Tumour Evolution in a Single Patient

Abstract

Tumour evolution underlies many of the most pressing challenges facing precision oncology today, indicating that a better understanding of tumour evolution has the potential to improve clinical care. This thesis focuses on investigating the programmes of evolution followed in many tumours from around the body of a single patient, and aims to provide useful knowledge about tumour evolution that may be missed in studies of multiple patients with less breadth. A patient with a lung neuroendocrine tumour and 90 metastases requested and consented to donate her tumour tissues for research after her death through rapid autopsy. We constructed an n=1 research autopsy programme to enable acceptance of this gift. 358 tissue samples were collected across 90 tumours, from which a representative subset of 42 tumour samples from 14 anatomical sites were analysed through complementary genomic technologies, including DNA whole-exome sequencing, linked-read whole-genome sequencing, RNA sequencing and RNA expression microarrays. Additionally, targeted DNA sequencing was completed on two clinical biopsies and one blood plasma sample. DNA sequence analysis elucidated the order of accumulation of putative driver variants and chromosomal-scale alterations, indicating that some chromosomal alterations preceded gene variants and likely played a critical role in tumour initiation. Cancer evolution tools highlighted the progressive accumulation of genomic alterations and identified two dominant tumour groups disseminated across the patient's body, while Bayesian evolutionary analysis enabled the estimation of evolutionary timings, each highlighting the potential for the improved cancer evolution tools. Sequencing a peripheral blood plasma sample taken while the patient was alive revealed clear detection of shared tumour variants and private variants from some but not all individual tumours, questioning the 'global representation' of circulating tumour DNA assays. RNA sequence and expression analysis highlighted the potential role of the immune system in shaping early tumour evolution, however the two expression datasets each featured a strong unexpected grouping, the source of which was rigorously investigated. The challenges in analysing this large and complex evolutionary dataset provided the motivation for developing and evaluating two tools to enhance cancer evolution research. PathNote is an iPad and iPhone app streamlining the documentation of complex tissue sample collections, ensuring tissues are suitably annotated for maximal downstream research use. An augmented reality model of our patient's tumours, how they changed through time, and their genomic profiles was also created, enabling collaborative multidisciplinary discussions around tumour evolution, where users could interact with the genomic data from 42 tumour locations in anatomical space and in the context of the clinical timeline. Overall, this unique patient donation has provided wide-ranging opportunities to better understand tumour evolution at the resolution of a single patient. It has opened doors to further improving our understanding and ultimately care for individual patients.