T1ρ-based dynamic glucose-enhanced (DGEρ) MRI at 3 T: method development and early clinical experience in the human brain

Purpose: The aim of this study was to translate the T-1 rho-based dynamic glucose-enhanced (DGE.) experiment from ultrahigh magnetic field strengths to a clinical field strength of 3 T. Although the protocol would seem to be as simple as gadolinium-enhanced imaging, several obstacles had to be addressed, including signal-to-noise ratio (SNR), robustness of contrast, and postprocessing, especially motion correction. Methods: Spin-lock based presaturation and a 3D gradient-echo snapshot readout were optimized for 3 T with regard to robustness, chemical exchange saturation transfer effect strength, and SNR. Postprocessing steps, including dynamic B-0 and motion correction, were analyzed and optimized in 7 healthy volunteers. The final protocol, including glucose injection, was applied to 3 glioblastoma patients. Results: With appropriate postprocessing, motion-related artifacts could be drastically reduced, and an SNR of approximately 90 could be achieved for a single dynamic measurement. In 2 patients with blood-brain barrier breakdown, a significant glucose uptake could be observed with a DGE rho effect strength in the range of 0.4% of the water signal. Thorough analysis of possible residual motion revealed that the statistical evidence can decrease when tested against pseudo effects attributed to uncorrected motion. Conclusion: DGE rho imaging was optimized for clinical field strengths of 3 T, and a robust protocol was established for broader application. Early experience shows that DGE rho seems possible at 3 T and could not only be attributed to motion artifacts. Observed DGE rho maps showed unique patterns, partly matching with the T-1-ce tumor ring enhancement. However, effect sizes are small and careful clinical application is necessary.