Abstract 15220: Development of a Prediction Model for Atrial Fibrillation in Patients With Heart Failure and Preserved Ejection Fraction: Secondary Analysis of Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist (topcat) Trial

Abstract

Background: Atrial fibrillation (AF) is a common comorbidity in heart failure with preserved ejection fraction (HFpEF) and portends an increased risk of cardiovascular events. We sought to identify predictors and develop a risk score of incident AF among patients with HFpEF. Methods: This was an exploratory, post-hoc analysis of the TOPCAT trial. Patients without known AF were included. Cox regression was used to identify independent predictors of incident AF. A risk score was derived from the weighed sum of the regression coefficients of each independent risk factor in the final model using Cox regression analysis. Results: A total of 2174 patients (mean age 67.0±9.4 years; female 55%) without known AF at baseline were included. During a median follow-up of 3 years, 102 (4.7%) patients developed new onset AF. Diabetes (HR=2.1, 95% CI 1.4-3.1; p=0.0002), peripheral arterial disease (HR=2.0, 95% CI 1.2-3.4; p=0.006), elevated (>144meq/dL) sodium (HR=2.1, 95% CI 1.4-3.1; p=0.0002) independently predicted incident AF, whereas current use of angiotensin-converting enzyme inhibitors/angiotensin receptor blockers was protective (HR=0.61, 95% CI 0.38-0.99, p=0.048). Based on the simplified risk score which included these 4 variables, annualized AF incidence rates were 0.8%, 1.8%, and 3.6% in the low (score=0), intermediate (score=1 or 2), and high-risk (score >2) groups, respectively (log rank P<0.0001; Figure). Compared to the low risk group, the intermediate and high risk groups had a 2.5-fold and 5-fold increase in the risk of incident AF, respectively (HR=2.5, 95% CI 1.5-4.0, p=0.0003 and HR=4.9, 95% CI 2.9-9.4, p<0.0001, respectively). Model discrimination was good (c-statistic=0.67; 95% CI 0.61-0.72). Conclusions: A simplified risk score derived from clinical and laboratory characteristics predicts incident AF in patients with HFpEF and, upon further validation, may be used clinically for risk stratification or for AF screening in high risk groups. Figure