Background: Variants at the 9p21 locus associate with the risk of coronary artery disease (CAD) or myocardial infarction (MI). However, atherosclerotic plaque deposition is distinct from MI (plaque rupture and thrombosis), and recent studies showed no association between these variants and MI in patients with preexisting CAD. We performed haplotype analysis at the 9p21 locus to test whether haplotypes at distinct linkage disequilibrium blocks predict these phenotypes.
Methods and results: Using 24 single-nucleotide polymorphisms genotyped in white patients without diabetes mellitus, we reconstructed haplotypes at the 9p21 locus. Patients with angiograhic CAD/MI had ≥1 epicardial stenosis >50% (n=2352), whereas controls were asymptomatic and over the age of 60 years (n=2116). For CAD patients, regression models examined the association of haplotypes with initial age of symptomatic CAD, number of diseased vessels, and history of MI. In the case-control study, only haplotypes at 1 block tagged by rs1333049 associated with CAD more so than MI. These haplotypes also associated with early onset of CAD (β=-0.13; P=1.37×10(-4)) and disease severity (β=0.1823; P=0.006) but not with prevalent MI among patients with CAD. In contrast, haplotypes at another block tagged by rs518394 associated with prevalent MI (β=0.239; P=2.05×10(-4)), but remarkably these are inversely associated with disease severity (β=-0.196; P=0.003). This MI association was replicated in the Cleveland Clinic GeneBank premature CAD cohort (n=1385; β=0.207; P=0.019).
Conclusions: Variants/haplotypes at 2 blocks are distinguished at 9p21; those at 1 block predispose to atherosclerosis, whereas those at the other predispose to MI among patients with preexisting CAD.
Keywords: atherosclerosis; coronary angiography; coronary artery disease; haplotypes; myocardial infarction.