Adam Himebauch, MD
Assistant Professor of Anesthesiology and Critical Care Medicine
The Children's Hospital of Philadelphia
Philadelphia, Pennsylvania, United States
Disclosure information not submitted.
Nadir Yehya, MD, MSCE
Children's Hospital of Philadelphia
Cherry Hill, NJ
Disclosure information not submitted.
Yan Wang, MD, RDCS, FASE
Project Lead Sonographer
Children's Hospital of Philadelphia, United States
Disclosure information not submitted.
Ryan Burnett, BS
Research Coordinator
Children's Hospital of Philadelphia, United States
Disclosure information not submitted.
Steven Kawut, MD, MS
Professor of Medicine, Biostatistics, and Epidemiology
Perelman School of Medicine at the University of Pennsylvania, United States
Disclosure information not submitted.
Francis McGowan, MD
Attending physician
Children's Hospital of Philadelphia, United States
Disclosure information not submitted.
Robert Berg, MD, MCCM (he/him/his)
Division Chief, Pediatric Critical Care Medicine
Children's Hospital of Philadelphia
Merion Station, Pennsylvania, United States
Disclosure information not submitted.
Laura Mercer-Rosa, MD, MSCE
Associate Professor of Pediatrics
The Children's Hospital of Philadelphia, United States
Disclosure information not submitted.
Title: NT-proBNP Correlates Poorly With Echocardiogram Measures of Myocardial Function in Early PARDS
Introduction/Hypothesis: N-terminal pro-brain natriuretic peptide (NT-proBNP) is released in response to hypoxic pulmonary vasoconstriction and myocardial stretch. The correlation of NT-proBNP with myocardial dysfunction in pediatric ARDS is unknown. Our hypotheses were that plasma NT-proBNP concentrations measured within 24 hours of ARDS onset would be elevated, correlate with echocardiographic measures, and be associated with mortality.
Methods: Single-center prospective cohort study of children (1 month to 18 years, 9/2018-11/2020) with PALICC-defined ARDS and invasive ventilation. Protocolized echocardiograms with paired blood sampling within 24 hours of ARDS onset were performed. Plasma NT-proBNP was measured by commercial assay. Echocardiograms were analyzed for right ventricular (RV) global longitudinal strain (GLS), RV free wall strain (FWS), RV fractional area change (FAC), tricuspid annular plane systolic excursion (TAPSE), left ventricular (LV) GLS, LV fractional shortening (FS), LV ejection fraction(EF), and LV lateral e’ velocity. Nonparametric statistics were calculated. Area under the receiver operating characteristic (AUROC) curve was calculated for the outcome of death with NT-proBNP as the predictor.
Results: 60 subjects were included: median age 5 years (IQR 1.2-10), 55% male, 20% immunocompromised, and 75% had either infectious pneumonia or sepsis as the cause of ARDS. PALICC categories were: 28 (47%) mild, 24 (40%) moderate, and 8 (13%) severe. Eight patients (14%) died. The median NT-proBNP concentration was elevated at 1268.2 pg/mL (IQR 397.0-4538.8) with no difference between PALICC categories (p=0.93). NT-proBNP correlated poorly with RV GLS (ρ=0.07), RV FWS (ρ= 0.12), RV FAC (ρ=-0.19), TAPSE z-score (ρ= 0.04), LV GLS (ρ= 0.03), LV EF (ρ= 0.06), LV SF (ρ= -0.22), and LV lateral e’ velocity (ρ=0.14). NT-proBNP discriminated those who died prior to discharge with an AUROC 0.84 (95% CI 0.69-0.99).
Conclusions: NT-proBNP was elevated early in the course of pediatric ARDS, correlated poorly with early echocardiographic measures of RV or LV systolic or diastolic function, and yet discriminated death. Given its association with mortality, NT-proBNP may have prognostic utility in pediatric ARDS as a predictor of later myocardial dysfunction or other mechanisms that warrant further investigation.