ESTRO 37 Abstract book

S345

ESTRO 37

high risk of selection bias. Moreover, the extreme heterogeneity in baseline characteristics of the patients and the different measures of the outcome further complicate the analysis of the literature findings. At the same time, severe toxicity assessment is very difficult, especially for radiotherapy since the differentiation between tumor recurrence and radionecrosis after Re-RT may be very difficult, but also because most authors did not report the grade of toxicity. So, given these uncertainties, the key issue may be an appropriate selection of the patients that should be based on the prognostic factors that were proven to be important in literature: both patient-related factors (good performance status, age and RPA class) and recurrent disease-related factors (long time from the first course treatment to progression, possibility of gross total removal for surgery and target volume for radiotherapy) may help in selecting patients. Recently, some prognostic score indices were developed both for reirradiation and repeat surgery, in order to define patients who may have a clearer benefit. Those indices are based on a combination of main prognostic factors (histotype, age, site of the disease, time from first course treatment, ependymal involvement) that may be used to predict patient survival . For patients with good prognostic factors, proper pre- retreatment risk estimate is fundamental to choose salvage local treatment: individual treatment decisions should not include only factors influencing the outcome, but also factors that may impact the potential morbidity of the treatment. On the one hand, the incidence of severe toxicity due to second surgery depends on recurrence site and, obviously, on comorbidities that may increase the anaesthesiologic risk. On the other hand, the incidence and severity of radiotoxicity can be increased by chemotherapy, age, diabetes. Additionally, disease- related factors may influence the risk of toxicity: lesion size, proximity to eloquent area or to organs at risk, overlapping with the target of the initial treatment are factors to take into account whenever a second irradiation is weighted against the surgical alternatives. An important advantage to consider surgery is the acquisition of tumor tissue at relapse. This may be valuable for differential diagnosis with radiation necrosis, confirmation of initial histology, definition of molecular markers of recurrent disease. By contrast, reirradiation may be a less invasive treatment. In this perspective, any attempt to reduce the expected toxicity of the treatment is extremely relevant. In this speech, existing data on the efficacy and toxicity of various Re-RT options including radiosurgery, hypofractionated stereotactic treatment, and conventionally fractionated RT will be reviewed. The literature will be thoroughly reviewed and the correlation between dose and response and between dose and toxicity will be herein explored in order to define the optimal fractionation and prescription dose in terms of efficacy and tolerability. Moreover, some practical considerations for radiation treatment planning will be provided in order to improve the therapeutic ratio of the salvage treatment.

Abstract text The increasing usage of MRI in the simulation process stems from its superior soft tissue contrast. This allows better delineation of tumor volume and OARs and thus contributes to overall improvement of treatment quality. As MRI does not provide direct information of the tissue electron density and bony anatomy for position verification, patients have to undergo two imaging exams in clinical practice: a CT and an MRI scan. Besides extra patient burden and medical costs, this introduces inevitable geometrical errors related to interscan differences and image fusion. This has been the rationale behind the development of MR-only simulation where all information needed for delineation, position verification and electron density is derived from MR images. A first prerequisite is of course the geometric fidelity of MR images. In the last two decades improvements in magnet design, gradient coils and image corrections have led to the possibility to perform geometrically accurate MR imaging for radiotherapy. Of course, this requires a quality assurance program to monitor geometrical fidelity and needs to be supplemented by proper sequence design (e.g. a high readout bandwidth). A second adaptation crucial for MR-only simulation is the possibility to scan the patient in treatment position. Both aspects have been acknowledged in the last years by MR vendors and they introduced wide bore MR systems and launched special MR-RT solutions such as flat table tops, QA phantoms+procedures, positioning lasers and special coil options. Much of the research towards MR-only has focused on different methodologies to generate socalled synthetic CT from MR images. A crucial aspect is the visualization of the bony anatomy with MRI. Due to the low proton spin density and very short T2*, cortical bone appears as a signal void on MRI. Simple segmentation of these signal voids to identify cortical bone on MR images will not be successful as inner air also appears a signal void. Dedicated MR sequences such as ultra short echo time (UTE) sequences to obtain signal from cortical bone have been applied with mixed success. Other methods have approached the problem as a quantitative MRI problem to convert quantitative MR images into synthetic CTs using signal-to-houndfields unit conversion models. Recently, various commercial solutions have become available that use rather standard MR sequences combined with clever image processing solutions to generate synthetic CT images. A very new technique, deep learning based synthetic CT generation is even more flexible in terms of requirements for image contrast. Deep learning methods like convolutional neural networks are able to classify bony or air voxels by a learning approach including local, contextual image information. With these new deep learning methods standard MR sequences that are primarily intended for delineation purposes can be utilized for synthetic CT generation. More general, deep learning methods offer great potential for other steps in the MR-only simulation process, e.g. for automatic contouring of organs-at-risk on MR images. A somewhat overlooked aspect of MR-only simulation is the need to generate reference images for position verification from MR images. The synthetic CT will suffice as a reference image for tumor sites where position verification is based on registration of kV or MV in-room images to a reference image. However, the situation is more challenging in the case of position verification for VMAT or IMRT prostate irradiation where implanted gold fiducials are used. Fiducials are easily localized on CT images due to their distinct local streaking artifacts. However, on MR images the appearance is less distinct and they manifest themselves as local signal voids in magnitude MR images. Correct manual classification of these signal voids as fiducials is feasible but sometimes complicated by the presence of calcifications that can

Symposium: Advances in MRI simulation (MRI-only treatment planning)

SP-0659 MRI techniques for MR-only simulation N. Van den Berg 1 , M. Maspero 1 , A. Dinkla 1 , M. Savenije 1 , G. Meijer 1 , P. Seevinck 2 , J. Lagendijk 1 , B. Raaymakers 1 1 UMC Utrecht, Department of Radiation Oncology, Utrecht, The Netherlands 2 UMC Utrecht, Centre for Image Sciences- Department of Radiology, Utrecht, The Netherlands