Pathophysiological topography of acute ischemia by combined diffusion-weighted and perfusion MRI

DG Darby, PA Barber, RP Gerraty, PM Desmond… - Stroke, 1999 - Am Heart Assoc
DG Darby, PA Barber, RP Gerraty, PM Desmond, Q Yang, M Parsons, T Li, BM Tress…
Stroke, 1999Am Heart Assoc
Background and Purpose—Combined echoplanar MRI diffusion-weighted imaging (DWI),
perfusion imaging (PI), and magnetic resonance angiography (MRA) can be used to
visualize acute brain ischemia and predict lesion evolution and functional outcome. The
appearance of a larger lesion by PI than by DWI quantitatively defines a mismatch of
potential clinical importance. Qualitative lesion variations exist in the topographic
concordance of this mismatch. We examined both the topographic heterogeneity and …
Background and Purpose—Combined echoplanar MRI diffusion-weighted imaging (DWI), perfusion imaging (PI), and magnetic resonance angiography (MRA) can be used to visualize acute brain ischemia and predict lesion evolution and functional outcome. The appearance of a larger lesion by PI than by DWI quantitatively defines a mismatch of potential clinical importance. Qualitative lesion variations exist in the topographic concordance of this mismatch. We examined both the topographic heterogeneity and relative frequency of mismatched patterns in acute stroke using these MRI techniques.
Methods—Acute DWI, PI, and MRA studies of 34 prospectively recruited patients with supratentorial ischemic lesions scanned within 24 hours of stroke onset (range 2.5 to 23.3 hours, 12 patients <6 hours) were analyzed.
Results—Ischemic lesions were predominantly in the middle cerebral artery (MCA) territory (94%), with DWI lesions most commonly affecting the insular region. Mismatched patterns with PI lesion larger than DWI lesion occurred in 21 patients (62% overall), in all 4 patients imaged within 3 hours, and in 44% of patients imaged after 18 hours. A patient with a large PI but no DWI lesion and severe clinical deficit at 2.5 hours after stroke onset recovered completely. Regional variations in DWI and PI lesion loci were found, inferring site of proximal MCA occlusion, embolic pathogenesis, and regional arterial reperfusion.
Conclusions—Analysis of the topographic concordance of PI and DWI lesions in acute stroke reveals regional PI lesions without concomitant DWI lesions, which do not necessarily progress to infarction but may suggest stroke pathogenesis and site of current arterial occlusion. Location of DWI lesions may suggest an earlier site of arterial occlusion and regions of maximal perfusion deficit.
Am Heart Assoc