Accuracy of advanced CT imaging in prediction of functional outcome after endovascular treatment in patients with large-vessel occlusion

Abstract

Background and purpose Computed tomography perfusion (CTP) and multiphase CT angiography (mCTA) help selection for endovascular treatment (EVT) in anterior ischemic stroke (AIS). Our aim was to investigate the ability of perfusion maps and collateral score to predict functional outcome after EVT. Patients and methods Patients with M1-middle cerebral artery occlusion, evaluated by mCTA and CTP and treated with EVT within six hours of onset, were enrolled. Perfusion parametric maps of cerebral blood flow (CBF), cerebral blood volume (CBV) and time to maximum of tissue residue function ( Tmax) were generated; areas of altered perfusion were manually outlined to obtain volumes CBFv, CBVv, [Formula: see text] and [Formula: see text]. Diffusion-weighted imaging (DWI) at 24–36 hours was used to manually outline the ischemic core (volume: DWIv). Collateral vessels were assessed on mCTA considering extent and delay of maximal enhancement (six-point scale). Functional outcome was evaluated by modified Rankin Scale score at three months. Volumes in good and poor outcome groups were compared by Wilcoxon rank-sum test t, and their discriminative ability for outcome was determined by receiver operating characteristic analysis. A logistic regression model, including Tmax, CBF and collaterals, was used to differentiate good and poor outcome. Results Seventy-one patients (mean age 75 ± 11 years, range 45–99 years) were included. [Formula: see text], [Formula: see text], CBVv, CBFv and DWIv were statistically different between the two groups. CBF had the best discriminative value for good and poor outcome (area under the curve (AUC) 0.73; 64.5% sensitivity; 74.4% specificity); the logistic regression model might be promising (AUC 0.79, 64.5% sensitivity, 82.1% specificity). Conclusions In patients with AIS, the combined use of CTP and mCTA predicts functional outcome of EVT and might allow better selection.