Noninvasive monitoring of evolving urinary metabolic patterns in neonatal encephalopathy

Abstract

Background: Infants with moderate and severe neonatal encephalopathy (NE) frequently suffer from long-term adverse outcomes. We hypothesize that the urinary metabolome of newborns with NE reflects the evolution of injury patterns observed with magnetic resonance imaging (MRI).

Methods: Eligible patients were newborn infants with perinatal asphyxia evolving to NE and qualifying for therapeutic hypothermia (TH) included in the HYPOTOP trial. MRI was employed for characterizing brain injury. Urine samples of 55 infants were collected before, during, and after TH. Metabolic profiles of samples were recorded employing three complementary mass spectrometry-based assays, and the alteration of detected metabolic features between groups was assessed.

Results: The longitudinal assessment revealed significant perturbations of the urinary metabolome. After 24 h of TH, a stable disease pattern evolved characterized by the alterations of 4-8% of metabolic features related to lipid metabolism, metabolism of cofactors and vitamins, glycan biosynthesis and metabolism, amino acid metabolism, and nucleotide metabolism. Characteristic metabolomic fingerprints were observed for different MRI injury patterns.

Conclusions: This study shows the potential of urinary metabolic profiles for the noninvasive monitoring of brain injury of infants with NE during TH.

Impact: A comprehensive approach for the study of the urinary metabolome was employed involving a semi-targeted capillary electrophoresis-time-of-flight mass spectrometry (TOFMS) assay, an untargeted ultra-performance liquid chromatography (UPLC)-quadrupole TOFMS assay, and a targeted UPLC-tandem MS-based method for the quantification of amino acids. The longitudinal study of the urinary metabolome identified dynamic metabolic changes between birth and until 96 h after the initiation of TH. The identification of altered metabolic pathways in newborns with pathologic MRI outcomes might offer the possibility of developing noninvasive monitoring approaches for personalized adjustment of the treatment and for supporting early outcome prediction.