E. Younus1,2, J. Qian2, D. H. Pafundi3, D. R. Johnson4, S. Chitsazzadeh5, T. Kaufmann1, D. M. Trifiletti3, J. L. Peterson3, M. M. Kim6, S. A. Vora7, U. Sener8, M. Ruff8, P. D. Brown2, J. N. Sarkaria2, A. Mahajan2, E. S. Yan2, N. N. Laack II2, R. O. Kowalchuk1, D. H. Brinkmann2, and W. Breen2; 1Mayo Clinic, Rochester, MN, 2Department of Radiation Oncology, Mayo Clinic, Rochester, MN, 3Department of Radiation Oncology, Mayo Clinic, Jacksonville, FL, 4Mayo Clinic, Division of Nuclear Medicine, Rochester, MN, 5Mayo Clinic, Phoenix, AZ, 6Department of Radiation Oncology, University of Michigan, Ann Arbor, MI, 7Department of Radiation Oncology, Mayo Clinic, Phoenix, AZ, 8Mayo Clinic, Department of Neurology, Rochester, MN
Purpose/Objective(s): Although gadolinium contrast-enhanced MRI is standardly used to identify the gross tumor volume (GTV) during target delineation for radiotherapy for glioblastoma, it has been established that biologically relevant, non-enhancing tumor extends beyond this volume. To identify non-enhancing tumor, advanced imaging techniques including 18F-DOPA PET/CT and advanced MRI (aMRI) utilizing dynamic contrast enhanced perfusion MRI and high b-value diffusion MRI have been employed at different institutions. This analysis is the first to compare these imaging modalities within the same patients and compare the tumor volumes identified to evaluate their impact on radiotherapy planning. We hypothesized advanced imaging would identify non-enhancing tumor in the majority of patients. Materials/
Methods: Patients enrolled on a prospective study underwent 18F-DOPA PET/CT and aMRI (perfusion and diffusion MRI) along with standard MRI sequences. GTV_PET was delineated using a predetermined threshold of mean SUV in normal contralateral brain. GTV_Diffusion was delineated by thresholding the FLAIR abnormality region on high b-value scans. GTV_Perfusion was delineated by thresholding the FLAIR abnormality region on the cerebral blood volume image obtained by fitting Tofts model to the perfusion data. A high dose GTV (HGTV) was then generated by adding GTV_PET, GTV_Diffusion, and GTV_Perfusion to the standard GTV (GTV_convMRI) comprised of T1 contrast enhancement and resection cavity. Isotropic 5- and 10-mm expansions were applied to the GTV_convMRI, and the volume of GTV_PET, GTV_Diffusion, and GTV_Perfusion outside of these expansions was quantified and compared. Results: Of the 52 patients enrolled to-date, 4 had GTV_Perfusion unavailable due to technical difficulties, 2 did not undergo aMRI due to MRI related contraindications, and 2 withdrew consent. The median (inter-quartile range) volume, in cm3, of GTV_convMRI, GTV_Diffusion, GTV_Perfusion, and GTV_PET were 24.9 (13.7-40.5), 2.1 (0.8-5.8), 11.2 (4.2-19.4) and 5.8 (1.6-13.1), respectively. In the 44 subjects included in the analysis, the HGTV was outside the GTV_convMRI+5mm in 39 (89%), while the HGTV was outside the GTV_convMRI+10mm in 25 (57%) patients. GTV_Diffusion and GTV_Perfusion were found to be outside of GTV_PET in 93% and 95% of patients, respectively. Conclusion: 18F-DOPA PET and advanced MRI techniques identify tumor over one centimeter beyond contrast enhancement in over half of cases. These imaging modalities appear to be complementary, rather than duplicative, with only partial overlap in the majority of patients. Additional prospective data is needed to demonstrate how altering target volumes impacts patient outcomes.
