ESTRO 36 Abstract Book

S853 ESTRO 36 2017 _______________________________________________________________________________________________

using a threshold of 10 mmHg. Assuming a clonogenic cell density of 10 8 per cm 3 in the CTV, the HTV doses required to achieve 95% local control (LC) were calculated based on a previously developed radiobiological model (Toma-Dasu et al 2009, 2012) accounting for the dynamic tumour oxygenation due to changes in acute hypoxia not visible in PET images. The total doses were calculated assuming that the treatment involves 24, 30 or 35 fractions. Results The non-linear conversion function and hypoxic threshold of 10 mmHg resulted in hypoxic subvolumes identified in five out six patients. Three out of six patients had a hypoxic subvolume > 3cm 3 . In two of the patients, the delineated HTV was not entirely confined within the primary CTV. For a treatment delivered in 30 fractions, the prescribed dose required to achieve 95% local control for the two patients with the largest HTVs of 32.74 and 38.29 cm 3 respectively were 75.52 and 75.67 Gy, both corresponding to an EQD2 of almost 79 Gy10. For the third patient with a smaller HTV of only 12.37 cm 3 , the total dose in 30 fractions for 95% LC was 72.35 Gy. If the total dose would be delivered in 35 fractions instead, the prescribed doses would increase with about 2.2% of the dose prescribed in 30 fractions for all three cases. The relative decrease in the total dose if the total dose will be delivered in only 24 fractions is about 3.5% for all three HTVs. For all patients and for all treatment fractionation schemes the dose levels required for achieving 95% tumour control probability accounting for local changes in the oxygenation of the tumour are below the levels of dose boosts proved to be feasible to be delivered without extra dose burden to the OARs on a previous study carried out on the same patients. Conclusion HX4-based delineation of hypoxic target volumes and calculation of required boost doses for a predefined tumour control probability appears to be feasible. HX4 is thus a potentially suitable tracer for the purpose of treatment individualisation in NSCLC. EP-1603 Atlas of complication incidence to explore dosimetric contributions to osteoradionecrosis L. Humbert-Vidan 1 , S. Gulliford 2 , V. Patel 3 , C. Thomas 1 , T. Guerrero-Urbano 4 1 Guy's & St Thomas' NHS Foundation Trust, Radiotherapy Physics, London, United Kingdom 2 The Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Joint Department of Physics, London, United Kingdom 3 Guy's & St Thomas' NHS Foundation Trust, Oral Surgery, London, United Kingdom 4 Guy's & St Thomas' NHS Foundation Trust, Clinical Oncology, London, United Kingdom Purpose or Objective Mandibular osteoradionecrosis (ORN) is one of the most severe complications in patients with head and neck cancer undergoing radiation therapy (RT). Potential risk factors include primary tumour site and stage, radiation dose, pre- and post-RT dental extractions and mandibular surgery, chemotherapy, dental hygiene, smoking or alcohol. This pilot study aims to assess the contribution of radiation dose to the mandible to the incidence of ORN and investigates the effect of different risk factors using the atlas of complication incidence (ACI) method to summarise dose-volume histograms and toxicity data. Material and Methods This retrospective study included 80 patients with head and neck cancer with a median age of 62 (range 35-86) treated with radical IMRT. Primary tumour sites included a majority of oropharynx cases (42), oral cavity (26), larynx (6), paranasal sinus (4) and two unknown primary cases. Half of the cohort had been diagnosed with ORN of the mandible within a median follow-up time of 3.2 years

(range 1.3-5.3) from the end of RT. The toxicity endpoint considered included ORN complication of any grading. The other 40 patients were control cases (no ORN observed) prospectively matched according to primary site and treatment option (PORT or primary RT and chemotherapy type). For given volume v j and dose d i levels, the numerical fraction in a cell of an ACI is composed of the number of patients (denominator) and the number of patients with ORN (numerator) that have a mandible percentage volume between v j and v j+1 exposed to a dose level between d i and d i+1 . Atlases were created for sub-sets of the entire cohort to investigate the effects of pre-RT surgery, chemotherapy, smoking or dental extractions. These risk factors were also tested with univariate statistical analysis. Dosimetric variables including d max and d mean were tested with ROC analysis. Results A dosimetric correlation with ORN incidence was observed in cases where treatment modality was primary RT (as opposed to post-operative RT). An increased ORN incidence was observed towards the large percentage volume and high doses region of the ACI, where a large percentage of the mandible volume had received doses of above 40Gy as well as smaller volume percentages receiving doses above 64Gy. A similar incidence pattern was observed for the ACI that included smoking patients only. The ACI for the sub-set of patients that had undergone dental extractions pre- or post-RT also showed a very similar incidence pattern; however, this sub-set included very few cases. Conclusion The ACI analysis carried out so far has shown a dose response in patients who received primary RT, patients who smoked at the time of diagnosis and patients who had dental extractions. The limited number of cases did not allow for any conclusive statistical significance of the logistic regression and ROC analysis results. This pilot study will be expanded to include cases from other centres to increase the cohort size. EP-1604 Ion induced complex DNA damage: In silico modelling of damage and repair using Geant4-DNA. J.W. Warmenhoven 1 , N.T. Henthorn 1 , M. Sotiropoulos 1 , R.I. Mackay 2 , K.J. Kirkby 1,3 , M.J. Merchant 1,3 1 University of Manchester, Division of Molecular and Clinical Cancer Sciences, Manchester, United Kingdom 2 The Christie NHS Foundation Trust, Christie Medical Physics and Engineering, Manchester, United Kingdom 3 The Christie NHS Foundation Trust, Manchester, United Kingdom Purpose or Objective This work uses Monte Carlo simulation to assess and understand the differences in biological response to various radiation qualities in the context of hadron therapy. The current clinical estimator for this is Relative Biological Effectiveness (RBE), offering a biological dose conversion between radiation qualities. A large variability in reported RBE measurements implies that this parameter does not give the full picture. This variability in RBE is a major source of uncertainty in ion therapy treatment planning. Recently, LET based biological effect models have been proposed, however, among those reviewed there are uncertainties in the behaviour of key biological