Randomized clinical trials have shown that postmenopausal hormone treatment (HT) is associated with either no cardiovascular benefit or an increase in cardiovascular risk. Because millions of women still use these agents and because the absolute risk of adverse cardiovascular events with HT use is small, it would be desirable to have markers of HT risk. The earliest stages of platelet thrombus formation in coronary arteries involve platelet adhesion and signaling responses mediated by glycoproteins (GP) Ibα and VI, respectively. Genetic variations in the corresponding genes have been associated with acute coronary syndromes, such as myocardial infarction (MI) and unstable angina (UA). Because of known gender differences in and estrogenic effects on platelet function, we sought to determine whether functional polymorphisms of GPIbα (−5T/C “Kozak”) and GPVI (13254T/C) were risk factors for CHD events and whether these functional polymorphisms interacted with HRT. The Heart and Estrogen/progestin Replacement Study (HERS) was a randomized clinical trial testing the effect of conjugated estrogens plus medroxyprogesterone acetate on CHD event risk among postmenopausal women with documented CHD. We genotyped 2145 patients in HERS, 537 of whom experienced CHD death/MI/UA, and assessed the event rate over 7 years of follow-up. In patients receiving placebo, there was an increased CHD event rate in carriers of the GPIbα −5C allele (CC〉CT〉TT, P=0.006; adjusted [adj] P=0.013) (adj for age, race, hypertension, diabetes, baseline LDL and HDL cholesterol, aspirin use and statin use). There was no significant risk associated with the GPVI genotype in the placebo group. Compared to placebo, HRT increased the relative risk (RR) of CHD death/MI/UA in patients with the GPIbα −5TT genotype by 16% and reduced the RR in patients with the TC+CC genotypes by 44% (interaction P=0.002; adj. interaction P=0.002). Compared to placebo, HRT reduced the RR in patients with the GPVI TT genotype by 13%, but increased the RR in patients with the TC+CC genotypes by 42% (interaction P=0.011; adj. interaction P=0.007). Furthermore, compared to placebo, HRT increased the RR of CHD events in patients with the GPIbα −5TT plus GPVI 13254TC+CC genotypes by 74% and reduced the RR in patients with the GPIbα −5 TC+CC plus GPVI 13254TT genotypes by 19% (interaction P=0.002; adj. interaction P=0.003). Similar results were obtained in all analyses after the first year of follow-up. In summary, in postmenopausal women: 1) the Kozak −5T/C polymorphism of GPIbα was a risk factor for recurrent CHD events, but the T13254C polymorphism of GPVI was not; 2) each polymorphism interacted individually with HRT to modify the risk for CHD events; and 3) both polymorphisms together interacted with HRT to modify the risk for CHD events (gene-gene-environment). Importantly, the lack of risk of a genetic variant for a clinical phenotype does not mean the variant cannot interact with environmental factors to influence the same phenotypic outcome (e.g., GPVI T13254C). Platelet pharmacogenetics may provide additional insights into the cardiovascular effects of hormone therapy.

Author notes

Corresponding author

Sign in via your Institution