Abstract 4018: Pretreatment MRI Markers May Predict Hemorrhagic Transformation in Patients Treated with IV tPA

Abstract

Objectives: Multiple MRI measurements have been investigated as possible markers of post-treatment imaging pathology. The objectives of this study were to understand how well measurements from pre-treatment MRI data including ADC volume, DWI volume, MTT volume, Tmax volume, ADC ratio, and CBV ratio predicted post-treatment hemorrhagic transformation. Methods: Patients were selected from the Lesion Evolution of Stroke and Ischemia On Neuroimaging (LESION) database if they: had an acute ischemic stroke, were treated with IV tPA only, had a pre-treatment MRI with evaluable diffusion (DWI) and perfusion (PWI) weighted imaging, and had measurable lesion volumes on both ADC and Tmax images. All ADC and Tmax maps were calculated in Perfscape™ (Olea Medical). A trained rater applied an ADC threshold of 615 mm 2 /seconds and Tmax threshold of >6 seconds using Neuroscape™ (Olea Medical) to produce the ADC and Tmax volumes that included any voxels across affected slices within the visually confirmed area of ischemia. All CBV maps were calculated from the PWI in PMA™ (ASIST-Japan). The regions of interest (ROIs) from the ADC volumes were used in MIPAV™ (National Institutes of Health) to measure the average voxel intensity (AVI) of the ADC and CBV lesions in both the ischemic (AVII) and contralateral hemispheres (AVIC). Logistic regression with covariates of age, baseline NIHSS, DWI volume, MTT volume, ADC volume, Tmax volume, ADC ratio and CBV ratio was performed. Results: A total of 74 patients met the study criteria with mean age of 67.9 years (range 32-95), median NIHSS of 12 (range 1-40), median DWI volume of 28.21 ml (IQR 11.05-66.52), and median MTT volume of 147.69 ml (IQR 52.9-250). Fifty patients had hemorrhagic transformation status available at 2 and\or 24 hours post-treatment. Fifty-four percent (n=27) had some degree of hemorrhagic transformation at 2 or 24 hours post treatment. The pre-treatment volumes reported as median (IQR 25-75) for the hemorrhagic transformation cohort (n=27) were 55.22 ml (20.32-70.53), 181.75 ml (65.36-252.88), 4.52 ml (1.07-7.39), 51.19 ml (32.67-118.23) for the DWI, MTT, ADC, Tmax and 0.52 (0.5-0.55) for the ADC ratios. The pre-treatment volumes for the non-hemorrhagic transformation cohort (n=23) were 26.07 ml (6.29-66.96), 133.78 ml (58.7-209.09), 5.27 ml (0.74-8.83), 61.33 ml (26.08-85.86) for the DWI, MTT, ADC, Tmax volumes and 0.53 (0.47-0.63) for the ADC ratios. However the CBV ratios using median (IQR 25-75) for the hemorrhagic transformation cohort versus the non-hemorrhagic transformation cohort were 0.27 (0.14-0.42) versus 0.46 (0.17-0.70). The logistic regression analysis demonstrated that patients with lowered CBV ratio on pre-treatment MRI significantly predicted hemorrhagic transformation (p<0.019). Conclusions: Lowered CBV ratio on pre-treatment MRI demonstrates potential as a predictor of hemorrhagic transformation post-treatment.