Abstract 407: Use of a Previously Validated Blood-based Test Demonstrates Increased Diagnostic Accuracy as Measured by AUC over Usual Care in the Evaluation of Obstructive Coronary Artery Disease in Males and Females.

Abstract

Background: Current evaluation of stable, non-acute patients presenting with symptoms suggestive of obstructive coronary artery disease (CAD) is costly and often exposes patients to radiation and contrast-dye side effects. These risks are coupled with relative poor diagnostic accuracy, as consistently demonstrated by low yields at invasive coronary angiography. We hypothesized that the use of a previously validated blood-based test incorporating age, sex and whole-blood gene expression, in conjunction with a clinician’s clinical assessment, may improve on usual care methods for the evaluation of these patients. Methods and Results: This analysis includes evaluable data from two prospective multicenter clinical studies [[Unable to Display Character: –]] PREDICT (NCT005617, N=523) and COMPASS (NCT1117506, N=431) where patients were assessed for both pre-test CAD risk according to Diamond-Forrester (D-F) criteria as well as for obstructive CAD using either invasive coronary angiography or cardiac computed tomography angiography (CCTA). All patients in COMPASS were also assessed by myocardial perfusion imaging (MPI); a subset of N=307 subjects were assessed by MPI in PREDICT. Previously, we demonstrated diagnostic superiority, as measured by AUC, for a score combining patient age, sex and whole-blood gene expression (ASGES), in a combined set of men and women from both PREDICT (ASGES = 70%, D-F = 66%, MPI = 54%) and COMPASS (ASGES = 79%, D-F = 69%, MPI = 59%). In this expanded analysis, we report results stratified by sex and demonstrate superiority of the ASGES, as measured by AUC, to MPI for males and females in PREDICT (male ASGES = 66%, MPI = 55%; female ASGES = 65%, MPI = 48%) and in COMPASS (male ASGES = 73%, MPI=60%; female ASGES 73%, MPI = 55% respectively). In addition, we demonstrate that ASGES improves CAD risk classification when compared to D-F criteria in females in both PREDICT (ASGES = 65%, D-F = 51%) and COMPASS (ASGES = 73%, D-F = 58%). Conclusions: We demonstrate that use of a gender-specific, blood-based test incorporating age, sex, and gene expression provides better diagnostic accuracy for patients considered for referral to cardiology and advanced cardiac testing, when compared to usual care methods of D-F type risk classification and MPI.