2645 - Radiation Necrosis, Pseudoprogression, and Enhancing Radiologic Changes after Photon or Proton Chemoradiotherapy for Grade 2-3 IDHmt Gliomas

Purpose/Objective(s): To characterize patterns and severity of MRI contrast enhancement, pseudoprogression (PsP), and radiation necrosis (RN) after photon vs proton chemoRT for IDH-mutated (IDHmt) grade 2-3 gliomas.

Materials/Methods: We reviewed patients receiving photon or proton RT at a single academic center. Inclusion criteria: WHO grade 2-3, IDHmt, known 1p19q status, and concurrent and/or adjuvant chemotherapy. Primary endpoint: New enhancing pseudoprogression or radiation necrosis (PsP/RN), categorized as Grade 1 (asymptomatic radiologic enhancement), Grade 2 (symptomatic), and Grade =3 (admission, bevacizumab, or surgery/biopsy showing isolated or fewer tumor cells). Degree of preop enhancement, extent of surgery, and clinical target volume (CTV) were quantified. All records and images were reviewed by both a resident and senior CNS radiation oncologist. Cox proportional hazards model was used to analyze freedom from =G1 and =G2 PsP/RN (pts censored at disease progression). Factors associated at p<0.1 level on univariate analysis were included in a multivariate analysis.

Results: 108 pts received photon (n=46, median 54 Gy for WHO grade 2 and 59.4 Gy for WHO grade 3) or proton (n=62, median 54 Gy for WHO grade 2 and 60 Gy for WHO grade 3) RT from 2012-2024. Median f/u was 3 yrs (IQR 1.5-6 yrs). 52% had WHO grade 3 tumors (photons: 57%, protons: 48%) and 49% had 1p19q co-deletions (photons: 50%, protons: 48%). Median CTV was 161 cc (photons: 176 cc, protons: 155 cc). 57% had non-enhancing tumors and 66% of the 91 pts with known MGMT status were methylated. Surgical extent was biopsy in 17%, subtotal in 41%, and near/gross total in 41%. There were 38 pts with Grade =1 and 13 with Grade =2 PsP/RN (12 had grade =3 PSP/RN, and 12 had received protons). 1-year freedom from PsP/RN was higher with photon vs proton (Grade =1: 88% vs 70%, p<0.01; Grade =2: 98% vs 87%, p=0.01). Factors associated with PsP/RN on univariate analysis are shown in the Table. Protons were associated with a higher incidence of both Grade =1 and Grade =2 PsP/RN, remaining significant on multivariate analysis. Most Grade =2 events involved the ventricles and/or corpus callosum.

Conclusion: In IDHmt grade 2-3 glioma, protons were associated with a higher rate of both asymptomatic and high grade PsP/RN. These findings are relevant to clinicians interpreting post-RT imaging, provide insights on radiobiologic differences, and support a need to further optimize proton planning. Future research is needed to examine whether the higher apparent biological effect or protons translates into improved disease control.