Meta-analysis of magnetic resonance spectroscopy in the diagnosis of hepatic encephalopathy

Abstract

ObjectiveVarious imaging modalities have been used to explore pathogenic mechanisms and stratify the severity of hepatic encephalopathy (HE). The hypothesis of this meta-analysis was that there is a progressive identifiable derangement of imaging measures using magnetic resonance spectroscopy (MRS) related to the severity of the HE.MethodsStudies with more than 10 cases and HE diagnosis were identified from the electronic databases PubMed, EMBASE, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Literatura Latino Americana em Ciências da Saúde (LILACS), and Cochrane Central Register of Controlled Trials (CENTRAL) through July 25, 2018. Participants were stratified into healthy controls and patients with non-HE (NHE) (cirrhosis without HE), minimal HE (MHE), and overt HE (OHE). Analyses were organized by metabolite studied and brain region examined. Statistical meta-analysis was performed using the metafor package in R (v3.4.1). Pooled standardized mean differences between patient groups were calculated using a random effects model.ResultsWe identified 31 studies (1,481 patients) that included data for cirrhosis-related HE. We found the parietal region to be the most reliable in differentiating between patients with and without MHE, with standard mean differences of +0.82 (95% confidence interval [CI] +0.49 to +1.15, p < 0.0001, I2 = 37.45%) for glutamine/glutamate, −0.36 (95% CI −0.61 to −0.10, p = 0.007, I2 = 20.00%) for choline, and−0.77 (95% CI −1.19 to −0.34, p = 0.0004, I2 = 67.48%) for myo-inositol. We also found that glutamine/glutamate was the metabolite that reliably correlated with HE grade in all brain regions.ConclusionsThe meta-analysis reveals that MRS changes in glutamine/glutamate, choline, and myo-inositol, particularly in the parietal lobe, correlate with the severity of HE. MRS may be of value in the assessment of HE.