Introduction: Ischemic injury on MRI is defined acutely by measuring the diffusion of water alone. Chemical exchange saturation transfer imaging (CEST) allows measurement of intracellular pH, and changes in protein structure and mobility using Nuclear Overhauser Enhancement (NOE). When identifying treatment targets, changes to protein integrity have the potential to complement diffusion-weighted imaging (DWI) in distinguishing irreversible from reversible intracellular processes that are amenable to therapeutic intervention.
Methods: Patients with non-lacunar ischemic stroke underwent serial MRI over 1 month. The imaging protocol included single slice CEST imaging. NOE was quantified using a 3-pool exchange model relative to the contralateral hemisphere (rNOE*). DWI, ASL perfusion imaging, T1-weighted and FLAIR imaging were used to define the tissue outcome: ischemic core, infarct growth, and oligemic tissue that survived. Infarct growth was classified as early or late (before or after 24 hours). Images and masks were registered to CEST native image space for voxelwise and patient-level serial analyses.
Results: 30 patients were included in the analysis. Within 6 hours rNOE* in early infarct growth was significantly lower than ischemic core and late infarct growth (p<0.001 and p<0.001, see Figure). In patient-level analyses, rNOE* dropped over the initial 24 hours in the ischemic core and early infarct growth regions, but not in regions of late infarct growth or oligemia.
Conclusion: rNOE* showed a discrete temporal profile in tissue that infarcted by 24 hours in contrast to that which survived or infarcted later. This is additive information to that provided by routine acute stroke MRI at presentation, and demonstrates that the imaging of intracellular protein structure and mobility may have a role alongside DWI in helping to predict tissue fate. Further work is required to understand the potential of NOE as a modifiable imaging biomarker.