Although proton beam therapy (PBT) results in decreased dose to non-target tissue compared to modern photon therapy, there are concerns of rare CNS injury occurring following PBT due in part to RBE/LET uncertainties. Such uncertainties can be further exacerbated when delivering PBT using pencil beam scanning (PBS) technology. The primary objective of this study was to report the incidence of CNS toxicities in a large cohort of adult brain tumor patients treated with PBS at a single academic proton center. We hypothesized that PBS can be safely delivered for treatment of CNS tumors.A single-institution retrospective IRB-approved analysis was conducted of all patients with CNS tumors treated with PBT at our institution between February 2016 and April 2020. Patients were excluded if follow up was less than one month. Kaplan-Meier estimates of treatment toxicities were calculated, accounting for death and local recurrence as competing risks. Multivariate analysis (MVA) was performed using the Cox proportional hazard model to determine risk factors associated with symptomatic toxicity.A total of 283 consecutive patients with CNS toxicities (116 [41%] males, 167 [59%] females) completed a course of definitive-intent PBT; 116 patients had meningiomas/pituitary adenomas, 92 patients had gliomas/GBM, and 75 patients had other histologies (chordoma, paraganglioma, ependymoma, medulloblastoma, pineal tumor). Median age was 52 years old (range 18-91). Median dose of PBT was 51.3 Gy (RBE = 1.1, [range 20.4 - 78.8 Gy RBE]). Median dose per fraction was 1.8 Gy/fx (range 1.1-3.8 Gy/fx). Sixty patients [21%] had received some form of prior intracranial RT. Concurrent systemic therapy was delivered in 49 patients [17%]. Median follow-up time was 20 months. The 2-year incidence of symptomatic-treatment toxicity was 8.7%. Symptomatic radiation necrosis occurred in 13 patients including optic neuritis [N = 2], seizure [N = 3], neurologic deficit [N = 9], stroke [N = 1] and brainstem necrosis [N = 2]. Two patients had symptoms with no radiation-related imaging changes including stroke [N = 1] and seizure [N = 1]. MVA identified prior intracranial irradiation (HR 3.901, 95% CI 1.42, 10.73, P = 0.008) and increasing EQD2 of proton radiation (HR 1.077 per Gray, 95% CI 1.00, 1.15, P = 0.036) as predictive factors for increased incidence of adverse events including symptomatic toxicity and asymptomatic radiation necrosis. No PBT treatment planning parameters were found to correlate with toxicity risk.This is the largest series to date reporting outcomes for patients with CNS tumors treated with pencil beam scanning PBT. Our analysis indicates that pencil beam scanning PBT in this setting is well-tolerated with a toxicity profile similar to modern photon therapy. Future studies correlating toxicity-risk with LET are warranted.B. Savla: None. G.S. Alexander: None. K. Sun: None. S.M. Bentzen: Travel Expenses; University of Copenhagen. S. Mossahebi: None. Y. Kwok: None. W.F. Regine: None. M.V. Mishra: Employee; Orthofix. Research Grant; ASTRO, Keep Punching. Advisory Board; Patient Centers Outcomes Research Institute (PCORI. Travel Expenses; Patient Centers Outcomes Research Institute (PCORI.
Read full abstract