Abstract TP38: Core Growth Speed as a Marker of Collateral Circulation Status in Acute Ischemic Stroke with Known Onset Time

Abstract

Objective: We aimed to evaluate the parameter core growth speed (CGS) as a marker of collateral circulation status (CC) in acute ischemic stroke, and to compare it with other brain perfusion-derived markers of collateral capacity. Methods: We retrospectively studied acute ischemic stroke patients who were evaluated with urgent computed tomography perfusion (CTP) and CT angiography. Inclusion criteria comprised known time of onset and anterior circulation proximal occlusion. Collateral circulation was assessed on CTP-source images and rated as poor (0-1) vs. good (2-3) following a previously published scale. CTP maps were computed using Neuroscape 2.0 software by Olea Medical. Infarct core volume was calculated as the brain tissue with >70% reduction in cerebral blood flow (CBF) as compared to the unaffected side. CGS was obtained by dividing core volume by the time from stroke onset to CTP acquisition. Relative cerebral blood volume (rCBV), relative CBF, and hypoperfusion index ratio (HIR = Tmax>10s/Tmax>6s) were used as comparators. Results: We included 41 patients (mean age 71 years; median NIHSS 17; median onset-CTP time 150 minutes). We observed a positive correlation between CGS and HIR (ρ= 0.517 p< 0.001), and negative correlations between rCBV and CGS (ρ= -0.669 p<0.0001), and rCBF and CGS (ρ= -0.749 p<0.0001). Collateral circulation was categorized as poor or good in 15 and 26 patients respectively. A gradual descend in CGS was seen as CC improved (p=0.0005). A logistic regression model adjusted by rCBV, rCBF and HIR identified CGS as independently associated with CC. The association of CGS with good CC in a ROC curve was highly significant (p=0.002, area under the curve 0.8). Conclusion: Core growth speed is robustly associated with collateral circulation status. This parameter can be directly obtained from infarct core volume without the need to process other perfusion or angiographic images, if the time of onset is well known.