Noninvasive Differential Diagnosis of Liver Iron Contents in Nonalcoholic Steatohepatitis and Simple Steatosis Using Multiecho Dixon Magnetic Resonance Imaging

Rationale and Objectives: The roles of iron stores in nonalcoholic fatty liver disease have not yet been clearly identified, and it is lack of uniform criteria and a standardized study design for assessing the liver iron content (LIC) in nonalcoholic steatohepatitis (NASH). This study was to compare LICs in biopsy-proven simple steatosis (SS) and NASH based on T-2*-relaxometry. Material and Methods: A total of 32 subjects divided to three groups, consisting of 10 healthy controls, 12 SS and 10 NASH. All MRI examinations were performed on a 3 T MRI with a 32-channel body coil. To measure T-2*-value, we used a gradient echo sequence with six multiechoes within a single breath-hold. Hepatic iron contents among three groups were compared using Kruskal-Wallis H test and Mann-Whitney's posthoc tests. Diagnostic accuracy was determined by calculating the area under the receiver operating characteristics curve. To identify the reliability of iron measurements in the different region of interests, coefficient of variance (CV) was calculated overall CV values for the variability of measurements. Interobserver agreement and reliability were estimated by calculating the intraclass correlation coefficient. Results: The variations of all LIC measurements are not exceeded 20%, as a mean CV value 18.3%. intraclass correlation coefficients were higher than 0.9. Mean T-2*-values at localized region of interests were healthy controls 45.42 +/- 7.19 ms, SS 20.96 +/- 4.28 ms, and NASH 15.49 +/- 2.87 ms. The mean T-2*-value in NASH is significantly shorter than that in SS (p = 0.008). The area under the receiver operating characteristics curve to distinguish NASH from SS was 0.908 (95% confidence interval 0.775-1.000, p = 0.001) at a cut-off of iron contents greater than 17.95 ms, and its diagnostic accuracy had 0.833 sensitivity and 0.800 specificity. Conclusion: This study demonstrates that the T-2* calculation can help to differentially diagnose NASH.