Cervical cancer is a major problem in low/middle income countries where 85% of the new cases/deaths occur. Secondary prevention measures reduced incidence and mortality in developed countries over the last 30 years, but cervical cancer remains a major cause of cancer deaths in women. Hypoxia, or low tumoural oxygenation, is a ubiquitous feature of solid tumours which drives disease progression and restricts treatment efficacy. Hypoxia targeting therapies have shown great promise, however there is an unmet need for hypoxia biomarkers to select patients most likely to benefit from hypoxia modification treatment. The thesis aimed to develop hypoxia biomarkers in patients with cervical cancer which may be used in future hypoxia modification trials. The investigated biomarkers were derived from gene expression and magnetic resonance imaging (MRI) data. The gene expression signature development was from cell line experiments, developed using a publicly available dataset and validated in an independent retrospective cohort of women treated at The Christie with 4-5 years of clinical outcome data. The de novo 31-gene signature was enriched for known hypoxia pathways and biological processes, and showed prognostic significance in the external validation cohort. Oxygen enhanced (OE) -MRI of the female pelvis was developed in healthy volunteers and translated onto the MR Linac for the first time. The results showed quantitative T1 values derived using the inversion recovery sequence to be comparable with published literature. The ÎR1 parameter was shown to be repeatable across the two imaging systems (a Diagnostic MR and MR Linac) and using two hyperoxic gases (100% O2 and 98% O2 / 2% CO2). Finally, I identified the uterine body as a quality control region for OE-MRI. In a prospective patient pilot study, I analysed data from locally advanced cervical tumours in three world firsts: a) tumour biopsies acquired following five weeks of chemoradiation assessed with the de novo hypoxia signature; b) serial assessments of patient tumours using OE-MRI during chemoradiotherapy; and c) exploratory imaging-genomic correlations using independently derived hypoxia biomarkers. Unfortunately, a batch processing error meant I could not analyse all the prospective patient biopsies. The data suggested intra-tumoural heterogeneity of the transcriptional and imaging biomarkers. The ÎR1 parameter was able to map, quantify and track whole tumour changes in hypoxia modification secondary to chemoradiation during the five weeks of external beam radiotherapy. In summary, this thesis presents important new insights on hypoxia associated gene expression and OE-MRI data acquisition and analysis. I highlight a potential role for the combined imaging-genomic evaluation of tumour hypoxia and highly targeted radiation delivery on the MR Linac system.
- Gene expression
- Magnetic resonance imaging
- Hypoxia
- Cervical cancer
Evaluation
Datta, A. (Author). 31 Dec 2023
Student thesis: Phd