Estimation of B-type Natriuretic Peptide Values from N-Terminal proBNP Levels

Abstract

Both brain natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) are established biomarkers that are necessary in the diagnosis and management of heart failure (HF). However, it is difficult to infer BNP concentration from NT-proBNP concentration for a clinician who is familiar with BNP. We investigated whether estimated BNP concentration from NT-proBNP has an equivalent prognostic strength compared with the actual BNP concentration in the prediction of future outcomes. We created a formula for estimating BNP concentration using multivariate analysis in a derivation cohort with known or suspected HF (n = 374). We determined whether the estimated BNP level had a similar prognostic power compared with the actual BNP and NT-proBNP levels in a validation cohort (n = 375). There was a strong correlation between log-transformed BNP and log-transformed NT-proBNP (r = 0.90) in the derivation cohort. We created two types of equation from the derivation cohort. During a median of 1 year of follow up, 49 major adverse cardiac events developed in the validation cohort. Cox proportional analysis revealed that the actual and estimated BNP levels represented equivalent and significant predictors of the future cardiovascular outcome. The estimated BNP levels calculated by our new formula showed a prognostic power similar to the actual BNP levels. This equation will be useful, especially for a physician who is not familiar with NT-proBNP testing.