Wide-Field OCT Angiography Investigation of the Relationship Between Radial Peripapillary Capillary Plexus Density and Nerve Fiber Layer Thickness

Abstract

Purpose: To characterize the density and distribution of the radial peripapillary capillary plexus (RPCP) and its relationship with retinal nerve fiber layer (NFL) thickness in healthy subjects.

Methods: Using spectral-domain optical coherence tomography (OCT), split-spectrum amplitude decorrelation angiography algorithm and automated montaging, wide-field OCT angiography (OCTA) was used to measure the RPCP capillary density (CD) and NFL thickness. Polar sector-average CD and thickness maps were also created on each eye.

Results: Wide-field OCTA (8 × 8 mm) in 10 healthy eyes from 10 subjects demonstrated the distribution of the RPCP throughout the posterior pole. RPCP-CD decreases with distance from the disc, but along the arcuate nerve fiber bundles relatively dense (> half maximum density) RPCP extends more than 5 mm from the disc and includes regions superior to and inferior to the macula. The RPCP-CD and NFL thickness are highly correlated (R2 = 0.85, P < 0.001) and fit well with a nonlinear stacked-layer model. The model fit suggests that the RPCP is present when the NFL is thicker than 17.9 μm and reaches a ceiling area density of 84% and that the RPCP has an apparent volume density of 19% at the current instrument transverse resolution. This indicates that capillary overlap can be expected to occur when NFL thickness reaches 40 μm.

Conclusions: The wide distribution of dense overlapping RPCP suggests that wider (up to 8 mm vertical and 7 mm horizontal) OCTA scans may be better investigate capillary loss in the early stages of glaucoma or other optic neuropathies.

Figure 1

A representative image from a left eye. (A) Montaged RPCP angiograms (∼8 × 8 mm). (B) Montaged NFL thickness map. The size of each individual scan is 4.5 × 4.5 mm.

Figure 2

(A) Montaged RPCP OCT angiograms (∼8 × 8 mm) representing the maximum projection of flow signal within the NFL. (B) Montaged en face OCT of the NFL representing the mean projection of the reflectance signal. (C) Binary RPCP angiogram (white, vascular; black, nonvascular). (D) Binary large vessel map. (E) RPCP-CD map. The unit size (9 × 9 pixels) used for capillary density measurement was shown by a small red box in C.

Figure 3

Polar coordinate sectors were used to analyze the RPCP-CD map and the NFL thickness map.

Figure 4

(A) Group-average RPCP-CD map. (B) SD map of RPCP-CD, showing the intersubject variation.

Figure 5

(A) Group-average NFL thickness map. (B) SD map of NFL thickness, showing the intersubject variation.

Figure 6

Scatter plot of NFL thickness and RPCP-CD in each polar sector in each eye.

Figure 7

(A) Averaged RPCP-CD divided into polar sectors. Sectors with greater than 42% CD are outlined in red. Commercial OCTA scan areas of 3 × 3, 4.5 × 4.5, and 6 × 6 mm² (gray solid lines) and the area of efficient rectangle (white dash line; 8 × 7 mm) were overlaid. (B) Averaged retinal NFL thickness divided into polar sectors. Sectors with greater than 48-μm NFL thickness are outlined in red.