The prognostic value of NT-proBNP variability measured during routine follow-up in outpatients with heart failure with reduced ejection fraction
I Can, H Kafes, B Demirkan, A TemizhanAbstract
Background
This study investigated whether serial variability in NT-proBNP levels during routine follow-up provides prognostic value for adverse outcomes in clinically stable HFrEF patients.
Methods
This single-center, retrospective observational study included 323 HFrEF patients (LVEF ≤40%) with ≥1 year of clinical stability and ≥4 NT-proBNP measurements obtained at ≥1-month intervals. NT-proBNP variability was calculated using the coefficient of variation (CV)(CV formula:standard deviation/mean× 100),and patients were classified as high or low variability based on the cohort median. The primary composite endpoint included all-cause mortality, heart transplantation or mechanical circulatory support, and HF-related hospitalization. Mean NT-proBNP levels during the stable period were also calculated, and patients were divided into high or low NT-proBNP groups according to the median. To examine the combined prognostic relevance of NT-proBNP level and variability, four subgroups were created: High mean NT-proBNP with high variability (H-NP/H-V),high mean NT-proBNP with low variability(H-NP/L-V),low mean NT-proBNP with high variability(L-NP/H-V),low mean NT-proBNP with low variability(L-NP/L-V). Chronological patterns were evaluated using average values of the first four sequential NT-proBNP measurements.
Results
No significant difference was found in the incidence of the composite endpoint between high- and low-variability groups (p = 0.84). The median NT-proBNP level was 1062 pg/mL, defining low (<1062 pg/mL, n = 162) and high (≥1062 pg/mL, n = 161) NT-proBNP strata. Outcomes were significantly worse in the high NT-proBNP group (p = 0.001), but variability did not influence prognosis within either stratum. Subgroup comparisons showed similar results: variability was not prognostic, whereas elevated NT-proBNP consistently predicted adverse outcomes. Kaplan–Meier curves for the four subgroups are shown in Figure 1.
Conclusion
NT-proBNP variability during a one-year stable period was not associated with adverse HF outcomes, whereas higher NT-proBNP levels strongly predicted risk. In the high-variability subgroup, baseline NT-proBNP levels were higher and declined progressively across sequential measurements (Figure 2). This downward trajectory may have increased variability while attenuating its association with adverse events. Nevertheless, persistently elevated NP levels remain clinically important, and the key determinant of prognosis appears to be the direction of change in NP concentrations over time and fluctuations in NP levels among stable patients with persistent HFrEF are not invariably associated with adverse outcomes and may, in some cases, reflect treatment response.Figure 1For image description, please refer to the figure legend and surrounding text.Figure 2For image description, please refer to the figure legend and surrounding text.