Comparison between MRI FLAIR vascular hyperintensity-DWI mismatch and perfusion based triage for thrombectomy in the late time window.

Background: The clinical impact of patient selection using FLAIR vascular hyperintensity (FVH)-diffusion-weighted imaging (DWI) mismatch for endovascular thrombectomy (EVT) in patients who have been symptomatic for over 6 h remains unclear. Herein, a retrospective study was conducted to compare the inter-rater reliability and clinical outcomes of patients selected for thrombectomy based on FVH-DWI mismatch with perfusion.

Methods: Patients with anterior-circulation large-vessel occlusion selected simultaneously with MRI and perfusion imaging in the late time window from a single-center retrospective study were categorized into EVT-applicable (FVH-DWI mismatch on MRI or perfusion imaging meeting the DEFUSE3 standards) and EVT-inapplicable groups based on MRI and perfusion imaging. The primary outcome was the 90-day functional independence rate. Safety outcomes encompassed symptomatic intracranial hemorrhage and mortality in 90 days. We assessed the consistency of the two profiles and compared the differences in functional independence rates of EVT patients among the EVT-applicable groups determined by MRI and perfusion.

Results: A total of 130 patients were enrolled, of which 114 were classified into the EVT-applicable group after triaging using MRI images. In this group, 96 patients underwent EVT, with 53 of them (55.2%) achieving functional independence. A total of 110 patients were divided into EVT-applicable group based on perfusion, among which 92 underwent EVT, with 49 of them (53.2%) achieving functional independence. The consistency of identifying EVT indication was moderate between two groups (κ = 0.42, 95% CI, 0.17-0.67). The functional independence rate was comparable between patients in the two EVT-applicable groups based on the two methods (55.2% vs. 53.2%,  = 0.789).

Conclusion: MRI triaging based on FVH-DWI mismatch showed moderate inter-rater reliability compared with perfusion-based triage and comparable efficacy in predicting clinical outcomes after EVT.