Comparison between immunoradiometric and fluorimetric brain natriuretic peptide determination in patients with congestive heart failure

This study compared two different methods, namely the immunoradiometric (IRMA) and fluorimetric (FIA), in order to determine plasma brain natriuretic peptide (BNP) in congestive heart failure (CHF) patients. Methods: CHF in-patients underwent echocardiography and plasma BNP determination using both two methods. The echocardiograms analysed left ventricular end-systolic (LVESV) and end-diastolic (LVEDV) volumes and systolic dysfunction [left ventricular ejection fraction (LVEF) <50%]. Results: Seventy-three (71% males, age 67±9.6 yr) patients were enrolled, 31.5% affected by valvular heart disease. The mean LVEF was 39.8±14.1%; in 26 (35%) a hypertensive etiology emerged. The immunoradiometric assay (IRMA) BNP was found to be significantly lower than the FIA determination 116.5±149 pg/ml vs 267.3±285.6 pg/ml; p=0.0001) and the two methods were closely correlated (r=0.89; p=0.00001). Logistic regression demonstrated a significant correlation between BNP, LVEF, and LVESV/LVEDV (r=−0.45, p=0.0003; r=−0.48, p=0.00001; r=0.22 p=0.003; r=0.34 p=0.0001; r=0.13 p=0.02; r=0.28 p=0.001 IRMA and FIA, respectively). IRMA BNP and FIA BNP significantly increased according to the worsening functional class [from 34.3±60.2 pg/ml in NYHA (New York Heart Association) I to 555.5±273.1 pg/ml in NYHA IV; from 86.1 ±162.1 pg/ml in NYHA I to 1070±42.2 pg/ml in NYHA IV, respectively]. In severe systolic dysfunction (LVEF<30%), receiver operating characteristic analysis revealed a satisfactorily sensitivity and specificity using a cut-off point of 50.6 pg/ml with IRMA and 243 pg/ml with FIA. In mild systolic dysfunction (LVEF<50%), a good sensitivity and specificity using a cut-off point of 42 pg/ml with IRMA and 182 pg/ml with FIA emerged. Conclusions: In CHF patients both BNP methods correlated with NYHA class, LVEF, and ventricular volumes.