Combined Prognostic Value of Follow-Up Ejection Fraction and Natriuretic Peptide Measurements in Heart Failure With Reduced Ejection Fraction Following Initiation of Pharmacotherapy.
Kodur N, Gunsalus P, et al. • Journal of the American Heart Association • 2026
Among patients with HFrEF who have initiated guideline-directed medical therapy, NT-proBNP may offer much greater prognostic value than LVEF, with even mildly to moderately elevated NT-proBNP levels portending meaningful incremental risk.
Key Findings
Results
NT-proBNP had a much stronger association with composite outcome and greater discriminatory ability than LVEF in HFrEF patients on guideline-directed medical therapy.
The difference in concordance index between NT-proBNP and LVEF was +0.129 (95% CI, 0.118–0.141), favoring NT-proBNP.
LVEF offered only minimal prognostic value beyond NT-proBNP.
Both LVEF and NT-proBNP offered independent prognostic information.
The composite outcome assessed was HF hospitalization or all-cause death.
Results
Even mildly to moderately elevated NT-proBNP levels (125–1000 pg/mL) were associated with incremental risk of the composite outcome.
Higher NT-proBNP levels were associated with higher adjusted 1-year risk of the composite outcome across the spectrum of LVEF.
The association between elevated NT-proBNP and risk was observed even at NT-proBNP levels of 125–1000 pg/mL.
The median NT-proBNP in the cohort was 2161 pg/mL (IQR, 610–7350).
The risk relationship was assessed across the full spectrum of observed LVEF values.
Methods
The study cohort consisted of 6168 patients with HFrEF who had both LVEF and NT-proBNP measurements approximately 9 months following diagnosis.
The study was a retrospective cohort study conducted from 2009 to 2024 at a large, integrated health system.
Inclusion required a diagnosis of HFrEF (LVEF ≤40%) along with available LVEF and NT-proBNP measurements at approximately 9 months following diagnosis, defined as the study baseline.
The median LVEF at baseline was 36% (IQR, 27–46).
A large majority of patients were on at least 2 guideline-directed medical therapy medications at baseline.
Results
The median LVEF at study baseline was 36% and the median NT-proBNP was 2161 pg/mL among HFrEF patients approximately 9 months after diagnosis.
Median LVEF was 36% (IQR, 27–46).
Median NT-proBNP was 2161 pg/mL (IQR, 610–7350).
These measurements were taken approximately 9 months following initial HFrEF diagnosis.
The wide interquartile ranges reflect substantial variability in both LVEF and NT-proBNP across the cohort.
Background
Few prior studies had directly compared the prognostic value of LVEF and NT-proBNP in HFrEF patients who have initiated guideline-directed medical therapy.
The study was designed specifically to address this gap by assessing combined prognostic value of follow-up LVEF and NT-proBNP measurements.
Measurements were taken following initiation of guideline-directed medical therapy rather than at initial diagnosis.
The study context is important because treatment-era prognostic relationships may differ from those observed at initial presentation.
What This Means
This research studied over 6,000 patients with heart failure with reduced ejection fraction (HFrEF) — a condition where the heart pumps less blood than normal — to understand which test better predicts serious outcomes like hospitalization or death. Two commonly used measurements were compared: the left ventricular ejection fraction (LVEF, a measure of how well the heart pumps, typically assessed by echocardiogram) and NT-proBNP (a blood protein released when the heart is under stress). Measurements were taken roughly 9 months after patients started standard heart failure medications.
This research suggests that NT-proBNP was a far stronger predictor of hospitalization or death than LVEF in these patients. Specifically, NT-proBNP provided substantially better ability to distinguish between patients at high versus low risk, with a concordance index improvement of 0.129 over LVEF alone. Notably, even mildly to moderately elevated NT-proBNP levels (125–1000 pg/mL) were associated with meaningfully increased risk, suggesting that the threshold for concern may be lower than sometimes assumed. While both measurements provided some independent prognostic information, LVEF added only minimal value once NT-proBNP was already accounted for.
These findings matter because they suggest that in patients already being treated for HFrEF, a blood test (NT-proBNP) may tell clinicians more about future risk than a cardiac imaging measurement (LVEF). This research suggests that ongoing monitoring of NT-proBNP levels — not just tracking whether the ejection fraction has improved — could be especially important for identifying patients who remain at high risk despite being on medications. This could potentially influence how clinicians prioritize follow-up testing and risk stratification in this patient population.
Kodur N, Gunsalus P, Milinovich A, Dalton J, Tang W. (2026). Combined Prognostic Value of Follow-Up Ejection Fraction and Natriuretic Peptide Measurements in Heart Failure With Reduced Ejection Fraction Following Initiation of Pharmacotherapy.. Journal of the American Heart Association. https://doi.org/10.1161/JAHA.125.047015