Perfusion computed tomography thresholds defining ischemic penumbra and infarct core: studies in a rat stroke model

Background: Perfusion computed tomography is becoming more widely used as a clinical imaging tool to predict potentially salvageable tissue (ischemic penumbra) after ischemic stroke and guide reperfusion therapies. Aims: The study aims to determine whether there are important changes in perfusion computed tomography thresholds defining ischemic penumbra and infarct core over time following stroke. Methods: Permanent middle cerebral artery occlusion was performed in adult outbred Wistar rats (n=6) and serial perfusion computed tomography scans were taken every 30 mins for 2h. To define infarction thresholds at 1h and 2h post-stroke, separate groups of rats underwent 1h (n=6) and 2h (n=6) of middle cerebral artery occlusion followed by reperfusion. Infarct volumes were defined by histology at 24h. Co-registration with perfusion computed tomography maps (cerebral blood flow, cerebral blood volume, and mean transit time) permitted pixel-based analysis of thresholds defining infarction, using receiver operating characteristic curves. Results: Relative cerebral blood flow was the perfusion computed tomography parameter that most accurately predicted penumbra (area under the curve=0·698) and also infarct core (area under the curve=0·750). A relative cerebral blood flow threshold of <75% of mean contralateral cerebral blood flow most accurately predicted penumbral tissue at 0·5h (area under the curve=0·660), 1h (area under the curve=0·659), 1·5h (area under the curve=0·636), and 2h (area under the curve=0·664) after stroke onset. A relative cerebral blood flow threshold of