Objective: To evaluate the use of fluid-attenuated inversion recovery (FLAIR) imaging as surrogate marker of lesion age within the first 6 hours of ischemic stroke.
Methods: e analyzed FLAIR and diffusion-weighted imaging (DWI) sequences performed within 6 hours of symptom onset in 120 consecutive patients with ischemic stroke with known symptom onset. The visibility of acute ischemic lesions on FLAIR images was judged in two steps (on FLAIR alone and with knowledge of DWI) and compared with DWI.
Results: egative FLAIR in the case of positive DWI allocated ischemic lesions to a time window 3 hours or less with a high specificity (0.93) and a high positive predictive value (0.94), whereas sensitivity (0.48) and negative predictive value (0.43) were low. Lesion visibility on FLAIR images alone (35.6%) and with knowledge of DWI (62.5%) was lower than on DWI (97.1%). The sensitivity of FLAIR increased with increasing time from symptom onset from 27.0/50.0% <or= 3 hours to 56.7/93.3% after 3 to 6 hours (FLAIR alone/with knowledge of DWI). Multivariate regression analysis spotted longer time from symptom onset and larger size of the ischemic lesion as independent predictors of lesion visibility on FLAIR images.
Interpretation: "mismatch" between positive DWI and negative FLAIR allows the identification of patients that are highly likely to be within the 3-hour time window. Within the first 6 hours of stroke, the sensitivity of FLAIR sequences for acute ischemic lesions increases with time from symptom onset elapsing, approximating 100% after 3 to 6 hours.