Project description:Lung cancer is the worldwide leading cause of death from cancer. DNA methylation in gene promoter regions is a major mechanism of gene expression regulation that may promote tumorigenesis. However, whether clinically relevant subgroups based on DNA methylation patterns exist in lung cancer is not well studied. We performed whole-genome methylation analysis using 450K Illumina BeadArrays on 124 tumors including 83 adenocarcinomas, 23 squamous cell carcinomas, one adenosquamous cancer, five large cell carcinomas, nine large cell neuroendocrine carcinomas (LCNEC), three small cell carcinomas (SCLC) and 12 normal lung tissues. Unsupervised class discovery was performed to identify DNA methylation subgroups with clinicopathological and molecular features. Subgroups were validated in two independent NSCLC cohorts. Unsupervised analysis identified five DNA methylation subgroups (epitypes). One epitype was distinctly associated with neuroendocrine tumors (LCNEC and SCLC). For adenocarcinoma, in both discovery and validation cohorts, remaining four epitypes were associated with differences in clinicopathological and molecular features, including global hypomethylation, promoter hypermethylation, copy number alterations, expression of proliferation-associated genes, association with unsupervised and supervised gene expression phenotypes, KRAS, TP53, KEAP1, SMARCA4, and STK11 mutations, smoking history, and patient outcome. Based on a multicohort approach we conducted a comprehensive survey of genome-wide DNA methylation in lung cancer, identifying a distinct neuroendocrine epitype and four adenocarcinoma epitypes associated with molecular and clinicopathological characteristics, and patient outcome. Our results bring further understanding of the epigenetic characteristics and molecular diversity in lung cancer generally and in adenocarcinoma specifically. Genome-wide DNA methylation analysis of 124 lung carcinomas and 12 normal lung tissues using Illumina Human Methylation 450K v1.0 Beadchips.

Project description:The success of targeted therapies creates a need to discriminate tumors accurately by their histological and genetic characteristics. Here, we used cDNA microarray analysis to identify gene expression profiles and single markers that recapitulate the pathological and genetic background (i.e., BRAF, EGFR, KRAS, LKB1, PIK3CA, and TP53 gene alterations) of non-small cell lung cancer (NSCLC). Gene expression profiles were determined for 6 normal lung tissues and 69 lung tumors from patients diagnosed with NSCLC. We performed an unsupervised hierarchical clustering with the most variably expressed transcript to investigate whether there was evidence for natural grouping samples based on similarity in gene expression profiles. We examined the relationship between the clusters of tumors and patient and tumor characteristics such as gender, smoking status, histological type, degree of differentiation, tumor size, lymph node involvement and genetic background. We used a supervised analysis to identify the genes that are important for distinguishing subgroups of tumors defined by histological type. Moreover, we used this supervised approach to search for markers that characterize those tumors carrying EGFR, KRAS, PIK3CA and TP53 alterations. We identify several genes with characteristic patterns of expression in each tumor histological type (adenocarcinoma and squamous cell carcinoma) and we found that the presence of EGFR mutations result in a particular expression profile. Keywords: Transciption profiling of tumors with different histological and genetic background. We analyzed 69 NSCLC tumors. All the samples were characterized by histological type and by the presence of alterations at genes that are known to be involved in lung carcinogenesis. Mutational analysis of BRAF, EGFR, KRAS, and LKB1 is provided only for adenocarcinomas and analysis of alterations at PIK3CA gene (mutational analysis or gene amplification by FISH analysis) is provided only for 47 of the tumors. We analyzed gene expression profiles to identify those genes that are differentially expressed between two subgroups by t-test: 1) adenocarcinomas vs. squamous cell carcinomas; 2) tumors with mutation at EGFR gene vs. wild type tumors; 3) tumors with mutation at KRAS gene vs. wild type tumors; 4) tumors with mutation at TP53 gene vs. wild type tumors; 5) tumors with alterations (mutation or gene amplification) at PIK3CA gene vs. wild type tumors.

Dataset Information

Homo sapiens

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure