B-type Natriuretic Peptide: Perioperative Patterns in Congenital Heart Disease

Abstract

B-type natriuretic peptide (BNP) has diagnostic, prognostic, and therapeutic roles in adults with heart failure. BNP levels in children undergoing surgical repair of congenital heart disease (CHD) were characterized broadly, and distinguishable subgroup patterns delineated. Prospective, blinded, observational case series. Academic, tertiary care, free-standing pediatric hospital. Children with CHD; controls without cardiopulmonary disease. Interventions. None. Preoperative cardiac medications/doses, CHD lesion types, perioperative BNP levels, intraoperative variables (lengths of surgery, bypass, cross-clamp), postoperative outcomes (lengths of ventilation, hospitalization, open chest; averages of inotropic support, central venous pressure, perfusion, urine output; death, low cardiac output syndrome (LCOS), cardiac arrest; readmission; and discharge medications). Median BNP levels for 102 neonatal and non-neonatal controls were 27 and 7 pg/mL, respectively. Serial BNP measures from 105 patients undergoing CHD repair demonstrated a median postoperative peak at 12 hours. The median and interquartile postoperative 24-hour average BNP levels for neonates were 1506 (782-3784) pg/mL vs. 286 (169-578) pg/mL for non-neonates (P < 0.001). Postoperative BNP correlated with inotropic requirement, durations of open chest, ventilation, intensive care unit stay, and hospitalization (r = 0.33-0.65, all P < 0.001). Compared with biventricular CHD, Fontan palliations demonstrated lower postoperative BNP (median 150 vs. 306 pg/mL, P < 0.001), a 3-fold higher incidence of LCOS (P < 0.01), and longer length of hospitalization (median 6.0 vs. 4.5 days, P= 0.01). Perioperative BNP correlates to severity of illness and lengths of therapy in the CHD population, overall. Substantial variation in BNP across time as well as within and between CHD lesions limits its practical utility as an isolated point-of-care measure. BNP commonly peaks 6-12 hours postoperatively, but the timing and magnitude of BNP elevation demonstrates notable age-dependency, peaking earlier and rising an order of magnitude higher in neonates. In spite of higher clinical acuity, non-neonatal univentricular CHD paradoxically demonstrates lower BNP levels compared with biventricular physiologies.