Reperfusion time window based on ischemic and hemorrhagic CBF thresholds measured by xenon-enhanced CT in hyperacute ischemic stroke

Background and Aims: Although the newly emerged thrombolytic treatment for stroke is effective, its therapeutic time window is extremely narrow, and it may trigger intracranial hemorrhage. Therefore, it would be beneficial, if we could apply some clinical techniques that are not time-consuming and useful for selecting appropriate candidates for treatment. It is well established in animal models that the viability of ischemic brain tissue depends on its residual cerebral blood flow (CBF) and the duration of the ischemic period. Moreover, lower CBF in the affected brain has been implicated in increased incidence of hemorrhagic transformation in humans. However, substantial clinical data are still lacking, especially in the hyperacute period. We therefore tried to determine the ischemic and hemorrhagic CBF thresholds at each ischemic time point in patients with recanalization after an ischemic insult. Methods: Forty-three sequential patients (≦85 year, mean age: 64.7+12.8, NIHSS: 17.1+9.3) with sudden-onset cortical ischemic stroke who had presented to our hospital within 2 hours (44.8+34.5 min) after the onset were studied. In the case series, local CBF was measured by xenon-enhanced CT (Xe-CT) (124.0+49.8 min). In 6 selected patients, intra-arterial thrombolytic treatment with urokinase was performed according to the procedural protocol of PROACT II. In other 5 patients, the neurological deficit abruptly disappeared, strongly suggesting spontaneous recanalization. We estimated the duration of the ischemic period in these 11 patients from the time point of intervention or symptomatic recovery. The infarction and hemorrhagic areas were determined from CT and MRI images which were obtained more than 72 hours after the onset. Mean CBF values in hemorrhagic infarction, non-hemorrhagic infarction, intact area on the ischemic side, and the symmetrical area to the infarction on the contralateral side were calculated by setting ROI on the CT images in a blind manner. Results: At 3.3 hours after the onset, an area with a low CBF of 9 (ml/100 g/min) could escape from infarction. However, the CBF thresholds for the non-hemorrhagic infarction in gray and white matter reached a plateau (19 and 12, respectively) after 3.3 hours. The CBF threshold for hemorrhagic infarction was initially found to be lower (5-9) than that for non-hemorrhagic infarction at 4.0-4.7 hours. However, the thresholds for hemorrhagic infarction in gray and white matter reached comparable levels (17 and 9, respectively) to those for non-hemorrhagic infarction at 5.5 hours. Conclusions: Although our cases were limited, these data suggested that the duration of the ischemic period, as well as residual CBF, is closely related to tissue viability and the risk of hemorrhage in patients with reperfusion. These data also support that thrombolytic therapy is reasonable within 3 hours, while it might trigger hemorrhagic transformation more frequently if it is performed after 6 hours. Xe-CT may be a useful decision-making tool in selecting candidates for thrombolytic therapy in the hyperacute phase, providing residual CBF values which are associated with the outcome.