Abstract LB-99: EGF receptor tyrosine kinase defines critical prognostic genes of stage IA lung adenocarcinoma

Abstract

Abstract Non-small cell lung cancer (NSCLC) is the commonest and the most fatal histological subtype of lung cancer. About 10-30% of patients diagnosed as stage IA (the earliest stage of a lung tumor, smaller than three centimeters in diameter with no evidence of regional lymph node and/or regional metastasis) and submitted to surgery die due to recurrence. Therefore, the identification of prognostic biomarkers for stage IA lung adenocarcinoma with poor prognosis is of great importance to select patients who will be benefited by adjuvant therapy. To the best of our knowledge, there is not a set of genes that identifies these patients. It is known that epidermal growth factor (EGF) signaling pathway is closely related to aggressive phenotypes of lung and other cancers. EGFR tyrosine kinase-specific inhibitor, namely gefitinib is expected to alter the gene expression patterns caused by EGFR tyrosine kinase activity. We performed microarray assays in order to obtain the entire gene expression time course profile of human primary lung epithelial cells that were stimulated with EGF in both the presence and absence of gefitinib. The time courses are composed of 19 time points along 48 hours. The data were subjected to a mathematical analysis, namely the State Space Model (SSM) in order to select genes with gene expressions altered by gefitinib. One hundred thirty nine genes were identified. These genes were used as expression signatures to train a risk scoring model that classifies patients in high- or low-risk (risk of dying in five years). This model was trained by using a data set composed of 253 North American patients with lung adenocarcinomas. Then, the predictive ability of the risk scoring model was examined in two independent cohorts composed of North American and Japanese patients. The model enabled the statistically significant identification of high-risk stage IA lung adenocarcinoma in both cohorts, with hazard ratios (HRs) for death of 7.16 (P=0.029) for North American and 10.98 (P=0.008) for Japanese. The set of 139 genes altered by gefitinib includes several ones that are involved in biological aspects of cancer phenotypes but are yet unknown to be involved in EGF signaling. This result strongly re-emphasizes that EGF signaling status underlies aggressive phenotype of cancer cells, and also suggests the first set of genes that are useful for the identification of stage IA lung adenocarcinoma patients with poor prognosis. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr LB-99. doi:1538-7445.AM2012-LB-99