Abstract
BACKGROUND: Calcific aortic valve disease (CAVD), and ensuing severe aortic stenosis (AS), is the foremost valvular disorder of aging, yet preventive therapies are lacking. A better understanding of the molecular underpinnings of aortic valve calcification (AVC) is necessary to develop pharmacologic interventions. METHODS AND RESULTS: We undertook large-scale plasma proteomics in a cohort study of adults ≥65 years old, the CHS (Cardiovascular Health Study), to identify individual proteins associated with echocardiographic AVC and incident moderate/severe AS. Proteomics measurements were performed with the aptamer-based SomaLogic platform of ~5000 proteins. Significant proteins were validated in a second cohort, the AGES-RS (Age, Gene/Environment Susceptibility-Reykjavik Study), which assessed AVC and AS by computed tomography. The potential causal associations of replicated proteins were tested in 2-sample Mendelian randomization using identified cis protein quantitative trait loci in consortia having computed tomography-quantified AVC or AS as outcomes. Six proteins showed Bonferroni-corrected significant relationships with AVC in CHS. Three of these, CXCL-12 (C-X-C chemokine ligand 12), KLKB1 (kallikrein), and leptin, replicated in AGES-RS, of which the former 2 are novel. Only 1 protein, CXCL6, which showed a near-significant association with AS in the replication cohort, was significantly (positively) associated with incident AS. Mendelian randomization analysis was conducted for KLKB1, CXCL12, and CXCL6, which supported a causal relationship for higher KLKB1 with lower AVC (beta=-0.25, P=0.009). CONCLUSIONS: This study of older adults newly identified and largely replicated associations of 3 circulating proteins with calcific aortic valve disease, of which the relationship of plasma KLKB1 may have a causal basis. Additional investigation is necessary to determine if KLKB1 could be harnessed for calcific aortic valve disease therapeutics.