Background: Imaging profiles such as mismatch ratio and core size have been used to select patients most likely to benefit from thrombectomy. Core blood-brain barrier (BBB) disruption (cBBBD), when severe, has been associated with an increased risk of bleeding complications. However, there is limited data on the role of cBBBD as an imaging profile for patient selection. Here we compare cBBBD with other established imaging profiles. Methods: Pretreatment MRI scans for a cohort of patients with large vessel occlusion who were subsequently transferred for thrombectomy were included in this study if they had perfusion weighted imaging of sufficient quality. RAPID software was use to identify the core and Tmax regions of interest (ROIs). Perfusion source images were processed to generate BBB maps based on contrast leakage. The variable cBBBD was calculated as the average of all leaky voxels in the core ROI. cBBBD was treated as a dependent variable and compared with core volume, Tmax volumes (Tmax6, Tmax10), mismatch ratio (mmratio) and hypoperfusion intensity ratio (HIR) as independent variables using linear regression. Results: The study included 297 patients with a median age 73 of which 50% were women. The median core volume was 16 mL, the median Tmax6 volume was 92 mL, and the median mmratio was 5.6. Comparing cBBBD with volumes the strongest correlation was with core (b=0.023, CI 0.018:0.029, r 2 =0.18, p<0.001, n=297), followed by Tmax10 (b=0.022, CI 0.016:0.028, r 2 =0.15, p<0.001, n=297) and then Tmax6 (b=0.014, CI 0.010:0.019, r 2 =0.13, p<0.001, n=297). HIR also correlated (b=4.67, CI 3.38:5.97, r 2 =0.15, p<0.001, n=293), however the strongest correlation was with mmratio (b=-0.030, CI -0.035:-0.024, r 2 =0.27, p<0.001, n=297). Conclusions: Larger ischemic lesions were associated with more severe BBB disruption in the core infarct. However, a more favorable mmratio was strongly correlated with less BBB disruption. BBB imaging maybe useful as an additional imaging profile when selecting patients for thrombectomy. Further studies are needed.
Leigh et al. (Thu,) studied this question.