Optimal Spot Scanning Treatment Plans in Infants With Brain Tumors in the Posterior Fossa

Abstract

Although treatments with scattered protons are currently available in proton facilities, there is limited experience with the use of spot scanning in the treatment of pediatric patients. Because even very low doses can have detrimental late effects in infants, optimal beam arrangement is critical to maximize the gains of proton radiation therapy in comparison to photon treatment. The purpose of this study was to determine the optimal proton spot scanning parameters for posterior fossa, infant brain tumors. In an IRB approved protocol, 9 consecutive, normal infants (<2 yrs) with head CTs obtained in the ER were selected for this study. For each patient, all relevant normal tissues along with a modeled, standardized posterior fossa GTV were contoured. A 5 mm GTV to CTV and a 3 mm CTV to PTV margin was used. The main parameters varied were spot size, number of fields (1 - 3), direction of fields, minimum MU per spot, and optimization technique: Single- vs. Multi-Field-Optimization (SFO vs. MFO). The treatment planning system was commissioned based on Monte Carlo simulations of a proton facility now under construction. The different spot size options were realistic values that will be selectable on a per patient basis. For each plan, the prescription dose was 54 Gy, the target coverage was held constant and the dose to normal tissues was minimized. A 5 field IMRT plan was created and included in the study. Items of comparison included: target coverage homogeneity, conformity index, dose to normal structures (including skin), integral dose, estimated treatment time, and plan robustness. Mean integral dose to non-target tissues was reduced by an average of 20% when comparing small to large spot, and by 60 and 50% respectively, when compared to IMRT. Volume of brain, hypothalamus, temporal lobes, pituitary, and hippocampus was noticeably reduced with small spot in comparison to large spot. Differences between volumes of OAR receiving 20-90% of prescription dose varied with proximity of the structure to the PTV and were highly dependent on beam angles. Noticeable trends did exist for some critical structures including the hippocampus. For example, the mean dose to the left hippocampus as a percent of prescription for the small spot with 1,2 or 3 fields planned with MFO or SFO ranged between 36.3 and 36.8% compared to 43.1 and 46.3% for large-spot. For IMRT, the doses to the left hippocampus averaged 53%. Our preliminary data suggest scanning proton plans allow better target coverage and normal tissue sparring than IMRT for posterior fossa targets. Regarding the proton inter-comparison, only spot size has shown a trend, with smaller being better. The number of fields and SFO vs. MFO did not show any meaningful difference. The current data is based on a subset of plans and plan robustness is yet to be analyzed.