Prognostic impact of new-onset atrial fibrillation in myocardial infarction with and without improved ejection fraction

Abstract

Aims: Improvement in left ventricular ejection fraction (impEF) often presents in contemporary acute myocardial infarction (AMI) patients. New-onset atrial fibrillation (NOAF) during AMI is an important predictor of subsequential heart failure (HF), while its impact on the trajectory of post-MI left ventricular ejection fraction (LVEF) and prognostic implication in patients with and without impEF remains undetermined. We aimed to investigate the prognostic impacts of NOAF in AMI patients with and without impEF.

Methods and results: Consecutive AMI patients without a prior history of AF between February 2014 and March 2018 with baseline LVEF ≤ 40% and had ≥1 LVEF measurement after baseline were included. ImpEF was defined as a baseline LVEF ≤ 40% and a re-evaluation showed both LVEF > 40% and an absolute increase of LVEF ≥ 10%. Persistently reduced EF (prEF) was defined as the second measurement of LVEF either ≤40% or an absolute increase of LVEF < 10%. The primary endpoint was a major adverse cardiac event (MACE) that was composed of cardiovascular death and HF hospitalization. Cox regression analysis and competing risk analysis were performed to assess the association of post-MI NOAF with MACE. Among 293 patients (mean age: 66.6 ± 11.3 years, 79.2% of males), 145 (49.5%) had impEF and 67 (22.9%) developed NOAF. Higher heart rate (odds ratio [OR]: 0.84, 95% confidence interval [CI]: 0.73-0.97; P = 0.015), prior MI (OR: 0.25, 95% CI: 0.09-0.69; P = 0.008), and STEMI (OR: 0.40, 95% CI: 0.21-0.77; P = 0.006) were independent predictors of post-MI impEF. Within up to 5 years of follow-up, there were 22 (15.2%) and 53 (35.8%) MACE in patients with impEF and prEF, respectively. NOAF was an independent predictor of MACE in patients with impEF (hazard ratio [HR]: 7.34, 95% CI: 2.49-21.59; P < 0.001) but not in those with prEF (HR: 0.78, 95% CI: 0.39-1.55; P = 0.483) after multivariable adjustment. Similar results were obtained when accounting for the competing risk of all-cause death (subdistribution HR and 95% CIs in impEF and prEF were 6.47 [2.32-18.09] and 0.79 [0.39-1.61], respectively).

Conclusions: The NOAF was associated with an increased risk of cardiovascular outcomes in AMI patients with impEF.

Keywords: Atrial fibrillation; Competing risk analysis; Improved ejection fraction; Myocardial infarction.

Figure 1

Flowchart of the study population. AF, atrial fibrillation; AMI, acute myocardial infarction; LVEF, left ventricular ejection fraction.

Figure 2

Selection of clinical characteristics related to post‐MI impEF using the LASSO regression model. (A) A total of 23 variables were introduced into the LASSO binary logistic regression analysis. (B) A coefficient profile plot was illustrated against the log (lambda) sequence. Selection of the optimal parameter by minimum criteria and 1‐s.e. criteria in the LASSO model. (C) Forest plot of 10 variables (other than the NOAF) selected by the LASSO regression analysis associated with the post‐MI impEF. (D) ROC curves and C‐indexes to assess the predictive performance of multivariable models with or without NOAF. ¹Variables in the multivariable model included admission HR, STEMI, and history of MI. HR, heart rate; impEF, improved ejection fraction; LASSO, least absolute shrinkage, and selection operator; ROC, receive‐operating characteristic curve; SBP, systolic blood pressure.

Figure 3

Kaplan–Meier curves of cumulative incidence of the primary and secondary outcomes stratified by rhythm status. Incidence of the primary endpoint, cardiovascular death, HF hospitalization, and all‐cause death in AMI patients with impEF and prEF. Patients with NOAF are indicated in red and those with sinus rhythm are in blue. The incidence rate was calculated using the total number of outcomes of interest divided by the person‐years at risk. prEF, persistently reduced ejection fraction.

Figure 4

Multivariable adjusted HRs and 95% CIs of the primary and secondary endpoints in AMI patients with (A) impEF and (B) prEF. ¹Variables in the fully adjusted multivariable model included GRACE score, sex, history of HF, CKD, primary PCI, NT‐pro BNP, and C‐reactive protein. CKD, chronic kidney disease; GRACE, Global Registry of Acute Coronary Events; HF, heart failure; MACE, major adverse cardiac events; PCI, percutaneous coronary intervention; SR, sinus rhythm.