Dosimetric evaluation of dose shaping by adaptive aperture and its impact on plan quality

Mevion's single-room HYPERSCAN proton therapy system employs a proton multileaf collimator called the adaptive aperture (AA), which collimates individual spots in the proton delivery as determined by the Treatment Planning System (TPS). The purpose of this study is to assess the dosimetric benefits of the AA, specifically in the dynamic aperture (DA) mode, and evaluate its impact on proton treatment plan quality as compared to a traditional pencil beam scanning (PBS) system (Varian ProBeam). The spot dose distributions with dynamic collimation (DA), a unique AA shape for each energy layer, and with static collimation (SA), a single AA collimation shape shared by all energy layers per field, were calculated and compared with the spot dose distribution of the Varian ProBeam proton therapy system. The lateral and distal dose falloff gradients and their dependence on air gap were evaluated quantitatively. Treatment plans for ten arbitrarily selected intracranial target image sets were created, and the HYPERSCAN and ProBeam beam models were compared. The spot sizes of the HYPERSCAN system are significantly larger than ProBeam system, especially at low energy. With the help of DA, the lateral dose penumbra of the HYPERSCAN is dramatically improved at lower energy and comparable at higher to ProBeam PBS beams. While the ProBeam spot size does not change with the air gap, beam penumbra of the HYPERSCAN with DA increases with the air gap. The distal dose falloff gradient for the HYPERSCAN with or without DA remains consistently around 4.8 mm through all energies due to the beamline design, not substantially varying with energy or air gap. Treatment plans of ten randomly selected intracranial cases demonstrated favorable OAR sparing but unfavorable dose uniformity for the HYPERSCAN with DA compared to ProBeam. Dose shaping by adaptive aperture substantially improves the lateral penumbra without a significant change in the distal dose gradient. The dose gradients of the multiple beam DA plans with layer-by-layer blocking are improved compared with SA plans and are close to the ProBeam plans for the ten randomly selected brain cases. With layer-by-layer DA blocking, the HYPERSCAN plans have similar plan conformality indices as the ProBeam plans, but the overall plan quality indices are lower than ProBeam plans, largely due to the lower dose homogeneity. In some cases, DA blocking was found to be superior in sparing OAR surrounding the target.