ESTRO 36 Abstract Book

S306 ESTRO 36 _______________________________________________________________________________________________

Respiratory motion models compliment real-time imaging in two ways: 1) motion prediction models overcome the inherent latency in tracked delivery associated with the real-time feedback loop, and 2) motion estimation models alleviate the need for full 3D coverage at each time-point thus allowing dose deposition mapping in both the tumor and surrounding organs at risk. The efficacy of motion modelling relies on the available information (input) and the redundancy in the data (dimensionality reduction). Optimal results are therefore only achieved when both the acquisition strategy and motion modeling are matched. In this talk I will go through the basic requirements of an online MR-guided workflow and discuss the use of respiratory motion models for 1) tumor tracking, and 2) dose accumulation mapping. An outlook in then given, in which the foreseen advances in imaging speed are discussed and the role of motion models in the future. SP-0576 Tracking: present status and what to expect in the near future T. Ravkilde 1 1 Aarhus University Hospital, Medical Physics- Department of Oncology, Aarhus, Denmark In recent years there has been a growing trend towards hypofractionation in radiation therapy. The recent surge of interest in stereotactic body radiation therapy is yet another stone on this path. With these treatment schemes large radiation doses are delivered in few fractions, making the correct delivery of every treatment fraction critical. Unfortunately, many sites in the thorax, abdomen and pelvis are subject to motion, which can significantly deteriorate the highly conformal dose distributions typical of today’s standard of care. One way to restore dose conformity is by adjusting the radiation delivery to the moving anatomy on-the-fly; so-called tumour tracking. Tumour tracking was originally proposed more than 15 years ago and has been heavily investigated since, yet is very sparsely available in clinics worldwide. While dedicated machines are now commercially available, tracking on the ubiquitous standard C-arm linear accelerators is still lacking. Despite a large body of research publications and many convincing results in phantom experiments at multiple institutions tumour tracking has proven difficult to push into the clinic, with very few clinical trials existing. Unfortunately, tumour tracking inherently presents a problem for existing pre-treatment plan-specific QA regimes as parameters of the treatment machine is continuously adjusted during beam delivery of each treatment fraction and thus cannot be known beforehand. This has shown to be a concern for many physicians and physicists with regards to the clinical introduction of tumour tracking. This presentation will present an overview of machinery and techniques and discuss pros, cons and opportunities. It will make a brief review of the history of research and development and attempt to sum up the present status of tumour tracking while finally sharing a few thoughts on the direction it could be headed in the near future. SP-0577 Developments in head and neck toxicity data, models, and treatment optimisation A. Van Der Schaaf 1 1 UMCG University Medical Center Groningen, Radiation Oncology, Groningen, The Netherlands Radiotherapy in the head and neck region plays a pivotal role in the treatment of cancer but it is also associated with a large spectrum of toxicities. In past years many studies have concentrated on xerostomia and dysphagia as Symposium: Modelling and treatment customisation

the toxicities with the most detrimental effect on quality of life, but the list of complications that are studied is expanding. Prospective data registration programs are increasingly used to collect large standardized data sets of good quality. Modelling techniques become more advanced by adopting methods from the field of (bio)statistics and machine learning, enabling data driven exploration of complicated dose response relationships and a large variety of other predictive factors, including emerging factors like genetics and image features. Also, biological experiments are narrowing down on the potential mechanisms, aiming to find anatomical targets, e.g., a stem cell rich compartment in the parotid glands. All these findings can be used to (automatically) optimise treatment planning, or to select patients that benefit most from advanced treatment modalities, such as protons. For this to be effective, however, it is required that the observed findings describe general and causal relationships, which is checked with validation studies and rapid learning methodologies. This talk will include characteristic examples from literature and details from recent developments in the Netherlands and at our department, illustrating current processes from data collection to toxicity reduction. SP-0578 New NTCP data in the thoracic region: probing 'dark toxicity' J. Deasy 1 , M. Thor 1 , J. Oh 1 , A. Rimner 1 1 Memorial Sloan-Kettering Cancer Center, Medical Physics, New York- NY, USA Following the shocking results of RTOG 0617, it has become clear that high dose radiotherapy to the thorax can be lethal within a few months to a few years following treatment. This previously unappreciated syndrome is greater than any expected treatment benefit due to dose escalation, although it goes the wrong way. Given this new perspective, we can freshly review old and emerging data related to irradiation of central thorax structures. Although heart irradiation appears to be the likely cause of this 'dark toxicity', there are many remaining questions, such as the sensitive anatomic sub-structures. In our search for safe 'tolerance' thresholds, we will review all the data currently available to better understand this unexpected lethal toxicity in lung cancer. We will also highlight new methods of interrogating imaging data to identify local damage. In addition, we will also review recent data and risk modeling results for severe esophagitis and pneumonitis. SP-0579 New NTCP data in the pelvic area C. Fiorino 1 1 San Raffaele Scientific Institute, Medical Physics, Milano, Italy The last years were very fruitful in improving our knowledge regarding the prediction of toxicities after pelvic radiation therapy (RT). Prostate cancer (PCa) patients (pts) mostly contributed to the picture; on the other hand, results came also from other fields such as gynaecological, rectal and bladder cancer RT. Compared to the pivotal reviews of the Quantec group, published in 2010, our knowledge dramatically increased for most organs. Concerning rectum, the prevalently serial behaviour when considering moderate/severe bleeding was confirmed. In addition, several clinical parameters were found to significantly modulate the risk, primarily previous abdominal/pelvic surgery and cardio-vascular disease: the existence of a relationship between acute symptoms and late bleeding was also corroborated. Other end-points were investigated such as faecal incontinence, loose stools, urgency and pain: overall, robust models were reported for faecal incontinence showing a prevalently parallel behaviour of the rectum and/or of the anal canal and a relevant impact of clinical factors such