An Introduction to OCT Angiography for Ophthalmology Residents

I remember back about a decade ago when spectral-domain optical coherence tomography (SD-OCT) was first being introduced to the market. The images were gorgeous—significantly higher resolution than the preceding technology, time-domain OCT.

However, pretty images alone were not enough to convince a doctor to fork over $70,000 for a new machine. I remember hearing a discussion about the potential utility of the machine and, in particular, whether it added enough information to affect clinical decision making for our patients. At the time we were really using OCT for black-and-white assessments: is there cystoid edema or not in an AMD patient, is a macular hole present or not, is there vitreomacular traction or not. These questions could easily be answered with the older technology.

This is a perfect example of one of my favorite Yogi Berra-esque sayings: “you don’t know what you don’t know.” Over the next few years as we used SD-OCT we realized its potential. People started describing subtle yet important findings that would never have been possible before. We’re seeing this again with the current emergence of OCT angiography (OCTA). Practices will be reluctant to make a large investment into a new technology with uncertain clinical benefit.

First, a quick step back. As a fast, non-invasive, and richly-informative imaging modality, optical coherence tomography (OCT) has revolutionized the diagnosis and management of macular diseases. Since OCT’s introduction in the early 1990s (and commercial availability in the early 2000s or so), improvements in hardware and software have dramatically increased scanning speeds and axial resolution. With these technological advances, new clinical applications such as OCT angiography (OCTA) have developed.

While there are several OCTA devices on the market, there are currently three companies with FDA-approved devices capable of OCTA: the Optovue Angiovue, the Zeiss AngioPlex, and Heidelberg Engineering’s Spectralis.

OCTA is a novel imaging modality that allows clinicians to visualize retinal microvasculature in exquisite detail in a variety of pathologies such as diabetic retinopathy, retinal vascular occlusions, and dry and exudative AMD. The fundamental idea of OCTA is that in a static eye, the only motion detected is that from erythrocytes (i.e., Brownian motion). With high numbers (usually several—for instance, Zeiss takes up to four) of OCT B-scans taken rapidly at the same position, a decorrelation signal can be obtained and presented as bright pixels, where white is assigned to movement and black to non-movement. These bright pixels, when viewed en face, form an image of the retinal vasculature, which can closely resemble the look of a fluorescein angiogram in which active vasculature appears white and non-active areas are black.

These data are presented as enface “slabs” which correspond to different anatomical layers and vascular plexi. Depending on the pathology that you are examining, you will focus on a particular slab.

As NPDR primarily affects the inner retinal vasculature, the superficial slabs should be examined for abnormalities such as microaneurysms and capillary non-perfusion. As OCTA does not show leakage, it can reveal strikingly detailed images that can be easily followed over time, particularly compared to dye-based fluorescein. You can see the loss of individual capillaries in the foveal avascular zone.

Some systems take this a step further, and have the capability to quantify vessel density, which can be used to objectively follow patients over time. This is increasingly important given the consequences of progressive macular ischemia and the role of antiVEGF injections in preventing this. Widefield OCTA will add even more information as this is developed.

OCTA may also reveal more advanced features of DR such as intraretinal microvascular abnormalities (IRMA) and, potentially growth of neovascular tissue off the surface of the optic nerve and retina, impacting treatment decisions.

Similar to diabetic patients, patients with retinal venous occlusive disease can also be easily evaluated for vascular changes using OCTA. Areas of nonperfusion can be seen with inner retinal enface slabs, and this can help determine if a vein occlusion is associated with significant ischemia much the same way that a fluorescein angiogram would help make that distinction. Perhaps better, given the lack of dye leakage which can obscure the underlying vascular changes. You can also follow areas of non-perfusion for re-perfusion and vascular remodeling, which helps to track the patient’s progress and provides important prognostic information.

When checking for neovascularization, I use the vitreoretinal interface enface slab so that I can isolate vessels on the retinal surface. It’s important to know that not all machines have this capability; however, the ZEISS CIRRUS OCT includes angioplex software that provides this particular slab. This capability can be largely beneficial.

Unlike standard macular scans, OCTA retinal scans can be montaged (in machines that support this function) to create larger images to better encompass posterior segment vascular disease. The montage option will guide the patient by moving the internal fixation target after each successful scan and then automatically combine the images into a widefield OCT angiogram with up to 50° of field of view. Due to the large areas that are often affected in retinal vein occlusions, montaging several OCTA images together can improve visualization of nonperfusion and neovascularization.

OCTA montage may provide an advantage over dye-based angiography where there is dye transit at different stages. Fluorescein angiography (FA), along with indocyanine green angiography (ICGA), have been the gold standard for the evaluation of retinal and choroidal vasculature, respectively. These imaging modalities visualize vasculature by taking advantage of dye diffusion dynamics and excitation properties of two injectables dyes, fluorescein sodium in FA and indocyanine green in ICGA.

OCTA is also very useful in the management of neovascular AMD, and particularly useful in distinguishing conditions that masquerade as wet AMD such as CSCR and vitelliform macular degeneration. Unlike DR, OCTA in AMD relies on imaging of the outer retina, and choriocapillaris slabs, where we can visualize choroidal neovascularization (CNV). This is especially useful for distinguishing small or occult choroidal neovascular membranes from drusen, fibrous scars, or nonvascularized pigment epithelial detachments.

With the rise of OCTA, our understanding of quiescent or nonexudative CNV has grown tremendously. Quiescent CNV and exudative CNV have similar vasculature patterns on OCTA but there is no retinal fluid associated with quiescent CNV. These lesions can grow, regress, stay stationary, or convert to exudative CNV. Treatment patterns for these inactive lesions have not been well explored, but OCTA is a great way to image these cases and tailor your management to individual lesion characteristics.

Though OCTA is still in its infancy, researchers are already pushing this technology beyond the posterior segment. OCTA imaging of anterior segment vasculature may prove to be useful in tumor detection, iris neovascularization, and corneal neovascularization—helping us guide both medical and surgical treatment.

For all its benefits, OCTA does have some drawbacks. As OCTA only demonstrates movement, it is unable to represent “leakage,” “staining,” or “pooling” of dye, which are important features on a fluorescein angiogram (FA). It is also not a dynamic test, and cannot evaluate blood flow. The tests may be a bit more difficult for patients as blinking or any head movement could obscure the results.

Regardless, I am confident that with time OCTA will become a mainstay in retina. We don’t know what we don’t know, and as we gain that knowledge—as we describe new clinical findings, important disease features and variants, and find that we can follow vascular diseases more appropriately and tailor care better for our individual patients—ten years from now we will wonder how we ever practiced without this technology.