Abstract 13259: Large-Scale Plasma Proteomics Identifies C1QTNF1 as a Potential Causal Risk Factor for Aortic Stenosis
Khaled A Shelbaya,Victoria Lamberson,Yimin Yang,Pranav Dorbala,Leo Buckley,Brian Claggett,Hicham Skali,Line Dufresne,Jamie C Engert,Ta-Yu Yang,George Thanassoulis,James S Floyd,Thomas Austin,Anna E Bortnick,Jorge R Kizer,Renata Caroline C Costade Freitas,Sasha A Singh,Elena Aikawa,Christie M Ballantyne,Ron C Hoogeveen,Bing Yu,Josef Coresh,Kuni Matsushita,Amil M Shah
DOI: https://doi.org/10.1161/circ.148.suppl_1.13259
IF: 37.8
2023-11-07
Circulation
Abstract:Background: Despite the increasing prevalence of aortic stenosis (AS) and associated morbidity, limited data exist regarding circulating risk factors for AS development. Methods: Among Atherosclerosis Risk in Communities study participants with available proteomics (Somascan v4) at study Visit 5 (2011-13; n=4,899; age 76 ± 5 years, 57% women), we used multivariable linear regression to evaluate the association of 4,877 plasma proteins with peak aortic valve (AV) velocity and AV dimensionless index. We then tested their associations, when assessed at study Visit 3 (1993-95; n=11,430; age 60 ± 6, 54% women), with incident AV-related hospitalization post Visit 3 (median follow-up 22, IQR 14 - 25 years) using multivariable Cox PH regression. For the resulting candidate proteins, we tested associations with incident adjudicated AS in the Cardiovascular Health study (CHS) and assessed for potentially causal effects on AS and AV peak velocity using two sample Mendelian Randomization (MR). pQTLs were obtained from 3 previously published European studies, summary statistics for AS from the recently completed TARGET GWAS, and summary statistics for AV peak velocity from an ARIC GWAS. Results: We identified 52 plasma proteins with consistent associations with AV peak velocity, AV dimensionless index, and incident AV hospitalization. Of these 52 proteins, genetic instrumental variables (IVs) for MR analysis were available for 42. After multiple testing correction, a multi-SNP IV for C1QTNF1 demonstrated a potential causal effect on AS, such that higher genetically determined levels of C1QTNF1 associated with greater odds of AS. That same multi-SNP IV also demonstrated concordant MR association with peak AV velocity. C1QTNF1 was also associated with AS events in CHS (Figure). Conclusion: Large-scale proteomics in a longitudinal community-based cohort identifies C1QTNF1, an adiponectin paralog associated with inflammation, as a potential novel risk factor for AS.
cardiac & cardiovascular systems,peripheral vascular disease