Metabolic concentration and flux maps from DGE-DMI in mouse GBM.

Example of a CT2A tumor (C1). A T2-weighted reference image (top-left) displaying the tumor region (dashed lines) and representative peri-tumor and tumor voxels (back dots), and respective spectral quantifications (right-side): bottom, raw spectrum (black) with overlaid estimation (purple); center, individual components for each metabolite peak (black - semi-heavy water, DHO (black); deuterated glucose, Glc (red); and glucose-derived glutamate-glutamine and lactate, Glx (green) and Lac (blue)); top, residual. B Time-course de novo concentration maps for each metabolite (mM) following Glc i.v. injection (red arrow). C Average concentration maps for each metabolite after Glc injection. D Time-course concentration plots for each metabolite (dots) and respective kinetic fitting (straight lines), displayed for the peri-tumor and tumor voxels shown in A (same color codes) and applied to all the voxels to generate glucose flux maps: maximum consumption rate (Vmax); and respective individual rates for lactate synthesis (Vlac) and elimination (klac), and glutamate-glutamine synthesis (Vglx) and elimination (kglx).

Histopathologic and immunohistochemical assessment in two mouse models of GBM.

A H&E-stained sections with high magnification to highlight annotations of tumor, infiltrative zones in the tumor rim (blue), and secondary lesion (red), in CT2A and GL261 tumors (subjects C4 and G4, respectively). B Ki67 immuno-stained sections with overlaid detection of positive (red) and negative (blue) cells; and high magnification to highlight annotations of tumor and tumor border (yellow lines), in CT2A and GL261 tumors (subjects C1 and G3, respectively); and GBM cohort differences in tumor/border ratios of cell density and cell proliferation (dots representative of average values for each subject). CT2A (n=5) vs GL261 (n=5): * p<0.05; *** p<0.001; unpaired t-test. Error bars: standard deviation.

Mouse GBM models with different histopathologic phenotypes underlied by regional differences in lactate metabolism.

A Metabolic maps of de novo lactate accumulation (mM) and respective consumption/elimination rates (mM/min), in tumor and tumor border regions (delineated by dashed lines) of CT2A and GL261 tumors (subjects C1 and G3, respectively). B GBM cohort differences in de novo lactate accumulation (Lac) and consumption/elimination rates (klac). C Strong linear correlations (indicated by the Person correlation coefficient, R) of lactate consumption/elimination rates with (top) time post-tumor inoculation in each cohort, and (bottom) tumor vascular permeability in pooled cohorts. CT2A (n=5) vs GL261 (n=5): * p<0.05; ** p<0.01; unpaired t-test. Tumor (n=5, each cohort) vs Border (n=5, each cohort): # p<0.05; ## p<0.01; paired t-test. Error bars: standard deviation. Bar plot dots representative of average pixel values for each subject.

Peritumoral metabolic changes consistent with glutamate-glutamine recycling mirror GBM infiltration and migration leading to secondary brain lesions.

Metabolic map (left-side: C1 tumor), histogram distributions (center), and bar plot comparison of mean values (right-side:), showing significant decreases in peritumoral glutamate-glutamine accumulation (Glx) and increases in its consumption/elimination (kglx) in pooled GL261 and CT2A cohorts displaying secondary brain lesions (n=4; vs n=6 without): * p<0.05; unpaired t-test. Error bars: standard deviation. Bar plot dots representative of average pixel values for each subject.