I007] Treatment planning in proton therapy

Abstract

Purpose To present the most important physics aspects of proton therapy (PT) treatment planning for scanning beams and compare them with photon-based radiotherapy. Methods The most important differences of protontherapy treatment planning with pencil beam scanning (PBS) with respect to what is considered now the norm in radiotherapy with photons (XRT) are the following: 1) The dosimetric inaccuracies in the map of proton stopping power (PSP) estimated via CT imaging are larger than similar error in estimating electron density for photons. This is the reason behind the development of more refined methods (e.g. dual energy CT or proton CT) and the current impossibility of MR-based protontherapy; 2) The pristine beam is in principle simpler to model than in photons, as no beam modifiers are used to shape the beam, with one significant exception: the preabsorber (aka “range shifter”); 3) Up until recently, dose calculation was performed with pencil beam-based algorithms. Now a transition is happening towards Monte Carlo-based algorithms, however for reasons that are different than in XRT; 4) In XRT geometrical uncertainties are handled during planning with ad hoc volume such as the planning target volume (PTV). In PT this approach is hardly satisfactory, albeit still common, and most advanced planning solutions are in fact based on PTV-less so called “robust optimization”, where the effect of range and setup uncertainties is essentially part of the cost function. This leaves open the issue of dose reporting, which is now approached in different ways at different PT centers; 5) PT treatment planning is performed assuming a constant 1.1 relative biological effectiveness (RBE) of protons with respect to XRT. This is an approximation, and the need of variable RBE in proton planning is one of the “hot topic” at the moment; 6) Adaptive therapy, and as a consequence adaptive treatment planning, is an area of development that has to be explored further. PT may be a field where the concept of “online adaptive” may have a clinical impact, but the times are still premature for a large scale implementation of these techniques.