Kombination biologischer Optimierungsverfahren für die Ionentherapie
Final Report Abstract
Treatment planning for radiation therapy of cancer patients employs computerized optimization algorithms to find the best treatment plan for each individual patient. The aim of this project was to develop a unified framework for biologically oriented treatment planning including the relative biological effectiveness (RBE) of particle beams as well as biological planning goals in terms of tumour control probability (TCP) and normal tissue complication probabilities (NTCP). We carefully analysed existing biological models to estimate the RBE, the oxygen effect (and the corresponding oxygen enhancement ratio / OER), TCP and NTCP and adopted them to carbon ion beam parameters where necessary. Combining radiosensitivity parameter predictions of different biological models with Monte Carlo generated carbon ion beam fragmentation spectra enabled us to implement fast treatment planning on real patient CTs including biological effects of carbon ion beams. To feature biologically driven treatment plan optimization for ion therapy, novel concepts were designed to integrate TCP, NTCP and the equivalent uniform dose (EUD) as a surrogate for TCP/NTCP into the optimization loop. Two concepts were proposed to model EUD in the framework of ion therapy: an EUD based on RBE-weighted dose and an equivalent uniform effect (EUE). Objective functions based on either of these approaches were implemented into a treatment planning platform to enable biological optimization of treatment plans. The implementation was carried out for two prominent cost function metrics (logistic and squared differences). The optimization routine was applied to the calculation of multiple treatment plans using various sets of optimization parameters and objective function contributions. Advantages and drawbacks of EUD- over voxel-based optimization techniques were analysed and the impact of potential uncertainties in used biological parameters on the resulting dose distributions was evaluated as well. In a subsequent step, the tool for EUD optimization was further expanded to enable a combination of EUD, TCP and NTCP optimization objectives making the treatment planning platform fully capable of an entire and very flexible biological optimization of carbon ion therapy treatment plans. This enables full biological treatment planning for ions and better steering towards clinically relevant goals in the planning process. Since large uncertainties are the most important drawback of the employed biological models, a second part of the project focused on the analysis of these biological uncertainties in the treatment planning process and on the development of a risk assessment tool to visualize their potential clinical impact. To access the impact of different uncertainties we performed a variance-based sensitivity analysis. Here we assigned uncertainties to all relevant parameters and were able to evaluate the relative influence of these parameters. We performed evaluations for various situations (including RBE, TCP and dose-volume indicators) in real patient geometries. User-friendly visualisations of these analyses were developed. This is important as already small changes in input variations can lead to large variations of the examined values. Overall, we were able to combine conventional biological optimization for ions (accounting for RBE effects) with biological optimization techniques commonly used for photon beam (TCP/NTCP, EUD) into a unified framework in a research treatment planning system, which enables full biological treatment planning for ion beams.
Publications
-
(2011) "Theoretical analysis of the dose dependence of the oxygen enhancement ratio and its relevance for clinical applications" Radiat. Oncol. 6:171
Wenzl T., Wilkens J.J.
-
(2014) "Variance-based sensitivity analysis of biological uncertainties in carbon ion therapy" Physica Medica 30(5):583-7
Kamp F., Brüningk S., Cabal G., Mairani A., Parodi K., Wilkens J.J.
-
EUD-based biological optimization for carbon ion therapy. Medical Physics, Vol 42 Issue 11, November 2015, Pages 6248-6257
Brüningk S., Kamp F., Wilkens J.J.
-
PhD Thesis “Uncertainties in biological dose response models
and their integration in treatment planning of carbon ion therapy”, TU München, 2015. III, 127 S.
Kamp F.
-
“Fast Biological Modeling for Voxel-based Heavy Ion Treatment Planning Using the Mechanistic Repair-Misrepair-Fixation Model and Nuclear Fragment Spectra”. International Journal of Radiation Oncology*Biology*Physics, Volume 93, Issue 3, 1 November 2015, Pages 557-568
Kamp F., Cabal G., Mairani A., Parodi K., Wilkens J.J., Carlson D.J.
