High Grade Serous Ovarian Cancer (HGSOC) can be classified by gene copy number signatures into 7 subtypes that have differing prognoses and treatment sensitivities and that show differences in the activities of various signalling pathways. An ongoing clinical study has demonstrated the feasibility of imaging hyperpolarized ¹³C pyruvate metabolism in ovarian cancer and has demonstrated inter and intra-tumoural metabolic heterogeneity, where metabolic differences were observed between patients and between different tumour deposits within the same patient. Here I compared the use of ¹³C magnetic resonance spectroscopic imaging (MRSI) of hyperpolarized [1-¹³C]pyruvate metabolism and positron emission tomography (PET) measurements of 2-Deoxy-2-[¹⁸F]fluoroglucose ([¹⁸F]FDG) uptake for detecting metabolic heterogeneity between HGSOC patient-derived xenografts (PDXs) that had different copy number signatures.

I showed that differences in glycolytic metabolism between the subtypes, as defined by their copy number signatures, could be detected using hyperpolarized [1-¹³C]pyruvate but not with [¹⁸F]FDG PET. Dynamic Contrast Enhanced MRI measurements showed that the metabolic differences between the subtypes were not due to differences in tumour perfusion. I also investigated whether differences in tumour metabolism could help to predict and detect early treatment response. I compared the use of metabolic imaging techniques (hyperpolarized [1-¹³C]pyruvate imaging, [¹⁸F]FDG PET/CT) with cell death imaging techniques (diffusion-weighted ¹H MRI (DWI) and ²H MRSI measurements of [2,3-²H₂]fumarate metabolism, measurement of circulating tumour DNA (ctDNA)) to detect early evidence of response to standard-of-care chemotherapy (Carboplatin). Both hyperpolarized [1-¹³C]pyruvate and [¹⁸F]FDG-PET detected response to treatment with Carboplatin, in a Carboplatin-sensitive tumour, before there was a change in tumour volume. Both metabolic imaging techniques were successful in discriminating responding from non-responding tumours. The techniques for detecting cell death were not as sensitive for detecting treatment response, which may reflect a slow accumulation of dead cells post treatment, a lack of knowledge of when the rate of cell death increases post treatment and also because of immune clearance of dead cells. These studies have shown that imaging with hyperpolarized [1-¹³C]pyruvate has the potential to be used in the clinic to detect the early response of HGSOC patients to treatment.

Finally, I explored the potential of imaging glyoxalase-1 (Glo-1) activity with [2-¹³C]Methylglyoxal for detecting metabolic heterogeneity between breast and ovarian cancer PDXs. However, Glo-1 activity did not differ between ovarian or breast cancer subtypes and therefore while ¹³C MRSI with [2-¹³C]Methylglyoxal has the potential to detect the presence of disease it may not be useful for differentiating between different HGSOC or breast cancer subtypes.