Abstract LB-362: A six-gene classifier stratifies low-grade glioma patients with wild-type IDH

Gliomas are the most common primary brain tumors presenting a heterogeneous group of malignancies. Particularly, there is a significant heterogeneity in outcomes of patients with grade II and grade III gliomas. There has been recent progress in reclassifying glomas by integrating molecular signatures such as IDH mutation status, 1p/19q co-deletion and alternative telomer maintenance to the established histopathological classification scheme. Patients with IDH wild-type (WT) gliomas are found to have the worst prognosis. Currently there is no prognostic or predictive biomarker within this group of patients. To this end, the goal of the current study was to develop a gene classifier for identifying a prognostic/predictive biomarker model for grade II and grade III gliomas with IDH WT, which would have the potential of identifying patient groups with distinct degrees of risk for disease progression, treatment options, and survival outcome. Our approach was to identify a robust list of candidate genes using multiple datasets and create a single score classifier based on the expression of identified genes. We analyzed multiple datasets comprising a total of 756 glioma patients using a meta-analysis approach which identified six prognostic genes. A linear model consisting of these six genes was trained in an institutional cohort by employing RNAseq and qRT-PCR techniques to measure the gene expression and validated in an independent grade II/III glioma set (TCGA LGG). The classifier performed significantly in both the institutional cohort of 85 grade II/III glioma patients (HR, 29.94; 95% CI, 3.96-226.36; p = 0.001) and in the TCGA LGG dataset consisting of 416 glioma patients (HR, 4.03; 95% CI, 1.82-8.92; p = 0.001, for low versus high risk group). Multivariate Cox regression analysis validated the risk classifier independent of other clinical factors. After analyzing patient populations separately, based on IDH mutation status, we discovered that the classifier performed with high significance in the patients with IDH WT both in the institutional cohort and the TCGA LGG dataset (HR, 1.93; 95% CI, 1.24-3.02; p = 0.004 and HR, 1.99; 95% CI, 1.38-2.87; p Citation Format: Alexander Huebner, Oliver Oehlke, Tareq A. Juratli, Wei Meng, Ahmed Ibrahim, Bin Li, Daniel Erny, Pierre Robe, Dietmar Krex, Gabriele Schackert, Marco Prinz, Anca Grosu, Arnab Chakravarti. A six-gene classifier stratifies low-grade glioma patients with wild-type IDH. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr LB-362.