Genome-Wide Association Study of Beta-Blocker Survival Benefit in Black and White Patients with Heart Failure with Reduced Ejection Fraction

Abstract

In patients with heart failure with reduced ejection fraction (HFrEF), individual responses to beta-blockers vary. Candidate gene pharmacogenetic studies yielded significant but inconsistent results, and they may have missed important associations. Our objective was to use an unbiased genome-wide association study (GWAS) to identify loci influencing beta-blocker survival benefit in HFrEF patients. Genetic variant × beta-blocker exposure interactions were tested in Cox proportional hazards models for all-cause mortality stratified by self-identified race. The models were adjusted for clinical risk factors and propensity scores. A prospective HFrEF registry (469 black and 459 white patients) was used for discovery, and linkage disequilibrium (LD) clumped variants with a beta-blocker interaction of p < 5 × 10^-5, were tested for Bonferroni-corrected validation in a multicenter HFrEF clinical trial (288 black and 579 white patients). A total of 229 and 18 variants in black and white HFrEF patients, respectively, had interactions with beta-blocker exposure at p < 5 × 10^-5 upon discovery. After LD-clumping, 100 variants and 4 variants in the black and white patients, respectively, remained for validation but none reached statistical significance. In conclusion, genetic variants of potential interest were identified in a discovery-based GWAS of beta-blocker survival benefit in HFrEF patients, but none were validated in an independent dataset. Larger cohorts or alternative approaches, such as polygenic scores, are needed.

Keywords: beta-blocker; genome-wide association study; heart failure with reduced ejection fraction; survival.

Figure 1

QQ plots in self-identified black patients (A) and white patients (B). The solid red lines are the p-values expected by chance alone. The black points are the observed p-values for each locus tested in the GWAS.

Figure 2

Manhattan plot in black patients. The horizontal red line indicates p-value = 1 × 10⁻⁸. The horizontal blue line indicates p = 5 × 10⁻⁵. Variants of interest were defined as p < 5 × 10⁻⁵. The red circle indicates p = 1.3 × 10⁻¹¹ for rs117032090 on chromosome 7.

Figure 3

Manhattan plot in white patients. The horizontal red line indicates p-value = 1 × 10⁻⁸. The horizontal blue line indicates p = 5 × 10⁻⁵. Variants of interest were defined as p < 5 × 10⁻⁵.