Abstract MP59: Retinal Vascular Metrics Predict Pial Collateral Status in Patients With Acute Ischemic Stroke

Abstract

Background: The extent of the pial collateral circulation may determine outcomes in an acute ischemic stroke. Experimental studies suggest that retinal vessel metrics and geometric patterning may predict the pial collateral status. We have undertaken a translational study to quantify and relate retinal vascular metrics to the grade of pial collaterals in patients with acute ischemic stroke. Method: 35 patients admitted with acute stroke underwent computed tomography angiography (OCT) and were graded as having good (n=20)(47.55 ± 10.65 years) or poor (n=15) (48.93 ± 10.91 years) pial collaterals and compared to healthy controls (n=21)(44.26 ± 10.15 years). Retinal images were generated using OCT and central retinal artery equivalent (CRAE), central retinal vein equivalent (CRVE), artery-to-vein ratio (AVR), segmented fractal analysis and lacunarity, tortuosity index and fractal dimensions (capacity D 0 , information D 1 and correlation D 2 , curve asymmetry, singularity length and f-alfa-max using MONA software) were quantified. Results: Age ( p =0.709), BMI ( p =0.451), total cholesterol ( p =0.845), triglycerides ( p =0.679), LDL ( p =0.953), HDL ( p =0.361) and HbA 1c ( p =0.210) were comparable but the national institute of health stroke scale ( p =0.031) and modified Rankin Scale ( p =0.048) were higher in patients with poor compared to good collaterals. CRAE ( p =0.114), CRVE ( p =0.946), AVR ( p =0.114), lacunarity ( p =0.442), tortuosity index ( p =0.681), fractal analysis ( p =0.656), curve asymmetry ( p =0.619) and singularity length ( p =0.944) did not differ between patients with poor compared to good collaterals. However, fractal capacity D 0 (1.673 ± 0.029 vs 1.654 ± 0.025, p =0.042), fractal information D 1 (1.610 ± 0.027 vs 1.591 ± 0.024, p =0.036), fractal correlation D 2 (1.581± 0.028 vs 1.564 ± 0.024, p =0.060), and f alfa max (1.674 ± 0.027 vs 1.654 ± 0.025, p =0.030) were higher in patients with poor compared to good collaterals. Conclusion: This study shows differences in retinal vessel fractal dimensions between acute stroke patients with poor compared to good pial collaterals. This represents a non-invasive imaging method to define the pial collateral status and develop personalized intervention management strategies in acute ischemic stroke patients.