Abstract 4355216: Nonlinear mixed effect modeling of natriuretic peptides and associations with clinical events in interstage infants with single ventricle heart disease
Abstract
Introduction: Infants with single ventricle disease experience high morbidity and mortality, especially in the interstage period between the first and second palliative surgery. Identifying high-risk infants may prompt interventions and improve outcomes. Myocardial stress or stretch triggers release of natriuretic peptides (N-terminal pro-brain natriuretic peptide [NTproBNP]; mid-regional pro-atrial natriuretic peptide [MRproANP]) that may be used as biomarkers. We modeled longitudinal biomarker trajectories and aimed to understand how these relate to clinical events in interstage infants. Methods: We prospectively collected biomarker samples from interstage infants enrolled in a prospective, multicenter study (NCT03877965), excluding preterm infants, those with creatinine >2mg/dL, and those on extracorporeal support. We performed nonlinear mixed effect modeling in NONMEM to describe the two phases of biomarker decline and evaluated covariate effects. We explored associations between model parameters, biomarker concentrations, and clinical events. Results: Across 10 sites, 50 infants contributed 142 MRproANP and 151 NTproBNP plasma samples in the interstage period. Table 1 shows cohort characteristics and clinical events. There was a >15x faster decline in biomarker concentrations in the first month after stage 1 palliation (S1P) which then slowed ( Figure 1A ). A biexponential model ( Table 2 ) with fast and slow elimination phases after S1P best described these trajectories. The coefficient of the slow elimination phase (ECH) of both biomarkers trended higher for in infants who died (median [range] MRproANP 610 [593, 628] v 492 [329, 1309] pmol/L; p=0.05; NTproBNP 13,199 [12,179, 14,219] v 8538 [2838, 27,523] pg/mL, p=0.14; Figure 1B/C ). Absolute NTproBNP concentrations were higher 15-28 days after S1P (14185 [12627, 15743] v 6822 [1672, 17997] pg/mL) in infants who died. Absolute MRproANP and NTproBNP concentrations at 29-60 days after S1P were lower in those with tachyarrhythmias, (219 [192, 411] v 508 [258, 2755] pmol/L; 2977 [939, 5177] v 8341 [726, 65078] pg/mL). Absolute MRproANP concentrations in the same timeframe were higher in those with bradyarrhythmias (704 [369, 2755] v 494 [192, 1191] pmol/L). Conclusion: A biexponential model described biomarker trajectories. An elevated ECH may identify interstage infants at increased risk of death. Utility of serial biomarker measurement to evaluate biomarker trajectories should be explored in future studies.
Article Details
Authors (16)
Elizabeth Thompson
University of Pennsylvania, Philadelphia, Pennsylvania, United States
Rodrigo Gonzalez Ramirez
Duke Clinical Research Institute, Durham, North Carolina, United States
Anil Maharaj
The University of British Columbia, Vancouver, British Columbia, Canada
Henry Foote
Duke University, Durham, North Carolina, United States
Karan Kumar
Duke University Medical Center, Durham, North Carolina, United States
Kiona Allen
Alexis Benscoter
CCHMC, Cincinnati, Ohio, United States
Chi Hornik
Duke University Medical Center, Durham, North Carolina, United States
Erica Del Grippo
Nemours Children's Hospital, Wilmington, Delaware, United States
Brett Anderson
Icahn School of Medicine, New York, New York, United States
Danyal Khan
John Costello
School of Physical Sciences Dublin City University Dublin 9 Ireland
Denise Suttner
Rady Children's Hospital, San Diego, California, United States
Pirooz Eghtesady
WASHINGTON UNIVERSITY ST LOUIS, Saint Louis, Missouri, United States
Nancy Halnon
UCLA MEDICAL CENTER, Los Angeles, California, United States
Christoph Hornik
Duke Clinical Research Institute, Durham, North Carolina, United States