Development of Useful Conversion Equation between B-type Natriuretic Peptide and N-terminal pro BNP Levels

Abstract

Introduction Measurement of B-type natriuretic peptide (BNP) and N-terminal pro B-type natriuretic peptide (NT-proBNP) for diagnosis and prognosis of Heart Failure (HF) is recommended as a Class I indication by the ACCF/AHA guidelines. Although both molecules are released in equimolar proportions, their levels differ by clinical conditions, such as age, sex, body mass index (BMI), renal function, anemia and so on.Purpose: To derive and validate a conversion equation between BNP and NT-proBNP. Methods The derivation cohort included 928 patients (mean 66 yrs., female 26%) from the Supplemental Benefit of Angiotensin Receptor Blocker in Hypertensive Patients with Stable Heart Failure using Olmesartan (SUPPORT) Trial at enrolment. The validation cohort included 1,160 consecutive cardiovascular patients aged >20 yrs., in whom both BNP and NT-proBNP levels were measured simultaneously at Tohoku University Hospital between April 2017 and October 2017 (mean 66 yrs., female 41%). To develop the conversion equation, we regressed log2 NT-proBNP onto log2 BNP and other covariates by non-linear regression model. In the multivariable models, all possible combinations of log2 BNP and the covariates including age, s(age), sex, BMI, s(BMI), creatinine clearance (CCr), s(CCr), hemoglobin (Hb), s(Hb), where s() indicated the spline transformation, were examined as independent variables. For the best conversion equation, we selected the set of the covariates that minimized the Akaike information criterion (AIC) in the derivation cohort first. Then, the root mean squared error (RMSE) was obtained in the validation cohort to examine the predictive accuracy of the selected model. Results In the derivation cohort, the model with the lowest AIC consisted of log2 BNP, age, s(CCr), s(Hb). Predicted NT-proBNP levels were calculated by the conversion equation. The figure showed the estimated NT-proBNP levels and the measured NT-proBNP levels in the validation cohort ( Figure ). In the validation cohort, RMSE of this conversion equation was smaller than that of previously reported conversion equations, demonstrating that this equation excellently converts BNP levels to NT-proBNP levels. Conclusion We have developed an excellent and reliable conversion equation between BNP and NT-proBNP levels with age, CCr, and Hb.