Investigating the relationship between hypoxia and the immune tumour microenvironment in muscle-invasive bladder cancer

Abstract

Introduction: Muscle-invasive bladder cancer (MIBC) has high rates of recurrence with poor survival rates. There is a need to improve response to radiotherapy and survival outcomes. Both hypoxia and the immune tumour microenvironment (TME) are therapeutically targetable in MIBC using hypoxia-modifying therapy and immune checkpoint inhibitors (ICIs), respectively. Recent data showed hypoxia drives an immunosuppressive TME via various mechanisms including upregulation of programmed death ligand 1 (PD-L1) and the inhibition and exclusion of cytotoxic CD8 T cells. Published literature also demonstrates hypoxia-driven upregulation of inflammation. There is currently no single biomarker that stratifies MIBC patients for both hypoxia-modifying therapy and ICIs. Therefore, the main thesis aims were to: 1) investigate if hypoxia is associated with upregulation of PD-L1 in MIBC, 2) investigate whether patients with low tumour CD8 T cells benefit from hypoxia-modifying therapy, 3) investigate the role of HIF and hypoxia in immune-related signalling in MIBC, 4) develop a hypoxia-driven immune gene signature that could be used to stratify MIBC patients for hypoxia-modifying therapy and ICIs. Materials and methods: 1) Three human MIBC cell lines (T24, J82, UMUC3) were cultured in normoxia (20% oxygen) or hypoxia (1% and 0.1% oxygen) for 24 h. Differences in PD-L1 expression were measured using Western blotting, qPCR and flow cytometry. Published gene signatures were used to correlate hypoxia with PD-L1 expression and IFNy signalling in BCON and TCGA MIBC cohorts. The BCON trial randomised bladder cancer patients to radiotherapy +/- hypoxia-modifying carbogen plus nicotinamide (CON). 2) Tissue microarrays of diagnostic biopsies from 116 BCON patients were stained using multiplex immunohistochemistry (IHC) for CD8, CD4, FOXP3, CD68, and PDL1. Hypoxia was assessed using CA9 IHC. Relationships with overall survival (OS) were investigated using Cox proportional hazard models. 3) ChIPseq and microarray data from MIBC cell lines and the TCGA cohort were used to analyse relationships between HIF/hypoxia and immune-related signalling and to identify prognostic hypoxia-driven immune genes. A LASSO regression model defined a final six-gene signature. Validation and analysis was performed on four bladder cancer cohorts. All in silico analyses were carried out using R and RStudio. Results: 1) Increasing seeding density decreased PD-L1 protein (p

Date of Award	1 Aug 2023
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Debayan Mukherjee (Supervisor), Catharine West (Supervisor) & Timothy Illidge (Supervisor)