Nutritional and hormonal biomarkers in prostate cancer epidemiology

Evidence from international comparisons and migrant studies suggest that environmental factors, such as a Western diet, may be important in prostate cancer development, possibly through effects on hormone and growth factor secretion and metabolism. However, despite considerable research, convincing associations between diet and risk for prostate cancer have not been established. Random and systematic measurement error in dietary assessment using traditional survey methods may contribute to inconsistent findings, particularly as they may not capture adequately specific nutritional constituents of the diet that may be associated with risk, such as fatty acids or vitamins. Validated biomarkers of nutritional factors and hormonal activity, as used in this thesis, provide more precise, objective and integrated measures of exposure, with the capacity to clarify potential mechanisms in the causal pathway of prostate cancer development. Nutritional and hormonal biomarkers investigated for their potential role in the development of prostate cancer include: folate and vitamin B12, which are essential for DNA methylation, repair and synthesis; phytanic acid, obtained predominantly from ruminant fat intake and associated with an enzyme (α-Methylacyl-Coenzyme A Racemase (AMACR)) that is consistently over-expressed in prostate cancer tissue; and insulin-like growth factor (IGF-I), a growth factor influenced by diet and involved in the regulation of cell proliferation, differentiation, and apoptosis. All work presented in this thesis is from the European Prospective Investigation into Cancer and Nutrition (EPIC) study of 500,000 European men and women, using prospectively collected diet and lifestyle data and biological samples. The large number of prostate cancer cases diagnosed during long-term follow-up of EPIC participants enabled investigation of heterogeneity in risk for prostate cancer by time from recruitment to diagnosis (of particular importance for a disease with a long pre-clinical phase) and cancer characteristics such as disease grade and stage. Plasma phytanic acid concentration was highly correlated with dietary intake of fat from dairy products (r = 0.46) and beef (r = 0.30); capturing differences between countries in consumption of fat from these foods. Although phytanic acid is a useful biomarker of ruminant fat consumption, there was little evidence to support the hypothesis that the association between dairy products and prostate cancer risk (as suggested by previous work in EPIC and other studies) is mediated by phytanic acid (OR for doubling in concentration 1.05; 95%CI 0.91 – 1.21; P trend = 0.53). There was strong evidence for an association between higher circulating IGF-I concentration and risk for prostate cancer (OR for highest versus lowest fourth 1.69; 95% CI: 1.35, 2.13; P trend = 0.0002). Furthermore, the positive association observed among men diagnosed with advanced stage disease and among men diagnosed more than seven years after blood collection, supports the hypothesis that high IGF-I concentration is associated with clinically significant prostate cancer many years before diagnosis. There was no evidence of an association between prostate cancer risk and dietary folate or vitamin B12 intake, or between circulating levels of folate (OR for doubling in concentration 1.05; 95%CI 0.95 – 1.15; P trend = 0.33) or vitamin B12 (1.05; 95%CI 0.92 – 1.21; P trend = 0.47) and only limited evidence for an increased risk associated with elevated vitamin B12 in a meta-analysis of six prospective studies, that included the present study. All of these analyses were based on a blood sample taken at one point in time, with the assumption that this reflects the ‘true’ underlying concentration over the long-term. The poor to modest reliability estimates (intra-class correlation coefficients ranging from 0.18 to 0.48) for circulating concentrations of folate, IGF-I, phytanic acid and vitamin B12 taken in samples approximately six years apart in a sub-sample of participants from EPIC Oxford, show that estimates of usual concentrations based on a single blood measurement weaken the ability to detect associations with disease risk. Where small effect sizes are anticipated, this may bias associations toward the null. In conclusion, there is convincing evidence that IGF-I is an important and potentially modifiable risk factor for prostate cancer many years before diagnosis. However, there is little evidence for an association between biomarkers of folate, vitamin B12 and phytanic acid concentrations and risk for prostate cancer. Future studies should, where possible, incorporate multiple blood samples taken several years apart to better characterise long term relationships between biomarkers of nutritional and hormonal exposure and disease risk and pool individual participant data from multiple prospective studies to strengthen the power to detect modest associations.This thesis is not currently available on ORA