Non-invasive quantification of triglyceride content in steatotic rat livers by (1)H-MRS: when water meets (too much) fat

The aims of this study were to explore the relationship between triglyceride (TG) and water in steatotic rat livers and to accordingly test the validity of the currently used steatosis calculation methods from magnetic resonance spectra. The approximations commonly used to derive steatosis degrees from magnetic resonance spectra include the generic types TG/water and TG/(TG + water). Hepatic fat and water content was quantitated by histology, magnetic resonance spectroscopy (MRS), gas chromatography, and dry/wet weight ratio analysis in increasingly (0%-95%) steatotic rats. Correlation analysis was performed to assess the statistical relationships among the steatosis quantification techniques. Subsequently, data were fitted with linear and nonlinear functions to determine the relationship between hepatic water fraction versus hepatic TG content and TG/water ratio versus macrovesicular steatosis degree to test the validity of commonly used steatosis calculation methods. Histologic analysis of macrovesicular steatosis correlated very strongly with TG content determined by gas chromatography and MRS. A strong positive correlation was also found between gas chromatography-derived and MRS-derived TG content. Biochemical analysis revealed a linear converse relationship between hepatic fat and water content. This relationship was nonlinear when determined by MRS. The MRS-based TG/(TG + water)-type approximations reflected the linear water-fat relationship better than the TG/water-type approximations, particularly when the calculations were performed with a maximum number of TG resonances. Hepatic fat approximations of the type TG/water overestimate hepatic steatosis degree because hepatic fat accumulation concurs with hepatic water exudation. Consequently, MRS-based approximations should be of the type TG/(TG + water) and contain a maximum number of TG resonances in the denominator