Sep 29

SS 27 - Advances in Multimodal Therapy for Glioblastoma: From Phase I to Phase III Trials

241 - Initial outcomes from a phase II study of multiparametric MR-guided high-dose response adaptive radiotherapy with concurrent temozolomide in patients with newly diagnosed glioblastoma

01:10pm - 01:20pm ET

Room 153

Presenter(s)

Michelle Kim, MD - University of Michigan, Ann Arbor, Michigan

M. M. Kim¹, M. P. Aryal¹, B. S. Rosen¹, H. A. Parmar², D. You¹, D. Leung³, N. Clarke², J. Fortunato², W. Al-Holou⁴, J. A. Heth⁴, D. Altshuler², T. Hollon⁴, M. Kalandjian², J. Sharp², D. M. Edwards¹, D. R. Wahl¹, T. S. Lawrence¹, K. Suresh¹, and Y. Cao¹; ¹Department of Radiation Oncology, University of Michigan, Ann Arbor, MI, ²University of Michigan, Ann Arbor, MI, ³Department of Neurology, University of Michigan, Ann Arbor, MI, ⁴Department of Neurosurgery, University of Michigan, Ann Arbor, MI

Purpose/Objective(s):

Adversely prognostic hypercellular and hyperperfused glioblastoma (GBM) tumor regions progressive during radiation therapy (RT) are identifiable with multiparametric MRI (mpMRI) and predict tumor recurrence and survival. We hypothesized that treating these putative emerging areas of treatment resistance using dose-intensified response-adaptive chemoRT would improve survival (OS).

Materials/Methods:

This phase II single-institution trial enrolled adult patients with newly diagnosed WHO grade IV glioma between 11/2020-10/2025. Maximum tumor diameter requirement was <5 cm, based on combined hypercellular tumor volume (TV_HCV, the high b-value diffusion-weighted MRI abnormality >mean contralateral normal brain + 2 SD) and hyperperfused tumor volume (TV_CBV, the DCE perfusion MRI abnormality >contralateral normal frontal lobe grey matter + 1 SD). Combined TV_HCV/ TV_CBV was treated with integrated boost to 50 Gy in 20 fractions of 2.5 Gy each, and following mid-RT reassessment, persistent and developing TV_HCV/ TV_CBV was boosted to 30 Gy in 10 fractions of 3 Gy each. Standard contrast enhancement/cavity was treated to 60 Gy in 30 fractions. A sample size of 30 patients provided 80% power to detect an improvement in 12-month OS from historical benchmark of 0.65 to 0.85, based on one sample binomial test at a 2-sided significance level of 0.10. Secondary endpoints included RANO-defined progression, toxicity per CTCAE v5.0, quality of life (QOL, per EORTC QLQ-C30/BN20) and symptom burden/interference per MDASI evaluated using reliable change indices censored at time of progression.

Results:

Median follow-up was 46 months (IQR 33-57). Among 30 patients enrolled, 20 (67%) were male, 7 (23%) were RPA class III, 18 (60%) class IV, and 5 (17%) class IV. Five (17%) had MGMT methylated tumors and 3 had IDH1 mutation. Prior to RT, median volume of residual enhancement + cavity was 30 cc, whereas mpMRI-defined hypercellular/hyperperfused TV_HCV/ TV_CBV was 7 cc (range, 2-43) and 31% (range 1-74%) was non-enhancing, with 2 cc (up to 11 cc) extending beyond enhancement/cavity. There were no treatment breaks for boost planning. By midRT, median TV_HCV/ TV_CBV was reduced to 5 cc (range, 0.1-34) and 46% (0-74%) was non-enhancing. The median TV_HCV/ TV_CBV persisting from pre- to midRT was 2 cc (range 0-24), with an additional 2 cc (range 0.1-21) developing outside of the initial region. Median OS was 18 months (95% CI 15-33), and median PFS was 11 months (95% CI 8-20). Grade 3 RT necrosis rate was 7%. No clinically meaningful declines in patient-reported global or brain-specific QOL, or symptom burden or interference were observed.

Conclusion:

Response-adaptive RT targeting adversely prognostic tumor regions defined using diffusion/perfusion MRI identifies a significant proportion of non-enhancing tumor before and emerging during RT in patients with GBM. Dose-intensification is well-tolerated with reasonable initial survival estimates and final analysis pending complete follow-up of all patients.