Managing Diabetic Retinopathy With OCT Angiography

A case presentation illustrates the potential value of this new imaging modality.

By Jenae Stiles, BS, and Mike Cymbor, OD, FAAO

The prevalence of diabetes in the United States continues to grow at alarming rates. In 2014, it was estimated that 29.1 million people, or 9.3% of the US population, has diabetes.1 The disease is projected to effect from one in five to one in three Americans by 2050.2

The exposure of eye care professionals to the eye-related complications of diabetes, a lifelong disease, is bound to increase. A number of diagnostic technologies for detecting and following diabetic eye disease are well established, but optical coherence tomography angiography (OCTA) shows promise to be an additional valuable diagnostic tool for these patients.

The OCTA platform used in the case presentation that follows, the AngioVue (Optovue), received clearance from the FDA last year for imaging the retinal vasculature.3 This technology compares backscattered OCT signal intensity to sequential OCT B-scans taken at the same cross-section to construct a blood flow map in a matter of seconds.4 With this technology, scans take approximately 3 seconds, with a maximum imaging size of 8 mm by 8 mm and axial resolution of approximately 5 µm.5 This allows imaging of vasculature in the superficial (internal limiting membrane to inner plexiform layer), deep (inner nuclear layer to outer plexiform layer), and outer (outer nuclear layer to Bruch membrane) retinal layers and the choroid. The optic nerve and surrounding vitreous can also be imaged and then compared to any retinal scan in a montage view to give a full picture of possible pathology.

Figure 1. OCTA of the right optic nerve. Note areas of capillary nonperfusion superonasally (A1-A3). The choroid OCTA does not appear to have any new vessel growth or dropout (A4), as would be expected considering the superficial nature of diabetic retinopathy.

There are obvious differences between OCTA and other imaging methods such as fluorescein angiography (FA) and indocyanine green angiography (ICGA). Notably, FA and ICGA are invasive techniques that use conventional imaging methods, and OCTA is noninvasive, with virtually no side effects and the ability to perform scans faster. OCTA also offers greater resolution of each vascular layer and the ability to image vasculature, pathogenic or not, of the retinal layers from the vitreous to the choroid. OCTA also gives us the ability to examine each layer individually to help determine where the pathology originates.

FA and ICGA do not offer the advantage of segmentation of pathology of different retinal layers, especially when the media are obstructed by hemorrhage, vitreous or lens opacity, without a solid understanding of leakage and blockage patterns. It should be noted that bleeding in the back of the eye will limit the amount of light reaching the retina to be imaged, which can obscure an OCTA scan as well. That being said, OCTA can offer size and boundary measurements for pathology that could otherwise not be delineated by FA or ICGA.6

Hwang et al reported that OCTA was able to identify enlargement and distortion of the foveal avascular zone (FAZ) and retinal capillary dropout similarly to FA.7 These authors also noted that retinal areas of capillary loss obscured by fluorescein leakage on FA were better defined with OCTA. Further, areas thought to be microaneurysms on FA were actually defined as early neovascularization on OCTA by observing the flow rate just above the internal limiting membrane. The limited field of view with OCTA used by Hwang and colleagues in this report led to failure of recognition of some microaneurysms present; however, this has been partially addressed by the use of the montage view now available on the AngioVue software (Figure 9), which can show two scans side by side for a full view of the posterior pole.

Figure 2. OCTA of the right macula. Dark areas diagonal to the fovea, most evident in frame B7, appear to be artifact.

Figure 3. OCTA of the left optic nerve. Capillary dropout can be seen along the NFL defect leading to the superior temporal disc (C1-C3). Note obvious areas of capillary nonperfusion at the site of previous PRP inferonasally.

Possible drawbacks of OCTA technology include nonrecognition of slow blood flow in large diameter vessels, limited field of view, motion artifact due to microsaccades during fixation, and nonrecognition of vessel leakage.6,8 The AngioVue system has addressed motion artifacts with the development of DualTrac Motion Correction to minimize this possible hindrance by employing motion correction during live scanning to correct motion from large blinks and patient movement, as well as during post-processing where pixel-level corrections are made to address motion from saccades.5


A 43-year-old woman with type 1 diabetes mellitus of 35 years’ duration treated with Novolog (insulin aspart [rDNA origin] injection; Novo Nordisk), presented with a self-reported A1C of 7.4% and fasting glucose level of 120 mg/dL. Concomitant conditions included thyroid dysfunction, hypertension, and hypercholesterolemia, all controlled with oral medications.

The patient had undergone FA approximately 12 years ago, which showed patchy capillary dropout in the temporal macula OD and capillary dropout with hyperfluorescence and leakage inferonasal to the disc OS. Approximately 4 years ago, the patient underwent panretinal photocoagulation (PRP) nasally OS due to retinal neovascularization.

Figure 4. OCTA of the left macula. Darker areas of nonperfusion seen inferior to the foveal area (D5-D6), corresponding to patchy areas of dropout (D1-D2).

The patient underwent lumbar puncture and magnetic resonance imaging 12 years ago due to bilateral disc edema. Both were reported to be normal, and it appears that a diagnosis of diabetic papillitis could be made, although records from that time indicate possible normal tension pseudotumor cerebri.

Figure 5. OCTA of the superficial retina and optic nerve head of the left eye in montage view. This view allows the two 6 mm by 6 mm scans to be placed together to provide a more complete clinical picture.

At this visit, the patient’s best corrected visual acuity was 20/25 OD and OS with high myopic astigmatism correction. Her intraocular pressures were 14 mm Hg OD and 16 mm Hg OS by noncontact tonometry, with history showing that her intraocular pressures have typically ranged from mid to high teens. Moderate nonproliferative diabetic retinopathy was noted on dilated fundus exam in each eye, with nasal scarring OS due to PRP.


The patient underwent extensive imaging and diagnostic testing at the time of this presentation, with Optovue OCT and OCTA, perimetry testing (Octopus; Haag-Streit), and fundus photography (Figure 6).

In analyzing the Angio Structure/Function view of the patients optic nerve OCTA scans OD and OS (Figures 1 and 3, respectively), we noted several patches of superficial capillary dropout, appearing as black patches scattered among the first three scans (Figure 1, A1-3; Figure 3, C1-3) and excluding the deep choroidal scan, as would be expected. Sequential comparison of optic nerve OCTA scans of this type over time can aid in determining progression of ischemia that would not be apparent on fundus examination.

Recently, studies have been conducted to generate an automatic algorithm allowing quantification of diabetes-related capillary dropout on OCTA.9 No neovascularization is evident on either disc, as would most likely be evident in the second frame vitreous scan (Figure 1, A2; Figure 3, C2).10 PRP can be observed OS inferonasal to the disc, and an area of obvious nonperfusion can be observed throughout the scans in Figure 3, from C1 to C8.

Figure 6. Fundus photos taken on same day as OCTA. Note the previous PRP inferonasally in the left eye. Moderate nonproliferative diabetic retinopathy was noted on this visit due to multiple hemorrhages and venous beading. Slight pallor of each optic disc can be attributed to previous diabetic papillitis.

Figure 7. Octopus 24-2 visual field completed on same day as OCTA. Superotemporal defect in the left eye can be attributed to previous PRP. Possible shallow defect inferonasally in the left eye corresponds to the NFL defect superiorly in the papillomacular bundle. The diffuse defects in the right eye do not appear to follow any pattern and will be monitored with serial visual fields.

Next, examining the Angio Structure/Function of the macular OCTA scans OD and OS (Figures 2 and 4, respectively), we found the first two frames, showing the superficial and deep capillary plexus, to be of most interest, as this is where pathology typically resides with diabetic retinopathy (Figure 2, B1-2; Figure 4, D1-2). The darker areas noted diagonal to the foveal area OD (best seen in Figure 2, B3 and B7) appear to be an artifact, also noted in the right eye scans of other patients. These dark areas also appear to be faintly evident on all four OCTA scans (Figure 2, B1-4), which would not be expected if superficial capillary dropout from diabetes is the pathologic etiology, as expected. The darker area that appears inferior to the foveal area OS (Figure 4, D5-6) suggests an area of nonperfusion. This small area appears as a patchy capillary section in the corresponding superficial and deep retinal OCTA scans (Figure 4, D1-2).

The OCTA image of superficial retina OS (Figure 4, D5) shows a well delineated nerve fiber layer (NFL) defect arching superiorly in the papillomacular bundle (Figure 9). The capillary dropout of this bundle can be best observed in the superficial vasculature frame (Figure 4, D1). This capillary dropout can also be faintly observed in montage view of the optic nerve and superficial retinal vasculature OS (Figure 5).

Due to the extensive capillary nonperfusion observed in both the optic nerve and macular OCTA, it can be concluded with reasonable certainty that ganglion cell ischemia due to microvascular damage led to the NFL defect, as no other risk factors exist that would suggest glaucoma. Baseline OCT of the optic nerve head was also obtained for this patient for further monitoring as well as Octopus 24-2 perimetry (Figure 7). When the area of NFL dropout is correlated with the visual field results, there appears to be a possible shallow inferonasal arcuate defect. The visual field OD also shows possible defects without a distinct pattern. This was the first visual field this patient had completed, and serial fields are planned for further monitoring along with OCTA. The denser area of visual field defect superotemporal OS corresponds to PRP scarring (Figure 7).

Enlargement of the FAZ is often seen in patients with diabetes of long duration as an increasing area around the macular region loses perfusion due to microvascular damage. Because this is baseline testing for this patient, it will be valuable with future scans to determine if in fact the FAZ is expanding. This might allow us to determine a hypoxic threshold after which visual loss will soon follow.11 New software algorithms by Optovue called AngioAnalytics is under development that will quantify areas of flow and non-flow as well as create a vessel density map, which may facilitate monitoring disease progression on subsequent visits.5

Figure 8. Fundus photo of the left posterior pole compared with superficial retinal vasculature OCTA. Microaneurysms present on both fundus photography and OCTA are circled in green.

Figure 9. Enlarged OCTA view of the left macular area. Note the superior NFL defect (red arrows). This scan correlates with superficial OCTA in Figure 4 (D5).

No recent FA was available for comparison for this patient, but we were able to compare the microaneurysms visible in the fundus photography with those visible on OCTA (Figure 8). The macular superficial vasculature OCTA image OS shows multiple microaneurysms, and corresponding areas of bleeding can be seen on the fundus photo (circled in green on both images, Figure 8).6 OCTA microaneurysms appear as areas of dilation at the terminal ends of blood vessels. With the AngioVue software, any desired scan can be obtained in varying sizes including 3 by 3 mm, 6 by 6 mm, or 8 by 8 mm which affects the image resolution. Scans for this patient were taken at 6 mm by 6 mm to facilitate a general overview, but perhaps scans at greater resolution would allow better comparison of microaneurysms.

OCTA can miss some microaneurysms found by FA, but OCTA can also detect other pathology that is not as apparent on FA, such as areas of retinal nonperfusion, reduced capillary density, and vessel tortuosity.12 This type of imaging could be used to monitor the resolution of retinal hemorrhages or to further grade the severity of nonproliferative diabetic retinopathy. Although not all hemorrhages may be visualized with OCTA, other clinical pictures can be used to support similar conclusions, and the additional information offered by this technology, as discussed above, could prove to be just as valuable as other imaging technologies in clinical decision-making.


OCTA is an emerging tool that may be helpful in managing diabetic patients. To date, we have been surprised by how many patients exhibit some level of capillary dropout on OCTA. In addition to capillary dropout, microaneurysms and NFL dropout can also be followed. All of this can be done without side effects and with less time per image than needed for conventional imaging methods. Perfusion density mapping may soon prove invaluable in management of diabetic retinopathy, providing early clues to disease. Eventually, medications may be developed to target these early changes, resulting in reduced morbidity.

Additional work must be done comparing the accuracy of OCTA versus conventional imaging methods. More study is also needed to further reduce artifacts. OCTA may prove to be another valuable diagnostic method for managing and following the multitude of diabetic patients that will inevitably end up in our offices in the coming years.

1. Number of Americans with diabetes projected to double or triple by 2050 [press release]. Centers for Disease Control and Prevention. October 22, 2010. Accessed February 14, 2017.

2. National Diabetes Statistic Report, 2014. Centers for Disease Control and Prevention. Accessed February 14, 2017.

3. Non-invasive technology that visualizes blood vessels of the retina available in US [press release]. Optovue. February 12, 2016. Accessed February 14, 2017.

4. de Carlo TE, Lee GD, Lally DR, Duker JS. What’s Up With OCTA? Retina Today. July/August 2015.

5. Products: Optovue. AngioVue. 2016. Accessed February 14, 2017.

6. de Carlo TE, Romano A, Waheed NK, Duker JS. A review of optical coherence tomography angiography (OCTA). Int J Retina Vitreous. 2015;1:5.

7. Hwang TS, Jia Y, Gao SS, et al. Optical coherence tomography angiography features of diabetic retinopathy. Retina. 2015;35(11): 2371-2376.

8. Matsunaga D, Yi J, Olmos LC. OCT angiography (OCTA) of diabetic retinopathy. Invest Ophthalmol Vis Sci. 2015;56: ARVO abstract 3335.

9. Li W. OCT angiography for the evaluation of diabetes-related capillary dropout. October 3, 2016. Practice Update.

10. Cho H, Alwassia AA, Regiatieri CV, et al. Retinal neovascularization secondary to proliferative diabetic retinopathy characterized by spectral domain optical coherence tomography. Retina. 2013;33(3):542-547.

11. Zheng Y, Gandhi JS, Stangos AN, et al. Automated segmentation of foveal avascular zone in fundus fluorescein angiography. Invest Ophthalmol Vis Sci. 2010;51(7):3653-3659.

12. Salz DA, de Carlo TE, Adhi M, et al. Select features of diabetic retinopathy on swept-source optical coherence tomographic angiography compared with fluorescein angiography and normal eyes. JAMA Ophthalmol. 2016;134(6):644-650.

Mike Cymbor, OD, FAAO
• Partner, Nittany Eye Associates, State College, Pennsylvania
• Financial disclosure: member of the Optovue Speakers’ Bureau

Jenae Stiles, BS
• Doctor of optometry candidate, 2017, Salus University, Philadelphia, Pennsylvania