Introduction: The efficacy of endovascular thrombectomy in an extended time window for acute ischemic stroke patients with Target Mismatch (TM) on perfusion imaging was shown in a recent study and the ongoing DEFUSE-3 trial is studying thrombectomy in a 6-16 hour window for TM patients. A limitation of TM is that perfusion imaging is not widely available. We sought to identify a tool to predict TM based on clinical factors and CT angiogram (CTA) imaging, which is available at most hospitals.
Methods: We reviewed acute ischemic stroke patients from 2010-2014 with proximal middle cerebral artery occlusion, CTA and CT perfusion (CTP) at hospital admission. TM was identified on CTP using the Olea Sphere volumetric analysis software with Bayesian deconvolution. TM was defined by the DEFUSE-3 criteria. ASPECTS was derived from the non-contrast CT head and the CTA source images (CTA-ASPECTS). Two collateral scores were derived from CTA source images.
Results: 61 patients met inclusion criteria. The mean±SD age was 61±18 years and 61% were male. Mean NIH Stroke Scale (NIHSS) was 14.1±8.0 and median (IQR) follow-up modified Rankin Scale was 3 (1,6). TM was present in 35/61 (57%), who had lower mRS at follow-up (z=3.5, p<0.001). The predictor variables are shown in Table 1. The best combination of predictors was CTA-ASPECTS >4 and NIHSS <16, which had a sensitivity of 80% and specificity of 85% for TM (Figure 1).
Discussion: We report a reliable, accessible, and clinically useful tool for predicting TM. This score warrants further study as a tool to guide transfer decisions from primary or secondary stroke centers to tertiary centers where endovascular intervention would be possible for selected patients.