RESUMEN
INTRODUCTION: Severity of vascular damage of white matter may predict hemorrhagic transformation (HT). We assess the relationship between leukoaraiosis (LA) severity and the type of hemorrhagic transformation in elderly patients treated with thrombolysis. METHODS: We retrospectively analyzed the clinical data and pretreatment magnetic resonance imaging (MRI) of 180 consecutive ischemic stroke patients aged over 75 years. LA severity was graded according to the Fazekas scale, and acute diffusion-weighted-imaging (DWI) lesion volumes were semi-automatically outlined. Predictors of hemorrhagic infarction (HI) and parenchymal hemorrhage (PH) were identified using logistic regression analysis and exact multinomial logistic analysis. RESULTS: HT occurred in 31 patients (17 %). Baseline National Institute of Health Stroke Score (NIHSS; p = 0.008), severe LA (p = 0.02), and diffusion lesion volume (p = 0.02) were predictors of HT in univariable logistic regression. Adjusted to lesion volume and baseline NIHSS score, exact multinomial logistic analysis showed that severe LA was the only independent predictor of parenchymal hemorrhage (p = 0.03). CONCLUSION: In elderly patients, LA severity better predicts parenchymal hemorrhage than infarct size.