ISSN: 2468-6530


First Published: 2016

Journal: 2468-7219

Volume: 3

Issues: 12

Ophthalmology Retina, a journal of the American Academiy of Ophthalmology, serves society by publishing clinical and basic science research and other relevant manuscripts that relate to the sense of sight. Excellence is pursued throgh unbiased peer-review, the advancement of innovation and discoery, and the promotion of lifelong learning.

Multiple Effects Intravitreal Aflibercept on Microvascular regression in Eyes with Diabetic Macular Edema



To evaluate the effects of intravitreal aflibercept (IVA) on the microaneurysms (MAs) and sizes of non-perfused areas (NPAs) in eyes with diabetic macular edema (DME).


Interventional, prospective study.


Twenty-five eyes of 25 DME patients (average age, 64.0±8.8 years) were treated with 3 consecutive monthly IVA injections.


Fluorescein angiography (FA) and optical coherence tomography were performed prior to the IVA injections(baseline) and at 1 week after the IVA treatment. The number of MAs and the ischemic index (ISI), a measure of the NPAs, were determined. The correlations between central retinal thickness (CRT) and number of MAs and the ISI were also determined.

Main Outcome Measures

The mean number of MAs and NPA evaluated as the ISI.


At the baseline, the mean central retinal thickness (CRT) was 485.7±90.6 μm. After treatment, the mean CRT was significantly reduced to 376.9±81.6 μm (P=0.1 x10-5; repeated ANOVA). The mean number of MAs was significantly decreased from 49.6±33.2 at the baseline to 24.8±18.1 at 3 months after the initial treatment. This was a 50.4±21.2% reduction (P=0.3 x10-5, paired t tests). The mean ISI was also significantly decreased from 55.5±20.4% at the baseline to 28.8±16.8% after the treatment (P=0.3 x10-5, paired t test). This was a reduction of 43.3±28.5%. A significant correlation was found between the CRT and number of MAs at both the baseline (r=0.56, P=0.004) and after the treatment (r=0.53, P=0.006). A significant correlation was found between the CRT and the ISI at the baseline (r=-0.39, P=0.03) but not after the treatment (r=-0.06, P=0.79).


The reduction in the number of MAs was correlated with the reduction in the CRT.

Key Words:

diabetic macular edema, diabetic retinopathy, ischemic index, microaneurysms, vascular endothelial growth factor


Diabetic retinopathy (DR) is a leading cause of blindness, and it is present in over one-half of patients with diabetes mellitus of over 20years duration.1 Various vascular changes occur during its progression especially in the microvasculature including the formation of microaneurysms (MAs) which are the most characteristic feature of DR.

Diabetic macular edema (DME) is a significant complication in eyes with DR. The excessive leakage of plasma in eyes with DR results in a thickening of the retina which can then damage the photoreceptors. Such leakage is commonly evaluated by fluorescein angiography (FA), and they are detected as points of hyperfluorescence or focal leakage. Their importance is not only their association with the severity of the DR but also as a cause of the DME. Histopathologic studies have demonstrated that MAs are small outpouchings of the capillaries with focal endothelial cell proliferation and loss of pericytes due to hyperglycemia which then weakens the walls of the retinal vessels. The MAs are caused by a loss of the barrier properties of the vascular walls leading to abnormal leakage.

Many surgical and medical therapies have been reported to be effective treatments for DME. Several studies have shown that vascular endothelial growth factor (VEGF) plays a major role in the vascular proliferation and hyperpermeability in eyes with DME. The results of several multicenter trials have also shown that intravitreal injections of anti-VEGF agents led to a rapid resolution of the DME and an improvement of vision.

At present, various anti-VEGF agents, e.g., bevacizumab (AvastinR: R, Genentech, South San Francisco, CA, USA), ranibizumab (LucentisR: R, Genentech), and aflibercept (EyleaR: R, Regeneron Pharmaceuticals, Tarrytown, NY, USA) are being used to treat DME, and they have become the first-line therapy for DME. However, 0.8 to 3.9% of patients in the RIDE and RISE study and 0.7 to 3.2% in the VIVID and VISTA study lost ≥15 letters from the baseline letters of the visual acuity chart. In addition, there is still the problem that anti-VEGF agents are not effective for all cases, and there exists other factors that appear to block the improvements.

Leakage from the MAs plays important roles in the DME, and the ETDRS recommended focal/grid laser photocoagulation for such leakage. However, an important complication of photocoagulation is its destruction of normal retinal tissue. Recently, anti-VEGF agents have replaced photocoagulation as a first line treatment for DME. The results of earlier studies showed that the anti-VEGF agents acted by reducing the abnormal leakage including that from the MAs. In addition, the DRCR.net Protocol V study reported that there was no significant difference in the vision loss whether eyes were initially managed with aflibercept or with laser photocoagulation in eyes with center-involved DME and good visual acuity. They stated that the patients should be given aflibercept only if the visual acuity worsened. It will be necessary to examine whether anti-VEGF therapy affects only some of the MAs or all of the MAs. The results of recent studies have shown that anti-VEGF agents can improve the stage of the DR, which is a new aspect of anti-VEGF treatments.

The aim of this study was to evaluate the effect of intravitreal aflibercept (IVA) injections on the number of MAs and the size of the non-perfused areas (NPAs) in eyes with DME.


This prospective study was registered at http://www.umin.ac.jp (No. UMIN 000018315). Its protocol was approved by the Institutional Review Board of Mie University Hospital (No. 2913), and the procedures used conformed to the tenets of the Declaration of Helsinki. Oral and written consents were obtained from the patients after an explanation of the procedures to be used and possible complications.

Patient Demographics 

The age, sex, level of creatinine, estimated glomerular filtration rate (eGFR), hemoglobin, hemoglobin A1c (HbA1c), systolic/diastolic blood pressure (BP), severity of the DR, and previous PRP were determined at the baseline and after the third injection. All patients had a complete ophthalmologic examination before beginning the loading phase of the IVA protocol. Monthly examinations started from 1-2 weeks before beginning of the loading phase and continued through the post-injection period. The examinations included measurements of the best-corrected visual acuity (BCVA) with Snellen charts, slit-lamp biomicroscopy, intraocular pressure measurements, indirect ophthalmoscopy, spectral domain optical coherence tomography (SD-OCT), and FA. FA was performed at the baseline and at within 1 week after the third IVA injection.

OCT examinations were performed with the Spectralis OCT instrument (Heidelberg Engineering, Heidelberg, Germany). The structural OCT minimum acquisition protocol which included 19 horizontal raster linear B-scans with each composed of nine averaged OCT B-scans (1024 A-scans per line) covering an area of 30degrees x 30degrees was used. The CRT in the central 1-mm-diameter circle of the Early Treatment Diabetic Retinopathy Study (ETDRS) thickness map was calculated by the Spectralis Software (Heidelberg Eye Explorer, version, Heidelberg Engineering, Germany).

Inclusion and Exclusion criteria

The inclusion criteria were; patients ≥20-years-of-age with type I or II diabetes, eyes with DR and DME with signs of severe leakage indicating abnormal vessels within the vascular arcade area by FA, BCVA ≥20/320, DME involving the fovea, and a CRT ≥300 μm measured as the mean retinal thickness in the central 1 mm diameter circle by OCT.

The exclusion criteria were; any retinal photocoagulation treatment in the study eye within 3 months preceding the initial IVA, eyes with an ischemic macular region involving the fovea, prior focal/grid photocoagulation, eyes with vitreomacular traction, any history or presence of other ocular diseases causing vision reduction such as age-related macular degeneration and severe proliferative DR, optic nerve atrophy, glaucoma or intraocular pressure (IOP) >24 mmHg, history of vitreous surgery, aphakia, anti-VEGF treatment on either eye within 3 months preceding the initial IVA, cloudy optic media including cataract through which high quality fundus photographs or OCT images could not be obtained, history of cataract surgery in the study eye within the previous 3 months, history of cerebrovascular accident, myocardial infarction or other systemic disease requiring medications that could affect the results, severe renal failure with creatinine ≥2.0 mg/dl or >Stage IIb of nephropathy defined by classification of diabetic nephropathy, poorly controlled hypertension with systolic BP >200 mmHg or diastolic BP >110 mmHg, poorly controlled diabetes mellitus with HbA1c ≥12.0%, and patients who were judged as ineligible for any other reason by the investigators.

Intravitreal Aflibercept Injections

Each eye received 2 mg of IVA (40 mg/ml) under topical anesthesia. The injections were made with a 30-gauge needle that was inserted 4 mm posterior to the corneal limbus under sterile conditions. All patients received topical levofloxacin hydrate (1.5% CravitR:R ophthalmic solution) for 3 days after the injection. Each eye was treated with 3 consecutive monthly IVA injections (the loading phase) as described.

Fluorescein Angiography (FA)

FA was performed with the Heidelberg Spectralis HRA+OCT module with the standard intravenous infusion of 5 ml of 10% sodium fluorescein. Images were obtained using the standard 35-degree field of view lens, and ultra-widefield FA (UWFA) images were also obtained with the noncontact attachment module which enabled the recording of a 150-degree field centered on the fovea.

Image Analyses

One or more images from the early arteriovenous phase was selected from each angiographic series. The images were recorded between 45 to 60 sec and also from the late phase between 5 to 6 minutes. The images were transferred to the Adobe Photoshop ES software (Adobe Systems Inc, San Jose, CA, USA) or ImageJ software (Image processing and analysis in JAVA1.46r, Wayne Rasband, National Institute of Health, USA).


Hyper-fluorescence pools due to leakage was considered to be MAs as reported.2 All FA images were evaluated by 3 masked graders (MS, AI, and DM). One evaluator counted each image 3 times, and the average value was used for the statistical analyses.

Number of MAs

We used the color merge technique to determine the number of MAs. To do this, an early phase baseline FA image (Fig. 1a) was converted to a green image (Fig. 1c) and a post-treatment image (Fig. 1b) was converted to a red image (Fig. 1d) using PhotoshopR:R. These two images were overlapped, and the number of high intensity yellow dot were counted as the leakage points in the merged image (Fig. 1e). Thus, the post-treatment MAs were designated as RG-MAs.

Figure 1

Fluorescein angiographic image used to count the number of microaneurysms (MAs). Images were selected from the early arteriovenous phase from each angiographic series at the baseline (a) and at a post-treatment time (b). The baseline image was converted to green (c) and the post-treatment image was converted to red (d). The images were merged, and the high intensity yellow dot signals were defined and counted as MAs that still had leakages (e). These MAs were counted as RG-MAs.

We also counted the number of MAs in the central area which was associated with the origin of the DME. The 6.0 mm diameter (outer ring of the ETDRS circle centered on the fovea) of the OCT thickness map was delineated by Photoshop (Fig. 2a). These OCT images and the FA images (Fig. 2b) obtained at the same time were merged (Fig. 2c). The hyper-fluorescent dot signals in the area with retinal thickness >400 μm (color coded as red or warmer) were identified and counted as leakage points related to the DME. We compared the baseline images (Fig. 2c) and the post-treatment images (Fig. 2d). These MAs were counted as map-MAs.

Figure 2

OCT image used to count abnormal leakages related to the origin of the DME. The OCT thickness color coded map images (a) were merged with the FA images obtained at the same time (b). Merged mages were created at the baseline (c) or at a post-treatment tine (d). The MAs were counted as Map-MAs. After the treatment, the mean number of map-MAs was decreased to 24.8±18.1 (e, paired t-tests, P <0.01). A significant and moderate correlation was found between the number of RG-MAs and d%-map-MAs (f; r = 0.51, P <0.01; Spearman test). DME: diabetic macular edema, FA: fluorescein angiography, MAs: microaneurysms, OCT: optical coherence tomography.


Ischemic Index (ISI) 

The degree of peripheral retinal ischemia of all eyes was quantified by the ischemic index (ISI) which was calculated as described in detail.25 The ISI has been used in other studies as an indicator of retinal nonperfusion. The degree of capillary nonperfusion was determined by masked graders in the macula-centered UWFA frames as described in detail.26 Briefly, the images were exported in a jpeg format without changes in the contrast, gamma, or brightness and imported to the Image software. The total retinal surface and capillary nonperfused areas seen in the late phase were manually outlined using the area measurement function and divided by the area of the total image area in pixels. The capillary NPA was defined as the area where a dropout of the retinal capillary bed was detected in the images (Figure 3, asterisk). The ISI was calculated as the ratio between ischemic retina and total retinal area. All results were evaluated by 3 masked graders (MS, AI, and DM). One evaluator counted each image 3 times, and the average value was used for the statistical analyses.

Figure 3

Ischemic index (ISI) changes during DME treatment. The degree of peripheral retinal ischemia was quantified by the ISI. The retinal surface and capillary nonperfused areas seen in the late phase of UWFA were manually outlined (baseline a; post-treatment b). The mean baseline ISI was 55.5 ± 20.4%, and after the treatment, the mean ISI was significantly decreased to 28.8±16.8 (c; P <0.01; paired t-test). ISI: Ischemic index, UWFA: ultra-wide field fluorescence angiography Asterisk: non-perfusion area.

Statistical Analyses

To evaluate the relative changes in the CRT, the number of MA and degree of ISI were determined. We used d-values that were expressed as a rate of change from the baseline values. The d-values or reduction rates were defined as follow:

d-values (%) = 100 – (value after IVA/value before IVA x 100).

The results are presented as the means ± standard deviations (SDs). The significance of the differences of the data was determined by two-way non-repeated ANOVA followed by Bonferroni post-hoc tests for the comparison of the means. Paired t-tests were used to determine the significance of differences between two groups. Spearman’s rank-order correlation coefficient was used to determine the significance of the correlations among the variables. The strength of the correlations (r-value) was classified as: 0.0 to 0.2 not correlated or very weak; 0.2 to 0.4 weak or low; 0.4 to 0.7 moderate; 0.7 to 0.9 strong or high; and 0.9 to 1.0 very strong. The results were considered statistically significant when P <0.05.


Thirty eyes of 30 patients met the inclusion and exclusion criteria and were studied between September 2015 to March 2017 in the Department of Ophthalmology, Mie University Hospital. There were no IVA-related ocular complications including intraocular pressure elevations, infections, or any adverse systemic events. Five patients among the 30 patients could not complete the entire injection protocol or the entire examination schedules due to financial reason (3 patients) or difficulty in transportation to the hospital.

Baseline characteristics

In the end, 25 eyes of 25 DME patients consisting of 14 men and 11 women were studied. Their mean age was 64.0 ± 8.8 years. Fifteen of the eyes had been treated by PRP. Twelve patients had moderate non-proliferative DR and 13 patients had severe non-proliferative DR.

The average creatinine was 0.8 ± 0.5 (range 0.4 to 2.7) mg/dl, the average eGFR was 74.0 ± 23.6 (range 20.3 to 106.7) ml/min/1.73 m2, average hemoglobin was 14.0 ± 2.2 (range 9.3 to 17.2) g/dl, and average HbA1c was 7.8 ± 1.3 (range 6.4 to 10.5) %. The average systolic BP was 145.8 ± 19.3 with a range 101 to 180 mmHg and the average diastolic BP was 72.4 ± 13.0 with a range 51 to 95 mmHg.

Ocular Examinations

The mean baseline BCVA was 0.45 ± 0.35 logMAR units and the mean baseline CRT was 485.7 ± 90.6 μm. After the loading phase, the mean BCVA was significantly improved to 0.40 ± 0.38 logMAR units (P = 0.02; repeated ANOVA), and the mean CRT was significantly reduced to 376.9 ±81.6 μm (P = 0.1 x 10-6).

Number of MAs

The mean number of map-MAs at the baseline was 49.6 ± 33.2, and the mean number after the loading phase was reduced to 24.8 ± 18.1 (P = 0.3 x 10-6, paired t-tests; Figure 2e). The reduction rate of the d%-map-MAs was 50.4 ± 21.2%. A moderate and significant correlation was found between the number of RG-MAs and the d%-map-MAs (r = 0.51, P = 0.009; Spearman test, Figure 2f).

ISI values

The mean ISI at the baseline was 55.5 ± 20.4%, and the mean was significantly reduced to 28.8 ± 16.8 (P = 0.3 x 10-6; paired t-tests, Figure 3c). The reduction percentage (dISI%) was 43.3 ± 28.5%.

Figure 4

Correlation between the baseline and post-treatment number of MAs and the ISI. Significant and moderate correlations were found between the CRT and MAs at both the baseline (a; r = 0.56, P <0.01; Spearman test) and post-treatment (b, r = 0.53, P <0.01). A weak but significant correlation is present between the CRT and the ISI at the baseline (c; r =-0.39, P <0.05,) but no significant correlation is present after the treatment (d, r =-0.06; P >0.05). CRT: central retina thickness, MAs: microaneurysms,

A moderate significant correlation was found between the CRT and number of MAs at both the baseline (r = 0.56, P = 0.004; Spearman test, Figure 4a) and after the IVA treatment (r = 0.53, P = 0.006; Spearman test, Figure 4b).

A significant correlation was found between the CRT and the ISI at the baseline (r = -0.39, P = 0.03; Spearman test, Figure 4c). But no significant correlation was found between them after IVA treatment (r = -0.06, P = 0.79; Spearman test, Figure 4d). No significant correlation was found between the reduction rate of the CRT (d% CRT) and the d% map-MA (r = 0.08, P = 0.53; Spearman’s rank correlation coefficient).


The results showed that the IVA injections reduced the number of MAs and the size of the NPA. These results suggest that the reduction of the ISI did not play a role in the reduction of the CRT.

The ETDRS showed that focal macular photocoagulation is effective in treating MAs.8 But focal photocoagulations do not necessarily work immediately, and there are complications due to the damage caused by the damaged tissues, e.g., creeping atrophy. In addition, only 3% of the patients have an improvement of ≥15 letters.27 At present, anti-VEGF therapy has largely replaced photocoagulation and has become the first line of therapy for eyes with DME.

There have been at least two reports of a reduction in the number of MAs after anti-VEGF therapy as we found. But they were not the results of more than 3-consecutive injections. Many other studies have shown that multiple injections of anti-VEGF agents can improve the severity of the DR. In addition, PANORAMA, a Phase 3, double-masked, randomized study was designed to test the efficacy and safety of intravitreal aflibercept injections in patients with moderately severe to severe non-proliferative diabetic retinopathy (NPDR) compared to sham injections (ClinicalTrials.gov Identifier: NCT02718326). The study reported that the proportion of patients with ≥2-step improvement of the severity of the diabetic retinopathy was significantly higher after aflibercept than the sham injected eyes (presented at Bascom Palmer Eye Institute’s Angiogenesis, Exudation, and Degeneration 2019 symposium in Miami, Feb 9, 2019, Miami). But all of these studies did not specifically quantify the changes in the number of MA during the treatment. Our results showed that 3 consecutive anti-VEGF injections reduced the number of MAs and the ISI. These changes occurred because of the biological aspects of the anti-VEGF agents. Thus, the alterations reduced the number of MAs by changes in the structural features, viz., a loss of pericyte and astrocytes, and they also altered the hemodynamic properties, viz., increased capillary intramural pressure and production of vasoproliferative factors such as VEGF.7

Our results showed that anti-VEGF therapy can reduce the number of map-MAs. Central macular thickening, quantified by OCT maps as red or gray areas, may be considered a variable DME outcome.28 Our results also showed that there were still nonresponding MAs even after 3 consecutive loading injections. Although it is not clear whether all of the MAs disappeared after the multiple injections, many earlier studies have reported an improvement in DR after multiple injections. The subgroup analysis of the RIDE and RISE and the VIVID and VISTA studies showed that the DR improved in the eyes with DME. The mean injection number was high compared to our study.

Our results showed that the residual number of MAs was significantly correlated with the CRT after the loading injections. Earlier studies have shown that consecutive injections will decrease the edema and improve the vision in eyes with DR progression, however only 3 consecutive injections were not sufficient and continued treatment during a maintenance phase may be needed to reduce the number of residual MAs.

When we consider the problems of anti-VEGF treatment, its costs, inconveniences for doctors and healthcare providers, and the poor results with inadvertent complications, the need for a proper treatment regimen leading to optimal visual outcomes with fewer injections is needed. A stabilization with combined laser photocoagulation or other agents could also lead to earlier visual gains and reduced anti-VEGF retreatment numbers. The NAVILAS system has demonstrated that there is a significantly higher accuracy in the laser spot applications, and it has a potential of reducing the retreatment rates. This indicated that it could be used for combination therapy. Therefore, the use of combination therapy for the non-responding MAs means that fewer anti-VEGF therapy might be effective with minimum stress.

Large NPAs are generally associated with the most resistant cases of DME. The hemodynamic changes due to the DR can lead to reduced production of VEGF which may lead to DME. A blockade of the excess VEGF by photocoagulation of the retina and RPE or anti-VEGF agents leads to substantial improvements in the edema,36 and targeted photocoagulation of the NPA can also reduce the edema. On the other hand, the DAVE-study, a 3-year randomized trial of combination therapy with ranibizumab and targeted photocoagulation, reported the lack of improvement of the visual outcomes.39 The disadvantages of PRP are the decreased vision, macular edema, and development of visual field defects due to damage to the retina and RPE. Our results showed that the correlation between the IVA treatments and ISI was not significant. Even though we found a reduction in the NPA, it was not significantly correlated with the reduction in the CRT. We suggest that the reduction in the NPA is a separate effect of the anti-VEGF agents.

There are some limitations in our study including the small sample size. Our study also lacked controls. Because anti-VEGF therapy has become the standard of care, it is difficult to recruit control DME group without any treatment for ethical reasons. The accuracy of the screening by human graders is not necessarily accurate.42 To resolve these difficulties, an automated system for the screening would be better. The MAs are dynamic lesions with a significant turnover of less than 6 months. These changes occur over time, disappearing by closing down due to thrombosis with new ones appearing at different locations of the vascular tree. This turnover indicates dynamic processes and changes in the disease activity. Leicht reported that the specific effects of anti-VEGF therapy on the MA turnover or resolution, and that is its value in DME patients.18 In addition, the disappearance of the MAs (50.4%) was higher than that reported by Allingham et al (17.9%).48 They reported that most of the hyper-fluorescence that disappeared in patients was due to an improvement of the diffuse edema and not to the focal loss of MAs. A control group consisting of untreated patients is needed to compare the turnover of the MAs to show the effects of aflibercept on the number of MA. We need to consider the effects of anti-VEGF on this turn over. Although we excluded patients that had received any focal or grid photocoagulation, PRP-treated eyes were included. From a post hoc analysis of the RIDE and RISE study,22 patients with prior PRP achieved limited DR improvement compared to no PRP patients. The patients received prior PRP as each doctor’s clinical discretion before entry. Because there was no specific guidance for prior PRP, this may have affected the results. Finally, although FA may be more sensitive than ophthalmoscopy in identifying areas of retinal NPA, small areas of neovascularization and vascular leakage, it has a risk of anaphylaxis. Technical improvements in the evaluations of the MAs, such as OCT angiography (OCTA), are needed. OCTA is a newly developed instrument which enables the clinician to evaluate the retinal vasculature without invasion. From enface images of different retinal or choroidal layers, OCTA has enabled clinicians to detect conformational changes in more detail.49 However, we did not evaluate our patients by OCTA.


Cheung, N., Mitchell, P., and Wong, T.Y. Diabetic retinopathy. Lancet. 2010; 376: 124–136

View in Article | Google Scholar

Klein, R., Meuer, S.M., Moss, S.E., and Klein, B.E. Retinal microaneurysm counts and 10-year progression of diabetic retinopathy. Arch Ophthalmol. 1995; 113: 1386–1391

View in Article | Google Scholar

Kohner, E.M., Stratton, I.M., Aldington, S.J., and Turner, R.C. Matthews DR Microaneurysms in the development of diabetic retinopathy (UKPDS 42). UK Prospective Diabetes Study Group. Diabetologia. 1999; 42: 1107–1112

View in Article | Google Scholar

Moss, S.E., Klein, R., and Klein, B.E.K. The incidence of visual loss in a diabetic population. Ophthalmology. 1988; 95: 1340–1348

View in Article | Google Scholar

Joussen AM, Gardner TW, Kirchhof B, Ryan SJ (2007). Retinal vascular disease. Heidelberg: Springer.

The Diabetes Control and Complications Trial Research Group. Color photography vs fluorescein angiography in the detection of diabetic retinopathy in the diabetes control and complications trial. Arch Ophthalmol. 1987; 105: 1344–1351

View in Article | Google Scholar

Flynn HW, Smiddy WE (2000). Diabetes and ocular disease, San Francisco: The Foundation of the American Academy of Ophthalmology.

Aiello, L.P., Avery, R.L., Arrigg, P.G., Keyt, B.A., Jampel, H.D., Shah, S.T. et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med. 1994; 331: 1480–1487

View in Article | Google Scholar

Pierce, E.A., Avery, R.L., Foley, E.D., Aiello, L.P., and Smith, L.E. Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neovascularization. Proc Natl Acad Sci U S A. 1995; 92: 905–909

View in Article | Google Scholar

Elman, M.J., Bressler, N.M., Qin, H. et al. Diabetic Retinopathy Clinical Research Network. Expanded 2-year follow-up of ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2011; 118: 609–614

View in Article | Google Scholar

Heier, J.S., Korobelnik, J.F., Brown, D.M. et al. Intravitreal Aflibercept for Diabetic Macular Edema: 148-Week Results from the VISTA and VIVID Studies. Ophthalmology. 2016; 123: 2376–2385

View in Article | Google Scholar

Nguyen, Q.D., Brown, D.M., Marcus, D.M. et al. Ranibizumab for diabetic macular edema: results from 2 phase III randomized trials: RISE and RIDE. Ophthalmology. 2012; 119: 789–801

View in Article | Google Scholar

Brown, D.M., Schmidt-Erfurth, U., Do, D.V. et al. Intravitreal aflibercept for diabetic macular edema: 100-week results from the VISTA and VIVID studies. Ophthalmology. 2015; 122: 2044–2052

View in Article | Google Scholar

Brown, D.M., Nguyen, Q.D., Marcus, D.M. et al. Long-term outcomes of ranibizumab therapy for diabetic macular edema: the 36-month results from two phase III trials: RISE and RIDE. Ophthalmology. 2013; 120: 2013–2022

View in Article | Google Scholar

Haritoglou, C., Kook, D., Neubauer, A. et al. Intravitreal bevacizumab (Avastin) therapy for persistent diffuse diabetic macular edema. Retina. 2006; 26: 999–1005

View in Article | Google Scholar

Early Treatment Diabetic Retinopathy Study research group. Photocoagulation for diabetic macular edema. Early Treatment Diabetic Retinopathy Study report number 1. Arch Ophthalmol. 1985; 103: 1796–1806

View in Article | Google Scholar

Cho, H.J., Rhee, T.K., Kim, H.S. et al. Intravitreal bevacizumab for symptomatic retinal arterial macroaneurysm. Am J Ophthalmol. 2013; 155: 898–904