Item

Abstract

Purpose: CEST allows for indirect detection of diluted molecules via their saturation transfer to the abundant water pool 1–3. At 3T, the frequency separation of different CEST effects is difficult and many applications of CEST use MTRasym evaluation. In this study, we use low-power
saturation to separate the two major contributors the MTRasym signal,
namely APT (at +3.5 ppm) and NOE effects (at –3.5 ppm), and investigate
correlations with 18F-FET PET enhancement in brain tumors.
Methods: 9 patients scanned on a 3T PET/MR system for suspected glioma or recurrent glioma were evaluated (1 radio-necrosis, 1 gliosis, 3 high grade, 4 low grade). Selective protein CEST Z-spectra were acquired with 4s saturation (100 Gaussian pulses, B1 = 0.6 μT, pulse duration 20ms, duty cycle 50%) at 53 frequency offsets ranging from
–100 ppm to +100 ppm. CEST Z-spectra were de-noised using principle
component analysis, retaining 15 principle components. Two-stage Lorentzian analysis was applied to estimate contributions from direct water saturation, magnetization transfer effects from macromolecules4,
and selective CEST effects.
Results: Selective NOE effects were stronger than APT CEST effects in both healthy and diseased tissues. Compared to contralateral tissues, tumor regions exhibit reduced NOE signals and mixed APT signals, consistent with results from previous studies 4,5. There was no correlation between PET uptake and CEST signals within the tumor.
Conclusion: NOE and amide CEST separation is possible at 3T and provide additional information compared to 18F-FET PET.