Investigating radioresistance: A systems biology and functional genomics approach to identify mechanisms of resistance to radiation and develop a predictive classifier

Thesis

Radiotherapy forms an integral part of modern cancer management, but despite advances in molecular medicine that have revolutionised the field of medical oncology enabling predictive and prognostic molecular markers and targeted therapies, these have not translated to the field of radiation therapy. The work in this thesis uses the existing literature to inform the design of cell line experiments to explore molecular factors associated with radioresistance, applies these findings to clinical cohorts, and investigates the predictive potential of pre-treatment diagnostic biopsies in prostate cancer clinical samples from patients treated with radiotherapy.

Gene expression and cell characterisation identifies that generated radioresistant cell lines exhibit cellular and gene expression changes consistent with senescence and altered methylation. Ecological modelling using experimentally derived growth and radiosensitivity parameters suggests a mechanism whereby a subpopulation of slower growing radioresistant cells through competition, can become the dominant population under the pressure of clinically relevant fractionated radiation schedules.

Genes that were significantly increased in expression in generated radioresistant lung cancer cell lines were applied to a clinical dataset and demonstrated to be prognostic in a radiotherapy treated cohort of patients with lung cancer, but not in patients who had not been treated with radiotherapy.

In a small cohort who underwent multi-omic sequencing, coherent gene expression and methylation patterns were demonstrated in pre-treatment biopsies of patients progressing following radical radiotherapy, similar to patients who presented with metastatic disease and distinct from patients who had stable disease. The significantly differentially expressed genes demonstrated the potential to predict for treatment failure in a large external dataset of radiotherapy treated patients.

Authors: Charlton, P

• DOI: 10.5287/ora-4jenbzdx2
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: predictive molecular markers; methylation; radiation; ecological modelling; radiation resposne; gene expression
• Subjects: Radiotherapy; Prostate; Methylation; Cancer; Gene expression