OCT Angiography: Could it Change Clinical Practice?

Obtaining angiography using OCT is mechanistically different than traditional dyebased angiography; those differences are significant for the data the images provide.

By Jay M. Haynie, OD, FAAO

Much can be learned about the status of the posterior segment under direct observation or via optical coherence tomography (OCT). For diagnostic or prognostic purposes , however, it is often useful to appreciate a patient’s retinal blood flow. The most commonly used technique for understanding the retinal vasculature is fluorescein angiography (FA), in which fluorescent dye that highlights blood vessels is injected into the blood stream. The back of the eye is photographed in sequential images over a given period of time. Deeper imaging of the choroid may be achieved by a similar mechanism using indocyanine green (ICG) dye.

Dye-based staining is limited to 2-dimensional imaging, it is time consuming, and invasive. Newly available imaging algorithms using an OCT platform provide an ability to capture angiographic information that differs substantially from FA or ICG. Emerging evidence suggests that OCT angiography (OCTA) may be superior to FA in certain settings, specifically with respect to imaging the radial peripapillary or the deep capillary networks.1 At the very least, the noninvasive nature of this new technology allows for more rapid image acquisition.

WHAT IS OCT ANGIOGRAPHY?

OCTA and FA may yield similar quantitative and qualitative output, but the mechanisms by which each acquires information results in significant differences.2 The images they produce are comparable but not necessary equal.

For instance, interpretation of FA images depends on several factors, and, in general, the presence of hyper- and hypofluoresence patterns that result from the dynamics of the dye passing through the vasculature are used to identify areas of pathologic interest. Hyperfluorescence seen with FA is most commonly secondary to vascular leakage; however, it can also be seen as staining of lipofuscin (or drusen) or as a window defect when the retinal pigment epithelium is absent. Hypofluorescence seen with FA is when there is absence of retinal or capillary blood flow (referred to as nonperfusion). Another cause of hypofluoresence is “blocked” fluorescence from hemorrhage, lipid, fibrosis, or pigment clumping.

Figure 1. A choroidal neovascular membrane as seen with OCT-A is seen to originate in the choriocapillaris layer of this patient with exudative AMD.

OCTA, on the other hand, depends on algorithmic reconstruction of images that fundamentally track the movement of blood cells (and not dye) through the vasculature. Although multiple mechanisms for acquiring angiography on OCT platforms have been described, currently available technologies capture angiographic information using one or a combination of two methodologies.3 In split-spectrum amplitude-decorrelation angiography (SSADA), the final image is reconstructed from numerous cross-sectional image frames that result from a splitting of the OCT image into narrow spectral bands. Each spectral band provides a different speckle pattern based on the location of blood cells and surrounding structures. The variation (or decorrelation) between speckle patterns on the multiple spectral bands in sequential images over time corresponds to movements of red blood cells within the vasculature. These differences are mathematically averaged by processing software to yield a structural representation of vessel location and an index of flow velocity.4 Phase-variance algorithms use a slightly different mechanism, wherein consecutive B-scans taken at the same location are analyzed for change over time, with the resulting phase variance in backscattered light corresponding to the location of vasculature in the retina and choroid.5 The Optovue AngioVue employs the SSADA mechanism, whereas the Ziess AngioPlex uses a proprietary Optical Micro Angiography (OMAGc) algorithm that combines elements of the SSADA and phase-variance methodologies to produce images.

WHY IS IT IMPORTANT?

As noted above, the mechanism by which OCTA produces images results in important differences relative to FA. Notable among these are resolution of the capillary beds, the ability for image segmentation, and the noninvasive nature of OCTA; as a result of these features, OCTA may be applied in the clinic in a much different fashion.

Figure 2. Microvascular changes consistent with aneurysmal lesions are noted in a patient with diabetic retinopathy. Also noted is an enlargement in the foveal avascular zone consistent with macular ischemia.

Resolution of the Capillary Beds

One of the apparent limitations of FA is the inability to fully appreciate the deep capillary network, perhaps due to a backscatter of light at the retina.6 OCTA provides much better definition of the deep vascular anatomy, which may have consequences for understanding features of retinal pathologies as well as, potentially, the ischemic implications of glaucoma. As demonstrated by Spaide et al,1 OCTA appears to be a superior mechanism for understanding the radial peripapillary capillary network around the optic nerve head in particular.

Segmentation

Imaging the deep capillary bed using OCTA is additive to the ability to capture the superficial and intermediate layers. Final images can be assessed cumulatively or in different sections, the latter of which may have important applications in certain disease states. For example, angiographs at different layers might reveal which plexus is most affected by choroidal neovascularization secondary to AMD, while segmentation of the choriocapillaris and outer retina may reveal the presence of feeder vessels that may otherwise go unnoticed or underappreciated.

Noninvasive

Although FA and ICG are generally safe, the use of dye can be associated with complications, including nausea, vomiting, and pruritus at the injection site. Severe reactions, including urticaria, pyrexia, thrombophlebitis, and syncope, are rare, and death due to anaphylaxis is extremely rare.7 FA is generally contraindicated in pregnant women.8,9

On the other hand, one of the greatest advantages of OCTA is that it allows noninvasive evaluation of the retinal vasculature, which not only permits rapid differential diagnosis of functional vision loss, but also a greater facility to repeat imaging over time to understand change and response to the management plan.

Convenience and Clinical Applicability

Because of the aforementioned features of OCTA, it provides important actionable clinical information that can be quickly acquired for patients who present with a number of retinal pathologies with potential vascular involvement. As this pertains to the practice of optometry, this may mean the ability to follow more patients in the clinic without having to refer for a fluorescein angiography any time vascular involvement is suspected. Although there are numerous potential clinical applications of OCTA, the most prominent disease states in which it may be useful are diabetic eye disease (diabetic retinopathy), AMD (Figure 1), retinal vein occlusions, central serous chorioretinopathy, and ischemic retinal disease where patients lose vision due to lack of blood flow or ischemia. Of course, because OCTA acquires the same information as dye-based angiography, it is equally useful for additional entities classically monitored with FA, such as uveitis and retinal vasculitis; and ICG, such as choroidal neovascularization, pigment epithelial detachments, polypoidal choroidal vasculopathy, retinal angiomatous proliferation, intraocular tumors, uveitis, and choroidal inflammatory conditions.

Whether a new diagnostic or prognostic modality will affect clinical decision-making is an important consideration. With OCTA this is certainly the case, for example, facilitating the ability to differentiate between wet and dry AMD, as well as the capacity to detect functional vision loss due to ischemia that may not be amenable to treatment. In patients with diabetic eye disease, viewing the capillary bed may highlight microvascular changes that precede dot and blot hemorrhages, and microaneurysms that indicate diabetic retinopathy (Figure 2). Repeat scanning studies in individuals with diabetic retinopathy is also starting to distinguish patterns of interest in the foveal avascular zone that may be prognostic for progressing disease.

Overall, these attributes of OCTA make it a potentially useful tool for identifying complications of systemic disease before it results in subjective decline in vision, thus allowing the clinician to be proactive in managing disease and counseling patients to gain better control of their systemic diseases.

IMPORTANT CAVEATS

Of practical consideration, OCTA requires extremely fast imaging capabilities, likely around 70,000 A-scans per second at a minimum, to reconstruct accurate B-scans for sequential analysis. This technological requirement may mean that bringing this technology into one’s practice requires a new capital investment. In some cases, however, a software upgrade may be all that is needed. For example, users of the Cirrus 5000 (Carl Zeiss Meditec) platform can add angiography capabilities to their existing device. Similarly, users of the Optovue Avanti can add AngioVue, which the company offers in two different configurations.

Of note, Topcon earlier this year introduced the first commercially available swept-source OCT device (DRI OCT Triton), which has an angiography feature. This device appears to use a 1050-nm light source, which could facilitate penetration into deeper choroidal layers. Heidelberg Engineering is also working on an angiography application based on its Spectralis platform using its propriety OCT2 module that increases imaging speed to 85,000 Hz. However, neither the Topcon nor the Heidelberg devices are available in the United States, and their development plans are not clear. The bottom line is that although OCTA generally offers advantages over FA and other dye-based imaging, it is not clear if one platform is necessarily superior to another.

Another thing to note about OCTA is that the technology inherently assumes all changes between sequential en face images are due to blood movement; therefore, the images are prone to imaging artifacts. This is an area of keen interest to the field and it is subject to much ongoing research. The current understanding of artifacts associated with OCTA is that several factors may be involved, including the mechanism of OCT image acquisition; intrinsic properties of the eye and eye motion; image processing and display; and OCTA projection artifacts, where the software is essentially fooled and cannot determine where the flow is in the depth of the retina.4 However, OCTA platforms contain a correction algorithm that removes the projections to clean up the image.

The caveats mentioned here are by no means insurmountable, and in fact, they are probably pretty typical of a nascent technology such as OCT angiography. As the technology continues to evolve, many of the points of differentiation from traditional FA and ICG angiography should become more apparent. Over time, it is also likely that additional research will add greater understanding of the applications, especially if reliable normative databases are developed. Further innovations in the technology, such as image stitching capability, which will allow operators to capture a widefield montage, will likely serve to expand the clinical applicability and add to its impact on clinical decision-making.

1. Spaide RF, Klancnik JM Jr, Cooney MJ. Retinal vascular layers imaged by fluorescein angiography and optical coherence tomography angiography. JAMA Ophthalmol. 2015;133(1):45-50.

2. de Carlo TE, Romano A, Waheed NK, Duker JS. A review of optical coherence tomography angiography. Int J Retin Vitr. 2015;1:5.

3. Novais EA, Roisman L, de Oliveira PR, et al. Optical coherence tomography angiography of chorioretinal diseases. Ophthalmic Surg Lasers Imaging Retina. 2016;47(9):848-861.

4. Spaide RF, Fujimoto JG, Waheed NK. Image artifacts in optical coherence tomography angiography. Retina. 2015;35(11):2163-2180.

5. Schwartz DM, Fingler J, Kim DY, et al. Phase-variance optical coherence tomography: a technique for noninvasive angiography. Ophthalmology. 2014;121(1):180-187.

6. Mendis KR, Balaratnasingam C, Yu P, et al. Correlation of histologic and clinical images to determine the diagnostic value of fluorescein angiography for studying retinal capillary detail. Invest Ophthalmol Vis Sci. 2010;51(11):5864-5869.

7. Yannuzzi LA, Rohrer, MA, Tindel LJ, et al. Fluorescein angiography complication survey. Ophthalmology 1986;93:611-617.

8. Halperin LS, Olk J, Soubrane G, Coscas G. Safety of fluorescein angiography during pregnancy. Am J Ophthalmol. 1990;09:563-6.

9. Greenberg F, Lewis RA. Safety of fluorescein angiography during pregnancy [letter]. Am J Ophthalmol. 1990;110:323-5.

Jay M. Haynie, OD, FAAO
• Executive clinical director of Retina & Macula Specialists in Tacoma, Washington
jayh@retina-macula.com
• Financial interest: none acknowledged