Excess long-term risk of adverse outcomes in heart failure patients with high and low levels of NT-proBNP: A 7-year follow-up study (NorthStar Trial)

Abstract

Background: This study investigated excess risk in patients with heart failure with reduced left ventricular ejection fraction (HFrEF) with or without elevated levels of NT-proBNP (N-terminal pro-brain natriuretic peptide).

Methods: Patients with HFrEF from the NorthStar cohort (n = 1120) were matched on age, sex, and presence of AF (atrial fibrillation/flutter) to five controls without HFrEF from The Danish National Patient Registries. Patients were compared with controls before and after stratification according to baseline NT-proBNP levels, with cutoffs defined as </≥ 600 pg/ml in patients with sinus rhythm and </≥ 900 pg/ml in patients with AF. The primary composite endpoint was a 7-year risk of cardiovascular death or HF admission.

Results: In the HFrEF cohort, 704 patients had high NT-proBNP (median age, 73; mean left ventricular ejection fraction (LVEF), 33%). 416 patients had low NT-proBNP (median age, 65; LVEF, 30%). Patients from both groups were in NYHA class I-III. The primary endpoint occurred in 531 patients (75.4%) with HFrEF and elevated NT-proBNP, and 748 controls (21.3%) (risk difference, 54.2%; 95% confidence interval (CI) 50.7-57.6%). In comparison, it occurred in 199 patients (47.9%) with HFrEF and without elevated NT-proBNP, and 185 controls (8,9%) (risk difference, 38.9%; 95% CI 34.0-43.9%). Risk differences for all secondary endpoints were significant, except for overall mortality in the low NT-proBNP group (risk difference, 3.8%; 95% CI, -0.4-8.0%).

Conclusion: This study identified a significant excess risk in patients with HFrEF across various endpoints, which persisted after stratification into high and low levels of NT-proBNP.

Keywords: Excess risk; Heart failure; NT-proBNP; NorthStar.

Graphical abstract

Fig. 1

Flowchart. Inclusion, exclusion, and distribution of patients and matched controls.

Fig. 2

Cumulative incidences of adverse endpoints. Comparison between patients with HFrEF and matched controls without HF. A) Time to CV death or HF admission. B) Time to CV death. C) Time to HF admission. D) Time to all-cause death. E) Time to non-HF admission. AR: absolute risk; RD: risk difference; RR: relative risk.

Fig. 3

Forest plot comparing risk differences between patients and controls. A) All patients with HFrEF compared with matched controls without HF. B) Patients with HFrEF and high levels of NT-proBNP compared with matched controls without HF. C) Patients with HFrEF and low levels of NT-proBNP compared with matched controls without HF.

Fig. 4

Age-stratified analysis of the risk of the primary endpoint, CV death or HF admission. A) Patients with HFrEF were divided into three age groups (<65, 65–75, >75) and compared with matched controls. B) Patients with HFrEF and high NT-proBNP compared to controls. C) Patients with HFrEF and low NT-proBNP compared to controls.

Fig. 5

Time until the first prescription of loop diuretics. A) All patients with HFrEF compared with matched controls without HF. B) Patients with HFrEF and high levels of NT-proBNP compared with matched controls without HF. C) Patients with HFrEF and low levels of NT-proBNP compared with matched controls without HF.