Evaluating the Association of Widefield Swept-Source Optical Coherence Tomography Angiography with Visual Function in Common Retinal Diseases

Abstract

Background: The clinical evaluation, diagnosis, and treatment of ophthalmic diseases relies on both a structural evaluation and a visual function assessment. Wide-field swept source optical coherence tomography (WF SS-OCTA) is a novel imaging technique that acquires fast, non-invasive, high-resolution images of the retinal microvasculature. Visual acuity (VA) is the most common measure of visual function. Contrast sensitivity (CS) is an adjunct functional metric that studies suggest is better associated with vision related quality of life and affected earlier in retinal disease. Yet few studies have evaluated WF SS-OCTA and CS associations. We aim to evaluate the association of WF SS-OCTA with visual function, including CS, for a variety of common retinal diseases. Methods: Patients were prospectively-recruited from retina clinics at Massachusetts Eye and Ear. The cohorts included diabetic retinopathy (DR, n = 140), diabetic macular edema (DME, n = 61), retinal vein occlusion (RVO, n = 63), retinal detachment (RD, n = 57), age-related macular degeneration (AMD, n = 72), and healthy eyes (n = 57). WF SS-OCTA (PLEX® Elite 9000, Carl Zeiss Meditec) 3x3, 6x6, and 12x12 mm images were acquired, from which vessel density (VD) and vessel skeletonized density (VSD) in the superficial (SCP) and deep (DCP) capillary plexus, and whole retina (WR), foveal avascular zone (FAZ) parameters, and non-perfusion area were calculated. VA was measured on Snellen charts. CS was measured using the quantitative contrast sensitivity function (qCSF) device (Adaptive Sensory Technology), resulting in the following qCSF metrics: Area under the Log CSF (AULCSF), contrast acuity (CA), and contrast sensitivity (CS) thresholds at 1 to 18 cycles per degree (cpd). Mixed effects multivariable linear regression models were used to evaluate for associations between WF SS-OCTA vascular metrics and visual function. Results: Across all the diseases evaluated OCTA vascular metrics demonstrated significant associations with both VA and CS. Further, the effect size of OCTA vascular metrics was consistently larger on CS compared to VA on nearly all associations. For example, in DME the standardized beta coefficients for VSD and CS at 3 cpd (βSCP = 0.76, βDCP = 0.71, βWR = 0.72, p 0.001) were larger than those for VA (βSCP = -0.55, p 0.001; βDCP = -0.43, p = 0.004; βWR = -0.50, p 0.001). Across each disease there were instances where VA failed to demonstrate significant associations with the vascular metrics, unlike CS. For example among the RD cohort, CS had multiple significant associations with VD and VSD on all three image sizes, unlike VA. In AMD, nearly every single CS metric showed significant associations with VD and VSD on 3x3 mm images, even though VA did not. Each disease cohort revealed unique structure-function associations, for example, in RVO and RD the vascular metrics appeared to have the largest effect size on AULCSF, in DR and healthy eyes the largest effect size was on CS at 12-18 cpd, while in DME and AMD it was on CS at 3-6 cpd. Conclusion: Multiple structure-function studies across various retinal diseases demonstrated that WF SS-OCTA vascular metrics were significantly associated with both CS and VA. However, the effect size of OCTA was larger on CS, compared to VA, suggesting that CS likely better reflects the microvascular changes found in WF SS-OCTA. Integrating WF SS-OCTA and CS into the clinical workflow of retina clinics and utilizing these measures in clinical trials may positively impact the management and outcomes among various ophthalmic diseases.