gms | German Medical Science

Background: In summer 2008 the carbon ion radiotherapy pilot project - a joint effort of GSI/Darmstadt, DKFZ/Heidelberg, University Clinic/Heidelberg and FZ Rossendorf - ended after eleven years of successful patient treatments.

All treatment plans in the pilot project were created with the TRiP98 treatment planning system (TPS).

At this stage it is appropriate to summarize its current capabilities as well as its future aspects as a research prototype for use in the upcoming ion-beam radiotherapy facilities.

Material and methods: The complexity of interactions of ion beams heavier than protons with living matter makes it difficult to provide purely experimental base data sets for treatment planning. Hence one has to rely on sufficiently accurate – and fast – calculations to obtain base data such as depth dose distributions and particle spectra.

For all ions heavier than protons the inclusion of radiobiological effects, i.e. RBE, survival probabilities, biologically effective dose, into regular treatment planning is indispensable. Since ab-initio calculations are neither reliable nor computationally feasible for years to come, we use the versatile Local Effect model (LEM) for the planning of all irradiations.

One of the main future challenges is 4D treatment planning, i.e. irradiation under presence of motion of both, the scanned beam as well as the target, taking biological effects fully into account. These developments took place on a separate branch of the TRiP98 TPS software. They were tested extensively and verified experimentally. Currently they are incorporated into the TPSs main trunk, so that 3D and 4D planning will be available seamlessly within the same product.

Results: We have incorporated semi-empirical model calculations for various ions such as 12C and 3He. For the sake of completeness and comparability of treatment plans, protons will be added as a supported modality.

Suitable approximations allow reasonably fast calculations of RBE-weighted dose even in complex treatment configurations, without sacrificing accuracy. TRiP98 is designed to seamlessly incorporate radiobiological base data as predicted by LEM. Recent improvements for low-LET radiation, such as protons and higher energy carbon ions will allow to generate and compare patient plans with different ion-beam modalities under realistic conditions.

Simultaneous optimization of multiple fields under constraints leads to a considerably enhanced target conformation and sparing of organs at risk (OAR). Consequently, almost all treatment plans were generated with these advanced methods in the last two years of the pilot project.

Conclusion: TRiP98 and following releases of this TPS are well suited to serve as research prototypes for scanned ion-beam radiotherapy planning.