Impact of premature birth and critical illness on neonatal range of plasma amino acid concentrations determined by LC-MS/MS

Background: The number of newborns and the number of disorders detected by large-scale screening programs has increased dramatically in the last decade. Newborn screening is a multi-step process requiring confirmatory testing to establish and refine a diagnosis. Whereas screening cutoffs are established to detect all cases of a specific disease, confirmatory testing is performed against a backdrop of many co-morbid conditions and interpretation of results is often not straightforward. The goal of the current study was to define the range of amino acid concentrations encountered in normal, premature, and acutely ill newborns as an aid to the interpretation of follow-up testing for suspicious newborn screens.

Materials and methods: Residual plasma samples from 354 neonates (age 1-10 days) were utilized. 206 samples were from neonates with uncomplicated birth histories and prompt hospital discharge. 98 specimens were derived from a population of children receiving intensive care with diagnoses that included sepsis, respiratory insufficiency, cardiac malfunction/malformation and gastrointestinal complications. 50 samples were from infants born after 32 but before 37 full weeks gestation that were discharged following uneventful hospital courses. 25 amino acids were quantitated by LC-MS/MS and reference intervals determined by non-parametric statistical methods. Distributions were compared using Kruskal-Wallis analyses for independent samples.

Results: The distributions of multiple amino acids in premature and critically ill infants were significantly different than those observed in healthy newborns. Differing distributions were found for phenylalanine, the branched chain amino acids, methionine, glutamine, glutamate, arginine, lysine, α-aminoadipic acid, and β-aminoisobutyric acid. In most cases median values were increased and distributions more positively skewed in premature or ill newborns compared to healthy newborns. In addition, we report neonatal homocitrulline reference intervals for these newborn populations determined by an LC-MS/MS technique that is not confounded by methionine interference.

Conclusion: These data provide a basis for improved detection of genetic metabolic disorders in newborns, particularly those born prematurely or with a variety of critical co-morbid conditions.