Hypoperfusion-related cerebral ischemia and cardiac left ventricular systolic dysfunction.

BACKGROUND Cardiomyopathy and low ejection fraction (EF) are associated with cardiac thrombi and cardiogenic embolism but may also be risk factors for hypoperfusion-related cerebral ischemia (HRCI). Current stroke subtype criteria do not include an HRCI category. METHOD To look for evidence of HRCI, we compared mean infarct volume between serial patients with EF < or =35% and high-grade (> or = 70%) carotid stenosis and serial patients with normal EF and high-grade carotid stenosis. We matched serial stroke patients with EF < or =35% with stroke patients with normal EF and compared the number and type of ischemic lesion (symptomatic or asymptomatic) and mean infarct volume on magnetic resonance imaging. We blindly compared stroke subtype in these groups using modified Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria, including an HRCI category. RESULTS In patients with carotid stenosis, ipsilateral infarct volume was greater with EF < or = 35% (74.7 mL, 95% CI, 17.3-132.1 mL) than in controls (17.1 mL, 95% Cl, 9.4-24.8 mL) (P<.05). There was no difference in the mean number of HRCI-compatible infarcts on computed tomography scan between patients with low EF and controls. Symptomatic HRCI occurred in 4 of 15 patients with low EF and in 0 of 15 controls. CONCLUSIONS Symptomatic HRCI occurs in patients with low EF. Severe arterial stenosis may interact with left ventricular systolic dysfunction to cause cerebral hypoperfusion. Modification of the TOAST criteria to include an HRCI subtype is feasible and HRCI should be included as a stroke subtype.