Optical coherence tomography angiography (OCTA) is a non-invasive imaging technique that visualizes blood vessels by applying the principle of optical coherence tomography (OCT). OCT was invented in the 1990s and generates high-resolution cross-sectional images based on low-coherence interferometry. It has become one of the most important diagnostic imaging methods in ophthalmology.
Conventional OCT has low contrast between capillaries and retinal tissue, limiting its ability to monitor vascular changes. Fluorescein angiography (FA) and indocyanine green angiography (ICGA) require intravenous administration of contrast agents, which carries a risk of anaphylactic shock. Additionally, they only provide two-dimensional images without depth information of lesions.
OCTA overcomes these limitations by three-dimensionally visualizing blood vessels at different depths without the use of contrast agents. It is useful for diseases causing retinal circulation abnormalities (e.g., diabetic retinopathy, retinal vein occlusion), optic nerve circulation abnormalities (e.g., glaucoma, ischemic optic neuropathy), and conditions involving choroidal neovascularization.
In neuro-ophthalmology, evaluation of the radial peripapillary capillaries (RPC) around the optic disc is particularly important. A decrease in vessel density is observed in areas corresponding to retinal nerve fiber layer defects (NFLD) seen on fundus examination.
OCTA does not require contrast agents, is non-invasive, and allows three-dimensional, layer-specific evaluation of the vascular network. FA/ICGA can detect fluorescein leakage and filling defects, but carries a risk of anaphylaxis due to contrast administration, and the images obtained are only two-dimensional. The two are complementary; OCTA cannot evaluate fluorescein leakage.
In neuro-ophthalmic diseases, the main vascular changes observed on OCTA are as follows.
MS / Optic Neuritis
Multiple sclerosis (MS): RNFL thinning is observed regardless of the presence of optic neuritis. The ONH blood flow index (ONH-FI) is significantly reduced in eyes with a history of optic neuritis. Combined use with structural OCT parameters improves detection accuracy.
Optic disc edema (papillitis): Inflammatory disc edema does not show vascular dropout. The radial distribution of peripapillary capillaries is preserved.
Ischemic Optic Neuropathy
Non-arteritic anterior ischemic optic neuropathy (non-arteritic): Peripapillary capillary blood flow impairment is observed in both acute and chronic phases. The temporal sector has the highest blood flow density and is most affected in non-arteritic anterior ischemic optic neuropathy eyes. In the chronic phase, ONH and peripapillary vessel density are directly associated with RNFL damage and visual field defects.
AAION (arteritic): Features include dilation of superficial peripapillary capillaries and focal non-perfusion of superficial and deep retinal capillaries.
Hereditary and Degenerative Diseases
LHON (Leber Hereditary Optic Neuropathy): Capillary dilation is observed during the pseudoedematous phase. A pattern of choriocapillaris loss by disease stage has been described: in the early subacute phase, it decreases from the temporal side, and in the chronic phase, it decreases in all sectors.
Optic Atrophy: Decreased peripapillary microvasculature. Blood flow decreases via autoregulatory mechanisms due to reduced metabolic activity.
Glaucoma
POAG (Primary Open-Angle Glaucoma): Decreased vessel density and blood flow index in the intraoptic, macular, and peripapillary regions.
NTG (Normal-Tension Glaucoma): Decreased peripapillary capillary density is observed similarly to primary open-angle glaucoma but is milder. In suspected glaucoma cases, a decrease in blood flow index may be detectable even before visual acuity loss.
OCTA is useful for differentiating between papilledema and pseudopapilledema.
In papilledema, the edema obscures the underlying capillaries, but capillaries can be visualized on the edema, and the peripapillary vessel density is maintained at the same level as the control eye. In contrast, pseudopapilledema (e.g., optic disc drusen) shows reduced vessel density, and this difference serves as the basis for differentiation.
In optic neuritis or AION with optic disc swelling, detection of acute axonal damage is difficult due to increased cpRNFL thickness from impaired axonal transport. Macular inner retinal layer analysis, such as the ganglion cell complex (GCC), can detect thinning earlier than cpRNFL analysis.
In optic nerve diseases causing central scotoma or cecocentral scotoma, a thinning pattern reflecting damage to the papillomacular bundle (PMB) is observed. OCTA shows decreased RPC density corresponding to the PMB thinning area.
OCTA repeatedly images the same location on the fundus and detects only moving parts (red blood cells) as random signal changes. It utilizes the fact that flowing red blood cells produce greater signal variation between scans than stationary tissue.
There are mainly two types of detection methods.
A representative algorithm is SSADA (Split-Spectrum Amplitude Decorrelation Angiography). It splits the OCT spectrum into narrower bands and averages the intensity decorrelation of each band, significantly improving the signal-to-noise ratio (SNR).
OCTA automatically generates en face images of four layers.
| Layer | Name | Main evaluation target |
|---|---|---|
| Superficial | Superficial capillary plexus (SCP) | Retinal nerve fiber layer to ganglion cell layer |
| Deep layer | Deep retinal capillary plexus (DCP) | Around inner nuclear layer |
| Outer retina | Outer retinal layers | Normally avascular |
| Deep | Choriocapillaris (CC) | 10–30 μm outside Bruch’s membrane |
Segmentation settings vary by device. The RPC and SCP are often combined and displayed as SCP. Some opinions suggest that SS-OCT is superior to SD-OCT for visualizing the choriocapillaris.
The advantages of OCTA are as follows:
On the other hand, attention must be paid to the following limitations and artifacts.
The main artifacts include: decreased signal intensity due to cataract or vitreous opacity (caution: may be confused with non-perfusion areas), white lines or distortion caused by eye or head movement, and projection of superficial signals onto deeper layers (projection artifact). It is important to confirm correspondence with B-scan images during interpretation.
The procedure and key points for OCTA evaluation are described below.
The characteristics of each examination method are summarized below.
| Characteristic | OCTA | FA | ICGA |
|---|---|---|---|
| Contrast agent | Not required | Required | Required |
| Depth information | Three-dimensional (layered) | Two-dimensional | Two-dimensional |
| Fluorescein leakage | Not detectable | Detectable | Detectable |
| Quantifiability | High | Low | Low |
| Repeatability | Easy | Difficult | Difficult |
The following are representative evaluation indices when using OCTA in neuro-ophthalmology.
OCTA plays a complementary role with laser speckle flowgraphy (LSFG). While OCTA evaluates vascular structure (density and morphology), LSFG quantifies blood flow velocity. Combining both enables a more comprehensive circulatory assessment4).
Repeated scans of the same area separate moving components (red blood cells) from static components (tissue). The SSADA algorithm reduces background noise caused by subtle eye movements using an averaging method (volume averaging).
Imaging of the choriocapillaris (CC) is performed at a depth of 10–30 μm outside Bruch’s membrane. The appearance of en face images is granular rather than mesh-like. This is attributed to limitations in lateral resolution, background noise, and vascular discontinuity.
The optic nerve head (ONH) receives blood supply from the short posterior ciliary arteries (SPC arteries), which are terminal branches of the ophthalmic artery. The SPC arteries branch off from the ophthalmic artery and divide into 10–20 branches.
The vascular supply of each region is as follows.
ONH blood flow depends on ocular perfusion pressure (OPP = mean arterial pressure − intraocular pressure). Regulation involves endothelin-1 and nitric oxide from the vascular endothelium, and animal studies indicate that autoregulation is effective when OPP ≥ 30 mmHg.
In optic atrophy, as the number of peripapillary nerve fibers decreases, metabolic activity declines, leading to reduced blood flow via autoregulation. This is observed as a decrease in vascular density.
Yoshimura et al. (2024) reported a case of congenital nasal optic nerve hypoplasia (NOH) in a 20-year-old woman 1). Quantitative evaluation of RPC density using OCTA (Nidek RS-3000 Advance 2, 4.5mm×4.5mm) showed a marked decrease in RPC density on the nasal side of the affected eye to 19% (superior 51%, temporal 58%, inferior 38%). This corresponded to areas of cpRNFL thinning and wedge-shaped visual field defects on Humphrey perimetry, demonstrating the utility of OCTA in elucidating the clinical features and pathophysiology of NOH.
Erba et al. (2021) reported a case of acute VKH in a 24-year-old man 2). OCTA (Topcon DRI OCT Triton Plus) detected flow void spots in the choriocapillaris, which correlated with hypofluorescent areas on ICGA. After treatment with prednisolone 60 mg/day tapered plus cyclosporine A 100 mg twice daily, best-corrected visual acuity recovered to 20/20 in both eyes, and choroidal thickness decreased from RE 712 μm and LE 750 μm at initial visit to RE 538 μm and LE 548 μm at 3 months. Flow voids also markedly decreased after treatment, indicating the usefulness of OCTA for monitoring disease activity.
Mehta et al. (2022) reported a case of granulomatosis with polyangiitis in a 61-year-old man 3). The right eye showed superficial capillary dropout and the left eye showed AION, with OCTA noninvasively detecting capillary dropout in the superficial capillary plexus and flow voids in the choriocapillaris. After immunosuppressive therapy with methylprednisolone 500 mg for 3 days plus a single dose of cyclophosphamide 500 mg, OCTA findings markedly improved at 1 month. OCTA was shown to detect clinically invisible choroidal lesions.
Tsai et al. (2023) evaluated circulation in a 50-year-old woman with optic disc melanocytoma using OCTA and LSFG 4). OCTA detected deep retinal vascular networks within the tumor, and the LSFG MBR (mean vessel density ratio) of the optic disc and macula in the affected eye was lower than in the healthy eye (disc MBR: affected eye 23.0±0.8 vs. healthy eye 26.5±1.9). FA could only assess via dye blockage, but OCTA overcame this limitation. Combined use with LSFG enabled more comprehensive circulation assessment.
In MOG-ON, decreased peripapillary and parafoveal vessel density has been confirmed compared to healthy controls. Vessel density reduction correlated with number of ON episodes, pRNFL thickness, and visual acuity, suggesting that retinal vessel loss may be due to decreased metabolic demand associated with retinal degeneration.
OCTA, being non-invasive and repeatable, is expected to have clinical applications in the following areas.
Promising areas include early detection of glaucoma (detecting blood flow changes before vision loss), monitoring progression of optic nerve diseases (quantitative assessment of vascular density over time), and noninvasive evaluation of retinal and choroidal lesions in systemic vasculitis. Application to evaluating the efficacy of immunosuppressive therapy using changes in flow voids after treatment as an indicator is also advancing.
