Abstract 3032: Molecular Fingerprints of Aortic Valve Diseases: Gene Expression Clustering Specific for Stenosis versus Regurgitation
Background: Aortic stenosis (AS) and aortic regurgitation (AR) cause characteristic changes in cardiac geometry, myocyte biology and extracellular matrix (ECM) composition. From experimental data, pressure loading in AS causes concentric left ventricular (LV) and myocyte hypertrophy with ECM collagen accumulation; in AR LV hypertrophy is eccentric and ECM fibronectin accumulates. As yet, assessment of the underlying molecular pathophysiology largely has focused on specific target genes and their products from cultures of isolated cells, almost exclusively in experimental models. To evaluate the integrated response of the human myocardium to AS and AR, gene expression profiling was performed on intra-operative LV biopsies..
Method: Tru-cut biopsies were obtained at the thickest portion of the LV free wall from pts undergoing aortic valve replacement for isolated, pure AS (n=5) or isolated, pure AR (n=4), and from “control” pts with coronary artery disease (CAD) but no infarction undergoing elective coronary artery bypass grafting (CABG) (n=3). RNA was prepared using Qiagen kits and used to synthesize labeled cRNA to probe Affymetrix microarrays (Human Genome U133 Plus 2.0). Data were analyzed using dCHIP 2006 and GenMAPP 2.1.
Results: Comparison of RNA expression levels (AS vs AR, AS vs CAD, AR vs CAD) revealed >1.2-fold relative expression increase or decrease (p<.05 each comparison) of 777 genes. Among them were genes encoding ECM components, integrins and degradative enzymes, as well as components of signal transduction pathways for TGFβ, PDGF and FGF. AS differed from AR in 250 genes, including adrenergic type 1 receptors and FGF receptors. Cluster analysis of expression levels among the,777 significantly up- or down-regulated genes grouped patients according to diagnosis with 100% accuracy.
Conclusion: Unique patterns of gene expression are associated with chronic exogenous overload states in AS and AR compared with CAD, and with AS vs AR. Hundreds of genes are uniquely and characteristically affected for each disease. Further analysis must define the relation of these findings to post-op outcome, and must identify potential targets for beneficial intervention in the remodeling process before and after surgery.