Feasibility of Dynamic Contrast Enhanced MR Perfusion for Glioblastoma On a Low-Field MRI-Linac System
D Maziero1*, R Stoyanova2, YC Chang2, A Breto2, JC Ford1, EA Mellon1, (1) Sylvester Comprehensive Cancer Center, University of Miami Health System, United States, Miami, FL, (2) University of Miami Miller School of Medicine, Miami, FL
Presentations
WE-IePD-TRACK 3-5 (Wednesday, 7/28/2021) 3:00 PM - 3:30 PM [Eastern Time (GMT-4)]
Purpose: The purpose of this study is to evaluate the feasibility of measuring dynamic contrast enhanced (DCE) perfusion in glioblastoma on an MRI-Linac at 0.35T.
Methods: Images were acquired on a glioblastoma patient prior to first radiotherapy treatment on a 0.35T MRI-linac system using 15-channel surface array coils. A 3D gradient recalled echo sequence was used to acquire continuous 45-volumes with 18 axial slices of the brain during gadoteridol infusion. Each volume and total acquisition time were 9.75s and 7min19s, respectively. The tumor lesion was segmented on post-contrast enhancement T1-weighted images. An extended volume of interest (VOI) was created by expanding the margins of the tumor contour by 8mm to incorporate peritumoral voxels to the analysis. Constrained Non-Negative Matrix Factorization (cNMF) method was applied to the DCE curves in the VOI. The cNMF was set to seek for 3 solutions with the condition that the weights of a given pattern should be more than 40% in a given voxel to label it accordingly. Finally, the signal-vs-time curves, from voxels in a given solution were averaged and analyzed using Tofts’ pharmaco-kinetic model.
Results: The average VOI signal intensity late in the DCE time series was 21% higher relative to the pre-contrast signal. The three temporal curves identified by cNMF and their spatial distribution were consistent with the patterns of: (VAS)-blood vessels, characterized by rapid contrast wash-in and wash-out; (TC1)-well perfused tumor tissue, contrast fast wash-in and gradual wash-out and (TC2)-possibly hypoxic/necrotic tumor tissue, characterized by slow and continuous contrast uptake. The transfer constant Ktrans for patterns TC1 and TC2 were 0.82 and 0.47 min-1, respectively.
Conclusion: To the best of our knowledge, this is the first report of DCE perfusion on a 0.35T MRI-linac system. This technique can be used to evaluate early tumor perfusion response for adaptive radiotherapy of glioblastoma.
ePosters
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email