The causal role of the nutritionally regulated insulin-like growth factor system in prostate cancer: Mendelian randomization study

  • Topic: Prostate Cancer
  • Institution: University of Bristol
  • Country: United Kingdom
  • Status: Ongoing

Scientific abstract

(View plain language abstract)

Prostate cancer mortality is several-fold greater in 'western' than Asian countries, generating speculation that 'westernisation’ (involving high meat, milk, alcohol intakes; low exercise; and obesity) is important in the aetiology of prostate cancer. A metabolic endpoint of westernisation is hyperinsulinaemia and perturbations of the insulin-like growth factor (IGF) system, and this is a potential mechanism linking diet and obesity to cancer. The nutritionally-regulated IGFs and their binding proteins (IGFBPs) play a key role in somatic growth, but they also activate intracellular signalling networks involved in carcinogenesis. There are positive associations of circulating IGFs/IGFBPs with prostate cancer but results are inconsistent - whether elevated IGFs cause prostate cancer or reflect confounding or reverse causality is unclear. The objective of this research is to clarify the causal role of the circulating IGF system in prostate cancer and its progression, using Mendelian randomization and to investigate associations of dietary and lifestyle exposures with those IGFs/IGFBPs shown to be causally associated with prostate cancer.

Employing genes associated with circulating IGFs/IGFBPs as surrogates for circulating ligands Mendelian randomization enables causal inference about whether IGF-I causes prostate cancer or is, instead, a tumour marker, because genetic association studies are not subject to reverse causality and confounding seen in observational studies that measure phenotypes. Population-based case-control study of the IGF-system in screen-detected disease (ProtecT), in which 2,000 prostate cancer cases and 2,000 controls have already been phenotyped for circulating IGF-I, IGF-II, IGFBP-2 and IGFBP-3 will be used. DNA from the samples will be used for targeted genotyping to perform Mendelian randomization instrumental variables analysis and establish causality. (SNPs) for the analyses will be selected through a two stage procedure. Firstly literature and web-based bioinformatics searches to identify potentially functional common variants in IGF-related genes, which are associated with circulating IGFs. Secondly by investigation of associations of the SNPs identified above with IGFs amongst the 2317 in ProtecT who have already been typed for > 2.5 million genome-wide SNPs (GWAS, test-sample); those SNPs associated with IGFs will then be genotyped amongst 1,683 additional men to determine whether associations are replicated (validation-sample). Relationships between replicated SNPs and prostate cancer, will be undertaken. Mendelian randomization analysis, in the whole sample will also be conducted. Data on diet and lifestyle will be used to investigate associations of dietary and lifestyle exposures with those IGFs that are causally associated with prostate cancer.

If the results shows that IGF-prostate cancer associations are causal, this would suggest that the IGF system may mediate associations of dietary, lifestyle and metabolic factors with prostate cancer and suggest that nutritional and/or behavioural interventions to reduce IGFs could be indicated for the prevention of prostate cancer or its progression.

Plain language abstract

Background

Hormones that regulate how tall we grow and the normal development of our organs in early life circulate in high concentrations in our bodies. Levels of these hormones vary considerably (due to differences in our genetic make-up, diet and other reasons), and partly explain differences in height between individuals. However, in adulthood, after we have finished growing, people who have raised levels of these hormones seem to have an increased risk of cancer, including prostate cancer. The possible reasons for this finding, which has been shown in many studies, are unclear and the picture is currently confusing. Some people assume that the finding occurs because high levels of these hormones cause imbalances in the normal growth and production of cells and tissues in adulthood, which ultimately leads to the development of abnormal cells that become cancers. Others argue that the hormones don't cause cancers to develop, but high levels can cause those that have already started growing to become more aggressive/serious and to spread around the body. Another explanation is that the cancers themselves produce these hormones. Thus, raised levels of these hormones may not cause cancer or its progression, but are an indication that a cancer may be present. The final explanation is that these findings are nothing to do with the hormones, but are due to other factors that are in some way related to them.

Aims and Goals

This research aims to understand why men with prostate cancer appear to have raised growth-related hormone levels.

How it will be done

The researchers will assess whether men with a genetic make-up that causes a rise in these hormones also have an increased risk of prostate cancer developing or progressing, or whether raised hormone levels are due to the presence of tumour. Hormone levels in 2000 men with prostate cancer and 2000 healthy men will be studied

Potential impact

The research is important, because if growth-related hormones cause prostate cancer or its progression, then treatments that lower hormone levels in order to prevent the initiation or progression of prostate cancer could be potentially designed. Nutritional interventions may be relatively straightforward to develop as the nutritional determinants (e.g. milk) of growth-related hormones as a number of them are already known. If, instead, it is the cancers that cause a rise in the levels of these hormones, then this information could be used to develop clinical blood tests to diagnose the presence and aggressiveness of prostate cancer and to monitor the response of prostate cancers to treatment.

Grant publications