Worldwide, prostate cancer is the second most common cancer among men, yet there are no known modifiable risk factors to inform prevention. The only well-established risk factors for prostate cancer are male sex, increasing age, positive family history, and a number of genetic markers. In EPIC, a wide range of potential risk factors has been examined, including lifestyle, dietary and hormonal, metabolic and genetic factors. The most important finding so far is that the risk of prostate cancer is positively associated with serum concentrations of insulin-like growth factor-I (IGF-I)1. The group has also shown that serum levels of IGF-I are positively associated with the protein content of the diet, and has shown a modest positive association of dairy protein consumption with prostate cancer risk2.
Through a co-ordinated programme of analyses, the EPIC working group aims to advance understanding of the aetiology of prostate cancer, particularly of aggressive disease, and provide the reliable data needed for the development of evidence-based strategies for prostate cancer prevention.
We have one of the largest prospective studies of prostate cancer in the world with extensive phenotypic data, blood samples, genetic data and ~3 million person years of follow-up. Participants were recruited between 1992 and 2000, so that average follow-up is now approximately 20 years and there will soon be over 8,000 incident cases of prostate cancer. Over the next five years, we will use recent developments in high-throughput technologies to provide detailed information on novel biomarkers and metabolic profiles, combined with the wealth of data already collected and with detailed data on tumour phenotype, to examine a wide range of molecular and clinical factors that may influence the development of high risk prostate cancer.
Previous EPIC prostate cancer analyses have suggested that men with higher adiposity may have a higher risk of aggressive prostate cancer and we are looking into the mechanisms underlying this possible relationship3. We also showed that higher plasma concentrations of microseminoprotein-beta (MSP) are inversely associated with risk of total, advanced and fatal prostate cancer, and results from our Mendelian randomisation analyses implied that this relationship is causal4. Recent analyses have also suggested blood metabolite patterns may be related to lower risk of more aggressive prostate tumours and prostate cancer death5, and we are currently extending this metabolomics resource. Other studies in EPIC-Prostate have shown that few aspects of lifestyle or diet6-9 are strongly related to total prostate cancer risk; overall in EPIC, prostate cancer risk is not clearly related to alcohol, smoking, physical activity or having had a vasectomy10, while men with type 2 diabetes have a lower prostate cancer risk11.
Studies of blood biomarkers in EPIC have examined a range of hormones, growth factors and nutritional biomarkers in relation to prostate cancer risk11 and subsequently these data have been combined with all the other data available in an international biomarker consortium (The Endogenous Hormones, Nutritional Biomarkers and Prostate Cancer Collaborative Group). We have found a strong positive relationship between blood levels of the growth factor insulin-like growth factor-1 and prostate cancer risk and have shown for the first time that men with very low free testosterone have a lower risk for prostate cancer, but possible an increase in risk for high grade disease. Studies of other candidate risk factors have produced largely null results, for example for selenium, carotenoids, and tocopherols11. EPIC Prostate also participates in the PRACTICAL and ELLIPSE international genetics consortia, the NCI Pooling Project on Prostate Cancer (PC3) and the Harvard Pooling Project.
Selected publications:
Contact details/Working Group leader
Ruth Travis, DPhil
Cancer Epidemiology Unit
University of Oxford
Oxford OX3 7LF
United Kingdom
ruth.travis@ndph.ox.ac.uk