Genomic characterization of non-small cell lung cancer - clinical and molecular implications

Sammanfattning: Lung cancer accounts for 1.6 million cancers annually and 1.3 million deaths per year make the disease the deadliest type of cancer. Smoking is the most prominent cause. High mortality rates are partly due to late diagnosis which limits the therapy options. Improved understanding of molecular and biological mechanisms of tumor development could guide current therapy selection or pave way for novel therapies. The aim of this thesis is to characterize non-small cell lung cancer (NSCLC) genomes, stratifying tumors and patient groups with the intent of improved therapeutic options. A framework for treatment predictive mutation testing was established and, in parallel, the NanoString technology was evaluated for fusion gene detection. In an expansion of the fusion gene detection method, we included the possibility for simultaneous prediction of NSCLC histology, i.e. a multicomponent assay, applicable to tissue amounts used in standard clinical diagnostics. As accurate distinction of the histological subtypes is crucial for clinical management of lung cancer, the World Health Organization (WHO) administer and continuously update guidelines for histological classification. The mutational and transcriptional findings of this thesis work support their 2015 revision of histological subgroups. In addition, comprehensive global methylation analysis of lung cancer was performed, resulting in the detection of five epigenetically distinct groups of lung tumors associated with histology, gene expression, copy number variation and survival. In summary, this thesis has focused on genomic characterization of NSCLC, contributing to molecular findings and clinical implications.