Synopsis

Chemical Exchange Saturation Transfer (CEST) offers unique contrasts sensitive to micro environmental information such as protein concentration and pH. Thus CEST could be a promising biomarker to investigate radiotherapy-induced changes in tissue. In this study we examine brain tumor patients in the course of definitive radiotherapy on a 7 T whole-body scanner using the relaxation compensated contrast AREX. We compare the results of CEST imaging to Single Voxel Spectroscopy and high-resolution T₂-weighted imaging.

Purpose

Radiotherapy is commonly used in the management of glioblastoma¹. However, reliable biomarkers to assess treatment response in early stages are urgently needed. Chemical Exchange Saturation Transfer (CEST) allows the indirect detection of mobile proteins. Several studies have shown that CEST contrasts depend on protein concentration, protein folding state and pH^2-5. Consequently CEST may be a promising contrast to assess radiotherapy-induced changes. Here we present the first results of a follow-up study of inoperable glioma patients in the course of radiotherapy using relaxation compensated CEST at a 7T whole–body scanner.

Methods

One patient with newly diagnosed glioblastoma was examined one day before definitive proton radiotherapy (total duration 6 weeks, 60 Gy(RBE) in 30 fractions, Figure 1), one week after the end of radiotherapy and a third time 6 weeks follow-up. The examinations and radiation treatments were approved by the local ethics committee of the Medical Faculty of the University of Heidelberg and written consent was received from the patient.

Imaging: Spectral highly-resolved Z-spectra were obtained by a centric-reordered 2D-GRE-CEST sequence (3 slices with resolution of 1.72x1.72x5mm³) implemented on a 7 T whole body scanner (Siemens Healthineers, Erlangen, Germany) using a 24-channel head coil for reception. Two CEST measurements with different B₁ were performend (150 Gaussian-shaped pulses with t_p=15ms, DC=60% and 59 adapted frequency offsets). T₁–mapping was achieved by adiabatic inversion recovery GRE. High resolution 3-dimensional T₂–weighted imaging was performed using a TSE sequence (resolution: 0.3x0.3x2mm³). The WASABI⁶ sequence was used to map B₀ and B₁ field inhomogeneities. Water-suppressed ¹H single-voxel spectra (SVS) of tumor tissue and contralateral white matter (CLWM) were acquired with a semiLASER sequence⁷ (voxel size: (2cm)³, TR/TE/α = 4000ms/144ms/90°, 48 averages).

Post processing: Image registration was performed to first measurement (pre treatment) using the MITK software⁸. Lorentzian fit analysis allowed the isolation of the CEST peaks in B₁-corrected Z-spectra⁹. Application of the inverse metric at -3.5ppm yields the relaxation compensated CEST contrast AREX_NOE¹⁰. The downfield rNOE suppressed AREX (dnsAREX²) effect was calculated using the ratio r_rNOE = 0.2:

$$dnsAREX (+3.5ppm) = AREX(+3.5ppm) - r_{rNOE} \cdot AREX(-3.5ppm)$$

Results

Application of an external registration transformation allows direct and robust comparison of 2D CEST images, which were acquired in different examinations (Fig. 2). The external registration transformation was determined by the co-registration of T₂-weighted 3D image datasets. Results for an exemplary slice are depicted in Figure 2. The dilated ventricles in the third measurement are due to a constriction caused by tumor expansion in a lower part of the cerebrospinal fluid system (not visible in the demonstrated slice).

Tumor volume apparent on T₂–weighted images increases after therapy. AREX_NOE is decreased in the tumor regions, whereby the likely necrotic part has values close to zero. The AREX_NOE hypointense region enlarges after therapy. dnsAREX before treatment shows no direct correlation to the tumor region on T₂, however it depicts a hyperintense ring structure in the edema around the necrotic part (see Figure 2 ROI). After treatment this hyperintensity is vastly reduced in size and intensity.

SVS (Figure 3) exhibits the expected signal intensities of N-Acetyl aspartate (NAA), Choline (Cho) and Creatine (Cr) in healthy tissue (green voxel). The Choline amplitude in the tumor (magenta voxel) is increased before treatment. After treatment the CLWM shows only slight changes, however the tumor region does not contain any significant resonances anymore.

Discussion

AREX_NOE values pre treatment are in accordance to reported values¹⁰. The hypointense AREX_NOE region enlarges after therapy, which indicates a waterlike composition. This finding is verified by SVS, where the absence of characteristic peak indicates that only a negligible amount of mobile metabolites is apparent. The reduction of the protein concentration dominated contrast dnsAREX² after treatment also confirms this hypothesis. Contrary to these findings tumor volume visible in the T₂-weighted images increases after treatment. Consequently CEST might be able to distinguish pseudo from true progression. For a definitive statement about the response the 6 month follow-up has to be examined.

Additionally dnsAREX pre treatment might be helpful for an identification of starting infiltration and malignization of the tumor using CEST¹⁰. Thus the longitudinal study design might allow to investigate the possible prediction of tumor growth on CEST images.

Conclusion

This is the first study to longitudinal investigate inoperable brain tumors in the course of radiotherapy with CEST at 7T. The mapping of AREX_NOE and dnsAREX provided distinct insights into glioblastoma substructure before and after treatment and might help to guide and improve the treatment planning and monitoring.

We will continue to explore the potential of CEST imaging to monitor effects due to therapy response or tumor progression.

Acknowledgements

References

1. Stupp, R. et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005, 352:987–996

2. Zaiss, M. et al. Downfield-NOE-suppressed amide-CEST-MRI at 7 Tesla provides a unique contrast in human glioblastoma. Magn. Reson. Med. 2016; doi: 10.1002/mrm.26100

3. Jones, CK. et al. Amide proton transfer imaging of human brain tumors at 3T. Magn. Reson. Med. 2006; 56:585–592.

4. Zaiss, M. et al. MR imaging of protein folding in vitro employing Nuclear-Overhauser-mediated saturation transfer. NMR Biomed. 2013; 26(12):1815-1822

5. Zhou, J. et al. Using the amide proton signals of intracellular proteins and peptides to detect pH effects in MRI. Nat Med. 2003;9:1085–1090.

6. Schuenke, P. et al. Simultaneous mapping of water shift and B 1 (WASABI)-Application to field-Inhomogeneity correction of CEST MRI data. Magn. Reson. Med. 2016;doi: 10.1002/mrm.26133

7. Scheenen, TW. et al. Towards 1H-MRSI of the human brain at 7T with slice-selective adiabatic refocusing pulses. Magn Reson Mater Phy. 2008; 21:95-101

8. Nolden, M. et al. The Medical Imaging Interaction Toolkit: challenges and advances. International Journal of Computer Assisted Radiology and Surgery. 2013 doi: 10.1007/s11548-013-0840-8

9. Windschuh, J. et al. Correction of B1-inhomogeneities for relaxation-compensated CEST imaging at 7 T. NMR Biomed. 2015; 28:529–37.

10. Zaiss, M. et al. Relaxation-compensated CEST-MRI of the human brain at 7T: Unbiased insight into NOE and amide signal changes in human glioblastoma. Neuroimage 2015; 112:180-188