Metabolic penumbra of brain infarction determined by PET and MR

The concept of the ischemic penumbra is recently in confusion with using the various types of imaging modality, which have been introduced into the diagnosis of the acute cerebral infarction. Although perfusion and diffusion of magnetic resonance imaging (MRI) have demonstrated the tissue at risk in the acute human ischemic stroke, the target of the imminent therapy could not be clearly determined by this method because of the unpredictability of the infarct growth in the mismatch area within several days after onset. We prospectively investigated relative cerebral blood flow (relCBF), relative cerebral metabolic rate of oxygen (relCMRO2) and relative cerebral oxygen extraction fraction (relOEF) by means of positron emission tomography (PET) in the perfusion–diffusion mismatch of MRI in 11 patients with acute embolic infarction within 6 h of onset. We applied H215O and 15O2 count-based PET to the present study, since the total examination time was less than 30 min, and the relative counts in the regions of interest to those in the reference regions were well correlated with the quantitative relative CBF, CMRO2 and OEF. The count-based relative PET parameters and the apparent diffusion coefficient (ADC) of water were estimated in the 3 areas of the perfusion–diffusion mismatch, namely, in the ischemic core (abnormal diffusion initially and on day 3), an evolving infarct (normal diffusion initially, but abnormal on day 3) and the periinfarct (normal diffusion initially and on day 3). The mean relCBF and relCMRO2 in an evolving infarct (0.42 ± 0.07 and 0.53 ± 0.07) were significantly less than those in the periinfarct area (0.57 ± 0.10 and 0.76 ± 0.10, p < 0.05 and p < 0.01, respectively). The best cutoff value to discriminate between the evolving infarct and the periinfarct was 0.49 for relCBF (p < 0.0001) and 0.62 for relCMRO2 (p < 0.0001). The mean relOEF for the evolving infarct and the periinfarct demonstrated a significant increase compared with unity (p < 0.05 and p < 0.01, respectively). The mean lesion-to-contralateral ratios of the quantitative ADC (relADC) for the ischemic core, evolving infarct and periinfarct were 0.66 ± 0.05, 0.93 ± 0.09 and 0.95 ± 0.09, respectively. There was no significant difference in relADC between the evolving infarct and the periinfarct. These results revealed several interesting pathophysiology in the acute ischemic human brain; (1) the volume expansion of a brain infarction during the first 3 days was preceded by a reduction in the cerebral oxygen metabolism as early as 6 h after onset, (2) disturbed cerebral oxygen metabolism in the evolving infarct was not associated with cytotoxic edema, indicating that adenosine triophosphate synthesis is still maintained in the area as early as 6 h after onset, (3) misery perfusion was observed even in the periinfarct area, which was a different phenomenon known in the chronic cerebrovascular disease. Although the various methodological limitations included in the present study, we elucidated ‘metabolic penumbra’ in the acute embolic cerebral infarction, which may become a critical target for the therapy by early reperfusion of the occluded arteries and/or neuroprotective agents to reduce the infarct growth of brain infarction.