Synopsis

Keywords: Tumors, Perfusion

The goal of this study was to assess the feasibility of pseudo-continuous arterial spin labeling (pCASL) at 7T on patients with intracranial tumors. We found that 7T ASL imaging was feasible and had a higher resolution than 3T ASL imaging and therefore, may enable in vivo assessment of subtle changes and provides a valuable tool for early detection and mechanism investigation of neurovascular function impairments in patients with intracranial tumors.

Introduction

Perfusion assessment is crucial in diagnosing, classifying, and grading intracranial tumors [1]. Arterial spin labeling (ASL) can measure perfusion non-invasively, quantitatively, and free from radiation [2] and has been widely used in the study of intracranial tumors [3].7T can benefit the practice of ASL from two aspects, the increased signal-to-noise ratio (SNR) and prolonged T1 relaxation time. With 7T ASL, the perfusion status can potentially be visualized more clearly and measured more accurately. However, the practice of ASL at 7T is problematic due to the deteriorated B1 and B0 field, the limited radio-frequency coverage, and the limitation of specific absorption rate (SAR), especially for the pseudo-continuous (pCASL) labeling approach. Recently, with rapid improvement, the practice of pCASL at 7T became feasible. This study performed 7T pCASL on patients with intracranial tumors and compared the perfusion images with those from 3T.

Methods

Ethical approval and informed content were acquired before the study. Four patients were enrolled in this study, one with multinodular vascular neuronal tumor, one with recurrent glioma resection, and two with glioma. The examinations were performed using 3T whole-body scanners (GE Discovery MR 750, GE Healthcare, Milwaukee, WI, or MAGNETOM Skyra, Siemens Healthcare, Erlangen, Germany), and a 7T whole-body scanner (MAGNETOM Terra, Siemens Healthcare, Erlangen, Germany). Routine protocols (T2w, T2 dark-fluid, T1w, and T1w+C) and ASL sequences were performed. For 3T, the commercialized 3D ASL was used, with the post-labeling delay (PLD) set to 2.0s. For 7T, all the patients received a prototype 2D pCASL. Additionally, the subject with glioma resection recurrence also received a prototype 3D pCASL at 7T and dynamic susceptibility-weighted perfusion imaging (DSC) at 3T. The 7T 2D pCASL were acquired with gradient echo sequence with the following parameters: 11 slices sequentially played out from inferior to superior, acquisition time per slice 150 ms, slice thickness 5 mm, voxel size 2.2x2.2x5.0 mm³, labeling duration 1.5 s, and PLD 2.0s. For 7T 3D acquisition, 24 slices were acquired, slice thickness 4 mm, voxel size 2.2 x 2.2 x 4.0 mm³, the labeling duration 1.5s, and PLD 2.0s.

Results

In all the cases, 7T and 3T ASL imaging showed roughly similar perfusion in the lesion region. ASL at 7T had a higher resolution compared with ASL at 3T. 7T ASL showed better details and a clearer margin of the cortical area and tumor region, while 3T ASL had more blurring (Figure1-3). In the case of glioma resection recurrence, the hyper-perfusion area was not distinct with 3T ASL imaging because of the hemorrhage and postoperative changes. But with 7T ASL imaging, the contrast between the lesion and the normal and hyper-perfusion margin were more apparent (Figure 2). Simultaneously, the surgical resection area was better presented with 7T ASL imaging (red arrow in Figure 2). In the case of glioma, the SNR significantly improved on 3D ASL compared with 2D ASL at 7T (Figure 3). Both 7T and 3T ASL showed hyper-perfusion in the lesion area of the right temporal lobe. However, 3T ASL also showed hyper-perfusion in the left occipital lobe, which was not shown on 7T ASL or 3D DSC imaging (Figure 3). 3T ASL may give a false appearance caused by the volume effect due to insufficient resolution.

Discussion and conclusion

7T ASL imaging was feasible and had a higher resolution than 3T ASL imaging, especially for cortex perfusion imaging. High-resolution ASL at 7T may enable in vivo assessment of subtle changes and provides a valuable tool for early detection and mechanism investigation of neurovascular function impairments in patients with intracranial tumors. In the future, more experiments will be performed as this study included only four patients due to its exploratory nature.

Acknowledgements

None.