Optical Coherence Tomographic Patterns in Diabetic Macula Edema Can Predict the Effects of Intravitreal Bevacizumab Injection as Primary Treatment

Abstract

Purpose:

To identify optical coherence tomography (OCT) patterns in diabetic macular edema (DME) that were predictive of visual outcomes after intravitreal bevacizumab (IVB) injection.

Methods:

This was a retrospective study. We examined 31 eyes (24 patients) with clinically significant macular edema that received IVB injections along with macular OCT data. The eyes were categorized into 4 groups by using OCT features: diffuse retinal thickening (DRT), cystoid macular edema (CME), serous retinal detachment (SRD), and vitreomacular interface abnormalities (VMIAs). Changes in retinal thickness, retinal volume, and visual acuity (VA) after IVB injection were compared on the basis of OCT patterns.

Results:

After IVB injections, changes in VA logarithm of the minimum angle of resolution were −0.06±0.36, −0.26±0.26, 0.09±0.13, and −0.08±0.15, respectively, for DRT, CME, SRD, and VMIA patterns. Central macular thickness decreased by 70.5±105.5, 110.67±97.28, 181±125.87, and 24.25±77.12 μm for the DRT, CME, SRD, and VMIA patterns, respectively. The CME group was associated with a greater reduction in retinal thickness (P=0.009) and volume (P=0.027) with superior VA improvement (P=0.012) as compared with the DRT, SRD, and VMIA groups.

Conclusions:

Patients with CME gained greater improvement in visual acuity and macular thickness and volume after IVB injection had been administered as the primary treatment for DME, as compared with other patients. The OCT patterns of DME may indicate the appropriate treatment; we consider these patterns to be prognostic of the response to IVB injection for macular edema.

Introduction

Diabetic macular edema (DME) can cause retinal structural changes and is one of the most common causes of visual impairment in patients with diabetes.¹ The gold standard of treatment for DME remains focal (direct/grid) laser photocoagulation, as was established by the Early Treatment of Diabetic Retinopathy Study (ETDRS). Focal laser photocoagulation can reduce the risk of moderate visual loss by 50%; however, 12% of treated eyes still lose vision due to persistent DME.^2,3 Other treatments for DME include intravitreal triamcinolone, vitrectomy, and intravitreal bevacizumab (IVB; Avastin^®; Genentech Inc., South San Francisco, CA). Although bevacizumab has shown promising results for the treatment of clinically significant macular edema (CSME),^4–6 the effects of IVB are transient, as its biologic life span in the vitreous is 4–5 weeks.^7,8

Traditional methods of ophthalmic examination, such as slit-lamp biomicroscopy, stereoscopic fundus photography, and fluorescein angiography, are limited in assessing retinal thickness. Recently, several novel ophthalmic imaging techniques have been developed, with optical coherence tomography (OCT) appearing to be the most promising among these. OCT can provide information regarding quantitative measurements for diabetic macular thickening and internal retinal structure.^9–12 Recently, DME has been categorized on the basis of OCT findings into 4 patterns: diffuse retinal thickening (DRT), cystoid macular edema (CME), serous retinal detachment (SRD), and vitreomacular interface abnormalities (VMIAs).^10,13–15 Further, Alkuraya et al. found a significant correlation between the OCT patterns of clinically significant DME and central macular thickness (CMT), severity of retinopathy, and visual acuity.¹⁴

In a previous study, patients with DRT achieved a greater reduction in retinal thickening and greater increases in visual acuity (VA) after focal laser photocoagulation than patients with other patterns.¹⁶ However, patients with CME showed a poor response to modified grid laser photocoagulation.^3,17 Subsequently, Roh et al. reported that patients with CME achieved better VA and greater changes in macular thickness than patients with DRT after the IVB injection.¹⁸

In this retrospective study, we examined the OCT findings of patients with DME to determine whether specific OCT patterns are predictive of visual outcome after the IVB injection.

Methods

This study retrospectively reviewed the medical records of patients examined between June 2007 and January 2010. Our study was approved by the Chang Gung Memorial Hospital Institutional Review Board, and all research procedures followed the tenets of the Declaration of Helsinki. Informed consent was obtained from participants, and the off-label use of the drug and its potential risks and benefits were extensively discussed with every patient.

The inclusion criteria were as follows: (1) CSME as defined by ETDRS²; (2) CMT of >250 μm as documented on OCT; (3) no previous IVB or intravitreal triamcinolone injections; (4) no previous macular focal laser therapy; and (5) age ≥18 years. The exclusion criteria were as follows: (1) ocular disease apart from diabetic retinopathy and cataract; (2) uncontrolled glaucoma; and (3) intraocular surgery, injection, or laser photocoagulation within 3 months of commencement of the study period.

At each visit, all patients underwent complete ophthalmic examination, including best-corrected Snellen visual acuity (BCVA), slit-lamp examination, intraocular pressure measurement, fundus photography, retinal thickness, and total macular volume (TMV) measurement by OCT. BCVA values were converted to the logarithm of the minimum angle of resolution (logMAR) scale for statistical manipulation. The main outcome measures were BCVA, CMT/TMV as measured by OCT, and retreatment rates.

The decision to perform reinjection was based on persistent or increased subretinal fluid on OCT with CMT >250 μm at the monthly control visits.

Every patient received 1 intravitreal injection of 1.25 mg of bevacizumab (0.05 mL) in the operating room. Under topical anesthesia using proparacaine drops for at least 30 s, the lids were scrubbed with povidone-iodine (10%), and a drop of 5% povidine-iodine was applied to the ocular surface. Under sterile conditions, a volume of 0.05 mL of bevacizumb was administered through the pars plana at 3.75 mm from the limbus with a 30-gauge needle. Antibiotic ointment was applied postoperatively for 3 days.

For each patient with DME, third-generation OCT (OCT 3 Stratus; Zeiss-Humphrey, San Leandro, CA) with software version 4.0 was performed. Patients were examined postoperatively at 1 week, and then monthly until a 3-month follow-up period was completed. The macula was scanned by using a standard, linear crosshair pattern, with a rapid macula scan option of 6 radial 6-mm-long scans for quantitative measures. CMT of 1 mm diameter and TMV of 3 mm diameter were measured by using OCT retinal mapping software as quantitative outcomes.

The eyes were classified into 4 groups based on the cross-sectional retinal morphologies. The DRT pattern was characterized by low intraretinal reflectivities and sponge-like swelling of the macula. The CME pattern was defined by intraretinal cystoid spaces of low reflectivity and highly reflective septa separating the cavities. The SRD pattern showed shallow detachment of the retina, with a clear space between the retinal pigment epithelium (RPE) and the retina. On the basis of the VMIA pattern, the OCT finding was characterized as epiretinal membrane (ERM) or vitreomacular traction (VMT). Both ERM and VMT presented as a highly reflective band or membrane on the retinal surface.

Group 1 included patients with only pure DRT. Group 2 included patients with CME without subretinal fluid or vitreous macular interface abnormalities. If DRT and CME coexisted, then CME was considered the dominant pattern. Group 3 patients comprised those with SRD without VMIA. When DRT, CME, and SRD were all present, then the eye was classified into group 3. Regardless of pattern combinations, eyes with ERM or vitreal macular traction were classified into group 4.

All analyses were computed by using SPSS software (v. 10.0, SPSS Inc., Chicago, IL). The changes in VA logMAR, IOP, CMT, and TMV were analyzed by using the paired t-test.

Results

Patient characteristics are summarized in Table 1. A total of 31 eyes of 24 individuals (10 women and 14 men) with nonproliferative (n=6; 19.4%) and proliferative diabetic retinopathy (n=25; 80.6%) were enrolled in the study. The mean duration of diabetes was 11.65±5.10 years. The mean age of the patients was 63.68±7.75 years (range: 51–81 years). Further, 26 eyes (83.9%) had previously undergone panretinal photocoagulation. No differences were noted among the 4 groups classified on the basis of OCT patterns with regard to the baseline characteristics of age (P=0.630), gender (P=0.813), duration of diabetes (P=0.504), stage (P=0.777), panretinal photocoagulation (P=0.502), and hypertension (P=0.921). Patients were required to complete 3 months of follow-up; patients who did not adhere to the 3-month follow-up period were not included in the analysis. IVB injection was repeated once for 3 eyes (9.7%) and repeated twice for 3 other eyes (9.7%).

Table 1.

Baseline Characteristics of the 31 Eyes with Diabetic Macular Edema

	DRT	CME	SRD	VMIA	All	P value
No. of eyes	10	10	3	8	31
Age (years)	64.60 (7.01)	61.20 (6.94)	63.00 (12.29)	65.88 (8.56)	63.68 (7.75)	0.630
Sex (Male/Female)	5/5	6/4	1/2	5/3	17/14	0.813
Durations of diabetes (years)	9.60 (3.41)	12.90 (6.39)	12.00 (6.93)	12.50 (4.63)	11.65 (5.10)	0.504
Stage (nonproliferative diabetic retinopathy/proliferative diabetic retinopathy)	1/9	2/8	1/2	2/6	6/25	0.777
Panretinal photocoagulation (no/yes)	4/6	3/7	1/2	2/6	10/21	0.921
Hypertension (no/yes)	9/1	7/3	3/0	7/1	26/5	0.502

Values are given as mean (SD) unless otherwise noted.

DRT was noted in 10 eyes (32.3%), CME in 10 (32.3%), SRD in 3 (9.7%), and VMIAs in 8 eyes (25.8%). Pure DRT was observed in 10 eyes. The remaining 21 eyes had mixed patterns; in particular, DRT along with CME was observed in 3 eyes (9.7%), DRT with SRD was noted in 2 eyes (6.5%), and DRT with ERM was observed in 3 eyes (9.7%). The common combinations of morphological subtypes of DME noted in our study are summarized in Table 2.

Table 2.

Common Combinations of Morphological Subtypes

OCT patterns	No. (%)
DRT
DRT alone	10 (32.3%)
CME
CME alone	70 (22.6%)
DRT/CME	30 (9.7%)
SRD
SRD alone	10 (3.2%)
DRT/SRD	20 (6.5%)
VMIA
ERM alone	50 (16.1%)
DRT/ERM	30 (9.7%)
All	31 (100%)

CME, cystoids macular edema; DRT, diffuse retinal thickening; ERM, epiretinal membrane; OCT, optical coherence tomography; SRD, serous retinal detachment; VMIA, vitreomacular interface abnormalities.

The effects of IVB injection therapy on retinal thickening, based on OCT patterns, are summarized in Table 3. Changes in CMT and TMV after IVB injection were evaluated. In all patients, the baseline CMT was 405±101.40 μm, and the baseline TMV was 3.69±0.72 mm³. The baseline macular thickness and volume were not significantly different between the 4 OCT patterns (P=0.459, P=0.266). The mean CMT and TMV after treatment were 322.22±81.16 μm and 3.15±0.48 mm³, respectively, and were significantly decreased at 3 months after IVB injection (P<0.001 each). The reductions in CMT and TMV values were 87.78±101.80 μm and 0.61±0.62 mm³, respectively. The baseline BCVA (logMAR) was 0.81±0.50 in all patients, and a significant decrease to 0.72±0.47 (P=0.033) was noted at the 3-month follow-up.

Table 3.

Change in Central Macular Thickness and Total Macular Volume from Baseline

	DRT Mean±SD	CMO Mean±SD	SRD Mean±SD	VMIA Mean±SD	All Mean±SD
Baseline CMT (μm)	362.9±126.63	420.8±96.99	446.3±143.58	422.4±38.66	405±101.40
3rd-month CMT (μm)	303.5±111.12	309.6±60.84	331.5±3.54	383.5±55.91	322.22±31.16
Change in CMT from baseline	−70.5±105.53	−110.7±97.28	−181±125.87	−24.25±77.12	−87.78±101.80
P value^a (3rd-month from baseline)	0.101	<0.010	0.291	0.574	<0.010
Baseline TMV (mm³)	3.38±0.089	3.77±0.60	4.27±1.02	3.76±0.35	3.69±0.72
3-month TMV (mm³)	2.93±0.55	3.24±0.37	3.60±0.72	3.22±0.45	3.16±0.49
Change in TMV from baseline	−0.56±0.67	−0.58±0.65	−1.23±0.25	−0.47±0.60	−0.61±0.62
P value^b (3-month from baseline)	0.049	0.027	0.090	0.215	<0.010

P value for difference between Baseline CMT and 3rd-month CMT.

P value for difference between Baseline TMV and 3rd-month TMV.

CMT, central macular thickness; TMV, total macular volume.

At 3 months after IVB injection therapy, the eyes with DRT, CME, SRD, and VMIA patterns had lower CMT measurements than the baseline values (−70.5, −110.67, −181, and −24.25, respectively). The reduction in retinal thickness was only significant in the CME group (P=0.009), and was not significant in the DRT, SRD, or VMIA group (P=0.101, P=0.291, P=0.574). Moreover, this was not significantly different between groups.

At 3 months after IVB injection therapy, the TMV values for DRT, CME, SRD, and VMIA patterns were also lower than the baseline values (−0.56, −0.58, −1.23, and −0.47, respectively). The reduction in retinal volume was more evident in the CME and DRT groups than in the SRD and VMIA groups (P=0.027, P=0.049); however, the differences were not statistically significant.

BCVA readings before and after IVB injection are shown in Table 4. At 3 months after IVB injection therapy, the mean BCVA was 0.72±0.47 logMAR, which was significantly different from the baseline value (P=0.033). The CME group showed better BCVA (0.65±0.49 logMAR) than the DRT, SRD, and VMIA groups (0.68±0.46, 1.2±0.71, and 0.78±0.38 logMAR, respectively). Eyes in the CME group showed significant improvement in BCVA at 3 months as compared with the baseline values (−0.26±0.26 logMAR). However, in the DRT and VMIA groups, the improvements were not significant; in fact, in the SRD group, the BCVA at 3 months had worsened. There were no statistically significant differences in the changes in mean BCVA between the 4 groups.

Table 4.

Change in Visual Acuity from Baseline

	DRT Mean±SD	CME Mean±SD	SRD Mean±SD	VMIA Mean±SD	All Mean±SD
Baseline VA (logMAR)	0.70±0.49	0.91±0.64	0.91±0.53	0.79±0.33	0.81±0.50
3-month VA (logMAR)	0.68±0.46	0.65±0.49	1.20±0.71	0.78±0.38	0.72±0.47
P value^a (3 months from baseline)	0.613	0.012	0.500	0.391	0.033
Change in logMAR from baseline	−0.06±0.36	−0.26±0.26	0.09±0.13	−0.08±0.15	−0.13±0.29

P value for difference between baseline TMV and 3-month TMV.

logMAR, logarithm of the minimum angle of resolution; VA, visual acuity.

In the CME group, a significant change was noted in logMAR (a decrease of 0.263±0.263, P=0.012), CMT (a decrease of 110.67±97.28 μm, P=0.009), and TMV (a decrease of 0.582±0.649 mm³, P=0.027) values from the baseline to 3-month follow-up. In DRT eyes, a less significant change in TMV (a decrease of 0.56±0.67 mm³, P=0.049) was noted; however, the changes in the logMAR (a decrease of 0.06±0.36, P=0.613) and CMT (a decrease of 70.5±105.53 μm, P=0.101) values were not significant at the 3-month follow-up examination.

No systemic complications, such as hypertensive episodes or thromboembolic events, occurred during the study period. At the final follow-up, no elevated intraocular pressure, endophthalmitis, or rhegmatogenous or tractional retinal detachment was reported.

Discussion

The pathogenesis of DME includes breakdown of the blood-retinal barrier, tissue hypoxia, circulatory changes in the retina and choroid, vitreoretinal traction, and increased levels of inflammatory factors. An animal study proved VEGF contributes to diabetic blood-retinal barrier breakdown, which is known to lead to macular edema.¹⁹ Macular focal photocoagulation remains the gold standard treatment in patients with CSME.^2,3,20 The recent findings of a 2-armed prospective, randomized, controlled BOLT study also indicated that bevacizumab had a greater long-term effect in patients with CSME than did macular laser therapy, irrespective of whether the patients had undergone laser therapy earlier.⁴ If the OCT patterns can predict the effects of a treatment, then it will be possible to choose the most effective treatment for each patient. Early intervention with appropriate therapy may prevent severe visual loss.

Campbell and associates found that OCT-based retinal volume and central foveal thickness display equal ability in detecting CSME,²¹ a previous study also adopted these 2 parameters to assess the treatment effects on DME.¹⁶ Since at present there is no technique to evaluate whether a patient had accurate fixation during OCT, we consider that measurements of both CMT and TMV can alleviate the effect of fixation loss. Previous studies have revealed that visual acuity is significantly correlated with foveal thickness and OCT type.^10,22

DRT is characterized by a thickened and sponge-like retina. Low reflective areas are located in the outer retinal layers, and the inner retinal layers are anteriorly displaced by the swollen outer retinal layers.¹ This characterization may suggest the primary location of edema. SRD is defined as detachment of the neurosensory retina, with a clear space between the RPE and the retina. The VMIA pattern presents traction bands and a thickened posterior hyaloid incompletely detached from the retinal surface or the optic nerve, and is characterized by a highly reflective membrane at the vitreoretinal surface in OCT. Karatas et al. found a high correlation between diffuse DME refractory to laser treatment and vitreopapillary traction.²³ The pathogenesis in these 3 subgroups seems to be less correlated to angiogenesis; this may partly explain the worse response to IVB in our study.

Histopathological examination of patients with CME has revealed widespread intracytoplasmic swelling and necrosis of Müller cells, with secondary neuronal degeneration.^24,25 This may explain why eyes with CME in our study had more profound visual loss than those with DRT at the baseline. The improvements in VA in CME eyes may be due to the presence of decreased intracytoplasmic fluid in Müller cells, which indicates that eyes with CME should be treated before the occurrence of Müller cell necrosis and secondary neuronal degeneration. Kang et al. found that eyes with CME and SRT had high incidence of diffuse retinal vascular hyperpermeability on fluorescein angiography, which may explain the trend that eyes with these 2 types were more refractory to macular photocoagulation.¹⁰ Recent evidence also demonstrates that the number of microaneurysms in the perifoveal capillary network is more significant in eyes with diabetic CME than in eyes with DRT or SRD.²⁶ It may explain why eyes with CME respond better to anti-VEGF therapy than eyes with DRT or SRD. Roh et al. also treated patients with refractory DME with IVB and found greater improvements in VA and macular thickness in patients with CME.¹⁸

Since it was difficult to determine the duration of the treatment effect of IVTA and macular grid laser, we decided to further evaluate the effects of IVB as the primary therapy for DME exhibiting different patterns and exclude cases treated with the macular grid laser.

The IVB reinjection may lead to improvements in VA and CMT in patients with DME,²⁷ but the indications and intervals for retreatment remain controversial. Roh et al. considered CMT >250 μm or deterioration of VA of at least 5 ETDRS letters score as compared with the previous value as the criteria for reinjection, with an interval of at least 12 weeks. Kook et al. performed a reinjection when the CMT changed by >100 μm with an associated VA deterioration of >5 ETDRS letters.²⁸ The criteria and interval for retreatment remain a matter of ongoing debate.

Our study has certain limitations. It is a retrospective, short-term, nonrandomized, and uncontrolled study. The number of patients in each subgroup was too small, in comparison to that in previous studies, to exclude the treatment effects of other therapies such as macular focal/grid photocoagulation or intravitreal triamcinolone. Although the prevalence of SRD in eyes with DME was relatively low (9.9%–26%) in previous studies,^{1,10,15,29,30} the SRD subgroup is especially small in our study. A previous study had noted that the distribution of OCT types had no significant correlation with the stages of diabetic retinopathy,¹⁰ we did not describe the correlation between them. The duration of DME was also difficult to accurately assess, because most patients presented with DME at the first visit. Due to the shortcomings of the current study, a larger prospective, randomized, controlled clinical trial is required. We included the fellow eye in the study or for simultaneous IVB injection, because much evidence has indicated the low concentration of bevacizumab in serum and the fellow eye after IVB injection.^7,31

Our findings indicate that the IVB injection could be a primary therapeutic modality for CME. OCT patterns in DME may predict the outcome and help choose the best treatment. Treatment with bevacizumab at an earlier stage of DME without retinal ischemia may have better visual outcomes.

Footnotes

Author Disclosure Statement

The authors have no proprietary interest and no financial support (grants) in any aspect of this study.

References

Otani

, Kishi

, Maruyama

Patterns of diabetic macular edema with optical coherence tomography. Am. J. Ophthalmol., 127:688–693. 1999.

Group ETDRSR. Photocoagulation for diabetic macular edema: early treatment diabetic retinopathy study report number 1. Arch. Ophthalmol., 103:1796–1806. 1985.

Group ETDRSR. Treatment techniques and clinical guidelines for photocoagulation of diabetic macular edema: early treatment diabetic retinopathy study report number 2. Ophthalmology, 94:761–774. 1987.

Michaelides

, Kaines

, Hamilton

R.D.

et al. A prospective randomized trial of intravitreal bevacizumab or laser therapy in the management of diabetic macular edema (BOLT study): 12-month data: Report 2. Ophthalmology, 117:1078–1086.e2. 2010.

Haritoglou

, Kook

, Neubauer

et al. Intravitreal bevacizumab (Avastin) therapy for persistent diffuse diabetic macular edema. Retina, 26:999–1005. 2006.

Arevalo

J.F.

, Fromow-Guerra

, Quiroz-Mercado

et al. Primary intravitreal bevacizumab (Avastin) for diabetic macular edema: results from the Pan-American collaborative retina study group at 6-month follow-up. Ophthalmology, 114:743–750. 2007.

Bakri

S.J.

, Snyder

M.R.

, Reid

J.M.

, Pulido

J.S.

, Singh

R.J.

Pharmacokinetics of intravitreal bevacizumab (Avastin)

Ophthalmology, 114:855–859. 2007.

Stewart

M.W.

Predicted biologic activity of intravitreal bevacizumab. Retina, 27:1196–1200. 2007.

Goebel

, Kretzchmar-Gross

Retinal thickness in diabetic retinopathy: a study using optical coherence tomography (OCT)

Retina, 22:759–767. 2002.

10.

Kang

S.W.

, Park

C.Y.

, Ham

D.I.

The correlation between fluorescein angiographic and optical coherence tomographic features in clinically significant diabetic macular edema. Am. J. Ophthalmol., 137:313–322. 2004.

11.

Ozdek

S.C.

, Erdinc

M.A.

, Gurelik

et al. Optical coherence tomographic assessment of diabetic macular edema: comparison with fluorescein angiographic and clinical findings. Ophthalmologica, 219:86–92. 2005.

12.

Ozdemir

, Karacorlu

Serous macular detachment in diabetic cystoid macular oedema. Acta Ophthalmol. Scand., 83:63–66. 2005.

13.

Panozzo

, Parolini

, Gusson

et al. Diabetic macular edema: an OCT-based classification. Semin. Ophthalmol., 19:13–20. 2004.

14.

Alkuraya

, Kangave

, Abu El-Asrar

A.M.

The correlation between optical coherence tomographic features and severity of retinopathy, macular thickness, and visual acuity in diabetic macular edema. Int. Ophthalmol., 26:93–99. 2005.

15.

Kim

B.Y.

, Smith

S.D.

, Kaiser

P.K.

Optical coherence tomographic patterns of diabetic macular edema. Am. J. Ophthalmol., 142:405–412. 2006.

16.

Kim

N.R.

, Kim

Y.J.

, Chin

H.S.

, Moon

Y.S.

Optical coherence tomographic patterns in diabetic macular oedema: prediction of visual outcome after focal laser photocoagulation. Br. J. Ophthalmol., 93:901–905. 2009.

17.

Lee

C.M.

, Olk

R.J.

Modified grid laser photocoagulation for diffuse diabetic macular edema: long-term visual results. Ophthalmology, 98:1594–1602. 1991.

18.

Roh

M.I.

, Kim

J.H.

, Kwon

O.W.

Features of optical coherence tomography are predictive of visual outcomes after intravitreal bevacizumab injection for diabetic macular edema. Ophthalmologica, 224:374–380. 2010.

19.

Qaum

, Xu

, Joussen

A.M.

et al. VEGF-initiated blood-retinal barrier breakdown in early diabetes. Invest. Ophthalmol. Vis. Sci., 42:2408–2413. 2001.

20.

Network DRCR. A randomized trial comparing intravitreal triamcinolone acetonide and focal/grid photocoagulation for diabetic macular edema. Ophthalmology, 115:1447–1459.e10. 2008.

21.

Campbell

R.J.

, Coupland

S.G.

, Buhrmann

R.R.

, Kertes

P.J.

Optimal optical coherence tomography-based measures in the diagnosis of clinically significant macular edema: retinal volume vs foveal thickness. Arch. Ophthalmol., 125:619–623. 2007.

22.

Yamamoto

, Yamamoto

, Hayashi

, Takeuchi

Morphological and functional analyses of diabetic macular edema by optical coherence tomography and multifocal electroretinograms. Graefes Arch. Clin. Exp. Ophthalmol., 239:96–101. 2001.

23.

Karatas

, Ramirez

J.A.

, Ophir

Diabetic vitreopapillary traction and macular oedema. Eye (Lond), 19:676–682. 2005.

24.

Fine

B.S.

, Brucker

A.J.

Macular edema and cystoid macular edema. Am. J. Ophthalmol., 92:466–481. 1981.

25.

Yanoff

, Fine

B.S.

, Brucker

A.J.

, Eagle

R.C.

Jr.

Pathology of human cystoid macular edema. Surv. Ophthalmol., 28,Suppl:505–511. 1984.

26.

Murakami

, Nishijima

, Sakamoto

et al. Foveal cystoid spaces are associated with enlarged foveal avascular zone and microaneurysms in diabetic macular edema. Ophthalmology, 118:359–367. 2011.

27.

Roh

M.I.

, Byeon

S.H.

, Kwon

O.W.

Repeated intravitreal injection of bevacizumab for clinically significant diabetic macular edema. Retina, 28:1314–1318. 2008.

28.

Kook

, Wolf

, Kreutzer

et al. Long-term effect of intravitreal bevacizumab (avastin) in patients with chronic diffuse diabetic macular edema. Retina, 28:1053–1060. 2008.

29.

Catier

, Tadayoni

, Paques

et al. Characterization of macular edema from various etiologies by optical coherence tomography. Am. J. Ophthalmol., 140:200–206. 2005.

30.

Koleva-Georgieva

, Sivkova

Assessment of serous macular detachment in eyes with diabetic macular edema by use of spectral-domain optical coherence tomography. Graefes Arch. Clin. Exp. Ophthalmol., 247:1461–1469. 2009.

31.

Miyake

, Sawada

, Kakinoki

et al. Pharmacokinetics of bevacizumab and its effect on vascular endothelial growth factor after intravitreal injection of bevacizumab in macaque eyes. Invest. Ophthalmol. Vis. Sci., 51:1606–1608. 2010.