We hypothesized that automated assessment of collaterals on computed-tomography (CT) perfusion can predict the rate of infarct growth during transfer from a primary to a comprehensive stroke center for endovascular stroke treatment. We identified consecutive patients (N=28) and assessed their collaterals based on the hypoperfusion intensity ratio (HIR) prior to transfer. Infarct growth rate was strongly correlated with HIR (r= 0.78, p<0.001). ROC analysis identified an HIR of =0.5 as optimal for predicting infarct growth. Patients with HIR = 0.5 had a median infarct growth rate of 10.1 mL/h (IQR: 6.4-18.4) compared with 0.9 mL/h (IQR: 0-2.8), p< 0.001) in patients with a HIR < 0.5. Patients with an HIR above = 0.5 had an 83% probability of significant core growth, whereas patients with HIR <0.5 had an 88% probability of core stability. These preliminary data have the potential to guide decision making regarding whether repeat brain imaging should be performed after transfer to a comprehensive stroke center. This article is protected by copyright. All rights reserved.
View details for PubMedID 30168180