Tumor oxygenation imaging biomarkers using dynamic susceptibility contrast imaging for prediction of IDH mutation status in adult-type diffuse gliomas

Abstract

Objectives To evaluate the role of tumor oxygenation imaging parameters in predicting the isocitrate dehydrogenase (IDH) mutation status in adult-type diffuse gliomas. Methods This retrospective study included 296 patients with adult-type diffuse glioma (237 IDH-wildtype, 59 IDH-mutant). The normalized cerebral blood volume (nCBV), cerebral metabolic rate of oxygen values (CMRO2), capillary transit time heterogeneity, and oxygen extraction fraction (OEF) values from dynamic susceptibility contrast (DSC) imaging and ADC values from DWI were obtained from autosegmented tumor masks. Logistic analyses were performed in entire patients and in a subgroup of 46 patients (15.6%) without CE tumors. Results In entire patients, 10th percentile of CMRO2 (odds ratio [OR] = 0.64, p < 0.001) and 10th percentile of OEF (OR = 0.32, p < 0.001) independently predicted IDH mutation, along with age, frontal location, presence of CE tumor, and 90th percentile of nCBV, with an area under the curve (AUC) of 0.92 (95% confidence interval [CI] 0.88-0.95). In the subgroup of patients without CE tumors, 10th percentile of CMRO2 (OR = 0.58, p = 0.044) was an independent predictor for IDH mutation, along with age, and 10th percentile of ADC with an AUC of 0.94 (95% CI 0.83-0.99). Conclusion Tumor oxygenation parameters, including CMRO2 and OEF, may predict IDH mutation independently of previously known clinical and quantitative imaging data. Lower 10th percentile of CMRO2 and OEF predicts IDH mutation in entire tumors, while lower 10th percentile of CMRO2 predicts IDH mutation in patients without CE tumors.