Predictive Value of Serum N-Terminal pro-B-Type Natriuretic Peptide and Troponin T for Incident Heart Failure: A Meta-Analysis of 9 International Cohorts.
Nisha Bansal, Morgan E Grams, Josef Coresh, Kunihiro Matsushita, Shoshana H Ballew, Yingying Sang, Aditya Surapaneni, Johan Ärnlöv, Samira Bell, Jarett D Berry, Kevin Damman, James A de Lemos, Mirela Dobre, Shih-Jen Hwang, Ron T Gansevoort
Abstract
Open AccessBACKGROUND: Prior prediction equations for heart failure (HF) omitted cardiac biomarkers and used select populations. We assessed the added value of NT-proBNP (NT-terminal pro-brain natriuretic peptide) and hsTnT (high-sensitivity troponin T) as predictors of HF, across a broad population, including participants with chronic kidney disease or atherosclerotic cardiovascular disease. METHODS: Among 41 427 individuals free of HF from 9 prospective cohort studies, we performed an individual-participant data meta-analysis, quantifying the associations of NT-proBNP and hsTnT with incident HF when added to a clinical model. Changes in Harrel's C-statistic with and without NT-proBNP or hsTnT were estimated within each cohort and then pooled using random effects meta-analysis. RESULTS: Over a mean of 11 years, 4599 incident HFs occurred. A 2-fold higher NT-proBNP was associated with incident HF (meta-analyzed hazard ratio [HR], 1.47 [95% CI, 1.38-1.56]), P<0.001, with no interaction between NT-proBNP and estimated glomerular filtration rate (HR, 0.97 [95% CI, 0.94-1.02], P=0.33) or atherosclerotic cardiovascular disease (HR, 0.97 [95% CI, 0.93-1.02], P=0.48). Adding NT-proBNP significantly increased the C-statistic by 0.030 (95% CI, 0.021-0.040, P<0.001) over the clinical model. Similar results were seen for hsTnT [meta-analyzed HR per 2-fold higher hsTnT 1.41 (95% CI, 1.30-1.54), P<0.001], but C-statistic increased by only 0.008 (95% CI, 0.005-0.011, P<0.001). In models that included NT-proBNP, addition of hsTnT had a smaller effect (change in C-statistic, 0.002 [95% CI, 0.001-0.003], P<0.001). CONCLUSIONS: NT-proBNP improved risk discrimination of incident HF when added to traditional HF risk factors, even in individuals with chronic kidney disease and atherosclerotic cardiovascular disease. The contribution of hsTnT was modest. Measurement of NT-proBNP may help identify individuals at risk of HF.