Nonproliferative Diabetic Retinopathy

Nonproliferative Diabetic Retinopathy

Thomas Hwang, MD; Yali Jia, PhD; Miao Zhang, PhD; and David J. Wilson, MD

The clinical features of nonproliferative diabetic retinopathy (NPDR) predict the risk of vision loss associated with the development of proliferative diabetic retinopathy (PDR).^1,2 Fluorescein angiography (FA) provides additional diagnostic and prognostic information by revealing microvascular details.³ Recently, angiographic evaluation in NPDR has become more important as studies have found that macular ischemia on angiography can predict the risk of progression to PDR when anti-vascular endothelial growth factor (VEGF) injections for diabetic macular edema (DME) have masked the clinical features.⁴ However, FA is not used in the daily management of NPDR because of the cost, the time, and the risk associated with the procedure.

Like FA, optical coherence tomography angiography (OCTA) can also assess macular ischemia and other vascular abnormalities, but noninvasively and more rapidly, making it more suitable for everyday use. The 3-dimensional (3D) nature of the technology can be leveraged to reveal details and anatomic relationships not seen on dye-based angiography. The consistently high contrast in OCTA allows objective quantification of nonperfusion. This chapter will review the advantages and limitations of OCTA in this setting in detail as well as the OCTA features of NPDR.

ADVANTAGES AND LIMITATIONS OF OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY IN DIABETIC RETINOPATHY

FA provides vascular details in a dynamic fashion. Depending on the time elapsed between the dye injection and image capture, there is differential contrast between retinal and choroidal fluorescence and dye excursion within and out of normal and abnormal vessels. Thus it is well suited to evaluate certain features of DR, such as hyperpermeability of vessels and abnormal transit of blood through diseased vessels. In addition, wide-field and ultrawide-field systems have been developed for FA, which allow global evaluation of the retinal vasculature. Furthermore, major clinical trials have validated its utility in treatment and risk stratification.⁵

The limitations of FA in DR, however, go beyond the obvious—the need for intravenous access, risk associated with dye injection, and the time and inconvenience of extended light exposure.⁶ Because the macular capillary details are best obtained during the early transit phase, it can evaluate only one eye for these details per study. The contrast is variable depending on the pigmentation of the retinal pigment epithelium (RPE) and the degree of choroidal fluorescence. Leakage-associated hyperpermeability caused by DR can obscure capillary details. It can visualize only the superficial vascular plexus, and ambiguities can remain about neovascularization even with stereoscopic angiograms because of limited objective depth information.

Figure 21-1. A 3 × 3 mm OCT angiogram is segmented as superficial (left) and deep (center) plexus. Dilated saccules at the end of capillary branches are identified in both layers and correlate with MAs on FA (right), identified with blue circles. Not all MAs seen on FA could be identified on OCT angiogram, and not all dilated saccules on OCT angiogram could be identified as MAs on FA.

OCTA addresses many of the shortcomings of FA. No dye injection or intravenous access is necessary. The infrared source light allows for a more comfortable patient experience. Because the image does not vary based on the time between dye injection and image acquisition, both eyes can be evaluated. Each image can be acquired in 2 to 4 seconds. Since the contrast is not dependent on the difference between the choroidal and retinal fluorescence, it can display the retinal capillaries with consistent contrast, which is especially important for automated quantification of nonperfusion areas.7,8

The 3D capacity of OCTA has opened new possibilities in DR.

The retinal vasculature consists of 4 distinct layers.⁹ The radial peripapillary capillaries are found within the nerve-fiber layer around the optic disc. The superficial plexus, which the FA shows in normal patients,¹⁰ occupies the ganglion cell layer throughout the macula and has the familiar branching pattern. The intermediate and deep plexuses are networks of capillaries of uniform density and caliber, concentrated in the inner and outer borders of the inner nuclear layer, respectively. Some authors consider these vessels flanking the inner nuclear layer as a single vascular layer.^5,11

These layers are distinct within the macula, but become less so in the peripheral retina.¹² Projection artifacts, which are reflectance-dependent axial tails from superficial flow signals, have prevented visualization of 3 distinct layers in the macula. A technique that resolves these projection artifacts have allowed visualization of these distinct layers, revealing previously unseen microvascular pathology in DR.¹³

Because OCTA obtains dense volumetric structural information simultaneously, it can present angiographic and structural information together. By demonstrating the relationship between intraretinal cysts in DME and vascular pathology, for example, it can provide additional insights about DR.¹⁴

Limitations of OCTA include a relatively small field of view and the need for excellent fixation. Motion artifacts can significantly degrade the image quality, and the patient’s cooperation for 2 to 4 seconds is necessary for a high-quality image. Extended field of view with OCTA is being explored but the commercial systems currently typically offer 3 × 3 mm, 6 × 6 mm, and 8 × 8 mm images. Like traditional imaging, OCTA is affected by media opacity. Finally, some dynamic information, such as accumulation of dye in microaneurysms (MAs) or hyperpermeability of vessels seen as leakage, do not have a direct equivalence in OCTA.

ANGIOGRAPHIC FEATURES

Microaneurysms

MAs are a hallmark of NPDR, and often the earliest sign of disease. Histologically, they are saccular outpouchings of the capillary wall. The lumen can be patent or occluded by thrombus. Clinically, they appear as small red dots of 25 μm to 100 μm. FA shows MAs as pinpoint hyper-fluorecent dots that have variable leakage and fading in the later frames.¹⁵

In OCTA, MAs appear as dilated saccules or focally dilated capillaries contiguous with other capillaries (Figure 21-1).¹⁶ They are better seen on 3 × 3 mm angiograms compared to 6 × 6 mm,¹⁷ and their visualization is enhanced by separating the vascular layers to superficial and deep plexuses, even without 3-layer separation with a projection-resolved approach.

Dye accumulation within MAs allows FA to detect MAs. But accumulation of dye within an MA is not equivalent to flow, which is what OCTA detects. Not surprisingly, vascular features consistent with MAs on OCTA do not always correlate with MAs on FA and MAs on FA cannot always be seen as a specific feature on OCTA.¹⁶ It is not clear whether OCTA would provide equivalent information as FA for focal laser treatment. Despite this limitation, MAs on OCTA are closer to histological appearance than in FA, which often show these lesions as disconnected points.