ESTRO 38 Abstract book

S635 ESTRO 38

for patients with GTV<15cc, 32% for patients with GTV between 15 and 70cc and 0% for patients with GTV>70cc. In Cox regression model, both volumetric staging, addition of chemotherapy and BED reached statistically significant p-value of p=0.00006 for VS, p=0.004 for BED and p=0.017 for addition of chemotherapy. Conclusion Volumetric staging could be used as a prognostic tool in tongue cancer patients treated with radical radiotherapy and proved to be more accurate than TNM staging system at predicting 5-year overall survival and 3-year disease- free survival. EP-1145 Xerostomia and volume and CT number changes of parotid glands during IMRT for head and neck cancer S. Ishikura 1 , T. Nakabayashi 2 , F. Kobayashi 2 , H. Fukuma 3 , Y. Shibamoto 1 1 Nagoya City University Graduate School of Medical Sciences, Department of Radiology, Nagoya, Japan ; 2 Accuray Japan K.K., Division of TomoTherapy Physics & Clinical Support, Tokyo, Japan ; 3 Nagoya City University Hospital, Department of Radiology, Nagoya, Japan Purpose or Objective The purpose was to quantify and analyze the volume and CT number changes of the parotid glands during helical tomotherapy (HT) using daily MVCT images and to evaluate the correlation among these changes, xerostomia grades, and doses to the parotid glands. Material and Methods Between September 2014 and February 2018, 34 patients (pts) with oropharyngeal squamous cell carcinoma were treated with HT at our institution. Patient characteristics were as follows: male/female, 28/6; median age, 68 years (range, 41-84); subsites anterior/lateral/posterior, 7/25/2; T1/2/3/4a, 4/17/8/5; N0/1/2a/2b/2c, 5/7/2/14/6; Stage II/III/IVA, 4/8/22. We used simultaneous integrated boost (n=18: 66 Gy/30 fr in 16; 70 Gy/35 fr in 2) and conventional cone-down boost (n=16: 70 Gy/35 fr in 13; others in 3). Therapies were induction chemotherapy (n=17: paclitaxel/cisplatin/fluorouracil, 15; others, 2) and concurrent biotherapy or chemotherapy (n=26: cetuximab, 20; cisplatin, 4; others, 2). During HT, pts underwent MVCT-guided radiotherapy at each session. We used an Accuray Precision™ research workstation with PreciseART™ Adaptive Radiation Therapy to analyze the volume and CT number changes of the parotid glands during HT. We examined correlations among the dose- volume histogram (DVH) values of the parotid glands, the volume and CT number changes for the first 20 fractions, and the grades of acute xerostomia (dry mouth, CTCAE version 4.0). For pts with a conventional cone-down boost, the DVH values of the initial plan were scaled to 70 Gy/35 fr. We analyzed the correlations of the xerostomia grades and the volume and CT number changes by box plots (significance level, p <0.05). We analyzed the correlations between the doses to the parotid glands and the volume and CT number changes by scatter plots and calculated correlation coefficients. This study was approved by our institutional review board. Results The median follow-up was 1.2 years. The 2-year overall survival was 89%. Eight pts had acute grade 2 xerostomia; none had grade 3. Of these 8 pts, 7 had stage IV disease; 6 received induction chemotherapy, and 6 received concurrent biotherapy. The average of the median dose to the right and left parotid glands was 33.7 Gy (range, 10- 63.6 Gy) and 26.2 Gy (7.6-56.9 Gy), respectively. There was no significant difference in the average of the median dose, the volume change, and CT number change between pts with grade 0-1 and 2 xerostomia. Weak-to-moderate correlations existed between the median dose to the parotid glands and both the volume change (Fig. 1, R =

Conclusion Savage treatment of locoregional failure is effective in disease control for DTC with high percentage of RFS and OS. Surgical approach represents the treatment of choice. In patients ineligible for surgery, I-131 and external beam radiotherapy offered a valid option of cure. EP-1144 Is volumetric staging an alternative to TNM staging system in radiotherapy of tongue cancer? M. MISZCZYK 1 , A. Napieralska 2 , B. Maciejewski 2 1 Centrum Onkologii – Instytut im. Marii Skłodowskiej- Curie Oddział w Gliwicach, Zakład Planowania Radioterapii, Gliwice, Poland ; 2 Centrum Onkologii – Instytut im. Marii Skłodowskiej-Curie Oddział w Gliwicach, Zakład Radioterapii, Gliwice, Poland Purpose or Objective The purpose of this study is to evaluate the prognostic value of GTV-based volumetric staging system (VS) in tongue cancer patients and compare it with the results based on the TNM staging system. Material and Methods The clinical material consists of 99 consecutive patients with anterior or base of tongue cancers, aged 25-83 (median 58,2), treated with radical radiotherapy or chemoradiotherapy as a primary treatment between 2003 and 2014 in a single institution (Table I). The study excluded patients with prior surgical treatment. Total dose of 64-78 Gy was delivered to the primary site. Neoadjuvant chemotherapy was administered in 7 cases, concurrent in 15 cases and both in 8 cases. Efficacy of the irradiation was evaluated using 5-year overall survival (5yOS) and 3-year disease-free survival (3yDFS). The study group was retrospectively assessed using TNM classification (7 th edition) and VS. The latter divided patients according to the gross tumor volume (GTV), that is combined total volume of the primary lesion and volume of metastatic lymph nodes. The cut-off values were 15 and 70 cubic centimeters (cc). The thresholds for volumetric staging system were adapted from works by Studer et al. based on retrospective analysis of the frequency of local failures in particular GTV ranges. The statistical analysis employed basic statistical tools and following tests: Shapiro–Wilk test, Kruskal–Wallis H test, Gehan-Wilcoxon test, logrank test, Chi-square test, Kaplan-Meier estimator and Cox proportional hazards model for single and multiple predictor analysis. Results TNM stage groups correlated well with mean GTV (p=0.0001) but the standard deviation of GTV overlapped between patients with TNM stage I-III, II-IVa and III-IVc. TNM stage groups did not correlate well with 5yOS (p=0.1). At 5 years patients with TNM stage group III presented highest OS and patients with TNM stage IVa-IVc proved to have only marginally worse 5yOS than patients with TNM stage I-II disease. The volumetric stage correlated well with 5yOS as presented in attached figure (p=0,001). There was no statistically significant correlation between 3yDFS and TNM stage group (p=0,05). On the other hand, VS correlated well with 3yDFS (p=0,0001). The 3yDFS significantly decreased with increasing GTV and it was 69%