Abstract 16357: Identification of Circulating Proteins that Predict the Presence of Obstructive Coronary Artery Disease in Symptomatic, Non-Diabetic Patients Referred for Invasive Coronary Angiography

Abstract

Background: The diagnosis of patients presenting for the first time with typical or atypical symptoms consistent with obstructive coronary artery disease (CAD) remains challenging; despite the existence of a number of non-invasive modalities for the assessment of CAD, the yield of obstructive CAD in patients who undergo invasive coronary angiography remains low. We sought to identify circulating protein biomarkers that might aid in the diagnosis of obstructive CAD in non-diabetic patients. Methods: 386 non-diabetic patients from the PREDICT study (NCT00500617) were divided into two independent case:control sets for initial marker discovery (Set 1, n=187) and preliminary model validation (Set 2, n=199). Cases were defined as patients with ≥50% stenosis in ≥1 major coronary artery; controls had < 50% stenosis in any major coronary artery as determined by quantitative coronary angiography (QCA). In 11% of the patients QCA was not available and 70% stenosis by clinical read was used as a case threshold. 135 protein markers were evaluated in plasma from Set 1 patients using an electrochemoluminescence-based platform (MesoScale); the top markers were analyzed in Set 2 and used to develop predictive models using penalized logistic regression (LASSO). Model performance was estimated via 2500 iteration of cross validation on random holdout sets of 14 patients. Results: In a sex and age adjusted analysis, nine protein markers obtained nominal significance (p < 0.05) in Set 1, with 9 additional markers trending towards significance (p 0.7); the average of these pairs was used for subsequent analysis. Top markers were evaluated in Set 2 and a five marker (Adiponectin, ApoA1, NT-proBNP, S100A8-MPO, PlGF) model was fitted via LASSO, resulting in a mean cross-validated AUC of 0.64. Conclusion: We have identified a set of plasma protein markers in subjects referred for invasive angiography, which in age and sex adjusted analysis are responsive to the presence of obstructive CAD. Use of such protein markers alone, or in combination with other genomic markers or clinical risk factors may aid in the identification of obstructive CAD in non-acute, symptomatic patients.