Computational Genomic Models For Spatio-Temporal Investigation of Early Lung Cancer Pathology

<p>Lung cancer, of which non-small cell lung cancer (NSCLC) is the most common form, is the second most prevalent cancer and the leading cause of cancer-related deaths. NSCLCs primarily comprise adenocarcinomas (LUAD) and squamous cell carcinomas (LUSC). Advances in early detection and prevention have been limited by the lack of early-stage biomarkers and targets. A comprehensive molecular characterization of premalignant lesions and tumor-adjacent normal tissue can aid in better understanding NSCLC pathogenesis. However, these investigations are further challenged by limited tissue availability and low cellular fractions of detectable somatic mutations.</p> <p>Therefore, there is a dearth of knowledge about the pathogenesis of premalignant lung lesions, especially for atypical adenomatous hyperplasia (AAH), the only known precursor to LUADs. We performed a cross-platform integrative analysis comprising targeted DNA sequencing, genotype array profiling and transcriptome sequencing of matched AAHs, LUADs and normal tissues from 23 early-stage patients. The study revealed potentially divergent pathways based on the mutation status of AAH (<em>BRAF</em> vs <em>KRAS</em>), recurrent chromosomal aberrations (17p loss) and the presence of immune deregulation early in the pathogenesis of AAHs.</p> <p>Molecular changes, characteristic of NSCLCs, might also occur in normal tissues, preceding identifiable premalignancy-associated morphological changes. We sought to comprehensively survey the somatic mutational architecture of the normal airway in early-stage NSCLCs. Targeted DNA sequencing allowed us to capture driver mutations at low cellular fractions, typical of these non-malignant tissues. Additionally, genotype array profiling helped characterize subtle chromosomal aberrations in these tissues. This multi-region study included tumor-adjacent and -distant airways, nasal epithelia and uninvolved normal lung (collectively cancerized field) along with matched multi-region NSCLCs and blood cells from 48 patients. Integrative computational analysis revealed genomic airway field carcinogenesis in 52% of cases. The airway field exhibited mutations in known drivers, that were present at lower frequencies compared to NSCLCs, suggestive of selection-driven clonal expansion. These driver events also comprised somatic “two-hit” alterations in matched airway field and NSCLCs.</p> <p>Our study design offers spatiotemporal insights into NSCLC development and suggests potential targets for early detection and treatment, in possibly less hostile environments of premalignancy. To validate and enhance the utility of the bioinformatic techniques devised and implemented for these investigations, I also provide methods to expand such analyses across multiple tumor sites.</p>

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation (PhD)

Year dc:date.available

2019

Authors dc:creator

• Sivakumar, Smruthy
• <p>0000-0001-7790-5615</p>

Contributors dc:contributor

• Paul Scheet, Ph.D.
• Humam Kadara, Ph.D.
• Eduardo Vilar Sanchez, M.D. Ph.D.

• Premalignant
• Lung cancer
• Field cancerization
• Bioinformatics
• Cancer genomics
• TCGA
• Genomic instability
• Allelic imbalance
• Computational Biology
• Genetics and Genomics
• Translational Medical Research

Repository record dc:identifier

https://digitalcommons.library.tmc.edu/utgsbs_dissertations/945

OAI identifier oai:identifier

oai:digitalcommons.library.tmc.edu:utgsbs_dissertations-1990