Introduction

Amide proton transfer chemical exchange saturation transfer (APT-CEST) imaging can detect increased levels of amides and peptides^1,2, which is potentially useful to detect and monitor brain tumors like gliomas^3-5. While APT-CEST has shown promising results, its reproducibility has not yet been extensively investigated. Good reproducibility is however crucial for clinical follow-up and cut-off value determination. This study evaluates the reproducibility of APT-CEST at 3T for both healthy volunteers and glioma patients.

Methods

Participants and Study design
21 healthy volunteers (age 39±11 yrs, 11 women) and 10 glioma patients (age: 55±15 yrs, 7 women) were scanned at 3T (Vida, Siemens, Erlangen, Germany) using a 20-channel head coil, including whole-brain 3D T1w and APT-CEST (2.8x2.8x2.8 mm resolution, 7 frequencies with MT pulses (10x 100ms at 2.5μT gaussian pulses) at ± 3.0, 3.5 and 4.0 ppm off-resonance, scan duration: 4m36s) plus a dual-echo GRE B0 map for geometric distortion correction.⁶
Healthy volunteers underwent 3 APT-CEST scan-sessions, while patients underwent one session. Each scan-session contains two consecutive APT-CEST scans and one B0-map. For a detailed description of the scan protocol see fig. 1.

Post-processing
Motion and distortion-corrected APT-CEST maps were calculated as magnetization transfer (MT) asymmetry (MTR_asym). APT-CEST unprocessed scans were registered to the 3D T1w images and transformed to MNI using ExploreASL and smoothed by a 3.5x3.5x3.5mm FWHM Gaussian filter. PyCharm was used to create voxelwise within-subject standard deviation (SD) maps of the difference between sessions/blocks (SD_diff) and further image processing.

Analysis
A neuroradiologist (VK), with 10 years of experience, manually delineated the hyperintense glioma regions-of-interest (ROIs) on the post-contrast T1w. The MTR_asym of tumor ROI was compared to a similarly-sized contralateral region and the between-days SD_diff.

SD_diff maps were calculated for the within-session, in-between-session, and between-day reproducibility. The within-session SD_diff maps of the volunteers were compared to the within-session SD_diff maps of the patients. Variance component analysis (Rstudio VCA) was performed for the whole brain, cortical gray matter, deep white matter, thalamus, and putamen. The normal APT-CEST values for healthy volunteers were also calculated for these masks and compared to mean tumor values. Finally, the scanning variability was related to the MTR_asym effect size between tumor and contralateral tissue. The effect size is given by mean_tumor_MTR_asym - mean_contralateral_MTR_asym.

Results

APT-CEST values in healthy volunteers
Normal whole brain MTR_asym ranged between -0.92% and 1.96% —5^th and 95^th percentile, respectively — with a mean of 0.49% (Table 1).

SD_diff map
The within-session SD_diff maps showed a low variance within each session (fig. 2), but increased with larger time periods between scans. The mean voxelwise within-session, in-between-session and between-days within-subject SD_diff were 0.033 (95^th percentile is 0.13%), 0.064 (95^th percentile is 0.25%), 0.093% (95^th percentile is 0.36%) respectively. The variance was highest in the skull base (within-session, in-between-session, and between-days SD_diff: 0.57, 1.24, and 1.72%) and lowest occipitally and centrally (within-session, in-between-session, and between-days SD_diff: 0.06, 0.1, and 0.13% and 0.08, 0.15, and 0.22% respectively for occipital and central regions).

Covariance analysis in healthy volunteers
The total variance was unexpectedly low and appears to be mainly explained by variability between participants rather than measurement error (Table 2). The variances corroborate the visual findings shown by the SD_diff maps (fig. 2). Moreover, the reproducibility decreased with increasing time between scans.

Findings in glioma patients
The within-session within-subject SD_diff for patients was 0.024% (n=10). Intratumoral mean MTR_asym ranged from 0.56 to 3.11%. Contralateral same-sized ROI MTR_asym in healthy-appearing tissue were 0.0-0.94% (fig. 3). For GBM (n=7), the mean MTR_asym value of the tumors was 2.27% whereas the mean contralateral MTR_asym value was 0.75. The effect size between GBMs and contralateral tissue is 1.52. The SD_diff is 6% of the effect size when considering the between-days SD_diff in the healthy volunteers.

Discussion

These findings show that APT MTR_asym measurements have good reproducibility for healthy volunteers even over longer periods of time in a clinical setting. Note that the mean between-days session SD_diff was much smaller than the absolute APT-CEST values. However, focal signal inhomogeneity in areas close to the skull base can be an issue.
The in-between scan variance can be caused by the correction of B0 inhomogeneity and potentially also by registration errors.
Tumors expectedly showed larger APT-CEST values than healthy tissue. The effect size between tumors and healthy tissue was large compared to the mean between-days within-subject SD_diff. Tagao et al. used a similar scanning protocol and found comparable results.⁷ However they looked at mean tumor APT-CEST values, while we also performed a voxel-wise reproducibility analysis on healthy volunteers, which allows for more general conclusions concerning broader clinical applications.

Conclusion

The current study demonstrates that APT-CEST results have a very high consistency between scan timepoints for the entire brain and masked regions including tumor at 3T. APT-CEST can therefore be recommended for clinical applications in the brain including monitoring and cut-off value determination.

Acknowledgements

No acknowledgement found.

References

1. Zhou J, Lal B, Wilson DA, Laterra J, van Zijl PCM. Amide proton transfer (APT) contrast for imaging of brain tumors. Magn Reson Med. 2003;50:1120–1126.

2. Zhou J, Payen J-F, Wilson DA, Traystman RJ, van Zijl PCM. Using the amide proton signals of intracellular proteins and peptides to detect pH effects in MRI. Nat Med. 2003;9:1085–1090.

3. Zou T, Yu H, Jiang C, Wang X, Jiang S, Rui Q, et al. Differentiating the histologic grades of gliomas preoperatively using amide proton transfer-weighted (APTW) and intravoxel incoherent motion MRI. NMR Biomed. 2017;31. doi:10.1002/nbm.3850

4. Jiang S, Eberhart CG, Zhang Y, Heo H-Y, Wen Z, Blair L, et al. Amide proton transfer-weighted magnetic resonance image-guided stereotactic biopsy in patients with newly diagnosed gliomas. Eur J Cancer. 2017;83:9–18.

5. Dou W, Lin C-YE, Ding H, Shen Y, Dou C, Qian L, et al. Chemical exchange saturation transfer magnetic resonance imaging and its main and potential applications in pre-clinical and clinical studies. Quant Imaging Med Surg. 2019;9:1747–1766.

6. Zhang Y, Yong X, Liu R, Tang J, Jiang H, Fu C, et al. Whole-brain chemical exchange saturation transfer imaging with optimized turbo spin echo readout. Magn Reson Med. 2020;84:1161–1172.

7. Togao O, Hiwatashi A, Keupp J, Yamashita K, Kikuchi K, Yoshiura T, et al. Scan-rescan reproducibility of parallel transmission based amide proton transfer imaging of brain tumors. J Magn Reson Imaging. 2015;42:1346–1353.