Interleaved trinuclear MRS for single‐session investigation of carbohydrate and lipid metabolism in human liver at 7T

Abstract

The liver plays a central role in metabolic homeostasis, as exemplified by a variety of clinical disorders with hepatic and systemic metabolic disarrays. Of particular interest are the complex interactions between lipid and carbohydrate metabolism in highly prevalent conditions such as obesity, diabetes, and fatty liver disease. Limited accessibility and the need for invasive procedures challenge direct investigations in humans. Hence, noninvasive dynamic evaluations of glycolytic flux and steady‐state assessments of lipid levels and composition are crucial for basic understanding and may open new avenues toward novel therapeutic targets. Here, three different MR spectroscopy (MRS) techniques that have been combined in a single interleaved examination in a 7T MR scanner are evaluated. 1H‐MRS and 13C‐MRS probe endogenous metabolites, while deuterium metabolic imaging (DMI) relies on administration of deuterated tracers, currently 2H‐labelled glucose, to map the spatial and temporal evolution of their metabolic fate. All three techniques have been optimized for a robust single‐session clinical investigation and applied in a preliminary study of healthy subjects. The use of a triple‐channel 1H/2H/13C RF coil enables interleaved examinations with no need for repositioning. Short‐echo‐time STEAM spectroscopy provides well resolved spectra to quantify lipid content and composition. The relative benefits of using water saturation versus metabolite cycling and types of respiratory synchronization were evaluated. 2H‐MR spectroscopic imaging allowed for registration of time‐ and space‐resolved glucose levels following oral ingestion of 2H‐glucose, while natural abundance 13C‐MRS of glycogen provides a dynamic measure of hepatic glucose storage. For DMI and 13C‐MRS, the measurement precision of the method was estimated to be about 0.2 and about 16 mM, respectively, for 5 min scanning periods. Excellent results were shown for the determination of dynamic uptake of glucose with DMI and lipid profiles with 1H‐MRS, while the determination of changes in glycogen levels by 13C‐MRS is also feasible but somewhat more limited by signal‐to‐noise ratio.