Determinants of leptomeningeal collateral flow in stroke patients with a middle cerebral artery occlusion.

INTRODUCTION

Poor leptomeningeal collateral flow is related to worse clinical outcome in acute ischemic stroke, but the factors that determine leptomeningeal collateral patency are largely unknown. We explored the determinants of leptomeningeal collateral flow and assessed their effect on the relation between leptomeningeal collateral flow and clinical outcome.

METHODS

We included 484 patients from the Dutch acute stroke study (DUST) with a middle cerebral artery (MCA) occlusion. The determinants of poor leptomeningeal collateral flow (≤50 % collateral filling) were identified with logistic regression. We calculated the relative risk (RR) of poor leptomeningeal collateral flow in relation to poor clinical outcome (90-day modified Rankin Scale 3-6) using Poisson regression and assessed whether the determinants of leptomeningeal collateral flow affected this relation.

RESULTS

Leptomeningeal collateral flow was poor in 142 patients (29 %). In multivariable analyses, higher admission glucose level (odds ratio (OR) 1.1 per mmol/L increase (95 % CI 1.0-1.2)), a proximal MCA occlusion (OR 1.9 (95 % CI 1.3-3.0)), and an incomplete posterior circle of Willis (OR 1.7 (95 % CI 1.1-2.6)) were independently related to poor leptomeningeal collateral flow. Poor leptomeningeal collateral flow was related to poor clinical outcome (unadjusted RR 1.7 (95 % CI 1.4-2.0)), and this relation was not affected by the determinants of leptomeningeal collateral flow.

CONCLUSION

Our study shows that admission glucose level, a proximal MCA occlusion, and an incomplete ipsilateral posterior circle of Willis are determinants of leptomeningeal collateral flow that represent a combination of congenital, acquired, and acute factors. After adjustment for these determinants, leptomeningeal collateral flow remains related to clinical outcome.