Volumetric Quantification of Retinal Ischemic Perivascular Lesions (RIPLs) in Coronary Artery Disease

Purpose: Retinal ischemic perivascular lesions (RIPLs) have been reported as correlates of several systemic conditions. This study aimed to assess the prevalence, number, area, and volume of RIPLs and to evaluate their association with coronary artery disease (CAD) using a contemporary imaged coronary angiography (CA) cohort. Methods: CA patients were imaged using an SS-OCTA device during the same visit. Structural volumes comprising 500 B-scans were pixel-wise annotated by three graders. RIPLs were quantified by presence, count, area and volume for the first time. CAD severity was assessed using the Gensini score (GS). Associations between RIPL metrics and CAD were analyzed using generalized estimating equation models. Multivariable and age and sex adjusted models were calculated. Multiple testing corrections were applied to primary analyses. Results: RIPLs were detected in 38.0% of eyes. In models with the Gensini score (GS) as outcome, RIPL metrics showed weak and inconsistent trends and were not statistically significant, especially after adjustment. The GS showed no significant associations with RIPL presence (β = 0.132, P = 0.102), area (β = 0.173, P = 0.093), count (β = 0.023, P = 0.165), or volume (β = 0.240, P = 0.085) either. Cardiovascular risk factors were not independently associated with RIPL metrics. Those negative findings remained consistent in sensitivity analyses using the SYNTAX II and SCORE2. Conclusions: In this contemporary coronary angiography cohort with high-density OCT imaging, RIPL burden was not significantly associated with CAD and no robust or proportional association with CAD severity was found. In this work, no other disease specific associations could be established.