Key Features
- •
OCTA can noninvasively detect the movement of red blood cells at capillary-level resolution.
- •
OCTA is particularly useful for detecting regions of impaired perfusion and neovascularization.
- •
OCTA has been used to evaluate many of the pathological macular changes in retinal vascular diseases, including diabetic retinopathy, retinal vein occlusion, macular telangiectasia, and neovascular age-related macular degeneration.
Introduction
Optical coherence tomography (OCT) is a noninvasive imaging method that has been used extensively in the field of ophthalmology since 2002. OCT generates high-resolution cross-sectional images of the retina based on the interference of back-scattered coherent light. Progress in OCT technology has facilitated new OCT-based imaging methods, such as polarization-sensitive OCT, spectroscopic OCT, phase-sensitive OCT, and spectral-domain (SD) OCT angiography (OCTA). OCTA is a functional extension of OCT and is being used increasingly to detect microvascular changes in many retinal diseases since approval by the U.S. Food and Drug Administration in 2016. We will briefly review the biological basis of OCTA imaging, highlight the various methods of generating OCTA images, and discuss the strengths and limitations of this novel method.
Biological Basis of OCTA
OCTA is based on the variation in OCT signal caused by moving particles, such as red blood cells (RBCs), in contrast to stationary surrounding neurosensory tissue. This variability is relatively analogous to the Doppler shift that moving particles impose on reflected light. Although there are several different methods of performing OCTA imaging ( Table 6.8.1 ), all of these methods differentiate moving particles from static retinal tissue by comparing multiple (two or more) OCT B-scans performed at the same location. This is in contrast to standard OCT, which performs a single B-scan at each location. Because RBCs are constantly moving, they generate a unique variation in the intensity and phase of the backscattered OCT signal compared with the nonmoving retinal tissue within each repeated B-scan at the same location. Several methods of analyzing this variance have been developed and can be divided into those that use the phase variance of light, the intensity variance of light, or both phase and intensity (see Table 6.8.1 ).
| Phase-Based OCTA | Intensity-Based OCTA | Phase + Intensity (Complex) OCTA |
|---|---|---|
|
|
|
OCTA images provide a map of retinal vessels with blood flow detectable on SD-OCTA devices, but do not provide the rate of blood flow at any locations ( Fig. 6.8.1 ). In vitro studies and some in vivo studies suggest that current SD-OCTA devices can detect blood flow in the range of 0.3–3.3 mm/sec. This is approximately the range that has been demonstrated by confocal scanning laser ophthalmoscopy. Flow rates above or below this range may artifactually appear as regions of “nonperfusion.” Also of note, although RBCs are the most mobile component of retinal tissue, any particulate motion can theoretically generate similar motion contrast signal. For example, lipid particulates in solution generate OCTA signals as a result of Brownian motion. Therefore, the possibility of artifactual signal should be considered when interpreting OCTA images.
OCTA Versus Dye-Based Angiography Methods
With the rapid adoption of OCTA, it becomes necessary to determine the appropriate roles of OCTA versus those of fluorescein angiography (FA) and indocyanine green angiography (ICGA). Table 6.8.2 summarizes many of the strengths, limitations, and practical applications of these methods. It is important to recognize that although OCTA and dye-based angiography methods provide somewhat similar en face images, they measure different biological phenomena. Specifically, OCTA is based on light scattering from RBCs and particulate debris, so there is no diffusion of dye on OCTA images. This fundamental difference is illustrated by the absence of “leakage” on OCTA images in subjects who have macular edema on FA. Another illustration of this difference is the variable rate of microaneurysm detection on OCTA in comparison with FA.
| Imaging Method | Strengths | Limitations a | Optimal Applications |
|---|---|---|---|
| Fluorescein angiography (FA) |
|
|
|
| Indocyanine green angiography (ICGA) |
|
|
|
| OCTA |
|
|
|
Many studies have demonstrated that OCTA does not directly detect hyporeflective pockets of intraretinal fluid as observed on OCT. Nor does OCTA demonstrate the hyperfluorescence that is observed in late phase FA in subjects with diabetic macular edema or cystoid macular edema. This is attributed to the notion that fluid within most cystoid spaces does not contain large particles that can backscatter light. In contrast, FA is based on the tissue distribution and fluorescence of dye molecules that leak into cystoid spaces. Although dye leakage highlights cystoid spaces as well as abnormal vessels, such as neovascularization ( Fig. 6.8.2 ), it also tends to obscure potentially relevant details both in pathological cases as well as normal cases. This is because there is still modest dye leakage from normal vessels which increases the background noise in FA and ICGA. For example, high background fluorescence can make it appear as though the macula is completely ischemic by obscuring fine capillaries in proliferative diabetic retinopathy, whereas OCTA can clearly demonstrate the persistence of capillaries and visual potential.
Some recent studies have highlighted the significant difference in appearance of microaneurysms on OCTA versus FA, whereas others have shown significant similarities. In some cases, the excellent depth resolution of OCTA images has revealed that lesions consistent with the appearance of microaneurysms on FA are actually small tufts of neovascularization. It has also been demonstrated that, at least in some cases, microaneurysms are not detected as frequently on OCTA as on FA. Even when microaneurysms are detected on OCTA, they are usually of different sizes and shapes compared with those on FA. It is likely that the reason for this and the discrepancy among studies is that some microaneurysms are sclerosed or clotted and without blood flow, whereas others are patent or partially patent. Because OCTA only detects the movement of RBCs, sclerosed or clotted microaneurysms will not appear at all on OCTA. In addition, the flow rate of blood within microaneurysms may be outside the detection speed of SD-OCTA devices. Microaneurysms that are partially sclerosed or partially recanalized will also appear much smaller on OCTA than on FA because only the region with RBC flow will be visualized on the former, whereas the whole lesion will stain with dye on FA.
Other considerations for the use of dye-based angiography versus OCTA are their practical limitations and strengths (see Table 6.8.2 ). These considerations include how each method is performed and the resolution of the images. The most important drawback of any dye-based method is the possibility of adverse reactions ranging from mild reactions to severe and possibly life-threatening reactions. An important strength of dye-based imaging methods is that current wide-field systems provide an almost comprehensive assessment of the central and peripheral retina. In addition, ICGA is generally superior in detection of choroidal neovascularization (CNV) compared with current OCTA systems because of the limited penetration of the OCTA signal through the retinal pigment epithelium (RPE). Lastly, compared with OCTA, dye-based angiography has relatively limited resolution. For example, although FA can delineate the foveal avascular zone very well in primates, <40% of the capillaries outside the foveal center are visualized on FA as compared with histology. Mendis et al. showed that FA assessment of capillary density is ≈50% less than histology-based assessments. Specifically, FA cannot resolve the radial peripapillary plexus or the capillaries in the deep retinal layers. In contrast, OCTA clearly and reliably resolves these capillaries in human subjects.
One final caveat with regard to OCTA is the potential to misinterpret images because of artifacts. In many cases, the artifacts on OCTA are similar to and derived from artifacts that are observed on standard OCT images. The most common in OCTA images is called a “projection artifact,” where vessels in the superficial retinal layers cast shadows or “projections” on deeper retinal layers, which results in the artifactual appearance of vessels where none exists. This is most clinically relevant in the interpretation of choroidal pathology and CNV. An excellent review of the subject is available, and several methods of removing these artifacts have been developed.
Key Applications of SD-OCTA
Detection of Impaired Perfusion (or “Nonperfusion”)
In this chapter, we refer to “nonperfusion” as “impaired perfusion.” Because it is very hard to absolutely demonstrate lack of blood flow with any current imaging method, the term “nonperfusion” is misleading and likely inaccurate in many cases. For both dye-based imaging studies as well as OCTA, there can be very slow blood flow rates that are either obscured by background noise or undetectable. Because of the relatively low resolution of FA, impaired perfusion is likely very severe by the time it is detectable on FA. In contrast, OCTA can reliably resolve individual capillaries with unprecedented depth-resolution in humans. OCTA images demonstrate capillary detail approaching the resolution of histology. This makes OCTA an ideal method for detecting and monitoring regions of impaired perfusion in subjects with retinal vascular disease ( Fig. 6.8.3 ). In addition, OCTA can be repeated many times to obtain the ideal image with essentially no risk to subjects, which is not possible with FA. Therefore, OCTA provides a whole new dimension of depth information regarding the severity of impaired perfusion that is not possible with FA. For example, studies have suggested that impairment of perfusion in the deep retinal layers is more severe in certain diseases, such as paracentral acute maculopathy, diabetic retinopathy, and retinal venous occlusion. In addition, detection of capillary loss in the macula of patients with diabetes and peripapillary capillary plexus of patients with glaucoma is now possible before clinical lesions are evident. The clinical relevance of this additional information remains to be determined, but at the least OCTA will allow for detection of impairment in blood flow (ischemia) much earlier than before and help assess the severity of the impairment with far more precision.
