Early infarct FLAIR hyperintensity is associated with increased hemorrhagic transformation after thrombolysis


Correspondence: A. Kufner, International Graduate Program Medical Neuroscience, Charité – Universitätsmedizin Berlin, Charitéplatz 1, Berlin 10117, Germany (tel.: +4903450560137; fax: +4903450560932; e-mail: anna.kufner@charite.de).


Background and purpose

Absence of FLAIR hyperintensity within an acute infarct is associated with stroke onset <4.5 h. However, some patients rapidly develop FLAIR hyperintensity within this timeframe. We hypothesized that development of early infarct FLAIR hyperintensity would predict hemorrhagic transformation (HT) in patients treated with tissue plasminogen activator (tPA) < 4.5 h after onset.


Consecutive acute stroke patients treated with intravenous tPA <4.5 h after onset who had MRI before and 1 day after thrombolysis were included. Two raters (blind to HT) independently identified FLAIR hyperintensity with reference to the diffusion-weighted image (DWI) lesion. HT was assessed using T2* MRI at 24 h. Hemorrhagic infarction (HI) was defined as petechial HT without mass effect, and parenchymal hematoma (PH) as HT with mass effect. Multivariable logistic regression analysis for HT included FLAIR status, baseline National Institutes of Health Stroke Scale and DWI lesion volume, leukoaraiosis (Wahlund score), serum glucose and reperfusion.


Of 109 patients, 33 (30%) had acute FLAIR hyperintensity. HT occurred in 17 patients (15.6%; 15 HI, 2 PH). HT was more common in FLAIR-positive patients than FLAIR-negative patients (33.3% vs. 9.2%, P = 0.009). Median time-to-scan and median time-to-thrombolysis did not differ significantly between patients with HT and without [97 IQR(68, 155) vs. 90 IQR(73, 119), P = 0.5; 120 IQR(99, 185) vs. 125 IQR(95, 150), P = 0.6, respectively]. In multivariable analysis, only FLAIR hyperintensity was independently associated with HT after thrombolysis (OR 18; 95% CI 2–175, P = 0.013).


Early development of FLAIR hyperintensity within the area of diffusion restriction is associated with increased risk of HT after thrombolysis in acute stroke patients.