BACKGROUND: Alcohol consumption is associated with increased risk of breast cancer (BC), and the underlying mechanism is thought to be sex-hormone driven. In vitro and observational studies suggest a mechanism involving peroxisome proliferator-activated receptor gamma (PPARγ) in a complex with peroxisome proliferator-activated receptor gamma coactivator 1-α (PGC-1α) and interaction with aromatase (encoded by CYP19A1). Use of non-steroidal anti-inflammatory drugs (NSAID) may also affect circulating sex-hormone levels by modifying PPARγ activity.

METHODS: In the present study we assessed whether genetic variation in CYP19A1 is associated with risk of BC in a case-control study group nested within the Danish "Diet, Cancer and Health" cohort (ncases = 687 and ncontrols = 687) and searched for gene-gene interaction between CYP19A1 and PPARGC1A, and CYP19A1 and PPARG, and gene-alcohol and gene-NSAID interactions. Association between the CYP19A1 polymorphisms and hormone levels was also examined among 339 non-HRT users. Incidence rate ratios were calculated based on Cox' proportional hazards model. Furthermore, we performed a pilot randomised controlled trial to determine the effect of the PPARG Pro(12)Ala polymorphism and the PPARγ stimulator Ibuprofen on sex-hormone levels following alcohol intake in postmenopausal women (n = 25) using linear regression.

RESULTS: Genetic variations in CYP19A1 were associated with hormone levels (estrone: P rs11070844 = 0.009, estrone sulphate: P rs11070844 = 0.01, P rs749292 = 0.004, P rs1062033 = 0.007 and P rs10519297 = 0.03, and sex hormone-binding globulin (SHBG): P rs3751591 = 0.03) and interacted with alcohol intake in relation to hormone levels (estrone sulphate: P interaction/rs2008691 = 0.02 and P interaction/rs1062033= 0.03, and SHBG: P interaction/rs11070844 = 0.03). CYP19A1/rs3751591 was both associated with SHBG levels (P = 0.03) and with risk of BC (Incidence Rate Ratio = 2.12; 95 % Confidence Interval: 1.02-4.43) such that homozygous variant allele carriers had increased levels of serum SHBG and were at increased risk of BC. Acute intake of alcohol decreased blood estrone (P = <0.0001), estrone sulphate (P = <0.0001), and SHBG (P = 0.009) levels, whereas Ibuprofen intake and PPARG Pro(12)Ala genotype had no effect on hormone levels.

CONCLUSIONS: Our results suggest that genetically determined variation in CYP19A1 is associated with differences in sex hormone levels. However, the genetically determined differences in sex hormone levels were not convincingly associated with BC risk. The results therefore indicate that the genetically determined variation in CYP19A1 contributes little to BC risk and to alcohol-mediated BC risk.

TRIAL REGISTRATION: NCT02463383 , June 3, 2015.