Diffusion-Weighted-Imaging infarct volume measurement tools show discrepancies leading to diverging thrombectomy decisions

Background and purpose: Recent clinical trials demonstrated the benefit of thrombectomy beyond 6h based on the automated measurement of infarct volume exclusively with the RAPID software. We aimed to compare eight tools commonly used for the measurement of infarct volume and see whether they would lead to similar thrombectomy decisions based on the Diffusion-weighted-imaging or computerized-tomography-perfusion Assessment with clinical mismatch in the triage of Wake-up and late-presenting strokes undergoing Neurointervention with Trevo (DAWN) trial imaging inclusion criteria.

Materials and methods: The diffusion-weighted-imaging (DWI) infarct volume of 36 patients was measured with 3 automated tools (including RAPID) and 5 non-automated tools. The agreement for the measurements of DWI infarct volume and the resulting thrombectomy decisions were assessed with intraclass correlation coefficient (ICC) and Fleiss' Kappa (K) statistics.

Results: The correlation for the measurement of DWI infarct volume between all 9 tools was excellent (ICC>0.8). After dichotomization, agreement was substantial for any of the cut-points used in DAWN trial. Discrepancies involving at least one of the tools for thrombectomy decisions based on DAWN criteria occurred in one third of cases. Compared with RAPID, the use of any other tool for treatment decision based on DAWN criteria would have led to contradictory decisions in 6% to 19% of cases.

Conclusion: There are several currently available tools for the measurement of DWI infarct volume with excellent correlation. Despite the high agreement demonstrated in our study, frequent discrepancies between measurements in some dichotomized configurations led to frequent diverging thrombectomy decisions when applying DAWN criteria.