Room: Karl Dean Ballroom C
Purpose: In a study comparing clinical outcomes, it was observed that meningioma patients treated with proton therapy presented more frequently with radiation-induced visual toxicity than those treated with photon therapy. This work aims to identify the factors that led to this observed outcome.
Methods: 82 patients were selected for analysis based on availability of dosimetric and clinical data. The cohort consisted of 42 proton-treated patients (10 with toxicity) and 40 photon-treated patients (3 with toxicity). All recorded visual toxicities were CTCAE grade 2 or 3. Multivariate logistic regression analysis with backward elimination (criteria of p < 0.05) was performed with radiotherapy (RT)-associated visual toxicity as the response parameter. Initial predictor variables included age at RT, gender, optic structure maximum dose, radiation modality, tumor site, and pre-RT surgical status. Optic structure maximum dose was compared across modalities. Utilizing Monte Carlo techniques, relative biological effectiveness (RBE)-weighted dose in optic structures was investigated for proton patients with visual toxicities.
Results: Multivariate analysis revealed that radiation modality (p = 0.02) and age at RT (p = 0.04) were the only significant predictors associated with RT-induced visual toxicity. The odds ratio for modality was 5.76 (95% CI: 1.49-29.6) while that of age at RT was 1.05 (95% CI: 1.00-1.10). For a constant proton RBE of 1.1, there was no significant association between radiation modality and optic structure maximum dose. Variable RBE-weighted dose resulted in a median increase in maximum optic structure dose of 8% (range: 4-20%) over constant RBE dose for patients with visual toxicity.
Conclusion: In this cohort of patients treated for meningioma with proton or photon therapy, radiation modality and age at RT were the only significant predictors for RT-associated visual toxicity. Further investigation of the influence of modality-specific factors is warranted.
Funding Support, Disclosures, and Conflict of Interest: Funding for this work was provided by National Institutes of Health grant 5U19CA021239 and Cancer Prevention and Research Institute of Texas grant RP160232.