Abstract Book

S302

ESTRO 37

scaling up radiotherapy is huge. It is estimated that if radiotherapy is fully scaled up by 2030, close to 27 million life years would be saved. This could be achieved over the next 20 years with an investment of $97 Billion. To realize this potential, not only radiotherapy but all other elements of cancer management must also be scaled up. Others have documented huge gaps in access to diagnostics, to surgery, and to chemotherapy. The Disease Control Priorities 3 rd edition project provided a comprehensive evaluation of the status of healthcare resources and recommended stepwise implementation of a number of key initiatives with the largest benefit for the investment made. With a large number of health conditions requiring attention, cancer control falls behind other priorities. However, cancer incidence and mortality are increasing, especially in low and middle income countries. Scaling up cancer control requires not only a substantial investment in healthcare infrastructure but also in human resources. This will take time. Therefore there is an urgent need to start the process, to support innovation in all aspects of health systems strengthening. A diagonal approach is recommended. Building population wide initiatives such as prevention, vaccination, and primary care should be matched by building vertical comprehensive cancer centres with diagnostic and treatment technologies matched by skilled human resources with a mandate to educate additional staff and enable new facilities to open. Further research effort is needed to define most cost effective measures, to create and use new technologies from new energy sources (e.g., solar energy), new information and communication technologies, and to overcome distance and budget barriers. In addition, ongoing advocacy for comprehensive approach to cancer globally is urgently needed. The advocacy at local, national, and global level raises awareness, stimulates action and mobilizes resources. Partnerships between developed and developing centres may provide substantial support for new and emerging cancer centres or programs. Numerous efforts by disparate networks of experts, volunteers, and advocates should be coordinated to create synergies and accelerate the progress. Professional societies, population based cancer programs, government bodies and UN agencies all must be mobilized to coordinate the action and create the global public goods in support of expanding access to radiotherapy and other cancer control needs. SP-0576 Comparison and limitations of DVH-based NTCP models derived from 3D-CRT and IMRT data M. Söhn 1 , M. Alber 2 1 Klinik und poliklinik für Strahlentherapie und Radioonkologie, Academic Physics, München, Germany 2 Heidelberg University Hospital, Department of Radiation Oncology, Heidelberg, Germany Abstract text The fundamental premise of common NTCP models is, that the same DVHs yield the same complication probability. Apart from the often voiced objection that all spatial information is removed when compiling a DVH from a dose distribution, there are other issues. Most prominently, a planned dose distribution differs from a delivered dose distribution, so that equivalent planning DVHs may result in different delivered DVHs, depending on how much the dose distribution outside the organ of interest differs. This is particularly true for comparing IMRT and 3D-CRT data. However, the pairing of 3D-CRT and IMRT data is Symposium: Normal tissue dose-response modelling across radiation modalities

problematic for other reasons, too: the data rarely stems from a randomized trial, so there could be a drift in patient population and treatment administration, and it is impossible to blind for treatment technique. In consequence, the desired and theoretically clean proof of superiority of IMRT over 3D-CRT is likely biased towards the more modern technique. As an example, an analysis of 457 prostate patients treated with 3D-CRT and 658 with IMRT, respectively, is presented. Both patient cohorts were treated according to the same adaptive, image-guided protocol at the same institution. Patient groups were analyzed for chronic GI toxicities grade >=2 (CTCAE v3). 3D-CRT patients were matched to IMRT patients based on various patient characteristics, using a propensity score-based algorithm. Cohort-DVH-distributions in both groups were statistically different, but by a relatively small amount in the high- and intermediate dose regime. In contrast to the moderate dosimetric improvement of IMRT as compared to the 3D-CRT group, there was huge difference in raw grade >=2 chronic GI toxicities of 7.8% vs. 28.7%, respectively. No contemporary NTCP model was capable of uniting these two patient groups. Concluding, the assumption that NTCP models from one era of radiotherapy can be extrapolated into another era may turn out naive due to a large number of factors, not including tissue dose response, that may play a role for the occurence of complications. SP-0577 Clinical evidence of spatially variable proton biological effectiveness in pediatric brain tumour patients D. Grosshans 1 , C. Peeler 2 , D. Mirkovic 2 , R. Mohan 2 1 MD Anderson Cancer Center, Radiation Oncology, Houston, USA 2 MD Anderson Cancer Center, Radiation Physics, Houston, USA Abstract text As the number of patients treated with proton therapy increases, there is growing debate regarding the relative biological effectiveness (RBE) of proton beams. Unlike photons, as a proton beam passes through tissue, the linear energy transfer (LET) increases. New evidence indicates that the capacity for proton beams to cause biological damage is substantially higher near the distal, high LET region. This is true both for tumors as well as normal tissues. Despite such data, in clinical practice the RBE of protons is assumed to have a constant value of 1.1. In support of the status quo, it is commonly cited that there are no clinical data to suggest that the proton RBE is not 1.1. We have previously documented increased rates of post-radiation MR imaging changes in pediatric ependymoma patients treated with proton therapy compared to those treated with photons. Such changes are indicative of early radiation injury and serve as an imaging biomarker of differential damage between radiation types. In search of clinical evidence supporting a variable RBE, we employed Monte Carlo techniques to compute both dose and LET distributions and found significant correlations between LET, dose and regions of imaging change. Our own group and others are now expanding research in this area conducting laboratory and clinical studies to better characterize the biologic effectiveness of particle beams. This is significant in that if the biologic effectiveness is known, with newer delivery modalities, in particular spot scanning proton therapy, intensity modulated proton therapy plans may be developed to preferentially divert high LET protons away from normal tissues into the target volume. This should reduce normal tissue damage while simultaneously increasing biologically effective doses to targets.