Are there time-dependent differences in diffusion and perfusion within the first 6 hours after stroke onset?

Background and Purpose-Stroke heterogeneity in computed tomography-based studies has been attributed as main cause for missing efficacy of intravenous tissue plasminogen activator (tPA) therapy within 3 to 6 hours. We investigated early time-dependent differences in acute stroke pathophysiology by multiparametric magnetic resonance imaging(MRI). Methods-Stroke MRI of 112 acute ischemic stroke patients within <6 hours were dichotomized into a <3-hour group (n=52) and a 3- to 6-hour group (n=60). Infarct volume was determined on days 5 to 8. Lesion volumes were determined for apparent diffusion coefficient (ADC_man) and the subregion with ADC values <550x10(-9) mm/s(2) (ADC<550), and for the time-to-peak (TTP) delay of 2 to 4 seconds, 4 to 6 seconds, 6 to 8 seconds, and >8 seconds. A subsample analysis was performed for occlusions of the middle carotid artery (MCA) trunk (n=36) and MCA branches (n=30), and for all patients treated by intravenous tPA (n=70). Results-ADC and TTP lesion volumes were not different within <3 hours compared with volumes at 3 to 6 hours. In patients receiving intravenous tPA (n=70), there were no significant differences in ADC_man, TTP >2 seconds, and infarct volume (days 5 to 8) between the 2 groups. There was a greater proportion of ADC <550/ADC_man, which was most pronounced in patients with MCA trunk occlusions after 3 to 6 hours and a larger mismatch in the <3-hour group compared with that of the 3- to 6-hour group. In MCA branch occlusions, there was a less severe TTP delay after 3 to 6 hours. However, all differences missed the significance level (P=0.05) after correction for multiple testing. Conclusions-We observed no significant time-dependent differences within 6 hours after stroke onset in degree and volume of diffusion and perfusion impairment. An exclusion from intravenous tPA solely based on a rigid 3- hour time window seems unjustified in MRI-confirmed ischemic stroke.