Purpose: Symptomatic hemorrhagic transformation is a severe complication of acute ischemic stroke which occurs at a higher frequency after thrombolysis. The present study was designed to analyze whether early DWI can be used for predicting the risk of hemorrhagic transformation with clinical worsening in MCA stroke patients.
Materials and methods: Of 28 patients with a middle cerebral artery (MCA) infarct and proven MCA or carotid T occlusion on DWI and MR angiography performed within 14 hours after onset (mean 6.5 +/- 3.5 hours, median 5.2 hours), 4 developed hemorrhagic transformation with clinical worsening, while 24 did not. For the 2 groups, we compared admission NIHSS score, site of arterial occlusion, volume of DWI abnormalities, and several apparent diffusion coefficient (ADC) measurements: ADC(infarct) (mean ADC value of the whole infarct), ADC(core) (peak ADC decrease as calculated in a 57 mm(2) circular ROI, manually centered on the ischemic area with the lowest ADC value on the ADC maps), ADC(superficial) and ADC(deep). Discriminant function analysis was used to determine the most accurate predictors of symptomatic hemorrhagic transformation.
Results: The best predictor was the ADC(core) (F=5.34, p=2.9%, cut-off value=300 x 10(-6) mm(2)/s). This monovariate model allowed to correctly classify all 4 patients (ADC(core) 300 x 10(-6) mm(2)/s) with subsequent symptomatic hemorrhage, and 17 of the 24 patients without symptomatic hemmorrhage (ADC(core)>300 x 10(-6) mm(2)/s) (100% sensitivity, 71% specificity).
Conclusion: Although preliminary, these results suggest that a simple measurement of minimum ADC values within an acute MCA stroke could be useful in targeting those patients with a high risk of severe hemorrhagic transformation.