2 Imaging triage of late window patients with acute ischemic stroke. A comparative study using multi-phase CT angiography vs CT perfusion

Background Current guidelines recommend the use of perfusion imaging for selection of patients for endovascular thrombectomy (EVT) beyond six hours from onset. The role of collateral imaging in this time window is not established. Methods We used data from a prospective multi-center observational study where all stroke patients with suspected large vessel occlusion underwent imaging with single- and multi-phase CT angiography (mCTA) as well as CT perfusion. For this analysis, we only included patients presenting beyond six hours from onset/last known well time. Two blinded reviewers judged patients’ eligibility for EVT using published collateral imaging (mCTA), compared to CT perfusion (using DAWN and DEFUSE-3 trials) selection criteria. All perfusion images were processed using an automated commercial software. The outcomes of patients eligible for EVT using mCTA, DAWN, or DEFUSE-3 criteria were compared using multivariable logistic regression modeling. Model predictive characteristics were assessed using c-statistic for the receiver operating curve, Akaike information criterion (AIC), and Bayesian information criterion (BIC). Results Of 614 patients, 86 patients presented beyond six hours from onset/last known well (median 9.6 hours, IQR 4.1 hours). Median age was 71 years (IQR 14 years), 48.8% were females, median baseline NIHSS was 12 (IQR=11). Thirty-five patients (40.7%) received EVT of which good functional outcome (90 day modified Rankin scale 0–2) was achieved in 47%. Collateral-based imaging paradigms significantly modified the treatment effect of EVT on clinical outcome i.e. 90-day mRS 0–2 (P interaction=0.007). The mCTA-based regression model best fit the data for 90-day outcome (C statistic 0.86, 95% CI 0.77 to 0.94) and was associated with least information loss (AIC 95.7, BIC 114.9) when compared to CTP based models. Perfusion imaging paradigm using DEFUSE-3 criteria had better predictive properties than the DAWN trial criteria. Conclusion Collateral-based imaging paradigm using mCTA compares well with CTP in selecting patients for EVT in the late time window. Disclosures M. Almekhlafi: None. W. Kunz: None. R. McTaggart: None. M. Jayaraman: None. M. Najm: None. S. Ahn: None. E. Fainardi: None. M. Rubiera: None. A. Khaw: None. A. Zini: None. M. Hill: None. A. Demchuk: None. M. Goyal: None. B. Menon: None.