Urinary 1H-NMR-based metabolic profiling of children with NAFLD undergoing VSL#3 treatment

BACKGROUND: Nowadays, non-alcoholic fatty liver disease (NAFLD) is one of the most common chronic liver diseases in children. Our recent clinical trial demonstrated that dietary and VSL#3-based interventions may improve fatty liver by ultrasound and body mass index (BMI) after 4 months. OBJECTIVES: As in this short-term trial, as in others, it is impracticable to monitor response to therapy or treatment by liver biopsy, we aimed to identify a panel of potential non-invasive metabolic biomarkers by a urinary metabolic profiling. METHODS: Urine samples from a group of 31 pediatric NAFLD patients, enrolled in a VSL#3 clinical trial, were analyzed by high-resolution proton nuclear magnetic resonance spectroscopy in combination with analysis of variance-Simultaneous Component Analysis model and multivariate data analyses. Urinary metabolic profiles were interpreted in terms of clinical patient feature, treatment and chronology pattern correlations. RESULTS: VSL#3 treatment induced changes in NAFLD urinary metabolic phenotype mainly at level of host amino-acid metabolism (that is, valine, tyrosine, 3-amino-isobutyrate or beta-aminoisobutyric acid (BAIBA)), nucleic acid degradation (pseudouridine), creatinine metabolism (methylguanidine) and secondarily at the level of gut microbial amino-acid metabolism (that is, 2-hydroxyisobutyrate from valine degradation). Furthermore, some of these metabolites correlated with clinical primary and secondary trial end points after VSL#3 treatment: tyrosine and the organic acid U4 positively with alanine aminotransferase (R = 0.399, P = 0.026) and BMI (R = 0.36, P = 0.045); BAIBA and tyrosine negatively with active glucagon-like-peptide 1 (R = -0.51, P = 0.003; R = -0.41, P = 0.021, respectively). CONCLUSIONS: VSL#3 treatment-dependent urinary metabotypes of NAFLD children may be considered as non-invasive effective biomarkers to evaluate the response to treatment.