**4. BNP in pediatric heart disease**

In comparison to the adult experience, there are far fewer data regarding BNP in pediatric cardiac disease(Das 2010). Knirsch and colleagues measured BNP levels before and during treatment in 522 pediatric patients (age of 6.4±5.2 years, range of 14 days to 18 years) with congenital heart disease, cardiomyopathies, or pulmonary arterial hypertension(Knirsch et al. 2011). They found that BNP levels were elevated in each type of heart disease, and that levels fell in all groups with therapy.

As opposed to adults with congestive heart failure, BNP levels in infants and children with congenital heart disease are quite varied, and are dependent in part upon the age of the patient and the specific physiology associated with the cardiac defect. For example, Law and colleagues performed a study that included 42 neonates and 58 older children between the age of 7 days and 19 years presenting in an acute care setting with symptoms potentially attributable to heart disease. BNP levels were higher in both age groups in those patients with heart disease compared to those without heart disease, but the cut-off values differed. A BNP level of 170 pg/ml was 94% sensitive and 73% specific for heart disease in neonates,

B-Type Natriuretic Peptide (BNP) in Neonates,

Infants, and Children Undergoing Cardiac Surgery 203

normal control patients, but interestingly, BNP levels were higher in those patients with single ventricles of right ventricular morphology compared to left ventricular morphology(Holmgren et al. 2008). In contrast, Koch and colleagues studied 48 patients with d-transposition of the great arteries, after arterial switch procedure, or congenitally corrected transposition of the great arteries. They found no difference in BNP levels between patients with systemic ventricles of right ventricular morphology compared to left ventricular morphology(Koch, Zink, and Singer 2008). They did, however, find correlations between BNP and the severity of tricuspid regurgitation and decreasing exercise capacity. Likewise, Inuzuka and colleagues measured BNP levels in 51 patients (mean age 1.1 years) with single ventricular defects undergoing cardiac catheterization before second stage palliation. Mean BNP levels were 90.4 pg/ml. BNP levels above 100 pg/ml were associated with increased Qp:Qs, end-diastolic volume, AV valve regurgitation, and lower ventricular mass to end-diastolic volume ratio, all consistent with an inadequate adaptation to volume overload(Inuzuka et al. 2011). In addition, multivariate regression analysis demonstrated that the BNP concentration was independently predictive of death or the need for cardiac

transplantation (with a hazard ratioof 3.05, CI: 1.06-8.83)(Inuzuka et al. 2011).

but with a ten-fold lower cut-off of 30 pg/ml (Shah, et al., 2009).

**5. BNP as a biomarker following cardiac surgery** 

**5.1 Early post-operative period** 

palliation of congenital cardiac defects.

Thus, it is clear that BNP determinations can provide physiologically relevant information about individual patients, but that finding cut-off values that can be generalized for clinical use across the spectrum of congenital heart disease may not be possible. Indeed, in adult patients the primary utility of BNP determinations is in establishing the diagnosis of congestive heart failure in acute care settings(Maisel et al. 2002). In a pediatric population, the questions are more diverse. They might include, when to repair a left-to-right shunt, when to proceed with staged palliation of single ventricle defects, or when to refer for cardiac transplantation. However, even studies focused on pediatric heart failure have revealed a wide range of BNP values, again likely related at least in part to age and cardiopulmonary hemodynamics. For example, in a study of infants and children with biventricular hearts and chronic left ventricular dysfunction, Price and colleagues found that a BNP level of ≥300 pg/ml was predictive of death, hospitalization, or listing for cardiac transplant(Price et al. 2006). Shah and colleagues also found that increasing BNP was associated with heart failure in their study of 29 patients with single ventricular physiology,

As a cardiac hormone with a relatively short circulating half life that is dynamically released in response to deranged myocardial performance, BNP appears perfectly suited for the perioperative management of pediatric patients undergoing cardiac surgery for repair or

A number of investigators have studied perioperative BNP levels in neonates, infants, and children undergoing cardiac surgery. Ationu and colleagues measured perioperative BNP levels in 9 children undergoing repair of congenital heart defects(Ationu et al. 1993). In that study, BNP levels decreased at 12 hours following surgery(Ationu et al. 1993). Most studies, however, have found increases in BNP and NT-proBNP levels after surgery. Costello and colleagues measured natriuretic peptide levels, including BNP, in 5 infants undergoing surgical repair of left-to-right shunts(Costello et al. 2004). As opposed to ANP and DNP, BNP concentrations increased after cardiopulmonary bypass. Sun and colleagues measured BNP levels before and after surgery in 27 infants and children undergoing biventricular

whereas a lower level of 41 pg/ml was 87% sensitive and 70% specific for heart disease in the older patients(Law et al. 2009). Koch and colleagues found in a study of 288 pediatric patients (mean age of 6 years) with congenital cardiac defects that normal BNP levels did not exclude cardiac pathology such as the presence of structural defects or ventricular hypertrophy, but rather were associated with the extent of ventricular impairment(Koch, Zink, and Singer 2006). Conversely, Cantinotti and colleagues studied 152 neonates with congenital heart disease and 154 healthy neonatal controls, and found the BNP levels were higher in neonates with congenital heart disease than controls with a diagnostic accuracy defined by the area under the curve (on receiver operating characteristic analysis) of 0.935 for neonates with congenital heart disease between 4 and 30 days of age(Cantinotti et al. 2010). Thus, it appears that age modifies the diagnostic utility of BNP for cardiac disease in pediatric patients. In fact, in a separate study Koch and colleagues demonstrated agedependent differences in the metabolic clearance of BNP and NT-proBNP(Koch, Rauh, et al. 2006).

A number of studies indicate that cardiopulmonary hemodynamics also modify the diagnostic utility of BNP in pediatric patients with congenital heart disease. For example, in a study of infants and children with ventricular septal defects, Suda and colleagues found that BNP levels correlated with the pulmonary-to-systemic blood flow (Qp:Qs) ratio and the mean pulmonary arterial pressure(Suda, Matsumura, and Matsumoto 2003). Koch and colleagues demonstrated similar correlations, in their study of 288 patients with various cardiac defects(Koch, Zink, and Singer 2006). They found that in patients with left-to-right shunts, BNP levels were increased and correlated with shunt volume, systolic right ventricular pressure, mean pulmonary artery pressure, and pulmonary vascular resistance(Koch, Zink, and Singer 2006). Likewise, Kunii and colleagues compared BNP levels between normal children (n=253), and children with ventricular septal defects (n=91), patent ductus arteriosus (n=29), and atrial septal defects (n=34). Like the studies of Suda and Koch, they found that BNP levels correlated with the Qp:Qs ratio, and also the left ventricular end-diastolic volume, and the right ventricular to left ventricular pressure ratio(Kunii et al. 2003). Mainwaring and colleagues also found a positive correlation between preoperative BNP levels and the Qp:Qs ratio in a study of 18 patients (2 months – 15.6 years of age) with ventricular septal defects(Mainwaring et al. 2007). Ozhan and colleagues found the same relationship between BNP and the Qp:Qs ratio in their study of 35 children (mean age 70±129 weeks) with ventricular or atrial septal defects. Receiver operating characteristic analysis found that a plasma BNP cutoff of ≥20 pg/ml was 69% sensitive and 79% specific for a Qp:Qs of greater than 1.5(Ozhan et al. 2007).

A number of studies of premature neonates found that BNP levels correlated with the degree of shunting across a patent ductus arteriosus and predicted hemodynamic significance as determined by echocardiography-based criteria(Choi et al. 2005; Flynn et al. 2005; Puddy et al. 2002; Sanjeev et al. 2005; Holmstrom and Omland 2002; da Graca et al. 2006). However, the precise cut-off value for BNP that was predictive varied widely. For example, Choi and colleagues reported that a BNP level of > 1110 pg/ml was 100% sensitive and 95% specific for the presence of a hemodynamically significant patent ductus arteriosus, while Sanjeev and colleagues reported a 92% sensitivity with a cut-off level of 70 pg/ml(Choi et al. 2005; Sanjeev et al. 2005).

Holmgren and colleagues studied 38 patients with single ventricular physiology. They found that BNP levels were higher in patients after first stage palliation (31.6 pg/ml) compared to patients after second and third stage palliation (6.7 and 9 pg/ml, respectively). In fact, BNP levels in patients after second and third stage palliation did not differ from

whereas a lower level of 41 pg/ml was 87% sensitive and 70% specific for heart disease in the older patients(Law et al. 2009). Koch and colleagues found in a study of 288 pediatric patients (mean age of 6 years) with congenital cardiac defects that normal BNP levels did not exclude cardiac pathology such as the presence of structural defects or ventricular hypertrophy, but rather were associated with the extent of ventricular impairment(Koch, Zink, and Singer 2006). Conversely, Cantinotti and colleagues studied 152 neonates with congenital heart disease and 154 healthy neonatal controls, and found the BNP levels were higher in neonates with congenital heart disease than controls with a diagnostic accuracy defined by the area under the curve (on receiver operating characteristic analysis) of 0.935 for neonates with congenital heart disease between 4 and 30 days of age(Cantinotti et al. 2010). Thus, it appears that age modifies the diagnostic utility of BNP for cardiac disease in pediatric patients. In fact, in a separate study Koch and colleagues demonstrated agedependent differences in the metabolic clearance of BNP and NT-proBNP(Koch, Rauh, et al.

A number of studies indicate that cardiopulmonary hemodynamics also modify the diagnostic utility of BNP in pediatric patients with congenital heart disease. For example, in a study of infants and children with ventricular septal defects, Suda and colleagues found that BNP levels correlated with the pulmonary-to-systemic blood flow (Qp:Qs) ratio and the mean pulmonary arterial pressure(Suda, Matsumura, and Matsumoto 2003). Koch and colleagues demonstrated similar correlations, in their study of 288 patients with various cardiac defects(Koch, Zink, and Singer 2006). They found that in patients with left-to-right shunts, BNP levels were increased and correlated with shunt volume, systolic right ventricular pressure, mean pulmonary artery pressure, and pulmonary vascular resistance(Koch, Zink, and Singer 2006). Likewise, Kunii and colleagues compared BNP levels between normal children (n=253), and children with ventricular septal defects (n=91), patent ductus arteriosus (n=29), and atrial septal defects (n=34). Like the studies of Suda and Koch, they found that BNP levels correlated with the Qp:Qs ratio, and also the left ventricular end-diastolic volume, and the right ventricular to left ventricular pressure ratio(Kunii et al. 2003). Mainwaring and colleagues also found a positive correlation between preoperative BNP levels and the Qp:Qs ratio in a study of 18 patients (2 months – 15.6 years of age) with ventricular septal defects(Mainwaring et al. 2007). Ozhan and colleagues found the same relationship between BNP and the Qp:Qs ratio in their study of 35 children (mean age 70±129 weeks) with ventricular or atrial septal defects. Receiver operating characteristic analysis found that a plasma BNP cutoff of ≥20 pg/ml was 69%

sensitive and 79% specific for a Qp:Qs of greater than 1.5(Ozhan et al. 2007).

pg/ml(Choi et al. 2005; Sanjeev et al. 2005).

A number of studies of premature neonates found that BNP levels correlated with the degree of shunting across a patent ductus arteriosus and predicted hemodynamic significance as determined by echocardiography-based criteria(Choi et al. 2005; Flynn et al. 2005; Puddy et al. 2002; Sanjeev et al. 2005; Holmstrom and Omland 2002; da Graca et al. 2006). However, the precise cut-off value for BNP that was predictive varied widely. For example, Choi and colleagues reported that a BNP level of > 1110 pg/ml was 100% sensitive and 95% specific for the presence of a hemodynamically significant patent ductus arteriosus, while Sanjeev and colleagues reported a 92% sensitivity with a cut-off level of 70

Holmgren and colleagues studied 38 patients with single ventricular physiology. They found that BNP levels were higher in patients after first stage palliation (31.6 pg/ml) compared to patients after second and third stage palliation (6.7 and 9 pg/ml, respectively). In fact, BNP levels in patients after second and third stage palliation did not differ from

2006).

normal control patients, but interestingly, BNP levels were higher in those patients with single ventricles of right ventricular morphology compared to left ventricular morphology(Holmgren et al. 2008). In contrast, Koch and colleagues studied 48 patients with d-transposition of the great arteries, after arterial switch procedure, or congenitally corrected transposition of the great arteries. They found no difference in BNP levels between patients with systemic ventricles of right ventricular morphology compared to left ventricular morphology(Koch, Zink, and Singer 2008). They did, however, find correlations between BNP and the severity of tricuspid regurgitation and decreasing exercise capacity. Likewise, Inuzuka and colleagues measured BNP levels in 51 patients (mean age 1.1 years) with single ventricular defects undergoing cardiac catheterization before second stage palliation. Mean BNP levels were 90.4 pg/ml. BNP levels above 100 pg/ml were associated with increased Qp:Qs, end-diastolic volume, AV valve regurgitation, and lower ventricular mass to end-diastolic volume ratio, all consistent with an inadequate adaptation to volume overload(Inuzuka et al. 2011). In addition, multivariate regression analysis demonstrated that the BNP concentration was independently predictive of death or the need for cardiac transplantation (with a hazard ratioof 3.05, CI: 1.06-8.83)(Inuzuka et al. 2011).

Thus, it is clear that BNP determinations can provide physiologically relevant information about individual patients, but that finding cut-off values that can be generalized for clinical use across the spectrum of congenital heart disease may not be possible. Indeed, in adult patients the primary utility of BNP determinations is in establishing the diagnosis of congestive heart failure in acute care settings(Maisel et al. 2002). In a pediatric population, the questions are more diverse. They might include, when to repair a left-to-right shunt, when to proceed with staged palliation of single ventricle defects, or when to refer for cardiac transplantation. However, even studies focused on pediatric heart failure have revealed a wide range of BNP values, again likely related at least in part to age and cardiopulmonary hemodynamics. For example, in a study of infants and children with biventricular hearts and chronic left ventricular dysfunction, Price and colleagues found that a BNP level of ≥300 pg/ml was predictive of death, hospitalization, or listing for cardiac transplant(Price et al. 2006). Shah and colleagues also found that increasing BNP was associated with heart failure in their study of 29 patients with single ventricular physiology, but with a ten-fold lower cut-off of 30 pg/ml (Shah, et al., 2009).
