ESTRO 2023 - Abstract Book

S574

Monday 15 May 2023

ESTRO 2023

well as the online adaptive process itself. We will then evaluate the clinical situations for which online adaptive radiotherapy is particularly indicated. We will detail some examples of adaptive radiotherapy management on MR-Linac and CBCT-Linac. Finally, we will evaluate the future development of online ART in the coming years and its possible evolution in the therapeutic arsenal of the radiation oncologist.

SP-0685 Physicists' challenges in clinical routines R. Dal Bello 1 1 Universitätsspital Zürich, Klinik für Radio-Onkologie, Zürich, Switzerland

Abstract Text Adaptive radiotherapy has recently drawn benefit from the introduction into the market of multiple certified hybrid linacs, which installations and number of treated fractions is steadily increasing. The main aim of treatment plan adaption is to compensate for the anatomical changes. Currently clinically implemented systems based on CBCT or MRI for image guidance are designed to re-optimize and update a base plan taking into account the inter-fractional changes, which are assessed during pre-treatment imaging. Intra-fractional motion is currently mitigated through gating, while volume tracking and additional techniques are still under development and in the research phase. In particular, MR guided radiotherapy (MRgRT) exploits the superior soft tissue contrast and the absence of imaging dose, which allows continuous imaging during beam delivery for monitoring intra-faction motion. The integration of such sophisticated imaging modalities with the linacs and the on-table adaptive workflows pose new challenges for medical physicists. The former requires dedicated procedures especially for hybrid MR-linacs, where the presence of the magnetic field does not allow the employment of conventional quality assurance (QA) phantoms and detectors. MR-compatible devices have to be developed and integrated in the routine QA program. Moreover, the effect of the magnetic field on the dose distribution and in the response of the detectors has to be quantified and taken into account. Additional challenges include the introduction of a QA program for the MR itself, which has not been historically part of radiation therapy departments. Special care should be dedicated to the evaluation of the geometrical distortion, which is critical for the accurate delivery of treatment plans in MRgRT. Even more challenging are the procedures related to the on-table adaptive workflows. Optimizing and limiting the patient on-couch time is of utmost importance for an even wider employment of adaptive radiotherapy. Therefore, automation and artificial intelligence (AI) are already playing and will play an always larger role within the clinical workflows. Automatic contouring and automatic re-planning are solutions that should aid clinicians and physicist in shortening and standardizing the treatments. The quantitative assessment of the quality of automatically generated contours and plans should also be included within the competences of medical physicist in adaptive radiotherapy. Furthermore, traditional workflows including pre-treatment patient specific QA (PSQA) cannot be applied in the case of adaptive radiotherapy when the new plan is generated and delivered while the patient is on-couch. Class solutions and independent verification algorithms should be implemented in the clinical workflow. Finally, adaptive radiotherapy is not only supporting developments aiming to automate human-based tasks (e.g. contouring, planning) but also driving towards the introduction of fully AI-based solutions (e.g. synthetic CT). In such cases, novel QA solutions have to be developed and clinically implemented, replacing the non-applicable traditional end-to-end testing. In conclusion, adaptive radiotherapy is generating increasing clinical evidence and is a stimulating and challenging field for new solutions to be developed and implemented by medical physicists. Abstract Text Online adaptive radiotherapy treatment allows for adapting the treatment plan to the anatomical changes between the reference CT scan it was based on and the actual treatment session(s). The first online adaptive treatments were MR-Linac based and for limited number of patients, treatment sites and/or fractions. Conebeam CT based options are currently also commercially available and patient, treatment sites and/or fraction numbers are increasing. Online adaptive radiotherapy started out as a resource heavy option with long time slots and a multi-disciplinary team at the treatment machine. This presentation will be focused on RTT-only online adaptive radiotherapy, and gives an overview of (1) local online adaptive radiotherapy, (2) workflow and responsibilities, (3) training to build competence and confidence for RTT-only online adaptive radiotherapy and (4) overview of literature. SP-0687 What is the role of offline adaptive radiotherapy in the area of online adaptive radiotherapy? L. Hoffmann 1 , A.I.S. Holm 1 , D.S. Møller 1 1 Aarhus University Hospital, Department of Oncology, section for Medical Physics, Aarhus, Denmark Abstract Text Curatively intended radiotherapy is predominantly delivered as a fractionated treatment lasting one to several weeks. The daily treatment is based on the CT or MR scanning obtained prior to treatment. However, anatomical changes may occur during the treatment course. The changes can be monitored by the cone-beam CT (CBCT) images used for patient positioning. In case of suspicion of target dose deterioration or normal tissue overdosage, the treatment plan can be adapted for the remaining fractions to reflect these changes. Adaptive radiotherapy may be performed either online just before each treatment delivery or offline after treatment delivery. In many patients, offline adaptation is sufficient to secure precise dose delivery during the treatment course and offline adaptation is less resource-demanding than online adaptation. However, offline adaption requires extra resources in the clinical routine. The daily CBCT scans need to be checked for deviations. This may be done by visual inspection using e.g. geometric measures either online by the RTTs before treatment delivery or offline with a selected schedule. Additionally, the daily dose delivered can be calculated based on the CBCT images. This requires validation of the structure propagation from the CT scan to the CBCT scan and calibration of the CBCT SP-0686 RTT-only online adaptive radiotherapy R. de Jong 1 1 Amsterdam UMC location AMC, Radiation Oncology, Amsterdam, The Netherlands