Evaluation of the Severity of Chronic Hepatitis C with 3-T1H-MR Spectroscopy

The purpose of this study was to compare the spectral characteristics of lipids, choline-containing compounds, and glutamine-glutamate complex assessed with (1)H-MR spectroscopy with the histologic findings in patients with chronic hepatitis C. Nine healthy controls and 30 patients with biopsy-proven hepatitis C virus-related liver disease participated in this prospective study. Degree of fibrosis and histologic activity were scored according to the METAVIR classification. The percentage of involved hepatocytes was used to grade steatosis. Hepatic spectra were obtained with a 3-T spectroscopic system. Tenfold cross-validated stepwise discriminant analysis was performed to classify disease severity on the basis of the spectroscopic findings. There was a strong correlation between (1)H-MR spectroscopically measured lipid concentration and the degree of steatosis at histologic examination (r = 0.9236, p < 0.0001). This finding enabled clear separation of groups according to degree of histologically determined steatosis. Variation in lipid concentration was consistent with the degree of steatosis (r = 0.7265, p < 0.0001) and stage of fibrosis (r = 0.8156, p < 0.0001). In univariate analysis, concentrations of both choline-containing compounds and glutamine-glutamate complex had a direct correlation with histologic grade (p < 0.0001) and degree of steatosis (p < 0.0001) but not with stage of fibrosis (p > 0.05). In multivariate analysis, the only factor independently associated with concentrations of choline-containing compounds and glutamine-glutamate complex was histologic grade. In cross-validated discriminant analysis based on choline-containing compound, glutamine-glutamate complex, and lipid resonance, 70% (21 of 30) of the histologic grade groups and 73% (22 of 30) of the steatosis groups were correctly classified. Hydrogen-1 MR spectroscopy can be an alternative to liver biopsy in the evaluation of steatosis and necroinflammatory activity in liver disease but is not useful for complete evaluation of hepatic fibrosis.