Cerebral glucose metabolism is of importance for brain function and maintaining electrophysiological activity including neuronal firing and signaling. Simultaneous assessment of cerebral glucose consumption rate (CMR_glc) and associated major metabolic fluxes, such as TCA cycle (V_TCA) and oxygen consumption rates, is crucial for understanding neuroenergetics under physiological and pathological conditions. Recently, we developed a novel in vivo Deuterium (²H) MR (DMR) imaging approach for noninvasive assessing glucose metabolism in rat brain at ultrahigh field (1-3). As for the CMR_glc measurement, this new approach has the advantage of eliminating the usage of radioactive tracer commonly employed by traditional ¹⁸FDG-PET imaging. Also when comparing with the classic ¹³C MRS method for V_TCA quantification, the much shorter T₁ relaxation times of quadruple ²H spins (e.g., ~0.05 s for deuterated glucose) allow more averages of the signal, thus, provide a substantial sensitivity gain for DMR detection (3). In this study, to evaluate its sensitivity in detecting altered energy metabolism under pathological conditions, dynamic localized DMR spectra were acquired using 3D-chemical shift imaging (CSI) technique in rat brains with gliosarcoma tumor at 16.4 T. Dynamics of the labeling enrichment into brain metabolites from normal appearing and tumor tissues were compared to investigate the glucose metabolic changes associated with the brain tumor interference.Three male Fischer 344 rats with appropriate size of the grown gliosarcoma (GS-9L cells, Sigma-Aldirch) were anesthetized by 2% isoflurane. Their femoral arteries and veins were catheterized for blood sampling, physiological monitoring and deuterated glucose infusion. All MR experiments were conducted at 16.4 T/26 cm scanner (Varian/VNMRJ) using a passively decoupled ¹H/²H surface coil. Dynamic 3D-²H CSI data of rat brains were acquired with 52 μL spatial and 69 s temporal resolutions for about 103 min to monitor the DMR signal changes before, during and after the deuterated glucose infusion. For each rat, 11.5 min baseline spectra were acquired that followed by a 2-min infusion of 1.3g/kg D-Glucose-6,6-d₂ (Sigma-Aldrich) dissolved in 2.5 mL saline. A 20 Hz linebroadening was used before Fourier transformation to enhance spectral SNR. All resonance signals (deuterated water, glucose, glutamate/glutamine (Glx) and lactate) were fitted using a MATLAB-based program, and the concentrations of metabolites were quantified as previously described (1-3). A short repetition time (TR=45 ms) was used in this study, thus, partial saturation effects on metabolites were also corrected for quantification.Figure 1 showed the anatomical image of a representative rat brain with a growing gliosarcoma. Localized in vivo DMR spectra obtained from the normal appearing and tumor tissues were also displayed for the brain after 35 min of deuterated glucose infusion (Fig. 1). The excellent spectral quality allows the detection of at least four well-resolved resonances (i.e., deuterated water, glucose, Glx and lactate) as shown in Fig. 1. Thus, their dynamic changes in interested brain region(s) could be monitored. As shown in Fig. 2, when compared with the normal appearing tissue (CON), accelerated glucose consumption (Fig. 2A) and lactate accumulation (Fig. 2C) were simultaneously observed in the brain region with tumor (TUM). Moreover, a much slower Glx turnover was also detected in the brain tumor (Fig. 2B).The findings of this work indicate that stimulated glycolysis (or increased CMR_glc) and inhibited oxidative metabolism (or slower V_TCA) occur in the brain tumor even in the presence of sufficient oxygen (Fig. 2). This corresponds to the Warburg effect of the cancer cells that shifts the fuel consumption towards glycolysis rather than the TCA cycle and oxidative phosphorylation (4). In summary, this pilot study demonstrates excellent spectral quality, adequate sensitivity and spatial/temporal resolution of the in vivo 3D-DMR imaging approach at ultrahigh field. It provides an opportunity for simultaneous studying glucose consumption, TCA cycle turnover and lactate accumulation in animal brain, potentially in human as well, under physiopathological conditions. This technique is particularly useful for monitoring tumor progression and treatment efficacy with improved specificity associated with cancer metabolism and biology. The same DMR data are being processed to generate quantitative metabolic images for identifying abnormal glucose metabolisms in tumor as compared to intact brain tissue.