Advances in hypoxia-modulated blood oxygen level-dependent MRI: a narrative review of physiological principles and applications for glioblastoma imaging.

Glioma imaging remains constrained by the limited ability of conventional MRI to capture biologically relevant tumor features beyond blood-brain barrier disruption. Hypoxia-modulated blood oxygen level-dependent (BOLD) MRI has emerged as a potential contrast-agent-free approach to probe oxygenation-related tumor physiology, using deoxyhemoglobin as an endogenous susceptibility contrast mechanism. By altering blood oxygenation during acquisition, the technique generates signal changes that may provide physiological information beyond structural imaging and resting perfusion measures. We performed a narrative review of 29 human and animal studies using hypoxic oxygen challenges during BOLD MRI. Substantial heterogeneity was identified in gas delivery, carbon dioxide control, MRI acquisition and analytical approaches. A recurring methodological issue was carbon dioxide control, since carbon dioxide is the dominant vasoactive stimulus and hypoxia-induced hyperventilation with secondary hypocapnia can confound BOLD signal interpretation. Preclinical brain tumor studies suggest sensitivity to spatial heterogeneity between viable and necrotic tumor regions. Early human glioblastoma studies support the feasibility and tolerability of hypoxia-modulated BOLD acquisitions and show spatially heterogeneous signal- and perfusion-related maps without exogenous contrast. Current evidence supports hypoxia-modulated BOLD MRI as a promising translational imaging approach, but clinical data remain limited and prospective validation in glioblastoma cohorts is required.