Introduction: CT Perfusion (CTP) has become an essential tool in determining candidates for endovascular stroke therapy (EST), particularly in later time windows. However, the reliability of CTP to define infarct core may vary based on time of onset and other clinical factors.
Methods: From our prospective institutional registry, we identified patients between 1/2014 - 3/2017 that underwent EST with successful reperfusion (TICI 2b/3), pre-procedure CTP and 24hr MRI. Source CTP data were analyzed using OleaSphere (La Ciotat, France) rCBF<30% and RAPID (IschemiaView, CA) rCBF<30%. Mismatch (MM) was defined as >10cc disparity between CTP core and 24hr DWI. Multivariate logistic regression assessed factors associated with MM (p<0.05 significant, p<0.10 trend).
Results: Among 109 patients (60 with RAPID) studied, mean age was 63±15, median NIHSS 19[15.5-22], median ASPECTS 8 [7-10], and occlusion location ICA in 19 (17%) and M1 in 90 (83%). CTA collateral grades were 0-2a in 70 (64.2%) and 2b/3 in 36 (35.8%). Median time from onset to CT (LKW-CT) was 128 [77-212] min, and from CTP to recanalization (CTR) of 142.5[111.5-169.5] min. MM occurred in 84 (77%, 8(7.3%) infarct shrinking and 76 (69.7%) infarct growth) using Olea and 38 (63.3%, 1(1.7%) infarct shrinking and 37(61.7%) infarct growth) using RAPID. In adjusted multivariable regression, MM was significantly associated with larger CTP core and lower NIHSS with RAPID (ORs 1.06, 0.87) and larger CT core and poor collaterals with Olea (ORs 1.04, 0.35). Infarct growth was significantly associated with core size, NIHSS and a trend towards LKW-CT in RAPID (ORs 1.05, 0.88, 1.78 and Figure), and collaterals in Olea (OR 0.31). CTR was not associated with infarct growth.
Conclusion: MM between CTP core and MRI infarct occurs frequently, with both infarct growth and shrinkage. The likelihood of infarct growth appears to be related to LKW-CT and collaterals, suggesting that CTP thresholding in later time windows may need to adjusted.