Comparison of stroke volume evolution on diffusion-weighted imaging and fluid-attenuated inversion recovery following endovascular thrombectomy

Background To compare the evolution of the infarct lesion volume on both diffusion-weighted imaging and fluid-attenuated inversion recovery in the first five days after endovascular thrombectomy. Methods We included 109 patients from the CRISP and DEFUSE 2 studies. Stroke lesion volumes obtained on diffusion-weighted imaging and fluid-attenuated inversion recovery images both early post-procedure (median 18h after symptom onset) and day 5, were compared using median, interquartile range, and correlation plots. Patients were dichotomized based on the time after symptom onset of their post procedure images (18h vs. <18h), and the degree of reperfusion (on Tmax>6s; 90% vs.<90%). Results Early post-procedure, median infarct lesion volume was 19ml [(IQR) 7-43] on fluid-attenuated inversion recovery, and 23ml [11-64] on diffusion-weighted imaging. On day 5, median infarct lesion volume was 52ml [20-118] on fluid-attenuated inversion recovery, and 37ml [16-91] on diffusion-weighted imaging. Infarct lesion volume on early post-procedure diffusion-weighted imaging, compared to fluid-attenuated inversion recovery, correlated better with day 5 diffusion-weighted imaging and fluid-attenuated inversion recovery lesions (r=0.88 and 0.88 vs. 0.78 and 0.77; p<0.0001). Median lesion growth was significantly smaller on diffusion-weighted imaging when the early post-procedure scan was obtained 18h post stroke onset (5ml [-1-13]), compared to <18h (13 ml [2-47]; p=0.03), but was not significantly different on fluid-attenuated inversion recovery (18h: 26ml [12-57]; <18h: 21 ml [5-57]; p=0.65). In the <90% reperfused group, the median infarct growth was significantly larger for diffusion-weighted imaging and fluid-attenuated inversion recovery (diffusion-weighted imaging: 23ml [8-57], fluid-attenuated inversion recovery: 41ml [13-104]) compared to 90% (diffusion-weighted imaging: 6ml [2-24]; p=0.003, fluid-attenuated inversion recovery: 19ml [8-46]; p=0.001). Conclusions Early post-procedure lesion volume on diffusion-weighted imaging is a better estimate of day 5 infarct volume than fluid-attenuated inversion recovery. However, both early post-procedure diffusion-weighted imaging and fluid-attenuated inversion recovery underestimate day 5 diffusion-weighted imaging and fluid-attenuated inversion recovery lesion volumes, especially in patients who do not reperfuse.