Background
Consuming a healthy diet is recommended by the World Cancer Research Fund (WCRF) for cancer prevention. This recommendation is notably supported by strong scientific evidence indicating that a diet rich in wholegrains and dietary fibre and/or poor in red and processed meat decreases the risk of colorectal cancer.
However, our understanding of the complex relationship between diet and cancer risk is still limited. In particular, the biological mechanisms by which diet may exert its protective or harmful effects on cancer risk are still largely unknown.
Metabolomics can measure hundreds to thousands of small molecules, termed metabolites, present in human blood. Some of these metabolites are likely to be influenced by diet, and some diet-related metabolites may influence cancer risk.
A few epidemiological studies have been successful in identifying metabolites, or combinations of metabolites (termed signatures), that were associated with both specific dietary exposures and cancer risk. These results were not only indicative of an association between dietary exposures with cancer risk, but also of the possible biological mechanisms underlying this association.
For example, using metabolomics data available in the European Prospective Investigation into Cancer and Nutrition (EPIC) and the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study (ATBC), we recently successively identified metabolites associates with alcohol intake, including one that was also strongly associated with pancreatic and liver cancer risk, suggesting that one particular amino acids pathway may be involved with cancer development.
Aims and Objectives
In MeDiCa, we will identify metabolic signatures for seven specific dietary exposures whose association with cancer risk has not been convincingly established yet or for which insight into the underlying mechanisms is needed: alcohol, dietary fibre, dairy products, coffee, red and processed meat, fruits and vegetables and Mediterranean diet.
We will then assess the association between the signatures and the risk of gastrointestinal cancers. Finally, we will investigate the biological roles of the metabolites involved in the signatures, in particular the metabolic pathways and networks they are linked to, and which could trigger cancer development.
How it will be done
We will use available data from three large prospective studies, namely the EPIC, EPIC Norfolk and ATBC studies. We will apply a refined version of the statistical workflow we developed for our successful study of habitual alcohol intake, combining different machine learning techniques.
Also, using available genotyping data, we will examine whether the metabolism of dietary components, and thus the metabolic signatures derived, is modulated by genotype. The biological roles of the identified metabolites will be characterized using modern bioinformatics tools that will combine our results with relevant information available in the literature and chemical databases.
Potential impact
This project will improve our understanding of the complex relationship between diet and cancer risk, and of the underlying biological mechanisms, by filling in some of the important knowledge gaps identified in the 2018 WCRF/AICR (American Institute for Cancer Research) Third Expert Report. Our results will eventually help prioritize targets for cancer prevention.
