ESTRO 37 Abstract book

ESTRO 37

S572

Conclusion The results of HDR Interstitial brachytherapy have shown an acceptable local control and overall survival rates along with tolerable toxicities and morbidity in recurrent H&N cancers. Based on these encouraging results, prospective clinical trials are warranted using HDR Interstitial Brachytherapy in recurrent H&N cancers to decrease late toxicity.

Results The recordings showed that the planner generally works iteratively on lowering OAR dose and increasing target dose. This process visualizes as a zig-zag pattern in the bi-objective problem formulation where both objectives are eventually improved (Fig.1a). In the bi-objective problem formulation, all clinical plans were inferior to the Pareto front plans obtained with the MOEA (Fig.1b). In the observer study, in all cases the observers chose a plan from the Pareto front as the best plan, and appreciated to directly see the trade-offs for that particular patient (Fig.1b). The clinical plan was in 4/20 cases considered to be the worst plan. The observers agreed on the best plan for 2 patients, which was the plan with the highest possible coverage while still satisfying all OAR criteria.

Poster: Brachytherapy: Physics

PO-1020 Better plans and easy plan selection via bi- objective optimization for HDR prostate brachytherapy S.C. Maree 1 , E.S. Kooreman 1 , N.H. Luong 2 , N. Van Wieringen 1 , A. Bel 1 , E.C.M. Rodenburg 1 , K.A. Hinnen 1 , G.H. Westerveld 1 , B.R. Pieters 1 , P.A.N. Bosman 2 , T. Alderliesten 1 1 Academic Medical Center, Radiation oncology, Amsterdam, The Netherlands 2 Centrum Wiskunde & Informatica, Amsterdam, The Netherlands Purpose or Objective In treatment planning for HDR prostate brachytherapy (BT), the trade-off between dose to the target volumes and dose to the organs at risk (OARs) is optimized. To achieve this with a traditional planning method (e.g., IPSA, HIPO), its parameters must be tuned iteratively. In our clinic, the plan is manually fine-tuned further using graphical optimization. These approaches are patient- dependent, require experience, and may lead to sub- optimal plans under clinical time pressure. Resulting plans are evaluated by the dose-volume indices (DVIs) of the clinical protocol, but also by visual inspection of the dose distribution. In this work, we analysed clinical planning sessions to formulate a bi-objective optimization model from which multiple plans follow that exemplify the trade-offs in treatment planning, and validated its usage by an observer study. Material and Methods HDR BT plan optimization sessions of 5 prostate cancer patients were filmed and changes in DVIs over time were recorded. After studying these recordings, a bi-objective optimization model was formulated that directly considers the DVIs of the clinical protocol. A coverage objective (i.e., least coverage index ) was formed from the criteria on the target volumes, and a sparing objective (i.e., least sparing index ) from the criteria on the OARs (Table 1). The clinical protocol is satisfied when both have a value >0. A recent multi-objective evolutionary algorithm (MOEA) was tailored to generate a large set of plans, a Pareto front , representing different high-quality trade-offs between coverage and sparing. The planning process is thereby intuitively facilitated without the need of parameter tuning. Finally, the clinical plan and 5 plans selected from the Pareto front (Fig.1b) were blindly presented to 1 BT technician and 3 brachytherapists in a retrospective observer study. Each observer was asked to select the best and worst plan.

Conclusion Our novel bi-objective brachytherapy planning method is parameterless and directly optimizes treatment plans on the evaluation criteria of a clinical protocol. It results in a set of high-quality plans to intuitively select from. In all cases, such a plan was preferred over the clinical plan. PO-1021 HDR Brachytherapy dosimetry: clinical use of micro-silica bead TLD & Gafchromic EBT3 film A. Douralis 1,2,3 , S.M. Jafari 2,3 , W. Polak 3 , A.L. Palmer 2,3 1 National Physical Laboratory, Medical Radiation Physics, Teddington, United Kingdom 2 University of Surrey, Physics, Guildford, United Kingdom 3 Portsmouth Hospitals NHS Trust, Radiotherapy Physics, Portsmouth, United Kingdom Purpose or Objective To develop a novel high resolution experimental method for validating Monte Carlo-derived TG-43 brachytherapy source data, and model-based treatment planning systems. Experimental verification is recommended in the “Report of the High Energy Brachytherapy Source Dosimetry (HEBD) Working Group, 2012”, however, the steep dose gradients with a wide dynamic dose range and rapid change in dose rate has been a limitation for the ability of common detectors to obtain accurate measurements. In this study, we tested a micro silica