Sex-specific inflammatory proteomic signatures predict aortic valve stenosis: evidence from multi-omics analyses
S Shu, Q Yang, D Wu, Z Zeng, W Yuan, T Yang, X Wang, Z Zhang, Z Luo, Y Ma, H Shi, Y Yang, J Song, C Yuan, S HuAbstract
Background
Aortic valve stenosis (AVS) is a growing cardiovascular burden with marked sex disparities. Inflammation is central to its pathogenesis, but whether circulating inflammatory proteins can serve as predictive biomarkers and therapeutic targets remains unclear.
Methods
Among 45 788 UK Biobank participants without valvular heart disease and with plasma proteomics measurements, we evaluated the predictive value of 792 plasma proteins for incident AVS in sex-stratified analyses. Case-control analyses identified sex-specific inflammatory alterations. Predictive modeling was performed using LASSO for variable selection and Cox regression for outcome estimation, yielding Inflammatory Protein Scores (IPS) that were assessed for discrimination, risk stratification, and 15-year absolute risk prediction, and deployed in an interactive web-based tool. Causality was examined using two-sample Mendelian randomization. Finally, single-cell, bulk, and spatial RNA sequencing were integrated to map the cellular, temporal, and spatial expression of genes encoding selected proteins in human valves.
Results
Both sexes demonstrated widespread proteomic alterations, with more than 60% of inflammatory proteins elevated in AVS cases. Over a median 13.8 years of follow-up, 315 men and 184 women experienced first AVS. In prospective analyses, 109 proteins in females and 18 in males showed significant predictive value for incident AVS. Based on these findings, Inflammatory Protein Scores (IPS) integrating 45 proteins in females and 17 in males achieved high predictive accuracy, with C-indices of 0.934 (95% CI, 0.900-0.968) and 0.890 (95% CI, 0.854-0.926), respectively. Two-sample Mendelian randomization further supported causal roles for 29 proteins in females and 5 in males. Transcriptomic profiling localized these proteins predominantly to valvular endothelial and interstitial cells, revealed dynamic expression trajectories across fibrotic and calcific stages, and demonstrated spatial co-localization with fibrocalcific markers.
Conclusion
Sex-specific inflammatory proteomic signatures provide powerful prediction of AVS risk and identify novel therapeutic targets, offering a precision-medicine framework for early detection and anti-inflammatory intervention.Figure 1Figure 2