ESTRO 36 Abstract Book

S847 ESTRO 36 2017 _______________________________________________________________________________________________

was 23.4 Gy(RBE) given by 13 fractions. The plan was optimized using a fixed RBE value of 1.1. The IMPT plans were imported into the FLUKA Monte Carlo code where dose and dose-average LET distributions (LET D ) were calculated. Subsequently, a published LET-based RBE model for proton therapy was applied to these distributions to quantify the biological doses for the heart, lungs and the PTV. Results The Monte Carlo simulations revealed elevated LET D values both in the PTV, heart and lung and correspondingly higher biological doses as compared to the doses obtained with an RBE of 1.1 (Figure 1). The dose volume histograms revealed small elevation in mean dose to the lungs (from 2.0 to 2.4 Gy) and the heart (from 0.7 to 1.0 Gy). However, the variable RBE model calculated a significantly higher max dose in both organs, with a shift from 21.8 Gy to 27.2 Gy for the lungs and 17.2 Gy to 21.2 Gy for the heart. Conclusion Including LET and biological dose models revealed an increased dose to heart and lung compared to doses calculated with RBE of 1.1, mainly due to the enlarged LET values at the distal dose falloff of the beam. The reproducibility of these results in a larger cohort of patients will be investigated. EP-1593 Accuracy of TCP model for nasopharyngeal cancer after more than five years average follow-up M. Avanzo 1 , J. Stancanello 2 , G. Franchin 3 , S. Barbiero 4 , R. Jena 5 , G. Sartor 1 , E. Capra 1 1 Centro di Riferimento Oncologico, Medical Physics, Aviano, Italy 2 Oncoradiomics, Mastricht, The Netherlands 3 Centro di Riferimento Oncologico, Radiation Oncology Department, Aviano, Italy 4 Casa di Cura S. Rossore, Radio-oncology, Pisa, Italy 5 University of Cambridge, Department of Oncology, Cambridge, United Kingdom Purpose or Objective A model for radiation therapy (RT) based tumor control probability (TCP) of nasopharyngeal carcinoma (NPC) which includes the effects of hypoxia and chemoradiotherapy was previously developed by fitting TCP to clinical local control data from published randomized studies and tested in a patient dataset with limited follow-up time. The purpose of this work was to validate the model by comparison of estimated TCP and average overall local control rate in a cohort of patients with long follow-up time. Material and Methods 96 patients treated with intensity modulated RT delivered by LINAC or Helical Tomotherapy and neoadjuvant chemotherapy for histologically proven nasopharyngeal carcinoma were followed for an average time of 5.5 years. TCP was calculated from individual GTV dose-volume histograms, fractionation and overall treatment duration data using a previously established model (Ref. 1) based on the linear-quadratic model and Poisson statistics modified to account for repopulation, chemotherapy, heterogeneity of dose to the tumor, and hypoxia. The model parameters were: α/β of 10 Gy, α of 0.396 Gy −1 with σ α of 0.07 Gy −1 , density of clonogens 10 7 cc -1 , fraction of patients showing hypoxia 22%, oxygen enhancement ratio (OER) of 1.417, chemotherapy enhancement factor 1.64, T pot of 3 days and T k of 28 days. The average TCP was compared to the local control rate (LCR), defined as the absence of clinical and radiological Electronic Poster: Physics track: (Radio)biological modelling

evidence of disease residuum in the GTV at the last follow- up visit. Results The average calculated TCP in patients treated at our institution was 86.9% with 95% confidence intervals (95% CIs) of 74%–95.0%. At a mean follow-up time of 5.5 years, 15 patients developed local failure in the nasopharynx and three had distant and local recurrences. Three patients developed recurrence outside the treated volume. Local control was therefore obtained in 81% (95% CI: 72-87.8%) patients, and was in agreement with calculated TCP. Conclusion The TCP model shows was in agreement with LCR in patients treated at our institution after a long follow-up time. References: 1. Avanzo et al. Med. Phys. 37 (4) 1533-1544, April 2010 EP-1594 Development of multivariable models for acute toxicities in nasopharyngeal cancer radiotherapy A. Cavallo 1 , T. Rancati 2 , A. Cicchetti 2 , N.A. Iacovelli 3 , F. Palorini 2 , C. Fallai 3 , E. Orlandi 3 , E. Pignoli 1 1 Fondazione IRCCS Istituto Nazionale dei Tumori, Medical Physics Unit, Milan, Italy 2 Fondazione IRCCS Istituto Nazionale dei Tumori, Prostate Cancer Program, Milan, Italy 3 Fondazione IRCCS Istituto Nazionale dei Tumori, Radiation Oncology 2, Milan, Italy Purpose or Objective To investigate the dose-response relationship for acute toxicities in nasopharyngeal cancer (NPC) RT and to build multivariable models aiming at the inclusion of other non- dosimetric factors. Material and Methods A series of consecutive NPC patients (pts) treated curatively with IMRT/VMAT + chemotherapy at 70 Gy (35- 33 fr, 2-2.12 Gy/fr) was considered. Clinical- tumor- and treatment-related data were retrospectively collected: age, gender, BMI, smoking history, histology, staging, comorbidities, RT technique, overall treatment time. Organs-at-risk (OAR) volumes and dose distribution for each pt were also considered. Acute toxicities were assessed according to CTCAE v4.0 at baseline and weekly during RT. Four endpoints were considered: mean grade (G) ≥1.5 and G≥3 oral mucositis (OM), G3 dysphagia and G≥2 salivary dysfunction (SD). The selected OARs were: oral cavity (OC) and parotid glands (PG) considered as a single organ for OM and SD; OC, pharyngeal constrictor muscles (PCM), supraglottic and glottic larynx (GL) for dysphagia. DVHs were reduced to Equivalent Uniform Dose (EUD): for each OAR the best volume parameter n was determined through numerical optimization. When this procedure did not converge, we chose to evaluate DVH cutpoint through t-test. EUD was inserted into a multivariable logistic (ML) model together with clinical/treatment features; variables selection was guided by LASSO. Goodness of fit was evaluated with Hosmer-Lemeshow test and calibration plot. Results Data were collected for 132 pts. MeanG≥1.5 and G≥3 OM were reported in 40 pts (30%), G3 dysphagia in 50 (38%) and G≥2 SD in 90 (68%). ML models (figures 1-2) consisted in: a single variable for meanG≥1.5 OM, i.e. OC EUD with n=1 (mean dose) (OR=1.07); 3 variables for G≥3 OM including OC EUD with n=0.05 (OR=1.02), PG EUD with n=1 (OR=1.06), BMI≥30 (OR=3.8, obese pts); 3 variables for dysphagia including PCM V50Gy (OR=1.02), GL and OC EUD with n=0.35 and 0.15 respectively (OR=1.02 and 1.04); 4 variables for SD including PG D98% (OR=1.04), OC EUD with n=0.05 (OR=1.11), age (OR=1.08, 5-year intervals), smoke (OR=1.37, yes vs no). Calibration was good in all cases.