Mapping the aetiological foundations of the heart failure spectrum using human genetics

Abstract

Summary paragraphHeart failure (HF), a syndrome of symptomatic fluid overload due to cardiac dysfunction, is the most rapidly growing cardiovascular disorder. Despite recent advances, mortality and morbidity remain high and treatment innovation is challenged by limited understanding of aetiology in relation to disease subtypes. Here we harness the de-confounding properties of genetic variation to map causal biology underlying the HF phenotypic spectrum, to inform the development of more effective treatments. We report a genetic association analysis in 1.9 million ancestrally diverse individuals, including 153,174 cases of HF; 44,012 of non-ischaemic HF; 5,406 cases of non-ischaemic HF with reduced ejection fraction (HFrEF); and 3,841 cases of non-ischaemic HF with preserved ejection fraction (HFpEF). We identify 66 genetic susceptibility loci across HF subtypes, 37 of which have not previously been reported. We map the aetiologic contribution of risk factor traits and diseases as well as newly identified effector genes for HF, demonstrating differential risk factor effects on disease subtypes. Our findings highlight the importance of extra-cardiac tissues in HF, particularly the kidney and the vasculature in HFpEF. Pathways of cellular senescence and proteostasis are notably uncovered, includingIGFBP7as an effector gene for HFpEF. Using population approaches causally anchored in human genetics, we provide fundamental new insights into the aetiology of heart failure subtypes that may inform new approaches to prevention and treatment.

Genetic and phenotypic architecture of human myocardial trabeculation

“…We used LocusZoom, eQTLs, and transcriptome-wide association study to prioritise genes at each locus. The identified associations include loci of definitive evidence cardiomyopathy-associated genes ( TNNT2, TTN, PLN ) and colocalise with recent HCM (e.g., TBX3, STRN, MTSS1 26 ), DCM (e.g., PKD1, STRN, MITF, CDKN1A 27 ), and heart failure GWAS 28 . For sensitivity analyses, we removed any individuals in the UK Biobank who had a diagnosis of any cardiomyopathy or heart failure.…”

“…Hyper-and hypo-trabeculation were deemed > or < 1.5 standard deviations from the mean. Mendelian randomization was assessed using GWAS Summary statistics from published literature [26][27][28] with mean global FD using the R packages TwoSampleMR and MVMR. Exposure variants were included if GWAS significant (P<5x10 −8 ).…”

“…Genetic correlations were assessed using –reml-bivar in GCTA for two traits ( Table S2 ). Polygenic risk scores for HCM and DCM were created previously from GWAS summary statistics and applied to the UK Biobank cohort 26,28,57 . Categorical variables were assessed using the Chi-Squared Test or Fisher’s Exact Test.…”

The Genetic Factors Influencing Cardiomyopathies and Heart Failure across the Allele Frequency Spectrum

“…Shah et al (2020) conducted a meta-analysis of HF GWAS using data from 29 studies encompassing 47,309 cases and 930,014 controls [39]. 11 genomic loci reached genome-wide significance (P < 5 × 10 −8 ; Table 1 [39][40][41][42]). Many of the loci identified have strong associations with cardiovascular disease risk factors such as FTO (fat mass and obesity-associated; associated with BMI), PITX2 (the PITX2 transcription factor; associated with atrial fibrillation [AF]), and LPA (lipoprotein(a); associated with coronary artery disease [CAD]).…”

“…Henry et al ( 2023) reported a GWAS meta-analysis on a cohort of 1.9 million ancestrally diverse people, including 153,174 cases of HF [42]. They identified 66 genetic susceptibility loci across HF subtypes (Table 1).…”