Texas Pediatric Society Electronic Poster Contest NTproBNP as a surrogate biomarker for diagnosis of pulmonary hypertension in premature infants with bronchopulmonary dysplasia Soham Dasgupta MD (PGY-3)a, Sunil K Jain MDb, Michael M Malloy MDb, Ashraf M Aly MD PhDc Department of Pediatricsa, Division of Neonatologyb, Division of Pediatric Cardiologyc University of Texas Medical Branch, Galveston Abstract Table 3: Mean NTproBNP values (pg/ml) in the BPD-PH, BPD and no-BPD groups at 28, 32 and 36 weeks CGA Background: N-terminal Pro B-Type Natriuretic Peptide (NTproBNP) has been proposed as a biomarker for pulmonary hypertension (PH) in extremely preterm infants with bronchopulmonary dysplasia (BPD). Objectives: Measure NTproBNP values in preterm infants and compare the values in infants diagnosed as having: 1) BPD with echocardiogram (ECHO) evidence of PH (BPD-PH), 2) BPD without ECHO evidence of PH (BPD) and 3) no BPD and no evidence of PH by ECHO (no-BPD). Study Design: Very low birth weight infants (birth weight <1500 gm & < 30 weeks gestational age [GA]) were included. BPD was defined as oxygen requirement at 36 weeks corrected gestational age (CGA). Presence of PH was determined by ECHO. Infants were divided into BPD-PH, BPD and no-BPD groups. NTproBNP levels and ECHO was obtained at 28, 32 & 36 weeks CGA. Results: The mean (sd) NTproBNP value at 28 weeks CGA was significantly elevated in the BPD-PH group vs the BPD group and in the BPD group vs the no-BPD group. These relationships persisted at 32 and 36 weeks CGA. ROC curves generated for NTproBNP at 28 weeks CGA provided a cut-point of 2,329 pg/ml for detection of BPD associated PH (sensitivity 87.5%, specificity 95%) and 578 pg/ml for BPD detection (sensitivity 89%, specificity 68%). Conclusions: NTproBNP appears to be a good screening tool to determine the onset of BPD-PH as early as 28 weeks CGA. Significant p value in BPD vs no-BPD and in BPD-PH vs BPD groups Figure 1: Mean NTproBNP values (sd) in the no-BPD, BPD and BPD-PH groups at 28, 32 and 36 weeks CGA Introduction PH is a well known complication of neonatal respiratory diseases including BPD with incidence as high as 18-25%1,2 and it increases the morbidity and mortality of BPD3,4. Early detection of PH is critical so that appropriate management can be initiated. NTproBNP is a biomarker, which is synthesized and secreted mainly by the ventricular myocardium under conditions of sustained overexpansion and pressure overload5. The relationship between NTproBNP and GA and the timing and rate of its increase in the course of BPD induced PH is not well documented. PURPOSE OF THIS STUDY Measure NTproBNP values in a cohort of preterm infants < 30 weeks gestation at 28, 32 and 36 weeks CGA and to compare those values in: Infants with BPD and ECHO evidence of PH (BPD-PH) Infants with BPD and no ECHO evidence of PH (BPD) Infants without BPD and no evidence of PH by ECHO (no-BPD) 28 weeks 32 weeks 36 weeks Figure 2: ROC curve comparing BPD-PH and BPD groups at 28, 32 and 36 weeks CGA Description of Study Study Design: Prospective, hypothesis driven. Study Period: February 2015 to February 2017. Inclusion Criteria: Very low birth weight infants (birth weight <1500 gm & < 30 weeks CGA). Exclusion Criteria: Babies with known chromosomal anomalies/congenital defects affecting the cardiorespiratory system. NTproBNP measurement: Performed at 28, 32 and 36 weeks CGA for serial measurement of NTproBNP. BPD was defined as oxygen requirement at 36 weeks CGA. Blood was collected within 12 hours of ECHO. Echocardiography: Performed at 28, 32 and 36 weeks CGA. PH was defined as 1) significant right ventricular hypertrophy, 2) flattening of the inter-ventricular septum and 3) pulmonary artery pressure > 30 mmHg estimated by tricuspid regurgitant velocity. Statistical analysis: SAS© procedures that included general linear models, repeated measures analysis of variance (ANOVA), and logistic modeling for receiver operator curve (ROC) generation. The set significance level was p<0.05. Figure 3: ROC curve comparing BPD and no-BPD groups at 28, 32 and 36 weeks CGA Results 54 infants met inclusion criteria. 8 infants (14.8%) were in the BPD-PH group, 28 infants (51.8%) were in the BPD group and 16 infants (29.6%) were in the no-BPD group. The mean (sd) NTproBNP value for infants in the BPD-PH group was significantly elevated as compared to the BPD group at 28, 32 and 36 weeks CGA (p<0.05) (Table 3 & Figure 1). The mean (sd) NTproBNP value for infants in the BPD group was significantly elevated as compared to the no-BPD group at 28, 32 and 36 weeks CGA (p<0.05) (Table 3 & Figure 1). In an ANOVA general linear model that included GA and NTproBNP values at 28, 32 and 36 weeks CGA, BPD-PH vs BPD remained significantly associated with the NTproBNP values, although the NTproBNP values for BPD-PH did not change significantly over time. NTproBNP showed a small, significant increase with time and GA. ROC curves generated for NTproBNP values at 28 weeks CGA provided a cut-point of 2,329 pg/ml for a BPD-PH detection sensitivity of 87.5% and specificity of 95% (Figure 2). Also generated a cut-point of 578 pg/ml for a BPD detection sensitivity of 89% and specificity of 68% at 28 weeks CGA (Figure 3). The cut-points were not significantly different at 32 and 36 weeks CGA. Conclusions SUMMARY OF RESULTS This is the first prospective study to perform serial measurements of NTproBNP to determine its role as a surrogate biomarker for PH in preterm infants with BPD. NTproBNP appears to be a good screening tool to determine the onset of BPD-PH as early as 28 weeks CGA. A value greater than 2,329 pg/ml appears to be 87.5% sensitive and 95% specific for the detection of BPD-PH at 28 weeks CGA. A value greater than 578 pg/ml appears to be 89% sensitive and 68% specific for the prediction of development of BPD as early as 28 weeks CGA. IMPLICATIONS It still remains uncertain whether therapeutic intervention at the earliest time of detection of BPD-PH determined by NTproBNP values, can modify the course of the disease process. As an extension of the study, we plan to follow the levels of NTproBNP with treatment of PH. Such an evaluation may establish NTproBNP as a marker of treatment success vs failure and may reduce the need for serial ECHO. Eventually, NTproBNP may replace ECHO as a more cost-efficient initial screening tool especially in areas where ECHO is not readily available. LIMITATIONS Small sample size. Although the NTproBNP values were obtained at fixed CGA, the actual age of the infants are different at these time points. The gold standard for diagnosis of PH is cardiac catheterization. In the absence of cardiac catheterization, ECHO is used most frequently to diagnose PH. ECHO diagnosis of PH may be subjective and operator dependent in the absence of a tricuspid regurgitant jet velocity. Table 1: Maternal demographic data CS: C-section; PPROM: preterm premature rupture of membranes; PIH: pregnancy induced hypertension, PE: preeclampsia, WB: with benefit, NS: not significant Table 2: Infant demographic data References 1. An HS, Bae EJ, Kim GB, Kwon BS, Beak JS, Kim EK et al. Pulmonary Hypertension in Preterm Infants With Bronchopulmonary Dysplasia. Korean Circ J. 2010;40(3):131-136. 2. Bhat R, Salas AA, Foster C, Carlo WA, Ambalavanan N. Prospective Analysis of Pulmonary Hypertension in Extremely Low Birth Weight Infants. Pediatrics 2012;e682-689. 3. Berkelhamer SK, Mestan KK, Steinhorn RH. Pulmonary hypertension in bronchopulmonary dysplasia. Semin Perinatol. 2013;37(2):124-131 4. Ambalavanan N, Mourani P. Pulmonary hypertension in bronchopulmonary dysplasia. Birth Defects Res A Clin Mol Teratol 2014; 100:240–24. 5. Suttner SW, Boldt J. Natriuretic peptide system: physiology and clinical utility. Curr Opin Crit Care. 2004;10(5):336-341. CPAP: continuous positive airway pressure; Sipap: synchronized inspiratory positive airway pressure; CMV: conventional mechanical ventilation; HFOV: high frequency oscillatory ventilation; TPN: total parenteral nutrition; NS: not-significant Texas Pediatric Society Electronic Poster Contest