Precise Characterization of the Penumbra Revealed by MRI: A Modified Photothrombotic Stroke Model Study

AIMS To precisely characterize the penumbra by MRI based on a modified photothrombotic stroke mouse model. METHODS The proximal middle cerebral artery was occluded by a convenient laser system in conjunction with an intravenous injection of Rose Bengal in mice. And the suture MCAO model was performed in seven mice as a comparison of the reproducibility. One hour after occlusion, the penumbra was defined in six random photothrombotic stroke mice by mismatch between perfusion-weighted imaging and the apparent diffusion coefficient map on a home-made workstation. After imaging, three random mice of them were chosen to perform the reperfusion surgery. And the other three mice were sacrificed to stain for several potential penumbra markers, such as c-fos and heart shock protein 90. In the remaining mice, the evolution of the lesions was detected on the apparent diffusion coefficient map, diffusion-weighted imaging and T2-weighted imaging at 1, 3, 6, 12 and 24 hours. After evaluating the neurological deficit scores, the brains were sectioned and stained by triphenyltetrazolium chloride and Nissl. RESULTS The mice subjected to photothrombosis showed significant behavioral deficits. One hour after occlusion, the low perfusion areas on the perfusion-weighted imaging interlaced with the hypointense areas on the apparent diffusion coefficient map, demonstrating that the penumbra was located both surrounding and inside the lesions. This phenomenon was subsequently confirmed by the c-fos and heart shock protein 90 staining. The final T2-weighted images of the mice subjected to the reperfusion surgery were also consistent with the penumbra images at one hour. At early stages, the lesions were clearly identified on the apparent diffusion coefficient map; the volumes of the lesions on the diffusion-weighted imaging and T2-weighted imaging did not reach a maximum until 12 hours. The coefficient of variation (CV) of the final lesions in the photothrombotic stroke mice was 21.7% (0.08 of 0.37) on T2-weighted imaging and 27.8% (0.10 of 0.35) on triphenyltetrazolium chloride, representing a high reproducibility (n = 7). While the CV of the lesions in the MCAO stroke mice was only 70% (0.24 of 0.34, n = 4). CONCLUSIONS This study has provided a precise imaging definition of the penumbra based on a reproducible photothrombotic stroke mouse model.