Genome-wide association study of platelet factor 4/heparin antibodies in heparin-induced thrombocytopenia

Abstract

Heparin, a widely used anticoagulant, carries the risk of an antibody-mediated adverse drug reaction, heparin-induced thrombocytopenia (HIT). A subset of heparin-treated patients produces detectable levels of antibodies against complexes of heparin bound to circulating platelet factor 4 (PF4). Using a genome-wide association study (GWAS) approach, we aimed to identify genetic variants associated with anti-PF4/heparin antibodies that account for the variable antibody response seen in HIT. We performed a GWAS on anti-PF4/heparin antibody levels determined via polyclonal enzyme-linked immunosorbent assays. Our discovery cohort (n = 4237) and replication cohort (n = 807) constituted patients with European ancestry and clinical suspicion of HIT, with cases confirmed via functional assay. Genome-wide significance was considered at α = 5 × 10-8. No variants were significantly associated with anti-PF4/heparin antibody levels in the discovery cohort at a genome-wide significant level. Secondary GWAS analyses included the identification of variants with suggestive associations in the discovery cohort (α = 1 × 10-4). The top variant in both cohorts was rs1555175145 (discovery β = -0.112 [0.018], P = 2.50 × 10-5; replication β = -0.104 [0.051], P = .041). In gene set enrichment analysis, 3 gene sets reached false discovery rate-adjusted significance (q < 0.05) in both discovery and replication cohorts: "Leukocyte Transendothelial Migration," "Innate Immune Response," and "Lyase Activity." Our results indicate that genomic variation is not significantly associated with anti-PF4/heparin antibody levels. Given our power to identify variants with moderate frequencies and effect sizes, this evidence suggests genetic variation is not a primary driver of variable antibody response in heparin-treated patients with European ancestry.

Graphical abstract

Figure 1.

Genome-wide association of anti-PF4/heparin antibodies in the discovery cohort (A) and replication cohort (B). P values were generated using linear regression adjusted for sex, age, and PCs 1 to 3 in an additive model. P values on the −log₁₀ scale are plotted on the left vertical axis, and the chromosomal position is plotted along the horizontal axis. The significance threshold of 5 × 10⁻⁸ is indicated by the solid red horizontal line, and the suggestive significance threshold of 1 × 10⁻⁴ is indicated by the dashed red line. Betas are indicated by dot color as described in the legend on the right.

Figure 2.

QQ plot for the genome-wide association of anti-PF4/heparin antibodies for discovery cohort (A) and replication cohort (B). QQ plots were generated from the primary GWAS analysis results. P values were generated using multivariate linear regression adjusted for age, sex, and principal components 1 to 3 in an additive model. Observed P values on the −log₁₀ scale are plotted on the left vertical axis, and expected P values on the −log₁₀ scale are plotted along the horizontal axis.

Figure 3.

Distribution of anti-PF4/heparin antibody titers. Antibody titer distribution by the genotype of the 2 most highly associated SNPs that passed the suggestive significance threshold in the discovery cohort and were associated in the replication cohort. Violin plots show the proportion of individuals across antibody titer values (optical density [OD]) within each genotype. The top panels show rs1555175145 in the discovery cohort (A) and replication cohort (B). The bottom panels show rs149530346 in the discovery cohort (C) and replication cohort (D). β, standard errors, and P values using linear regression adjusted for age, sex, and principal components 1 to 3 are listed at the bottom of each figure.