Abstract
The aim of the research presented in this thesis is to improve understanding of immune cells, predominantly T cells, in the context of cancer, mainly gastro intestinal tract cancers.Gastro-intestinal tract cancers are a major health problem and a leading cause of death worldwide. In addition to surgery, chemotherapy and radiotherapy, immunotherapy has emerged as a useful option for cancer treatment. Immunotherapy aims to reinvigorate the immune system to fight against tumour cells, typically by targeting either cancer antigens or antigens that switch off T cells. Patients respond unequally to immunotherapy, however, leading to cancer relapse and toxicity. Immunotherapies currently lack reliable predictors of efficacy, and the role of immune system infiltration in cancer remains to be clarified. The future development of immunotherapies and the prediction of patient-specific responses to treatment may require multiple approaches involving different platforms and strategies that provide accurate and rapid data interpretation. Methods such as gene expression array (Nanostring) and long read next-generation sequencing (Nanopore) have the potential to detect genetic mutations associated with treatment failure and cancer progression, as well as to identify novel therapeutic targets, thus enabling improved diagnosis, treatment and outcome. Progress in machine learning and tumour organoid engineering can also improve the efficiency and accuracy of immunotherapy treatments.
I first investigated the correlation between 12 pancreatic cancer patient prognoses and the presence of tumour-infiltrating-lymphocytes (TILs) and expression of immune genes using the PanCancer Immunoprofile Panel (Nanostring) (Chapter 2). I then compiled immune transcription profiles from multiple studies to identify a pan-cancer immune gene expression profile associated with patient survival across 515 patients representing ten cancer types, including solid and blood malignancies (Chapter 3). With the aim of establishing an individualised ex vivo model system to support T cell-based therapy, I demonstrated the feasibility of obtaining tumour reactive T cells by co-culturing murine peripheral blood mononuclear cells (PBMCs) with matched murine pancreatic tumour organoids (Chapter 4). Using both Nanopore and Illumina sequencing technologies, I then conducted one of the few investigations to date of expression levels and structural variations in colorectal cancer tumour-infiltrating T cells, focusing on immune checkpoints (Chapter 5). Employing the same Illumina data, I applied numerical, statistical and machine learning analyses to sequences of complementarity determining region 3 of the β chain (CDR3β) of the T cell receptor (TCR) from CD8+ and CD4+ T cells to classify and predict CDR3 repertoire properties according to sample conditions (Chapter 6).
The work presented here demonstrates how gene expression analysis can help to identify the best responders to immune checkpoint inhibitors and reveal transcriptional profiles associated with poor survival. It shows also that organoid techniques can be used to develop patient-specific treatments. Next generation sequencing can help to identify novel structural variations in inhibitory immune checkpoints. Sequencing data coupled with machine learning can be used to profile CDR3 sequences that could lead to a better understanding of immunotherapy toxicity and treatment resistance. Collectively, the multiple methods employed here can advance knowledge of tumour microenvironment interactions and contribute in the clinic via new strategies for diagnosis and prognosis, therapeutic decisionmaking, and identification of new therapeutic targets.
| Date of Award | 17 Jan 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Stefan Bagby (Supervisor), Araxi Urrutia (Supervisor) & Stephen Ward (Supervisor) |