A Monte Carlo study on the collimation of pencil beam scanning proton therapy beams.

PURPOSE:The lateral edge of a proton therapy beam is commonly used to achieve conformality to the treatment volume where critical structures reside close to the target. However, when treating shallow depths, the lateral edge of a pencil beam scanning (PBS) system may be broader than that of a double scattered (DS) system. Use of a range-shifter to degrade the beam and allow treatment of very shallow depths further blurs the lateral edge. The authors investigate the potential use of a collimator with a PBS system for delivery of 3D uniform dose-volumes to a water-tank phantom, identifying the key factors controlling the width of the lateral edge. METHODS:The geant4 application for tomographic emission (gate) Monte Carlo (MC) environment was used, following validation against previously published data. Key parameters for PBS beams were investigated to assess their impact on the lateral edge of both monoenergetic beams and uniform dose-volumes. These parameters included nozzle-to-surface distance (NSD), vacuum window-to-surface distance (VSD), use of a range-shifter, and spot optimization parameters. RESULTS:The lateral edge of an uncollimated PBS beam is particularly sensitive to VSD and NSD. While use of a range-shifter blurs the lateral edge, collimation allows the edge to be sharpened to between 2 and 4 mm depending on the depth of the target. Optimization of the spot weightings alone can provide a penumbral width close to that of a single spot, but also leads to poorer uniformity near the edge of the target volume. CONCLUSIONS:Collimation of PBS beams should be considered for superficial targets particularly for beams delivered through a range-shifter, since the resultant sharpening of the lateral edge will allow improved sparing of adjacent normal tissues. Further work is needed to develop collimators which are integrated into both nozzle designs and planning system optimization algorithms.