Magnetic resonance imaging of hypoxia in acute stroke compared with fluorine-18 fluoromisonidazole-positron emission tomography: A cross-validation study?

Abstract

Acute ischemic stroke results in an ischemic core surrounded by a tissue at risk, named the penumbra, which is potentially salvageable. One way to differentiate the tissues is to measure the hypoxia status. The purpose of the current study is to correlate the abnormal brain tissue volume derived from magnetic resonance-based imaging of brain oxygen saturation (S_t O₂ -MRI) to the fluorine-18 fluoromisonidazole ([¹⁸ F]FMISO) positron emission tomography (PET) volume for hypoxia imaging validation, and to analyze the ability of S_t O₂ -MRI to depict the different hypoxic tissue types in the acute phase of stroke. In a pertinent model of stroke in the rat, the volume of tissue with decreased S_t O₂ -MRI signal and that with increased uptake of [¹⁸ F]FMISO were equivalent and correlated (r = 0.706; p = 0.015). The values of S_t O₂ in the tissue at risk were significantly greater than those quantified in the core of the lesion, and were less than those for healthy tissue (52.3% ± 2.0%; 43.3% ± 1.9%, and 67.9 ± 1.4%, respectively). A threshold value for S_t O₂ of ≈60% as the cut-off for the identification of the tissue at risk was calculated. Tissue volumes with reduced S_t O₂ -MRI correlated with the final lesion (r = 0.964, p < 0.0001). The findings show that the S_t O₂ -MRI approach is sensitive for the detection of hypoxia and for the prediction of the final lesion after stroke. Once validated in acute clinical settings, this approach might be used to enhance the stratification of patients for potential therapeutic interventions.

Keywords: FMISO-PET; MRI; hypoxia; stroke; tissue oxygenation.