1034 - Patterns of Failure after 18F-DOPA PET and MRI-Guided Hypofractionated Proton Therapy in Older Adults with Glioblastoma

Purpose/Objective(s): We have previously reported the primary endpoint from a phase II prospective trial using short course hypofractionated proton therapy and advanced medical imaging for elderly patients with glioblastoma (GBM). Here, we present patterns of radiographic failure defined at first progression.

Materials/Methods: Patients aged =65 years with an ECOG of 0-2 and newly diagnosed WHO grade 4 GBM were enrolled. From 39 patients enrolled and treated per protocol, 31 had disease progression at last follow-up (median FU 13 months) and were included. GTVs were defined by 18F-DOPA-avid regions with tumor/normal brain SUV ratio =2.0 (GTV-PET), and by contrast-enhancing areas on MRI (GTV-MRI). The CTV was an isometric 1-cm expansion from the combined GTVs. Patients received either a 5-fraction (=65 cc GTV) or 10-fraction (>65 cc GTV) RT regimen with concurrent and adjuvant oral temozolomide for 6 cycles. Using a simultaneous integrated boost technique, doses of 35/30/25 Gy (5 fractions) or 40/35/30 Gy (10 fractions) were delivered to the GTV-PET, GTV-(MRI+PET), and CTV, respectively. For the patterns of failure analysis, DICOM files from the contrast-enhanced MRI brain with perfusion from the date of progressive disease (PD) were imported and fused with the planning CT, RT structure, and RT plan on MIM Software. Contrast-enhanced regions felt to represent PD by Radiology report, aided by perfusion, were contoured by a radiation oncologist. PD was classified via two different methods; a) according to the percentage of recurrent tumor overlap with target volumes; b) according to the lowest radiation isodose line interval involved by PD. Both methods were graded as either central (=95%), in-field (80 - <95%), marginal (20 - <80%), or distant (<20%).

Results: Seventeen (55%) patients were treated over 10 fractions, whereas 14 (45%) patients were treated over 5 fractions. The PD volume overlapped by any amount with GTV-PET, the GTV-MRI, and the CTV in 25 (81%), 29 (94%), and 30 (97%) patients, with a mean volume overlap of 24% (95%CI, 16-32), 63% (95%CI, 53-73), and 95% (95%CI, 92-98), respectively. Most PDs were central relative to CTV volumes (69%), marginal to GTV-MRI (55%), whereas PDs were predominantly distant (58%) relative to GTV-PET (Table 1).

Conclusion: A dose-escalated hypofractionated proton therapy regimen guided by 18F-DOPA PET and MRI provides promising central and in field disease control advantages in older adults with GBM.