The microbiome and bowel cancer

Colorectal cancer (CRC) is the second commonest cause of UK cancer-related deaths and the fourth commonest cause of global cancer-related deaths, with a rising incidence in non-Western countries. Research has demonstrated a CRC-associated microbiome, and the potential for microbiome testing to improve CRC screening accuracy. Currently the majority of studies analyse the microbiome from whole stool, transported and stored refrigerated/frozen; this limits study size and restricts research to Western countries with cold-chain facilities. This thesis investigated the potential to use guaiac faecal occult blood test (gFOBT) cards or faecal immunochemical test (FIT) samples to collect faeces for 16S ribosomal ribonucleic acid (16SrRNA) analysis. Screening potential was assessed by analysing the microbiome of gFOBT samples collected routinely by the NHS Bowel Cancer Screening Programme (NHSBCSP). The microbiome of non-Western countries was investigated by analysing gFOBT samples collected from healthy volunteers and CRC patients in Argentina, Chile, India and Vietnam. The microbiome was successfully analysed from processed NHSBCSP gFOBT samples stored for prolonged periods at room temperature and FIT samples for which NHSBCSP conditions were simulated. CRC-associated taxa demonstrated minimal temporal variation, but Escherichia-Shigella demonstrated marked variation, negating its potential as a screening biomarker. Microbiome-based models improved screening accuracy; combining gFOBT and microbiome results produced areas under the receiver operating characteristic curve of 0.855 (95% confidence interval (CI): 0.832- 0.877) for the detection of CRC and 0.868 (95% CI: 0.848-0.886) for the detection of CRC/adenoma. The microbiome of gFOBT samples stored and transported from abroad at ambient temperature was stable. The combined non-Western CRC-associated microbiome contained CRC-associated bacteria described in Western populations, suggesting that certain taxa may be universally associated with CRC. The results confirm that gFOBT is suitable for conducting large-scale national and global microbiome research. Clinical application is demonstrated with the development of novel microbiome-based CRC screening models with improved accuracy.