Acute heart failure despite preserved ejection fraction after STEMI: structural and microvascular injury as dominant determinants
M Zdravkovic, M Popovic, P Zafirovska, A Jovic, D Radovanovic, A Klasnja, S Klasnja, A Milosevic, O Nedeljkovic-Arsenovic, R Maksimovic, M Polovina, D Trifunovic-Zamaklar, G Krljanac, G Stankovic, V PopadicAbstract
Background
Early acute heart failure (HF) following ST-elevation myocardial infarction (STEMI) is conventionally attributed to impaired systolic function. However, a clinically relevant subset develops acute HF despite preserved left ventricular ejection fraction (LVEF), suggesting alternative pathophysiological drivers.
Purpose
To determine independent determinants of early acute HF among STEMI patients with preserved LVEF (>50% by CMR) and evaluate the incremental predictive value of structural myocardial injury.
Methods
Within the multicenter SINERGY-ACUTE cohort, patients with preserved LVEF (>50% by CMR) were analyzed (n=143 of 268 patients). Early acute HF was defined as Killip class ≥II and/or elevated NT-proBNP at admission. Structural parameters included infarct size (LGE%), LV volumes, microvascular obstruction (MVO), and intramyocardial hemorrhage (IMH). Penalized multivariable logistic regression was applied. Hierarchical models were constructed: (1) clinical; (2) clinical + LVEF; (3) clinical + LVEF + structural CMR parameters. Interaction between preserved EF and infarct size was formally tested. Discrimination was assessed by AUC (95% CI) with DeLong testing. Incremental value was evaluated by ΔAUC, Net Reclassification Improvement (NRI), and Integrated Discrimination Improvement (IDI). Internal validation used 1000-bootstrap resampling with calibration and decision curve analyses.
Results
Among patients with preserved LVEF (n=143), early acute HF occurred in 20.3%, 54.5 were male and 45.5% female with the median age of 60 years. Within the preserved EF range, LVEF was not independently associated with HF after adjustment. Independent determinants were infarct size (OR 1.27 per 5% increase; p<0.001), LV end-systolic volume (OR 1.21 per 10 mL; p<0.001), MVO (OR 1.93; p<0.001), IMH (OR 2.18; p<0.001), and log NT-proBNP (OR 1.51; p<0.001). The interaction between preserved EF and infarct size was not significant. Clinical model AUC was 0.69, increasing to 0.71 with LVEF and to 0.86 with structural CMR integration (ΔAUC +0.15; p<0.001). Reclassification improved significantly (NRI 0.41; IDI 0.18; both p<0.001). Bootstrap-corrected AUC was 0.84 with preserved calibration.
Conclusions
In STEMI patients with preserved LVEF, early acute HF is driven by structural myocardial injury and microvascular damage rather than systolic dysfunction. CMR-derived infarct size, MVO, and IMH provide substantial incremental discrimination beyond EF-based assessment, challenging reliance on preserved LVEF as a marker of low early HF risk.