Abstract book - ESTRO meets Asia

S92 ESTRO meets Asia 2018

1 Fudan University Cancer Hospital, Department of Radiation Oncology, Shanghai, China

management of resectable esophageal cancers. Gastric fundus is the most commonly used tissue for anastomotic conduit in radical esophagectomy. An increase in the dose and volume of irradiation to the fundus, might impair the healing of the anastomosis. Anastomotic leak is an important and possibly fatal surgical complication. The current study aims to evaluate the dosimetric parameters of the gastric fundus as an Organ At Risk (OAR) in NACRT for Carcinoma (Ca) Esophagus and study the occurrence of anastomotic leak. Material and Methods Records of 25 patients diagnosed with Ca. Esophagus, who have received NACRT were available for retrospective analysis. The gastric fundus was retrospectively contoured on the simulation Computerized Tomographic images; Dose Volume parameters received by the fundus were recorded. The occurrence of anastomotic leak was documented from surgical records. Results The mean age of the study population was 51 years. 21/25 (84%) patients had squamous cell Ca., while 4/25 (16%) had adenocarcinoma. 2/25(8%) had lesion in proximal esophagus, 8/25 (32%) in middle esophagus and 15/25(60% ) in the distal esophagus. Stage wise distribution was 6/25 (24%) with T3; 19/25 (76%) with T4; 12/25 (48%) with N0; 11/25 (44%) with N1; 2/25 (8%) with N2. Total dose received was 41.4 – 45 Gy/23-25 fractions with weekly Paclitaxel 50mg/m 2 + Carboplatin AUC=2. 20/25 (80%) of the cases were planned by Volumetric Modulated Arc Therapy (VMAT); 5/25 (20%) were planned by 3 Dimensional Conformal Radiotherapy (3DCRT). The mean + standard deviation of doses received by gastric fundus for proximal, middle and distal esophageal lesions are given in table 1. The mean volume of contoured fundus was 16.89 + 3.73cc, 14.85 + 6.02cc, 22.70 + 7.53cc for proximal, middle and distal esophagus respectively. 9/25 patients (36%) had post-operative anastomotic leak, among whom six patients (66%) and three patients (33%) had distal and middle esophageal lesions respectively. 4/5 (80%) patients treated by 3DCRT developed leak, whereas 5/20 (25%) by VMAT developed leak. Table 1. Dose volume parameters of Gastric Fundus in NACRT - Ca. esophagus Conclusion Gastric fundus may be routinely contoured as an OAR in esophago-gastric radiation with documentation and analysis of its dose volume parameters. Occurrence of anastomotic leak is more common in cases with distal esophageal lesions, with larger volume of gastric fundus being irradiated. Further correlation with rate of anastomotic leak using a larger sample size and restriction of doses to the fundus may be considered. PO-226 Modeling dosimetric parameters for upper gastro- intestinal tumors treatment toxicity reduction N. Dinapoli 1 , E. Lattanzi 2 , S. Longo 3 , L. De Filippo 4 , M. Massaccesi 1 , A. Re 3 , F. Cellini 1 , G.C. Mattiucci 3 , V. Valentini 3 1 Fondazione Policlinico Universitario A. Gemelli IRCCS, Dipartimento Scienze Radiologiche- Radioterapiche ed E matologiche, Roma, Italy 2 Università degli Studi di Parma, Scuola di Specializzazione in Radioterapia, Parma, Italy 3 Fondazione Policlinico Universitario A. Gemelli IRCCS - Università Cattolica del Sacro Cuore, Dipartimento Scienze Radiologiche- Radioterapiche ed E

Purpose or Objective Skeletal muscle depletion (Sarcopenia) or ongoing muscle wasting have been reported in previous studies to be associated with higher chemotherapy-related toxicities and decreased survival. The aim of this study was to investigate whether changes in body composition during chemoradiotherapy, as well as radiomics features on CT scans can predict toxicities and outcomes in gastric cancer patients receiving adjuvant chemoradiotherapy. Material and Methods A total of 109 patients who received radical resection and chemoradiotherapy (CRT) between January 2009 and December 2015 with available CT scans obtained before and after chemoradiation were retrospectively examined. The skeletal muscles at the level of the third lumber vertebral were contoured on images from CT scan and the radiomics features were extracted using MIM software TM . Comparisons between pre- and post-CRT images and its impact on toxicities and long-term outcomes were studied. Binomial logistic regression model was built for toxicities and survival using clinical factors and pretreatment radiomics features. Concordance (c) statistics were used to test predictive accuracy of the survival model. Results The skeletal muscle index decreased from 42.26cm 2 /m 2 pre-CRT to 41.32cm 2 /m 2 post-CRT (p=0.001). Relative muscle change during CRT was categorized into three groups: tertile one, <1.0%; tertile two, 1.0% to 3.8%; and tertile three, >3.8%. Skeletal muscle depletion was independently associated with overall survival (95% CI: 1.06 to 4.55; P=0.035). A cutoff value of muscle loss>10% seemed to be a threshold for dose-limiting toxicity of concurrent chemotherapy (P=0.047). A total of 253 radiomics features were calculated using in-house radiomics software after segmentation. Three radiomic features (GLRMS_RP, HL_GLRMS_LRE, and LH_absolute_std) was associated with disease-free survival and a prediction model containing radiomics features gave a c statistic of 0.73.

Conclusion Loss of skeletal muscle mass during chemoradiotherapy is an independent prognostic factor for overall survival and radiomics features of skeletal muscle can predict disease- free survival for gastric cancer patients receiving adjuvant chemoradation. PO-225 Gastric fundus as an Organ At Risk in Neo-adjuvant Chemo-Radiotherapy for Carcinoma Esophagus K.R. Nikhila 1 , S.P. Shyama 1 , R. Kalayarasan 2 , P. Prasanth 3 1 JIPMER, Radiation Oncology, Pondicherry, India 2 JIPMER, Surgical Gastroenterology, Pondicherry, India 3 JIPMER, Surgical Oncology, Pondicherry, India Purpose or Objective Neo-adjuvant chemo-radiotherapy (NACRT) followed by definitive surgery has been the standard of care for