ESTRO 36 Abstract Book

S581 ESTRO 36 2017 _______________________________________________________________________________________________

Purpose or Objective Intensity Modulated Radiotherapy (IMRT) to the head and neck cancer has been proven to reduce the incidence of long-term xerostomia and thereby improve quality of life (QOL) of survivors. However, it is also well known that there are ongoing changes in the dose intended to the parotids during radiotherapy often resulting in higher parotid doses. Parotid sparing adaptive radiotherapy (PSART) provides dosimetric corrections for such unintended higher doses. Our study evaluates the clinical benefits of PSART and also calculates the resource intensiveness. Material and Methods Thirty-nine of the planned 90 patients of head and neck cancer were screened if to at least one or both parotid (index parotid/s) were receiving a mean dose (MD) of between 25 to 30Gy and were recruited. The index parotid was delineated on the verification images acquired on 14 th and 19 th day and the MD was determined by overlaying the fused verification image on the planned CT. Dosimetric comparison was done using adaptive planning. If the MD had increased by 2% of the initial intended dose, an adaptive plan (AP) was attempted with an aim to reduce MD by 2% without compromising PTV coverage; this plan was then used to deliver the remaining treatment. The time required and number of personnel involved during each step was recorded and person hours (PH) were calculated using the formula: (Minutes x Personnel involved)/60. Xerostomia was assessed by a questionnaire (XeQOLS) at baseline, at 3 and 9 months after completion of treatment. Results Eighteen patients underwent radical radiotherapy with remaining receiving adjuvant treatment. Thirty were treated on Tomotherapy whilst others were treated on Novalis Tx. The median increase in parotid dose was 1.1Gy corresponding to a median reduction in the parotid volume of 1.1cc. Twenty-three patients required an AP with fifteen requiring it after the 14 th day. An acceptable adaptive plan, which met the criteria as described above, was achieved for 19 of these 23 patients. A median of 7.5 fractions were delivered with the adaptive plans. Median PH required for normal RT of a patient was 26PH while an additional 14.34PH was required in those undergoing PSART. All components of the XeQOLS (physical, pain, social and personal) were worse at 3 months compared to baseline and improved over time at 9 months in all patients irrespective of whether they underwent PSART or not (Figure 1). However, early data does not reveal any significant difference in QOL for those who underwent PSART. (Table 1) XeQOLS Score at 9 months No PSART Patients(Medi an Score) PSART Patients(Medi an Score) Significanc e, p(Mann- Whitney)

Conclusion The results confirm that PSART, which is resource intensive procedure, definitely reduces dose to the parotid. However, it is still unclear if such plans improve clinical QOL parameters further to the planned IMRT plans. Completion of this study could give us further confirmation on the clinical benefits of PSART. EP-1065 Prediction of Dysphagia and Xerostomia based on CT imaging features of HNSCC Patients K. Pilz 1,2 , S. Leger 1 , A. Zwanenburg 1 , C. Richter 1,2,3,4 , M. Krause 1,2,3,4,5 , M. Baumann 1,2,3,4,5 , S. Löck 1,2,4 , E.G.C. Troost 1,2,3,4,5 1 OncoRay - National Center for Radiation Research in Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus- Technische Universität Dresden- Helmholtz-Zentrum Dresden – Rossendorf, Dresden, Germany 2 Department of Radiation Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus- Technische Universität Dresden, Dresden, Germany 3 German Cancer Research Center DKFZ, Germany and German Cancer Consortium DKTK partner site Dresden, Dresden, Germany 4 Helmholtz-Zentrum Dresden – Rossendorf, Institute of Radiooncology, Dresden, Germany 5 National Center for Tumor Diseases partner site Dresden, Dresden, Germany Purpose or Objective Radiochemotherapy (RCT) for patients with head and neck squamous cell carcinoma (HNSCC) frequently causes xerostomia and dysphagia, which may be alleviated by treatment adaption, e.g., modulation of dose distribution to the salivary glands. Current clinical models, which are based on dosimetric parameters, mostly achieve moderate prediction accuracy. Therefore, we aimed to improve the prediction of xerostomia and dysphagia by using additional imaging biomarkers based on computed tomography (CT) scans. Material and Methods In this study 46 patients with UICC stage III/IV advanced head and neck squamous cell carcinoma (HNSCC) were considered (NCT00180180, [1]). All patients received primary RCT and underwent a pre-treatment CT scan without intravenous contrast agent. Patient-reported xerostomia and dysphagia were evaluated at baseline, every week during RCT, four weeks after treatment and three monthly thereafter. 5040 imaging features were extracted from the parotid and submandibular glands. Feature reproducibility tests based on the RIDER re-test data set [2] were performed leading to 1513 imaging features in total. The most informative features were selected by a univariate logistic regression analysis. The developed radiomic signature was used to train and validate multivariate logistic regression and random forest models using repeated 5-fold cross validation. The predication accuracy was assessed by the area under the curve (AUC). Results

Physical Domain Pain Domain

1.00

2.00

0.64

1.00

2.00

0.64

Physiologic al Domain 1.00

2.13

0.92

Social Domain

0.67

2.17

0.64

Total Score 0.93

2.07

0.77