Comparison of computed tomography and magnetic resonance imaging in acute ischemic stroke: choosing a hyperacute stroke imaging protocol for proper patient selection and time efficient reperfusive treatment

Acute ischemic stroke (AIS) is a clinical syndrome of rapid onset of focal cerebral deficit, lasting 24 hours or leading to death. The rationale behind intervention in hyperacute stroke is based on a three-compartment model of brain parenchyma following vascular occlusion. The infarct core represents non-viable brain that cannot be salvaged even with very prompt treatment (non-viable tissue, NVT). The ischemic penumbra is defined as functionally impaired yet still viable tissue surrounding the ischemic core. The penumbra includes ischemic areas that recover spontaneously (benign oligemia) and areas that progress to irreversible changes, unless effective treatment is used (referred to as tissue at risk, TAR). The penumbra is the most clinically relevant target for treatment by intravenous thrombolysis with intravenous tissue plasminogen activator (IVT rt-PA) and mechanical thrombectomy (MT). The goal of perfusion imaging is not only the diagnosis of ischemia, but also the detection of the penumbra. In order to quantify and more precisely detect brain perfusion, several standard flow parameters are calculated. The objective of the present study is to assess the utility of computed tomography perfusion (CTP) and/or magnetic resonance imaging (MRI) protocols for selection of patients affected by AIS for reperfusive treatments and to better characterize the reliability of the two methods in predicting follow-up infarction. We retrospectively reviewed consecutive AIS patients evaluated for the selection for reperfusive treatments at two comprehensive stroke centers (CSC), using two different hyperacute stroke imaging protocols (HSIP) for proper patient selection. The two CSC employ a FAST MRI based HSIP, including ASL and DWI sequences, and a multimodal CT based HSIP, including mCTA and CTP, respectively. We enrolled 158 consecutive patients admitted for AIS, undergoing mCT-based hyperacute stroke imaging. CTP presented a prognostic accuracy (ACC) of 75.9%. In patients with perfusion deficit in anterior circulation territory, CTP-Tmax > 16s proved relatively reliable to identify the NVT, with an overestimation in patients with favorable clinical-instrumental outcomes. Similarly, CTP-Tmax > 9.5s proved reliable to identify the area of total hypoperfusion, overestimating TAR in the vast majority of patients untreated or with reperfusion treatment failure. On the other hand, we recruited 349 consecutive patients admitted for AIS, undergoing FAST-MRI based hyperacute stroke imaging. MRI with DWI and pcASL presented ACC of 97.4%, decreasing to 83.9%, when only hypoperfusion on pcASL was considered. In patients with perfusion deficit in anterior circulation territory, DWI confirmed the high reliability to identify the NVT in patients with favorable clinical-instrumental outcomes. Similarly, pcASL identified the area of total hypoperfusion, with an overestimation of TAR in patients with unfavorable outcome, but also a higher proportion of complete overlap with the final infarct. In conclusion, our study showed that CTP-Tmax maps for the detection of AIS achieved a sufficiently reliable ACC, but this was not optimal. In patients with perfusion deficit in anterior circulation territory, CTP-Tmax > 16s proved relatively reliable to identify the ischemic core area, with a tendency, however, to overestimate NVT compared to FIA, in patients with favorable clinical-instrumental outcomes and, in particular, in patients undergoing early recanalization. Similarly, CTP-Tmax > 9.5s proved reliable to identify the area of total hypoperfusion, but the degree of correlation with FIA was in favor of an overestimation of TAR in the vast majority of patients untreated or with reperfusion treatment failure. On the other hand, our study showed that DWI-pcASL for the detection of AIS achieved a highly reliable ACC. In patients with perfusion deficit in anterior circulation territory, DWI confirmed the high reliability to identify the NVT predicting FIA, in patients with favorable clinical-instrumental outcomes and, in particular, in patients undergoing early recanalization. Similarly, rCBF on pcASL proved reliable to identify the area of total hypoperfusion, with a higher proportion of cases showing a complete overlapping with the FIA, in the vast majority of patients untreated or with reperfusion treatment failure. However, the degree of correlation with FIA was in favor of an overestimation of TAR in many cases.