Application of Cluster Analysis of Time Evolution for Magnetic Resonance Imaging -Derived Oxygen Extraction Fraction Mapping: A Promising Strategy for the Genetic Profile Prediction and Grading of Glioma

Abstract

Background: The intratumoral heterogeneity of oxygen metabolism and angiogenesis are core hallmarks of glioma, unveiling that genetic aberrations associated with magnetic resonance imaging (MRI) phenotypes may aid in the diagnosis and treatment of glioma. Objective: To explore the predictability of MRI-based oxygen extraction fraction (OEF) mapping using cluster analysis of time evolution (CAT) for genetic profiling and glioma grading. Methods: Ninety-one patients with histopathologically confirmed glioma were examined with CAT for quantitative susceptibility mapping and quantitative blood oxygen level–dependent magnitude-based OEF mapping and dynamic contrast-enhanced (DCE) MRI. Imaging biomarkers, including oxygen metabolism (OEF) and angiogenesis [volume transfer constant, cerebral blood volume (CBV), and cerebral blood flow], were investigated to predict IDH mutation, O6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation status, receptor tyrosine kinase (RTK) subgroup, and differentiation of glioblastoma (GBM) vs. lower-grade glioma (LGG). The corresponding DNA sequencing was also obtained. Results were compared with DCE-MRI using receiver operating characteristic (ROC) analysis. Results: IDH1-mutated LGGs exhibited significantly lower OEF and hypoperfusion than IDH wild-type tumors (all p < 0.01). OEF and perfusion metrics showed a tendency toward higher values in MGMT unmethylated GBM, but only OEF retained significance (p = 0.01). Relative prevalence of RTK alterations was associated with increased OEF (p = 0.003) and perfusion values (p < 0.05). ROC analysis suggested OEF achieved best performance for IDH mutation detection [area under the curve (AUC) = 0.828]. None of the investigated parameters enabled prediction of MGMT status except OEF with a moderate AUC of 0.784. Predictive value for RTK subgroup was acceptable by using OEF (AUC = 0.764) and CBV (AUC = 0.754). OEF and perfusion metrics demonstrated excellent performance in glioma grading. Moreover, mutational landscape revealed hypoxia or angiogenesis-relevant gene signatures were associated with specific imaging phenotypes. Conclusion: CAT for MRI-based OEF mapping is a promising technology for oxygen measurement and along with perfusion MRI can predict genetic profiles and tumor grade in a non-invasive and clinically relevant manner. Clinical Impact: Physiological imaging provides an in vivo portrait of genetic alterations in glioma and offers a potential strategy for non-invasively selecting patients for individualized therapies.