Williams, L Keoki

BACKGROUND: Heart failure remains a global health burden, and patients hospitalized are particularly at risk, but genetic associates for subsequent death or rehospitalization are still lacking. METHODS: The genetic substudy of the ASCEND-HF trial (Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure) was used to perform genome-wide association study and transethnic meta-analysis. The overall trial included the patients of self-reported European ancestry (n=2173) and African ancestry (n=507). The end point was death or heart failure rehospitalization within 180 days. Cox models adjusted for 11 a priori predictors of rehospitalization and 5 genetic principal components were used to test the association between single-nucleotide polymorphisms and outcome. Summary statistics from the 2 populations were combined via meta-analysis with the significance threshold considered P <5×10 − 8 . RESULTS: Common variants (rs2342882 and rs35850039 in complete linkage disequilibrium) located in FGD5 were significantly associated with the primary outcome in both ancestry groups (European Americans: hazard ratio [HR], 1.38; P =2.42×10 −6 ; African ancestry: HR, 1.51; P =4.43×10 − 3 ; HR in meta-analysis, 1.41; P =4.25×10 −8 ). FGD5 encodes a regulator of VEGF (vascular endothelial growth factor)-mediated angiogenesis, and in silico investigation revealed several previous genome-wide association study hits in this gene, among which rs748431 was associated with our outcome (HR, 1.20; meta P <0.01). Sensitivity analysis proved FGD5 common variants survival association did not appear to operate via coronary artery disease or nesiritide treatment ( P >0.05); and the signal was still significant when changing the censoring time from 180 to 30 days (HR, 1.39; P =1.59×10 −5 ). CONCLUSIONS: In this multiethnic genome-wide association study of ASCEND-HF, single-nucleotide polymorphisms in FGD5 were associated with increased risk of death or rehospitalization. Additional investigation is required to examine biological mechanisms and whether FGD5 could be a therapeutic target. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT00475852.