Genomic Modeling of Atherosclerosis in Peripheral Arterial Disease and Its Variant Phenotype in Patients with Diabetes

Abstract

Microarrays can be used to discover candidate genes associated with peripheral arterial disease (PAD) and develop models that predict patient clinical status. We hypothesize that multiple phenotypes of PAD with distinct patterns of gene expression exist. We histologically characterized and extracted ribonucleic acid from 31 arterial samples collected from the lower extremities of patients undergoing amputation or free fibular grafting. Analysis using the Affymetrix U133A microarray identified 335 genes with twofold or greater differences in expression between normal and diseased arteries ( p < .01) and 104 genes with twofold or greater differences between diabetic and nondiabetic atherosclerotic arteries ( p < .1). Many genes identified have known roles in inflammatory and lipid uptake pathways. Predictive models were developed that could predict PAD and the associated diabetic phenotype with an accuracy of 71 to 90%. Developing distinct genomic models of PAD will serve as the first step toward understanding the molecular and genetic basis of PAD and subsequent application of novel therapeutics to this condition.