0011

High-resolution T₁ Mapping of High-grade Glioma

Zhibo Zhu¹, Jay Acharya², Yannick Bliesener¹, R. Marc Lebel^3,4, Richard Frayne^3,5, and Krishna S. Nayak^1,2
¹Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, United States, ²Department of Radiology, University of Southern California, Los Angeles, CA, United States, ³Radiology and Clinical Neuroscience, Hotchkiss Brain Institute, University of Calgary, Calgray, AB, Canada, ⁴Global MR Applications & Workflow, GE Healthcare, Calgary, AB, Canada, ⁵Seaman Family MR Research Centre, Foothills Medical Centre, Calgary, AB, Canada

Synopsis

Native T₁ mapping is necessary for quantitative DCE-MRI of brain tumor and may have independent predictive value. Previous investigations using coarse spatial resolution have found tumor to have longer T₁ compared to normal white matter. In this work, we evaluate a recent millimeter-resolution whole-brain T₁ mapping approach in patients with high-grade glioma. T₁ values in tumor and peritumoral regions were higher than that of normal white matter, consistent with literature. We also observed T₁ spatial heterogeneity in these regions, further supporting the need for high resolution pre-contrast T₁ mapping for quantitative DCE-MRI.

Purpose

To apply millimeter-resolution whole-brain pre-contrast T₁ mapping to patients with high-grade glioma at 3T. To report T₁ values and spatial heterogeneity in tumor regions of interest.

Introduction

Native T₁ mapping before contrast agent injection is necessary for quantitative DCE-MRI of brain tumor (BT) and may have independent predictive value^1,2. Previous investigations have utilized coarse spatial resolution to avoid long scan time and accepted the possible errors that would propagate to the DCE-MRI analysis³. We recently demonstrated millimeter-resolution whole-brain T₁ mapping using sparse sampling and model-based reconstruction that required 2.5 minutes at 3T⁴. Here, we apply this approach to patients with high-grade glioma and investigate T₁ values and spatial heterogeneity in tumor regions.

Method

Data were acquired from 13 patients (4M/9F, age 42-80) each with three scan sessions (time points, TP) on a clinical 3T MRI scanner (MR750, GE Healthcare) with a 12-channel Head-Neck-Spine receiver coil. T₁ mapping was performed using a sparsely sampled variable flip angle (VFA) approach paired with a model-based reconstruction⁴. Relevant imaging parameters: TR=4.9ms; TE=1.9ms; FOV=24x24x24cm³; matrix size=256x240x120; resolution=1x1x2mm³; scan time=147s. B₁+ maps were acquired using Bloch-Siegert approach⁵.

Tumor regions of interest (ROIs) were manually drawn by a board-certified neuroradiologist with 10-years of experience. For all TPs, we report mean (M) and standard deviation (SD) of T₁ values within each ROIs. We report coefficient of variance (COV) as a coarse measure of tumor T₁ heterogeneity. These values were compared against the measured values and literature values⁶ of healthy white matter.

Results

All 39 T₁ maps were of high diagnostic quality despite one case (M49, TP1) showing unexpected inhomogeneity in cerebral spinal fluid. Individual tissues are clearly differentiable on T₁ maps.

Figure 1 shows M₀ (left) and T₁ (right) maps from three orthogonal orientations centred on the tumor in three representative patients at the TP1. Tumor ROIs are delineated as red contours on the M₀ maps. T₁ values in tumor regions for each patient were 1749±140ms, 1765±156ms and 1792±128ms, and in contralateral white matter were 1011±71ms, 1102±87ms and 1163±83ms. Cavities filled with fluid are identified by either observing long T₁ values (>3000ms) or fluid collection between skull and brain parenchyma.

Figure 2 shows zoomed-in versions of Figure 1, focusing on the tumor ROIs. The T₁ colorscale is tightened to highlight heterogeneity. The results show T₁ heterogeneity inside and around tumor. Light green arrow: T₁≈2000ms. Green arrow: T₁≈1500ms. Dark green arrow: T₁≈1300ms.

Figure 3 shows the zoomed-in tumor T₁ maps of one representative patient with tumor progression. Averaged T₁ values increased from 1983ms to above 2300ms within the ROIs. There was no substantial change in COV between TP2 (11.1%) and TP3 (11.0%). The cavity appears squeezed and the area with longer T₁ increased in size since TP1.

Figure 4 and Table 1 report T₁ values (M±SD ms, COV %) within ROIs for different patients at different TPs. Tumors in six representative patients were analyzed. The reported values show a good agreement with literature values (1392-3601ms^7–9), except that one case had smaller mean (1372ms) thus larger COV (17.0%) than others. There is more variation in T₁ mean and COV among patients than that between TPs. Mean COV within tumor ROI’s (10.4%) was found to be larger than that measured within WM ROI’s (8.4%) in this dataset and literature value (4.2%)⁶.

Discussion

This work used a model-based sparse VFA T₁ mapping approach that was designed to achieve T₁ mapping with spatial resolution that matched those of recent millimeter-resolution whole-brain DCE-MRI methods^10,11. We report T₁ values that are consistent with literature values and distinct from healthy tissues. We observed longitudinal changes in native T₁ of suspicious voxels and longitudinal changes in area.

The results suggest that glioma has slightly stronger T₁ heterogeneity compared to healthy tissues. The finding highlights the necessity of pre-contrast T₁ mapping with matched spatial resolution as high-resolution voxel-wise DCE-MRI analysis. The heterogeneity could be related to tumor cellular environment, for example, different T₁ values can be observed from cystic tumor, solid tumor and peritumoral regions. Its potential diagnostic values in a more accurate quantitative DCE-MRI will be of interest.

The results did not show clear correlation between tumor T₁ and progression. In the case of obvious progression (Figure 3), glioma T₁ heterogeneity seemed to behave independently from the progression. Further details about the patient, e.g., treatment, should help exploring and understanding progression and/or treatment response of T₁ values and strengthens the potential diagnostic value of high-resolution T₁ mapping.

The results show more T₁ variation among patients than between different TPs. The variation might result from various factors, e.g., age, cellular heterogeneity and treatment. This study would benefit from taking a closer look at the correlation between the variation and potential causes such as inter-patient variability.
There are also several limitations of this work. Due to scan time limit, it was impractical to acquire fully sampled reference T₁ values from patients. This work also ignored any bias in VFA T₁ mapping (relative to Inversion Recovery approaches)¹² due to magnetization transfer effects^13,14.

Conclusion

We evaluated a novel millimeter-resolution whole-brain T₁ mapping method in patient with high-grade glioma. Glioma T₁ value is found to be longer and more heterogeneous compared to normal appearing white matter.

Acknowledgements

We acknowledge National Institute of Health grant support (#R33-CA225400).

References

1. Sourbron SP, Buckley DL. Classic models for dynamic contrast-enhanced MRI. NMR Biomed. 2013;26(8):1004-1027. doi:10.1002/nbm.2940

2. Bane O, Hectors S, Wagner M, et al. Accuracy, repeatability and interplatform reproducibility of T1 quantification methods used for DCE-MRI: results from a multicenter phantom study HHS Public Access. Magn Reson Med. 2018;79(5):2564-2575. doi:10.1002/mrm.26903

3. Schabel MC, Parker DL. Uncertainty and bias in contrast concentration measurements using spoiled gradient echo pulse sequences. Phys Med Biol. 2008;53(9):2345-2373. doi:10.1088/0031-9155/53/9/010

4. Zhu Z, Bliesener Y, Lebel RM, Nayak KS. Direct Estimation of pre-contrast T1 for DCE-MRI. ISMRM 28th Scientific Session. https://cds.ismrm.org/protected/20MPresentations/abstracts/3468.html. Published 2020. Accessed December 7, 2020.

5. Sacolick LI, Wiesinger F, Hancu I, Vogel MW. B 1 mapping by Bloch-Siegert shift. Magn Reson Med. 2010;63(5):1315-1322. doi:10.1002/mrm.22357

6. Stanisz GJ, Odrobina EE, Pun J, et al. T1, T2 relaxation and magnetization transfer in tissue at 3T. Magn Reson Med. 2005;54(3):507-512. doi:10.1002/mrm.20605

7. Müller A, Jurcoane A, Kebir S, et al. Quantitative T1-mapping detects cloudy-enhancing tumor compartments predicting outcome of patients with glioblastoma. Cancer Med. 2017;6(1):89-99. doi:10.1002/cam4.966

8. Hattingen E, Müller A, Jurcoane A, et al. Value of quantitative magnetic resonance imaging T1-relaxometry in predicting contrast-enhancement in glioblastoma patients. Oncotarget. 2017;8(32):53542-53551. doi:10.18632/oncotarget.18612

9. Badve C, Yu A, Dastmalchian S, et al. MR fingerprinting of adult brain tumors: Initial experience. In: American Journal of Neuroradiology. Vol 38. American Society of Neuroradiology; 2017:492-499. doi:10.3174/ajnr.A5035

10. Guo Y, Lebel RM, Zhu Y, et al. High-resolution whole-brain DCE-MRI using constrained reconstruction: Prospective clinical evaluation in brain tumor patients. Med Phys. 2016;43(5):2013. doi:10.1118/1.4944736

11. Guo Y, Lingala SG, Bliesener Y, Lebel RM, Zhu Y, Nayak KS. Joint arterial input function and tracer kinetic parameter estimation from undersampled dynamic contrast-enhanced MRI using a model consistency constraint. Magn Reson Med. 2018;79(5):2804-2815. doi:10.1002/mrm.26904

12. Stikov N, Boudreau M, Levesque IR, Tardif CL, Barral JK, Pike GB. On the accuracy of T1 mapping: Searching for common ground. Magn Reson Med. 2015;73(2):514-522. doi:10.1002/mrm.25135

13. Malik SJ, Teixeira RPAG, Hajnal J V. Extended phase graph formalism for systems with magnetization transfer and exchange. Magn Reson Med. 2018;80(2):767-779. doi:10.1002/mrm.27040

14. Lee N, Zhu Z, Nayak KS. Variable flip angle T1 mapping with MT-balanced RF pulses. In: ISMRM 28th Scientific Session. Virtual conference; 2020:3793. https://cds.ismrm.org/protected/20MPresentations/abstracts/3793.html. Accessed September 20, 2020.

Figures

Figure 1: M₀ (left) and T₁ (right) maps for three representative patients (F71, F61, F67) at the first time point. Maps are volumetric, and axial, coronal, and sagittal slices centered on the tumor section are shown. Tumor ROIs are delineated by red contours on M₀ maps. Longer T₁ can be observed in corresponding regions on the T₁ maps. In addition, T₁ maps reveal the locations of surgical cavities and extra-axial fluid collection (blue arrows) and evidence of craniotomy (green arrows). White dashed-line boxes indicate the zoomed-in region in Figure 2.

Figure 2: Closeup of T₁ maps from the axial, coronal and sagittal views on three patients shown in Figure 1. Maps are zoomed into the tumor region (delineated by white dashed lines in Figure 1), with constrained display range. The proposed method captures T₁ heterogeneity. Light green: high T₁ values, green: medium T₁ values and dark green: low T₁ values. Coefficients of variation for T₁ inside the ROIs are 10.84%, 9.96%, and 7.31% for the top, middle, and bottom rows, respectively.

Figure 3: Zoomed-in T₁ map from one representative patient (F, 56) showing tumor progression over three time points. At time point 1 (TP1), the post-surgical cavity was noticeably filled with fluid, but no obvious tumor residuals were identified. Tumor showed substantial progression in size from time point 1 (TP1) to time point 3 (TP3) based on visual radiographic assessment. This coincided with an increase in averaged T₁ values within the ROI from 1983ms (TP1) to >2300ms in the respective time range, and shrinkage of cavity volume.

Figure 4: Tumor T₁ and COV from 6 patients that had tumor present at all three time points. (a) Tumor ROI T₁ plotted as M±SD for three time points (left to right). Temporal changes in tumor T1 did not exceed the intra-TP variation, with the lone exception being P6. We did not detect any statistically significant trends from this limited sample. (b) Tumor T₁ COV as a function of TP. The COV was consistent across patients and time points except for P6 at TP1.

Table 1: Patients demographics and T₁ mean, SD and COV within tumor ROIs. Reported T₁ values are longer than healthy white matter literature value (1084ms³) and agree with literature values (1392-3601ms^4–6). Mean COV is 10.4%. Other notes: ^*No tumor residuals: Patient was right after a surgery thus no obvious tumor residuals were identified. ^**Early stage: The tumor was probably in early stage thus not identified. Its progression was later observed in TP2 and/or TP3. ^***Invalid data: Disordered data from other patients thus invalid.

Proc. Intl. Soc. Mag. Reson. Med. 29 (2021)

0011