Quadrant Asymmetry in Optical Coherence Tomography Angiography Metrics in Ischemic Versus Non-Ischemic Central Retinal Vein Occlusion Eyes

Abstract

Purpose: To determine whether quadrant asymmetry (QA) of optical coherence tomography angiography (OCTA) metrics differs between non-ischemic versus ischemic central retinal vein occlusion (CRVO).

Methods: Fifty-eight eyes (21 non-ischemic, 10 ischemic CRVO, and 27 contralateral control eyes) underwent 3 × 3 mm spectral-domain OCTA scans with quantification of the superficial retinal layer vessel length density (VLD) and perfusion density (PD). QA, defined as the maximum-minus-minimum value among four parafoveal Early Treatment Diabetic Retinopathy Study (ETDRS) quadrants, was compared by linear regression including fixed effects for each eye.

Results: Mean age was 73.6 ± 11.4 (range 39-88), 73.8 ± 12.4 (range 39-91) and 77.2 ± 9.83, (range 60-88); and QA was 3.46 ± 1.76, 3.14 ± 1.57, and 4.88 ± 2.42 for VLD and 0.072 ± 0.038, 0.062 ± 0.036, and 0.11 ± 0.056 for PD for control, non-ischemic, and ischemic, respectively. QA was significantly higher in ischemic (0.109 ± 0.056) than non-ischemic CRVO eyes (0.062 ± 0.036; P = 0.02) and control eyes for PD (0.072 ± 0.038; P = 0.03). QA was also greater in ischemic (4.875 ± 2.418) than non-ischemic CRVO (3.141 ± 1.572) for VLD (P = 0.04). In terms of identifying which particular quadrant is most affected by ischemia, multivariate regression analysis comparing intra-quadrant effect on the presence of ischemia versus non-ischemia showed no quadrant was significantly affected (P > 0.05 for all quadrants).

Conclusions: Ischemic CRVO increases intraeye QA of OCTA metrics when compared to non-ischemic CRVO and control eyes. No specific ETDRS quadrant appears to be more affected.

Translational relevance: This work uses an intraeye method to delineate between ischemic and non-ischemic CRVO by OCTA imaging, overcoming inter-eye variables encountered in clinical care.

Figure 1.

QA of VLD for control, non-ischemic and ischemic central retinal vein occlusion eyes. Quantitative measurements of VLD and PD in the parafoveal ETDRS subfields in spectral domain OCTA images of the SRL. QA values for control (left-most column; A, D), non-ischemic CRVO (middle column; B, E) and ischemic CRVO (right-most column; C, F).

Figure 2.

CDF graph of QA (Maximum-minus-Minimum) for control and non-ischemic group compared to ischemic group. QA was calculated by subtracting the quadrant with the minimum value from the quadrant with the maximum value for each individual eye. CDF plot showed that ischemic group had greater QA than control and non-ischemic group across SRL VLD and PD. PD is defined as the total area covered by perfused vasculature per unit area (%); VLD is defined as the total length of perfused vasculature per unit area (mm/mm²).

Figure 3.

Intra-quadrant comparison of images for the superficial retinal layer vessel length density and perfusion density for control and non-ischemic group compared to ischemic group. Asterisk denotes significance set at P < 0.05. PD is defined as the total area covered by perfused vasculature per unit area (%); VLD is defined as the total length of perfused vasculature per unit area (mm/mm²).