Magnetic Resonance Angiography–Diffusion Mismatch Reflects Diffusion–Perfusion Mismatch in Patients with Hyperacute Cerebral Infarction

Abstract

We evaluated whether clinical-diffusion mismatch (CDM) or magnetic resonance angiography (MRA)-diffusion mismatch (MDM) is useful in detecting diffusion-perfusion mismatch (DPM) in hyperacute cerebral infarction within 3 hours after stroke onset. Among patients with cerebral infarction who arrived within 3 hours after stroke onset at our hospital between May 2007 and December 2010, we included 21 patients (16 men and 5 women; mean age 70 ± 7.8 years) with cerebral infarction of the anterior circulation, and in whom magnetic resonance imaging (diffusion-weighted imaging)/MRA and computed tomograpic perfusion of the head were performed at the time of arrival. DPM-positive status was defined as a difference between DWI abnormal signal area and mean transit time prolongation area (≥ 20% on visual assessment). CDM-positive status was defined as a National Institute of Health Stroke Scale score ≥ 8 and DWI-Alberta Stroke Program Early CT Score (ASPECTS) ≥ 8. MDM-positive status was defined as a major artery lesion and DWI-ASPECTS ≥ 6. Ten of 21 patients had DPM. In all DPM-positive patients, MRA revealed a major artery lesion. Of the 10 DPM-positive patients, 6 were CDM-positive. CDM detected DPM with a sensitivity of 60% and a specificity of 64%. The positive likelihood ratio was 1.65. Of the 10 DPM-positive patients, all were MDM-positive. MDM detected DPM with a sensitivity of 100% and a specificity of 82%. The positive likelihood ratio was 5.5. In hyperacute cerebral infarction within 3 hours after onset, MDM, as compared with CDM, was able to detect DPM with higher sensitivity and specificity. This suggests that MDM is more reflective of DPM.