Preview

PDF, English - main document Download (10MB) | Lizenz: Next-Generation Sequencing Analysis of Cell-Free DNA Identifies Actionable Alterations and Genomic Features in Pediatric Cancers by Puranachot, Pitithat underlies the terms of Creative Commons Attribution-NonCommercial 4.0

Abstract

Pediatric cancer is the third leading course of death among children and adolescents in the USA despite its low incident and high survival rate. Next-generation sequencing technologies allow the profiling of tumor genetics and the prediction of disease progression and response to therapies. However, tumor temporal and spatial heterogeneity could complicate the success of the selected therapy. Serial sampling of tumors at multiple time-points can accurately track the dynamics of clonal evolution during treatment. Multiple sampling of tumors at different locations can reveal all clonal genetic structures of the tumor. Nevertheless, both strategies might post discomfort or critical risk to the patient. Liquid biopsy has become an attractive strategy for obtaining tumor biomarkers non-invasively. Sequencing of cell-free DNA (cfDNA), DNA fragments in the liquid sample such as blood, has become a strategy to detect tumor-derived genetic markers known as circulating tumor DNA. Recently, cfDNA has been extensively evaluated its clinical value with different high-throughput sequencing technology in many adult cancers. Hence, cfDNA could also have a potential benefit to the management of pediatric cancer patients.

In this thesis, we developed bioinformatics workflows for analyzing cfDNA derived from an extensive group of pediatric cancer patients. The workflow aims to detect genetic alterations from three sequencing strategies, including low-coverage whole-genome sequencing (lcWGS), whole-exome sequencing (WES), and deep gene-panel sequencing (Panel-seq). The capabilities of detecting copy-number aberrations and point mutations have been compared between those strategies. We also compared the detectability of plasma cfDNA across tumor entities, including brain tumors, sarcomas, and other pediatric cancers. Sequencing strategy and tumor location have influences on the success of cfDNA in detecting tumor genetic alterations. An R package, cfdnakit, was developed to extract the length of cfDNA fragments and perform genome-wide fragment-length analysis using lcWGS dataset. The fragment-length analysis shows that the enrichment of short-fragment cfDNA is correlating with copy-number aberrations. In addition, this package calculates a comprehensive copy-number aberration (CPA) score that combines copy-number aberration and short-fragmented cfDNA ratio. This CPA-score is correlating with a higher level of ctDNA and could suggest the use of subsequent detection methods such as WES to detect actionable mutations with more sensitivity. Moreover, we applied TelomereHunter, a telomeric DNA analysis tool. It showed that telomeric DNA exists which opens an opportunity to detect telomeric aberration in plasma cfDNA. Analyzing plasma cfDNA of the pediatric cohort has shown the declining of telomere content. However, elongation and integration of telomeric variant repeats were found among brain tumor and sarcoma patients.

Finally, we demonstrated the utility of liquid biopsy cfDNA in the management of pediatric cancer. cfDNA reveals heterogeneous mutations possibly shred by tumor at metastasis site in a child with bilateral nephroblastoma. This finding supports the utility of cfDNA as a comprehensive source of genetic information derived from the tumor population in the body without invasive multiple tumor biopsies. In addition, we found that cfDNA can detect tumor temporal heterogeneity in several sarcoma patients through serial biopsy. This finding supports the idea of utilizing cfDNA to follow-up patients during the course of therapy.