P1.01-18 Integrated Genomic and DNA Methylation Analyses of Non-Small Cell Lung Cancer Patients with Brain Metastases

Brain metastases (BM), with a dismal prognosis, are a common and lethal complication of non-small cell lung cancer. Approximately, 10% patients present with BM at their initial diagnosis. Although, surgery and/or radiation therapy remain to be the mainstay treatment, targeted therapies are finding increasing application in treating BM. However, due to the very limited accessibility of brain lesions, its genomic and epigenomic landscape remain elusive. Capture-based targeted sequencing for somatic mutation profiling was performed on 27 treatment-naïve advanced NSCLC patients with paired lung primary and BM lesions using a pane consisting of 520 cancer related genes. DNA methylation analyses was performed on same samples using a DNA methylation panel consisting of 100,000 CpG sites. Collectively, we identified 370 (291 SNVs+Indels, 78 CNVs and 1 rearrangement) and 574 (245 SNVs+Indels, 327 CNVs and 2 rearrangements) mutations from lung primary lesions and BM, respectively. Among them, 242 mutations were shared; 128 were lung primary-specific and 332 were BM-specific. Among the BM specific mutations, a majority of them (82%, 272/332) were copy number variations (CNVs). Only 16% of CNVs were shared by lung lesions and BM. The concordance for SNVs and indels were much higher-54% between the two sources of tissues. Furthermore, we observed a much higher concordance rate (79%) in TP53 and classic lung cancer driver genes than other genes (p<0.001), indicating that they might be stem mutations. Next, we performed pathway analysis of genes that were only mutated in BM and revealed an enrichment of genes participating in PI3K-AKT and focal adhesion pathways. We also compared tumor mutation burden (TMB) between them and revealed comparable TMB (p=0.1). Our DNA methylation analysis revealed distinct methylation patterns with 268 blocks that are significantly differentially methylated between primary lung lesions and BM. Among them, 211 blocks were hypermethylated in BM and the remaining 57 blocks were hypermethylated in lung lesions. These blocks were enrichment in genes participating in cell adhesion, Rap1 signaling and calcium signaling pathways. We revealed diverse somatic mutation and DNA methylation profiles between lung primary lesions and BM. BM had significantly more unique CNVs. A great concordance was observed for classic lung cancer driver genes and TP53. Our study provided a comprehensive view of genomic and DNA methylation profiling for lung primary lesions and BM, paving the avenue for the development of targeted therapies for treating BM.