ESTRO 38 Abstract book

S376 ESTRO 38

was developed using a multivariate logistic regression analysis with backward selection. Results The median follow-up time for all patients was 36 months. In total 18 out of 740 cases (2,4%) developed ORN grade ≥ 2 and the median time to ORN was 12 months. The cumulative incidence of ORN over 6, 12, 18 and 24 months was 1.2, 2.1, 2.6 and 3.1%, respectively. To identify the most important DVH-parameter, the mean values of the DVH-parameters between cases and non- cases were compared first. The mean and maximum dose to the mandible were higher among patients with ORN vs. non-cases (31.7 vs 45.3 Gy, p= 0.003 and 53.1 vs 71.8 Gy, p< 0.001, respectively ) . DVH bins from 25 to V70 were all significantly higher in the ORN cohort compared with controls (figure 1). The best predictor for mandibular ORN was the V60 (volume receiving 60 Gy). No other significant prognostic factors or confounders for ORN were identified. The p-value for this model was 0.002, with an OR of 16.21 (95% CI 2.79-94.28) and an AUC of 0.78 (95% CI 0.71-0,85). See figure 2 for the NTCP model.

Conclusion The adoption of a GE expansion modality with resulting smaller HR CTVs was not associated with a detrimental impact on outcome in OPC. PO-0732 NTCP model for osteoradionecrosis after definitive radiotherapy in head and neck cancer patients. N. Den Haan 1 , L. Van den Bosch 1 , A. Van den Hoek 1 , H. Bijl 1 , R. Steenbakkers 1 , M. Dieters 1 , H.P. Van der Laan 1 , H. Langendijk 1 1 University Medical Center Groningen, Radiation Oncology, Groningen, The Netherlands Purpose or Objective The purpose of this prospective cohort study was to develop a normal tissue complication probability (NTCP) model for predicting the risk of mandibular osteoradionecrosis (ORN) in patients with head and neck cancer (HNC) treated with definitive The population of this study was composed of 740 patients with HNC treated with definitive (chemo)radiotherapy in the period from January 2007 to June 2016. Mandibular ORN was scored prospectively as part of a standard follow- up program according to CTCAEv4.0. Mandibular dose- volume histograms (DVH) were extracted from the treatment planning system. Dosimetric parameters were compared using the Wilcoxon rank-sum test. The model (chemo)radiotherapy. Material and Methods

Conclusion This is the largest prospective cohort study ever done on the development of mandibular ORN in patients treated with definitive radiotherapy or chemoradiation. In comparison to the incidence of ORN in the postoperative setting, the incidence of ORN is relatively low. The main risk factor is the dose to the mandible and in particular the high dose volume. The V60 turned out to be the most important prognostic factor. PO-0733 Non-invasive imaging for tumor hypoxia: a novel validated CT and FDG-PET-based Radiomic signature. S. Sanduleanu 1 , A. Jochems 2 , T. Upadhaya 3 , A. Even 2 , R. Leijenaar 2 , F. Dankers 4 , R. Klaassen 5 , H. Woodruff 2 , M.