Diffusion-weighted MRI and proton MR spectroscopic imaging in the study of secondary neuronal injury after intracerebral hemorrhage.

BACKGROUND AND PURPOSE Cerebral ischemia has been proposed as contributing mechanism to secondary neuronal injury after intracerebral hemorrhage (ICH). Possible tools for investigating this hypothesis are diffusion-weighted (DWI) and proton magnetic resonance spectroscopic imaging ((1)H-MRSI). However, magnetic field inhomogeneity induced by paramagnetic blood products may prohibit the application of such techniques on perihematoma tissue. We report on the feasibility of DWI and (1)H-MRSI in the study of human ICH and present preliminary data on their contribution to understanding perihematoma tissue functional and metabolic profiles. METHODS Patients with acute supratentorial ICH were prospectively evaluated using DWI and (1)H-MRSI. Obscuration of perihematoma tissue with both sequences was assessed. Obtainable apparent diffusion coefficient (Dav) and lactate spectra in perihematoma brain tissue were recorded and analyzed. RESULTS Nine patients with mean age of 63.4 (36 to 87) years were enrolled. Mean time from symptom onset to initial MRI was 3.4 (1 to 9) days; mean hematoma volume was 35.4 (5 to 80) cm(3). Perihematoma diffusion values were attainable in 9 of 9 patients, and (1)H-MRSI measures were obtainable in 5 of 9 cases. Dav in perihematoma regions was 172.5 (120.0 to 302.5)x10(-5) mm(2)/s and 87.6 (76.5 to 102.1)x10(-5) mm(2)/s in contralateral corresponding regions of interest (P=0.002). One patient showed an additional area of reduced Dav with normal T(2) intensity, which suggests ischemia. (1)H-MRSI revealed lactate surrounding the hematoma in 2 patients. CONCLUSIONS DWI and (1)H-MRSI can be used in the study of ICH patients. Our preliminary data are inconsistent with ischemia as the primary mechanism for perihematoma tissue injury. Further investigation with advanced MRI techniques will give a clearer understanding of the role that ischemia plays in tissue injury after ICH.