NT-ProBNP in Outpatients After Myocardial Infarction: Interaction Between Symptoms and Left Ventricular Function and Optimized Cut-Points.

Abstract: Background: N-terminal pro-brain natriuretic peptide (NT-proBNP) allows us to rule out left ventricular dysfunction (LVD) in the general population at a recommended cut-off concentration of 125 pg/mL. It was our objective to reassess this cut-point in outpatients after myocardial infarction. Methods and Results: NT-proBNP was assessed in 418 randomly selected outpatients who had experienced myocardial infarction and 352 siblings who had not experienced myocardial infarction (control). Left ventricular ejection fraction (LVEF) and mass-index (LVMI) were assessed by echocardiography. NT-proBNP was elevated in outpatients after myocardial infarction (mean [±SEM], 305 ± 25 pg/mL vs control, 84 ± 8 pg/mL; P < .01) and was correlated inversely with LVEF (P < .001). When patients were stratified according to the presence or absence of heart failure, NT-proBNP was elevated significantly throughout all LVEF strata (each P < .05). On regression analysis, NT-proBNP was correlated independently with LVEF, LVMI, heart failure, and glomerular filtration rate (all P < .01). In patients with heart failure, the optimal cut-point for the detection of an LVEF <35% was 348 pg/mL (sensitivity 80%; specificity 69%) and for the detection of an LVEF <45% was 260 pg/mL (sensitivity 60%; specificity 60%). The relative risk for LVD in the presence of elevated NT-proBNP increased from 2.7 to 7.7 (EF < 35%) and from 1.4 to 2.4 (EF < 45%) when these cut-points were applied instead of the 125 pg/mL cut-point. An LVEF of <35% could be ruled out in symptomatic outpatients after myocardial infarction with a negative predictive value of 97% (cut-point 348 pg/mL) and in asymptomatic outpatients after myocardial infarction with a negative predictive value of 98% (cut-point 157 pg/mL). Conclusion: NT-proBNP is higher in outpatients after myocardial infarction than in the general population. In symptomatic patients, a cut-point of 348 pg/mL yields satisfactory sensitivity and specificity for the detection of significant LVD (EF < 35%). Furthermore, significant LVD can be virtually ruled out in symptomatic and asymptomatic outpatients after myocardial infarction at below-threshold concentrations. [Copyright &y& Elsevier]