Room: AAPM ePoster Library
Purpose: Quantitative MRI (qMRI) has shown significant promise for defining tumor margins, assessing invasion, and quantifying tumor changes after treatment. By implementing qMR imaging on a 0.35T MR-linac, acquisition of serial images during the course of radiation therapy (RT) is possible. This work evaluates the feasibility of using onboard qMRI to detect changes within the tumor and brain tissue.
Materials: A multi-echo gradient echo MR method (STAGE: Strategically Acquired Gradient Echo, TR: 40ms, TE: 5, 20.63, 34.14 ms, FA: 10°, 50°) was optimized at 0.35T to generate T1, R2* and proton density (PD) brain maps in 10 minutes. Ten patients with Glioblastoma were consented to an Institutional Review Board approved prospective trial. Imaging was performed at simulation (Sim), weekly during RT, and 3 months post RT (Follow-up). To assess qMRI stability, the coefficient of variation (CV) was calculated for normal white and gray matter (WM and GM, respectively). Temporal changes of qMRI map voxels within the clinical target volume (CTV) were evaluated for response assessment.
Results: Average T1 values in the normal sub-cortical WM across all patients were 384.50±15.70ms (range: 351.53-405.00ms). CV of average T1 values across all timepoints in the normal region was 0.73%-2.49% (range: 1.28±0.56%), indicating stable qMRI performance in normal tissue at low field. Conversely, in the CTV, the CV of T1 values was 5.15 ±4.81% (range: 1.44-16.03%). In the R2* and PD maps for the CTV, the CVs were 6.22±4.59% (range: 1.70%-13.89%) and 2.44±1.18% (range: 1.40-4.84%) respectively. The most substantial T1 change (62.67%) occurred within a CTV between Sim and Follow-up.
Conclusions: STAGE qMRI maps acquired on a 0.35T MR-linac were stable in normal brain tissue while changes within the CTV were substantial yet patient-specific. Future work will correlate long term survival data to variations in qMRI maps within the CTVs to highlight potential imaging biomarkers.
Funding Support, Disclosures, and Conflict of Interest: The submitting institution holds research agreements with Philips Healthcare, ViewRay, Inc., and Modus Medical. Research partially supported by the National Cancer Institute of the National Institutes of Health under Award Number R01CA204189 and ASTRO-AAPM Physics Resident/Postdoctoral Fellow Seed Grant. The PI is on the Philips Healthcare Advisory Board.