision Improved by More Than 1 Line (start < 20/40)
Vision Unchanged (%)
Vision Worse by 3 or More Lines (%)
Total Eyes
65 (16) 88 (11)
321 (77) 595 (73)
30 (7) 129 (16)
416 812
ETDRSfl Treated eyes Control eyes
* Eyes reported as worse if visual acuity worse by one line or more; all other eyes pooled as same or better.
t Two-year data for comparison with other randomized trials provided by the courtesy of the Early Treatment Diabetic Retinopathy Study (ETDRS)
Research Group.
retina but doesn't directly affect the retinal vasculature) showed induced proliferation of endothelial cells in the retinal capillaries and venules. On this basis, the grid treatment might work by stimulating an endothelial repair process and perhaps restore thereby the inner BRB. It is also important to note that diffuse macular edema may spontaneously wax and wane or may respond to correction of systemic abnormalities. For example, resolution of diffuse macular edema has been described following diuresis, dialysis, or kidney transplantation in some diabetic patients with systemic fluid retention. 6 ,15,27 (Fig 5). Similarly, diffuse macular edema in pregnant diabetic women may resolve after parturition. 5 Because of these vagaries, evaluation of new treatments for diffuse edema requires adequate control groups and attention to the patient's systemic status. RESULTS OF CLINICAL TRIALS
There is now a total of five published randomized clinical trials of photocoagulation for diabetic macular edema that can be used as guidelines for patient management. 8,II,18-21 Although the treatment protocols and case selections of each differ, the conclusions are similar: treatment tends to stabilize vision in most eyes by slowing further deterioration, but relatively few eyes improve in vision. Except in the British Multicenter Trial, only eyes with NPDR were entered (Table 2).19,20 In the British
Trial, only mild PDR was allowed, and the visual acuity results are somewhat confounded by inclusion of eyes in both treated and untreated categories that lost substantial vision from vitreous hemorrhage. The type of treatment used in each trial is listed in Table 2. All but one trial employed the argon laser; the British Multicenter Trial used xenon photocoagulation. 19,20 Among the argon laser trials, treatment with a small spot (50-100 J,L) was applied to areas of focal leakage or microaneurysms, and in the three most recent trials a grid pattern (1 00-200-J,L spots) was used in areas of diffuse leakage. 8 ,II,18,21 In addition, the Early Treatment Diabetic Retinopathy Study (ETDRS) used a grid pattern in areas of capillary nonperfusion when diffuse edema was present. 21 It should be emphasized for patient management that many eyes received multiple treatment sessions in these trials. The compulsiveness with which the need for retreatment was assessed and then carried out, however, varied considerably from trial to trial. In the trial by Olk, II fluorescein angiography was repeated at each four-month follow-up visit, and additional treatment was applied for persistent leakage. In the ETDRS, eyes with persistent "clinically significant macular edema" (see below for definition) also had mandatory follow-up angiography, and additional treatment was required if treatable lesions were present. 21 In the British Trial, the initial and follow-up treatment protocol was not prescribed in a standard fash995
OPHTHALMOLOGY
•
JULY 1986 •
ion. In the early trial of Patz et al, angiography was not required and the treatment protocol was not standardized; in the trial of Blankenship, a follow-up routine also was not specified. 8, 18-20 The results of each of the randomized trials are summarized in Table 3 in the form of two-year visual acuity data. While the reporting techniques vary somewhat, the results are really quite comparable. A statistically significant reduction in vision loss could be attributed to treatment except in the smallest trial, where only a trend in this direction was found. 8 Most treated eyes showed no change in vision, and, except for one trial, relatively few treated eyes improved in vision. II The study by Olk, which selected eyes with predominantly diffuse edema, showed a surprisingly high rate of improvement (45%) in treated compared with untreated eyes (8%). II In general, however, the most important treatment effect was a reduction in the proportion of eyes that showed a further loss of vision over the two-year follow-up periods. Because of the large scope of the ETDRS and the recent publication of the study's photocoagulation results for macula edema, a detailed discussion of this trial will be presented. The first ETDRS report described the results of focal laser treatment in the subgroup of eyes with mild to moderate NPDR and macular edema.21 Eyes were randomized to either immediate photocoagulation or deferral of photocoagulation. (The immediate treatment group was further randomized to either immediate focal treatment for macular edema or immediate panretinal photocoagulation followed by focal treatment if the macular edema did not improve.) The report compares only the results of immediate focal treatment of macular edema (754 eyes) with the results of deferral of treatment unless DRS-High Risk Characteristics develop (1490 eyes)?8 Either the argon blue-green or green-only laser was used; both focal treatment of microaneurysms and grid treatment of areas of diffuse leakage and nonperfusion were applied (as described above). The immediate treatment eyes were less likely than the deferred treatment eyes to have significant visual acuity loss, defined as a loss of three or more lines on the ETDRS visual acuity chart, equivalent to a doubling of the visual angle. At three years' follow-up, 12% of the immediate treatment eyes and 24% of the deferred treatment eyes had significant visual acuity loss. While the major treatment effect was to prevent visual acuity loss, a modest improvement in visual acuity also occurred more frequently in treated eyes. Among eyes with less than 20/40 vision before treatment, approximately 20% of eyes assigned to deferral of treatment and 40% of eyes assigned to immediate treatment had gained more than one line of vision after two years of follow-up; improvement of three or more lines was uncommon in both groups. When eyes were subdivided by the severity of macular edema at baseline, the treatment effect (reduction of significant vision loss) was much more pronounced among eyes with clinically significant macular edema than among eyes with less severe edema. The ETDRS defined "clinically significant macular edema" as anyone of the following criteria: (I) thickening of the retina located
996
VOLUME 93
•
NUMBER 7
=:;;500 II- from the center of the macula; (2) hard exudates (with thickening of the adjacent retina) located =:;;500 II-
from the center of the macula; (3) a zone of retina thickening I disc area or larger in size, located =:;; I disc diameter from the center of the macula. In eyes with clinically significant macular edema at baseline the treatment effect was statistically significant by the eight-month follow-up visit and remained significant thereafter. Eyes with less severe edema at baseline did not show a significant treatment effect until after almost three years of follow-up; the proportion of eyes with vision loss before that time was relatively low in both treatment groups. In view of these results, the ETDRS recommended immediate focal argon laser treatment of eyes with clinically significant macular edema; before the development of clinically significant macular edema the risk of vision loss is low, and there is no evidence that treatment at this earlier stage is of added benefit. It should also be emphasized that a beneficial treatment effect was demonstrated for eyes with all levels of starting visual acuity; eyes with very good starting acuity were more likely to maintain that level if treated, and eyes with starting acuity less than 20/40 were more likely to show moderate improvement of acuity as well as more likely to maintain vision if treated. Thus, strong arguments can be made to treat eyes with clinically significant macular edema and good visual acuity to prevent future loss of vision, as well as to treat eyes with reduced acuity to prevent further loss.
DISCUSSION The randomized clinical trials for diabetic macular edema consistently demonstrate a beneficial treatment effect of photocoagulation therapy. Treatment reduces the rate of vision loss, and in a minority of eyes treatment may cause a modest improvement of vision. On the basis of these results, eyes with clinically significant macular edema, as defined in the ETDRS, and NPOR should receive prompt laser treatment. The results of the ETDRS also indicate that eyes with macular edema that is not Clinically significant can be watched and treated if the edema progresses to the clinically significant category. The ETDRS21 and other recent trials8, II have achieved a beneficial effect with the combined use of focal spot treatment of focal leaks and grid treatment for diffuse leakage (and for areas of capillary drop-out in the ETORS). Repeat fluorescein angiography and follow-up treatments for persistent clinically significant macular edema may be needed to achieve results comparable to those in the more recent trials. I 1,21 Since eyes with POR were not included in the randomized trials reported (except in the British Multicenter Trial), the results of treatment of macular edema with and without POR cannot be compared. There is some evidence that PRP for PDR can initiate or exacerbate macular edema,16,17 but there are limited data available to determine the efficacy of focal laser treatment of macular edema when it is combined with PRP for POR. The
BRESNICK •
DIABETIC MACULAR EDEMA
ETDRS has randomized eyes with severe NPDR or mild PDR to immediate laser treatment or deferral of treatment unless DRS-High Risk Characteristics develop. The eyes assigned to immediate laser treatment have PRP as well as focal laser treatment for macular edema if indicated. The results of this part of the ETDRS when available should provide useful data on the effectiveness of laser treatment of macular edema in the presence of more extensive treatment (PRP) for severe NPDR and mild PDR.
REFERENCES 1. Klein R, Klein BEK, Moss SE, et al. The Wisconsin Epidemiologic Study of Diabetic Retinopathy. IV. Diabetic macular edema. Ophthalmology 1984; 91 :1464-74. 2. Klein R, Klein BEK, Moss SE. Visual impairment in diabetes. Ophthalmology 1984; 91:1-8. 3. Bresnick GH. Diabetic retinopathy. In: Peyman GA, Sanders DR, Goldberg MF, eds. Principles and Practice of Ophthalmology. Philadelphia: WB Saunders, 1980; 1237-41. 4. Bresnick GH, Engerman R, Davis MD, et aI. Patterns of ischemia in diabetic retinopathy. Trans Am Acad Ophthalrnol Otolaryngol 1976; 81 :694-709. 5. Kearns M, Hamilton AM, Kohner EM. Excessive permeability in diabetic maculopathy. Br J Ophthalmol1979; 63:489-97. 6. Bresnick GH. Diabetic maculopathy; a critical review highlighting diffuse macular edema. Ophthalmology 1983; 90:1301-17. 7. Tso MOM, Cunha-Vaz JGF, Shih CY, Jones CWo A clinicopathologic study of blood-retinal barrier in experimental diabetes. ARVO Abstracts. Invest Ophthalmol Vis Sci 1979; 18(Suppl):169. 8. Blankenship GW. Diabetic macular edema and argon laser photocoagulation: a prospective randomized study. Ophthalmology 1979; 86:69-78. 9. Whitelocke RAF, Kearns M, Black RK, Hamilton AM. The diabetic maculopathies. Trans Ophthalmol Soc UK 1979; 99:314-20. 10. McDonald HR, Schatz H. Grid photocoagulation for diffuse macular edema. Retina 1985; 5:65-72. 11. Olk RJ. Modified grid argon (blue-green) laser photocoagulation for diffuse diabetic macular edema. Ophthalmology 1986; 93:938-48. 12. Ferris FL III, Patz A. Macular edema. A complication of diabetic retinopathy. Surv Ophthalmol1984; 28:452-61.
13. Bresnick GH, Condit R, Syrjala S, et al. Abnormalities of the foveal avascular zone in diabetic retinopathy. Arch Ophthalmol1984; 102: 1286-93. 14. Ticho U, Patz A. The role of capillary perfusion in the management of diabetic macular edema. Am J Ophthalmol1973; 76:880-6. 15. Bresnick GH, de Venecia G, Myers FL, et al. Retinal ischemia in diabetic retinopathy. Arch Ophthalmol1975; 93:1300-10. 16. Meyers SM. Macular edema after scatter laser photocoagulation for proliferative diabetic retinopathy. Am J Ophthalmol1980; 90:210-6. 17. McDonald HR, Schatz H. Macular edema following panretinal photocoagulation. Retina 1985; 5:5-10. 18. Patz A, Schatz H, Berkow JW, et al. Macular edema-an overlooked complication of diabetic retinopathy. Trans Am Acad Ophthalmol 010laryngol1973; 77:0P34-42. 19. Multicentered Controlled Study. Photocoagulation in treatment of diabetic maculopathy. Lancet 1975; 2:1110-3. 20. Townsend C, Bailey J, Kohner E. Xenon arc photocoagulation for the treatment of diabetic maculopathy; interim report of a multicentre controlled clinical study. Br J OphthalmoI1980; 64:385-91. 21. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular edema; Early Treatment Diabetic Retinopathy Study report number 1. Arch OphthalmoI1985; 103: 1796806. 22. Trempe CL, Mainster MA, Pomerantzeff 0, et al. Macular photocoagulation; optimal wavelength selection. Ophthalmology 1982; 89: 721-8. 23. Eisner A, Chenoweth RG. A low-cost contact lens for green-only argonlaser photocoagulation. Retina 1984; 4:126-8. 24. Wallow IH. Repair of the pigment epithelial barrier following photocoagulation. Arch Ophthalmol1984; 102:126-35. 25. Weiter JJ, Zuckerman R. The influence of the photoreceptor-RPE complex on the inner retina; an explanation for the beneficial effects of photocoagulation. Ophthalmology 1980; 87:1133-9. 26. MarShall J, Clover G, Rothery S. Some new findings on retinal irradiation by krypton and argon lasers. Doc Ophthalmol Proc Ser 1984; 36:2137. 27. Aiello LM, Rand LI, Briones JC, et al. Nonocular clinical risk factors in the progression of diabetic retinopathy, In: Little HL, Jack RL, Patz A, Forsham PH, eds. Diabetic Retinopathy. New York: Thieme-Stratton, 1983; 21-32. 28. Diabetic Retinopathy Study Research Group. Photocoagulation treatment of diabetic retinopathy; clinical application of diabetic retinopathy study (DRS) findings, DRS report number 8. Ophthalmology 1981; 88:583-600.
997