ESTRO 38 Abstract book

S357 ESTRO 38

3. Guckenberger M, et al. Radiother Oncol 2013;109:13-20 4. Shuryek I, et al. Radiother Oncol 2015;115:327-334 5. Metha N, et al. Pract Rad Onc 2012;2:288-95 6. Guckenberger M, et al. Radiother Oncol 2016;118:485- 91 7. Ohri N, et al. Int J Rad Onc Biol Phys 2012;84:e379-e384 8. Fischer JJ, Moulder JE. Radiology 1975;117(1):179-84. doi: 10.1148/117.1.179 9. Stavreva NA, et al. Med Phys 2005;32(3):720-5 10. Ruggieri R. Phys Med Biol 2004;49:4811-23 11. Ruggieri R, et al. Phys Med Biol 2013;58(13):4611-20. 12. Alite F, et al. Radiother Oncol 2016;121:9-14 SP-0681 Online adaptive planning in pancreatic cancer O. Bohoudi 1 1 VUMC, Radiotherapy, Amsterdam, The Netherlands Abstract text In recent years there has been growing interest in using stereotactic body radiotherapy for treatment of pancreatic cancer. Upper abdominal tumours such as pancreatic cancer are particularly suitable for performing adaptive treatment planning, because of the dynamic nature and proximity of several critical normal organs such as the duodenum, stomach and bowel. Recently, MR-guided RT systems have been clinically implemented, offering exceptional options for stereotactic and adaptive radiotherapy which necessitates fast and robust online planning. The aim of this talk is to describe and discuss an adaptive online strategy which can be performed within minutes, and only requires limited (re-)contouring by the physician. SP-0682 Future developments in adaptive strategies U.Oelfke 1 1 The Institute of Cancer Research, Joint Department of Physics, Belmont,United Kingdom 1 The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, Radiotherapy and Imaging, London, United Kingdom Abstract text Radiation technologies permitting high quality on-line imaging have demonstrated that the target for radiotherapy is dynamic. Many dosimetric studies have illustrated that accommodating for the individual's actual anatomical change offers opportunity for margin reduction with subsequent improved target coverage and normal tissue sparing. Clinical feasibility of this approach has also now been shown. There are many examples in medicine where pressures and demands for high-tech treatment have led to widespread implementation of innovation before, or even without, robust evidence has been generated. Arguably, well- designed clinical trials remain the optimal method to demonstrate outcome benefit for patients. At present, how best to utilise this adaptive strategy in order to maximise the clinical gains for our patients remains to be determined. Head to head comparative studies of adaptive techniques may not be ambitious enough to demonstrate true potential. Instead it may be necessary to leverage the improved therapeutic ratio to deliver radiotherapy in circumstances that would have been otherwise challenging. Symposium: Plan of the day - present status and future aims Abstract not received SP-0683 Clinical results of PotD strategies S. Hafeez 1

Automation, and computer aided image interpretation are emerging to address these issues. Conclusion SBRT has enriched the practice of radiation oncology. This has not been without challenges and technical and workflow issues highlight the need to engage the whole multidisciplinary team for successful SBRT implementation. SP-0680 Hints on optimal dose and fraction number from lung SBRT R. Ruggieri 1 1 Sacro Cuore Don Calabria Hospital, Radiation Oncology, Negrar, Italy Abstract text Initial lung SBRT treatments [1] suggested that the LQ- model for cell survival (CS) might be inadequate at as high doses per fraction ( d ) as 20 Gy. However, when compared with more complex LQL-models, the LQ-model produced equivalent goodness-of-fit for in vitro CS data even at high d -values, if a dose-range dependence from α and β parameters was recognised [2]. When different CS-models were next used for BED, and hence TCP, computations, equivalence in goodness-of-fit resulted also for local- control data of early-stage NSCLC [3]; thus supporting the LQ-model, given its least number of free parameters, as the most appropriate CS-model for fitting TCP from SBRT of early-stage NSCLC. Further, when intra-tumour α- heterogeneity was also included in TCP modeling of local- control data from lung SBRT, to deal with the hypothesis that tumor response is determined from the most radioresistant tumour clonogen sub-population, the goodness-of-fit from the LQ-model was even superior to the alternative LQL-models [4]. Once meta-analyses supported the existence of a dose- response relationship for SBRT of early-stage NSCLC [5], and then for oligometastatic lung lesions [6], with saturation of the effects over some threshold dose, the prescription of a minimum total BED of 100Gy 10 became the golden rule. Such threshold, however, might be modulated according to the tumour volume [7]: by reducing the dose to the smaller tumours (<3cm), so as by increasing the dose to the larger ones (>5 cm). Further, already from the Metha’s analysis [5] it was deducible a wide increase in the necessary dose to get from the 3- fractions schedule the same level of local-control of the schedules with 4-5 fractions. The above two observations, that a radioresistant clonogen sub-population might play a key role in severe hypofractionation, and that schedules with 4-5 fractions may be isoeffective although the use of a reduced dose (BED) with respect to 3-fractions schedules, strongly suggest that tumor hypoxia and its reoxygenation may be pivotal to lung SBRT. In support of this hypothesis, in vivo evidence was reported since 1975 of an ‘inverse’ dose behaviour between 3 and 5 fractions: i.e., D 50 ( n =5) lower than D 50 ( n =3) [8]. Models including tumour hypoxia were able to reproduce such ‘inverse’ dose behaviour [9-10], by the re-sensitizing effect of reoxygenation, thus supporting the use for lung SBRT of schedules with ≥ 5 fractions, instead of 3 fractions [11]. By modeling reoxygenation in terms of a mono-exponential time factor [11], it is also possible to explain the wide (30%) observed difference in 3-year local control after lung SBRT, if alternate instead of consecutive days for treatment were used [12]. In conclusion, without the need to invoke any different cell- killing effect of radiation as a function of the d -value, tumour hypoxia seems to be the likely determinant of the optimal dose, number of fractions, and inter-fractions time-interval for lung SBRT. References: 1. Timmerman R, et al. Chest 2003;124:1946-55 2. Wang JZ, et al. Science Transl Med 2010;2(39):39ra48. doi: 10.1126/scitranslmed.3000864