THU0530 HIGHLY-SENSITIVE CARDIAC TROPONIN I AND BETA-2-GLYCOPROTEIN-I IGA ANTIBODIES INFORM THE UTILITY OF SCREENING AND FOLLOW-UP NON-INVASIVE CORONARY ATHEROSCLEROSIS EVALUATION AND OPTIMIZE CARDIOVASCULAR RISK ASSESSMENT IN RHEUMATOID ARTHRITIS

Background:Occult coronary atherosclerosis burden predicts mid-term cardiovascular disease (CVD) events in rheumatoid arthritis (RA) above and beyond Framingham D’Agostino cardiac risk score (FRS-DA). Highly-sensitive cardiac troponin I (hs-cTnI) levels in blood associate with coronary plaque burden and event risk in RA. Moreover, IgA antibodies against beta2-glycoprotein-1 (a-b2GPI-IgA)- an atherosclerotic plaque antigen- in RA promote coronary plaque progression and moderate the effect of inflammation on CVD events. It is currently unclear when to recommend a screening, non-invasive coronary atherosclerosis evaluation in asymptomatic RA patients and whether such an assessment should be repeated.Objectives:To explore whether either biomarker alone or their combination improved prediction of plaque presence on an initial coronary CT angiogram (CCTA) beyond FRS-DA score; to evaluate whether either biomarker predicted progression to extensive or obstructive plaque on a follow-up evaluation.Methods:One hundred fifty RA patients underwent a baseline CCTA; 101 had repeat evaluation within 6.9±0.3 years. Hs-cTnI and a-b2GPI-IgA were assessed at baseline; the latter were confirmed 12 weeks later, if positive. Lesions rendering greater than 50% luminal stenosis were considered obstructive. Extensive plaque was defined as >5 coronary segments with plaque, or stenosis score>5, or coronary artery calcium score (CAC)>100. The diagnostic accuracy of FRS-DA alone vs. with hs-cTnI or a-b2GPI-IgA individually or combined for plaque or CAC at baseline was evaluated as area under the curve (AUC). Improvement in prediction accuracy between constructs was further assessed as integrated discrimination improvement (IDI). Similar AUC and IDI constructs evaluated the transition to obstructive or extensive atherosclerosis at follow-up in patients with baseline non-extensive or non-obstructive disease.Results:High hs-cTnI (>1.5pg/ml) added to FRS-DA increased AUC from 0.717 to 0.731 (Figure 1A) and improved prediction accuracy for baseline plaque [IDI=0.041 (SE)=0.017, p=0.015]. In contrast, a-b2GPI-IgA did not [IDI=0.005 (0.006), p=0.47] and the combination offered no added benefit to the hs-cTnI model alone. Similar observations were made for CAC. Presence of a-b2GPI-IgA independently associated with coronary plaque progression (IRR=1.67 [95%CI 1.04-2.67]), whereas hs-cTnI did not. Likewise, a-b2GPI-IgA associated with transition to extensive or obstructive disease independently of FRS-DA (OR=13.48 [95%CI 2.09-86.99]). Notably, 71.4% of a-b2GPI-IgA positive patients with high hs-cTnI progressed to extensive or obstructive disease compared to 7.7% of a-b2GPI-IgA negative subjects with high hs-cTnI (p=0.008). Addition of a-b2GPI-IgA to FRS-DA in patients with prevalent non-extensive non-obstructive plaque increased AUC from 0.785 to 0.900 (Figure 1B) and significantly improved the prediction for development of obstructive or extensive atherosclerosis at follow-up [0.387, (0.13), p=0.003].Figure 1.(A) Diagnostic accuracy for prediction of occult coronary atherosclerosis at baseline. FRS-DA alone is the base model followed by addition of hs-cTnI or a-b2GPI-IgA individually or combined.(B) Diagnostic accuracy for progression from non-obstructive and non-extensive plaque at baseline to obstructive or extensive atherosclerosis at follow-up.Conclusion:High hs-cTnI improved the risk of baseline plaque presence beyond clinical risk score and may trigger an initial non-invasive coronary atherosclerosis evaluation. A-b2GPI-IgA presence may justify a follow-up evaluation in patients with non-extensive, non-obstructive plaque at baseline.Disclosure of Interests:George Karpouzas Grant/research support from: Pfizer, Consultant of: Sanofi-Genzyme-Regeneron, Janssen, Speakers bureau: Sanofi-Genzyme-Regeneron, BMS, Sarah Ormseth: None declared, Elizabeth Hernandez: None declared, Matthew Budoff: None declared