The stroke patient who woke up: clinical and radiological features, including diffusion and perfusion MRI.

BACKGROUND AND PURPOSE Time of stroke onset is uncertain for patients who wake from sleep with stroke. Functional imaging techniques may allow estimation of benefit and risk of acute stroke therapy. We compared the clinical and multimodal MRI findings of patients with uncertain stroke onset with those with known onset time. METHODS Patients imaged within 24 hours of ischemic stroke onset between January 1997 and June 2000 were identified from a prospective stroke registry. Clinical and imaging data from patients with known stroke onset (group I) were compared with those who woke with stroke (group II). RESULTS A total of 364 patients were identified, of whom 100 (27%) woke from sleep with stroke. Group I and group II did not differ in age, gender, National Institutes of Health Stroke Scale, or TOAST (Trial of Org 10172 in Acute Stroke Treatment) diagnoses. Time from stroke onset was shorter in group I (mean 6.0 versus 13.3 hours, P<0.001); time from detection did not differ between groups (6.0 versus 5.9 hours). Within 3 hours, diffusion-weighted imaging (DWI) and perfusion-weighted imaging (PWI) lesion volumes were similar in both groups; DWI-PWI mismatch was present in 82% of group I and 73% of group II patients. Mean apparent diffusion coefficient of water (ADC) of group I patients was negatively associated with DWI volume (beta=-0.324, P=0.004) and time from stroke onset (beta=-0.238, P=0.031) in multivariate analysis. The mean ADC of group II patients was lower than that of group I patients within 3 hours of stroke detection (mean 556 versus 665 microm2/s, P<0.01), but individual group II patients had ADC values as high as 742 microm2/s, in addition to a DWI-PWI mismatch pattern. CONCLUSIONS Onset time is uncertain in over one quarter of acute ischemic stroke patients. Clinical features of these patients do not differ significantly from those with known onset time. Some patients who wake with stroke seem to have favorable imaging characteristics for acute stroke therapy. Further study is needed to determine whether criteria for therapy based on imaging parameters can safely be applied to these patients.