ESTRO 2020 Abstract book

S6 ESTRO 2020

treatment plans, but, at the same time, it can be used to go toward a patient-specific plan selection, allowing to easily run (in the background) multiple plans "for-free". Automated planning also helps the medical physicist in performing robust treatment planning studies based on a high number of consistent plans, e.g. to compare various treatment approaches (protons vs. photons, coplanar vs. non-coplanar, etc.). Automation of the treatment planning process however is time consuming in the initial phase of workflow development, and it has to be checked during time (validation proposal not defined yet). These aspects need a change of mind in the working daily activity (i.e. it needs more time investment at the beginning and less later) and an extra expertise is needed compared to conventional planning. Moreover automated planning can lead to a reduced planning expertise, among planners, which can be a problem in the long term. A review of current autoplanning systems will be provided focusing on strengths and limitations. SP-0029 Automation of QC/QA process N. Jornet 1 1 Hospital de la Santa Creu i Sant Pau, Medical Physics, Barcelona, Spain Abstract text The complexity of the radiotherapy process continues to increase which requires a high level of safety and constant quality monitoring and improvement. During the last decades, the number of quality controls performed both patient and equipment specific has increased enormously. We generate a large amount of data that is too often poorly analysed to redesign Quality Controls or redefine their frequencies. Therefore, as a consequence, performing and analysing routine quality controls takes a lot of working time of Medical Physics staff. Most of the routine QC could be partially or fully automated. Automation should encompass all patient quality controls from treatment planning to delivery as well as the periodic controls on treatment units, imaging systems and treatment planning systems. The benefits of automating QC are not only related to time saving of both personnel and treatment machines but also on the consistency and robustness of the Quality controls. An automated Quality Assurance platform could also help in data sharing (alerts in case data is out of the shared data for that machine), and predictive modelling of equipment break downs or suboptimal treatment plans. Treatment plan evaluation could also be partly automated by the use of automated checklists and automated comparison with standards, again this will improve both the time needed for plan preparation, but also will have a high impact in reduction of errors. In order to automate most of the QC tests there is a need of integrating information provided by different systems. Therefore, therefore vendors ought to provide stable APIs, and full access to data so that even with changes of versions of the systems automation still works. This is still not the case for major equipment vendors and compromises the potential of automation. The implementation of automation involves a considerable initial investment. Automation may also lead to a loss of flexibility, as any workflow modification may involve revisiting automation codes. Another potential disadvantage of automation is loss of knowledge of the sensitivity and specificity of the different QC tests which could lead to physicists over relaying on automated QA and not understanding the limitations of some tests. Most departments, due to the scripting opportunity given by major treatment planning systems vendors, have implemented local solutions for automation of their QA processes. At the same time some measurement equipment vendors are providing QA platforms that integrated with measuring equipment, treatment planning systems and linac allow for the automation of QC both in

SP-0026 Against the motion: Upfront radiotherapy is not mandatory in patients with brain metastases treatable with immunotherapy or targeted therapies? A. Grosu 1 1 Universitatsklinik Freiburg, Dept. for Radiation Oncology, Freiburg, Germany Abstract text Advances in targeted and immunotherapies have been considerable in the last decade and have most certainly led to a significant improvement in survival for selected tumor entities. Especially in the case of malignant melanoma and non-small cell lung cancer, the introduction of the immune checkpoint blockade and of third generation tyrosine kinase inhibitors has revolutionized therapy and prognosis. The permeability of the blood-brain barrier for these substances could be demonstrated and an impact on brain metastases was therefore noticed. Understandably, approaches to exclusively employ systemic therapies for the treatment of brain metastases were undertaken. However, published data are still scarce, on average on no more than 100 patients, often previously irradiated, and with few prospective studies. In general, the reported overall response rate varies between 18 and 66%, which is insufficient when compared with the local tumor control offered by stereotactic radiation therapy. Moreover, data for solitary immunotherapy are only available for malignant melanoma in patients with asymptomatic brain metastases with no need for steroid use or intervention. A potential benefit could indeed be seen in the combination of immunotherapy and stereotactic radiation therapy or whole brain irradiation, leading to highest response and overall survival rates. Furthermore, immunotherapy as well as selected targeted therapies may temporarily allow delays in radiation therapy in particular cases, such as patients with asymptomatic brain metastases with high systemic tumor burden and urgent need of systemic treatment. In conclusion, radiation therapy remains mandatory for patients with brain metastases in spite of new systemic treatment modalities. A limited number of metastases should be treated by radiosurgery or stereotactic fractionated radiation therapy and whole brain radiotherapy should only be used if metastases are disseminated. The combination with immunotherapy is potentially beneficial in melanoma patients with asymptomatic brain metastases and high systemic tumor burden. In specific patients, immunotherapy or targeted therapy initiation before local cerebral treatment can be taken into consideration.

Symposium: Where automation can or cannot help the medical physicist

SP-0027 Automation in software and hardware commissioning S. Mutic Washington University School of Medicine, USA

Abstract not received

SP-0028 Automation of treatment planning process L. Rossi 1 1 Erasmus MC - Cancer Institute, Department of Radiation Oncology, Rotterdam, The Netherlands Abstract text Automation of treatment planning can substantially improve plan quality or match the quality of manual plans. This is possible together with a reduced workload, generating time for manual planning of more complex cases, or out of protocol treatments. Automation of treatment planning can increase standardization in