A next-generation -omics analysis of the radiotherapy response in rectal cancer
File(s)
Author(s)
Poynter, Liam Robert
Type
Thesis
Abstract
Rectal cancer (RC) accounts for over a third of all cases of colorectal cancer (CRC), itself the second-leading cause of cancer-related death worldwide. Fundamental to the treatment armamentarium in locally advanced RC is the application of neoadjuvant radiotherapy (RT). The exact response to this therapy is highly individualised and unpredictable. As we enter an era where the role for RT in RC is set to increase, the critical unmet need remains both the lack of a predictive or prognosticating RT biomarker, and an incomplete understanding of the mechanisms that govern radioresistance in RC. Multiple, Next-Generation ‘-omics’ platforms now provide us with the tools to obtain a more holistic snapshot of the biology at play, in ways that until very recently were not feasible.
This body of work represents the design and implementation of a novel multi-omics pipeline to address a greater understanding of the contributing and complementary mechanisms at play in RC radioresistance. Firstly, I have demonstrated the value in application of machine-learning algorithms to the vast amount of existing experimental data to reveal previously untapped areas likely to be of biological significance. I have subsequently utilised a soft-ionisation mass spectrometry imaging platform to gain maximally representative biological insights into the tumour lipidome in RC in response to RT. I have further complemented this with a holistic whole-genome sequencing (WGS) approach of pre- and post-RT RC specimens, with a targeted view on the likely genomic aberrations that may be at play. My findings have demonstrated that:
(1) Gene Ontology (GO)-based network mapping of published, investigated, candidate biomarkers demonstrates areas of significant promise in the research of this problem. GO domains of cellular metabolism, response to stimuli and cell communication are not only the areas most implicated as being significantly correlated to RT response, but they are all heavily influenced by the actions of complex cellular lipids.
(2) Gene Set Enrichment Analysis (GSEA) demonstrates previously unidentified enrichment in gene sets associated with glycerophospholipid (GPL) metabolism in radioresistant RC in a legacy microarray dataset
(3) Desorption electrospray ionisation mass spectrometry imaging (DESI MSI) reveals increased abundances of phosphatidylethanolamines (PE), phosphatidylserines (PS) and phosphatidylglycerols (PG) in radioresistant RC, with increased integration of polyunsaturated fatty acids (FAs) into these GPLs as compared to RT-naïve tissues
(4) Low-pass whole-genome sequencing (lpWGS) reveals a significant burden of somatic copy number aberration (SCNA) in pre-RT RC, correlating with a CMS2 or CMS4 phenotype. Furthermore, there is significant heterogeneity in SCNA burden in separately sequenced macroscopic regions of tumour prior to RT, and an increased genomic divergence correlates with a superior response to RT.
This body of work represents the design and implementation of a novel multi-omics pipeline to address a greater understanding of the contributing and complementary mechanisms at play in RC radioresistance. Firstly, I have demonstrated the value in application of machine-learning algorithms to the vast amount of existing experimental data to reveal previously untapped areas likely to be of biological significance. I have subsequently utilised a soft-ionisation mass spectrometry imaging platform to gain maximally representative biological insights into the tumour lipidome in RC in response to RT. I have further complemented this with a holistic whole-genome sequencing (WGS) approach of pre- and post-RT RC specimens, with a targeted view on the likely genomic aberrations that may be at play. My findings have demonstrated that:
(1) Gene Ontology (GO)-based network mapping of published, investigated, candidate biomarkers demonstrates areas of significant promise in the research of this problem. GO domains of cellular metabolism, response to stimuli and cell communication are not only the areas most implicated as being significantly correlated to RT response, but they are all heavily influenced by the actions of complex cellular lipids.
(2) Gene Set Enrichment Analysis (GSEA) demonstrates previously unidentified enrichment in gene sets associated with glycerophospholipid (GPL) metabolism in radioresistant RC in a legacy microarray dataset
(3) Desorption electrospray ionisation mass spectrometry imaging (DESI MSI) reveals increased abundances of phosphatidylethanolamines (PE), phosphatidylserines (PS) and phosphatidylglycerols (PG) in radioresistant RC, with increased integration of polyunsaturated fatty acids (FAs) into these GPLs as compared to RT-naïve tissues
(4) Low-pass whole-genome sequencing (lpWGS) reveals a significant burden of somatic copy number aberration (SCNA) in pre-RT RC, correlating with a CMS2 or CMS4 phenotype. Furthermore, there is significant heterogeneity in SCNA burden in separately sequenced macroscopic regions of tumour prior to RT, and an increased genomic divergence correlates with a superior response to RT.
Version
Open Access
Date Issued
2021-04
Date Awarded
2022-02
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Darzi, Ara
Takáts, Zoltan
Mirnezami, Reza
Sponsor
National Institute for Health Research (Great Britain)
Publisher Department
Department of Surgery & Cancer
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)