Abstract LB-169: Comprehensive analysis of the expression of eight genes as a prognostic marker in head and neck cancer

Head and neck cancer (HNC) is the 6th most common cancer worldwide, and the survival rate of HNC is about 50%. There has been no significant improvement in survival in HNC for over 50 years. Many HNC cases are not identified until an advanced stage. Thus, prevention, early detection, and better treatment are critical factors to reduce morbidity and mortality in HNC. Genome-wide next generation sequencing techniques allow investigation of patient-specific molecular signatures that may provide insight into the design of customized cancer therapies. Our previous studies have identified critical genetic signatures for HNC initiation and progression, and these signatures are strongly associated with progression-free survival (PFS) of HNC patients. To improve current molecular markers, a survival predictive model was built using ElasticNet Cox regression model with different sets of molecular markers (313 unique genes) from our three published studies and TCGA RNAseq expression data (n = 260). Deregulated expression of eight genes (FADD, PPFIA1, RYK, SMARCA2, SNX10, DCAF13, C18orf21 and THBS1) was identified and showed strong association with PFS. This eight-gene prediction model was validated by an independent oral squamous cell carcinoma (OSCC) specific mortality dataset from MD Anderson Cancer Center (n = 71 OSCC, AUC = 0.857). Next, we analyzed the association between deregulated expression of these eight genes and the top 5% (n = 420) most commonly mutated genes in HNCs (n = 260) via pairwise Pearson correlation coefficient with p value. This analysis revealed that deregulated expression of these eight genes is strongly associated with mutations in TP53 and NOTCH1. This result was confirmed by additional analysis with a Regularized Multivariate Regression for Identifying Master Predictors (RemMap) model. Finally, to investigate the association of copy number variation (CNV) with expression of the 8-gene panel, pairwise Pearson correlation with p-value calculation was performed between the copy number data and the expression of the eight genes, and the most correlated regions were identified to be in chromosome arms 3p, 5q, and 11q. Further analysis using only HPV-negative OSCC samples revealed an association of eight-gene deregulation with CNV in 11q13. We conclude that this new eight-gene marker panel is strongly correlated with PFS of HNC patients. Further evaluation using new HNC data sets is thus justified to establish the utility of this new eight-gene panel as a prognostic marker. Citation Format: Heuijoon Park, Yuzheng Zhang, Chang Xu, Chu Chen, Pei Wang, Eduardo Mendez. Comprehensive analysis of the expression of eight genes as a prognostic marker in head and neck cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr LB-169.