Failure to demonstrate peri-infarct depolarizations by repetitive MR diffusion imaging in acute human stroke.

BACKGROUND AND PURPOSE Peri-infarct depolarizations (PIDs) have been demonstrated with diffusion-weighted MRI (DWI) in experimental stroke and are regarded as an important mechanism of ischemic injury. We tested the hypothesis that PIDs are of relevance for the early enlargement of human brain infarcts. METHODS Ten stroke patients were investigated by repetitive imaging of the apparent diffusion coefficient (ADC) in the acute phase (7 patients) or subacute phase (3 patients) of developing cortical infarction. In each patient, 20 ADC maps were obtained from serially measured echo-planar DWI (interval of 45 seconds). Data analysis focused on the potential spatial and temporal ADC changes, including structured qualitative analysis, calculation of subtraction images, serial analysis of regions of interest positioned in the infarct core and border, and calculation of hemispheric lesion areas, depending on various ADC thresholds ranging between 0 and 800 microm(2)/s. RESULTS Data analysis was unable to disclose any time-dependent changes in ADC that would resemble PID. In ischemic regions, the ADC reduction significantly progressed from the infarct border (555+/-96 microm(2)/s) to the infarct core (431+/-104 microm(2)/s, P:<0.01). CONCLUSIONS By using an MRI protocol with high temporal resolution and elaborated postprocessing, we were unable to demonstrate a pattern of diffusion changes that would be indicative of PID in human stroke. Experimental infarction and human stroke may differ in the detectability of PID.