Background: There is increasing evidence that children or young adults having acquired liver disease in childhood display neurocognitive impairment which may become more apparent as they grow older. The molecular, cellular and morphological underpinnings of this clinical problem are incompletely understood.

Aim: Therefore, we used the advantages of highly-resolved proton magnetic resonance spectroscopy at ultra-high magnetic field to analyze the neurometabolic profile and brain morphometry of children with chronic, compensated liver disease, hypothesizing that with high field spectroscopy we would identify early evidence of rising brain glutamine and decreased myoinositol, such as has been described both in animals and humans with more significant liver disease.

Methods: Patients (n = 5) and age-matched controls (n = 19) underwent 7T MR scans and short echo time ¹ H MR spectra were acquired using the semi-adiabatic SPECIAL sequence in two voxels located in gray and white matter dominated prefrontal cortex, respectively. A 3D MP2RAGE sequence was also acquired for brain volumetry and T₁ mapping. Liver disease had to have developed at least 6 months before entering the study. Subjects underwent routine blood analysis and neurocognitive testing using validated methods within 3 months of MRI and MRS.

Results: Five children aged 8-16 years with liver disease acquired in childhood were included. Baseline biological characteristics were similar among patients. There were no statistically significant differences between subjects and controls in brain metabolite levels or brain volumetry. Finally, there were minor neurocognitive fluctuations including attention deficit in one child, but none fell in the statistically significant range.

Conclusion: Children with chronic, compensated liver disease did not display an abnormal neurometabolic profile, neurocognitive abnormalities, or signal intensity changes in the globus pallidus. Despite the absence of neurometabolic changes, it is an opportunity to emphasize that it is only by developing the use of ¹ H MRS at high field in the clinical arena that we will understand the significance and generalizability of these findings in children with CLD. Healthy children displayed neurometabolic regional differences as previously reported in adult subjects.