ESTRO 37 Abstract book

ESTRO 37

S401

14 Centro di Riferimento Oncologico Regionale, Department of Radiation Oncology, Rio Nero in Vulture- Potenza, Italy 15 Radiation Oncology Center- University of Bologna, Department of Experimental- Diagnostic and Specialty Medicine - DIMES, Bologna, Italy Purpose or Objective Currently, computed tomography (CT) is the standard imaging for pancreatic tumor delineation in radiation treatment planning. A previous study, evaluating inter- observer variability of Clinical Target Volume (CTV) delineations of two different cases of pancreatic cancer treated by postoperative and preoperative radiotherapy on CT scan, was conducted. Greater variability was observed for Boost CTV (Gross Tumor Volume, GTV, plus margins) rather than the Elective CTV, in both setting, suggesting that the availability of contouring guidelines might help to reduce observational variability. Due to the high soft tissue resolution, Magnetic Resonance Imaging (MRI) could improve the accuracy of tumor delineation compared to CT scan. Since recommendations for pancreatic cancer contouring using MRI have been recently provided, a new multi-institutional contouring dummy-run study was proposed by the AIRO study group for gastrointestinal cancers, to evaluate the role of MRI imaging in inter-observer variability of GTV delineation in pancreatic cancer. Material and Methods Two clinical case about borderline resectable (Case1) and locally advanced unresectable (Case2) pancreatic cancer were selected by two radiation oncologists and two radiologist expert in pancreatic cancer. In two sequential steps, CT scan and 3T MRI sequences, both with or without contrast enhancement, were send to the adherent centers for the GTV and duodenum delineation. The mean volumes and Left-Right (L-R), Antero-Posterior (A-P) and Cranio-Caudal (C-C) shifts direction were calculated. The Dice Similarity Index (DICE) was used to evaluate the spatial overlap accuracy of boths volumes compared to a benchmark from a national high volume of treatment reference centre. Results Thirty four radiation centers joined the study. The study is still ongoing regarding the delineation of required volumes and final inter-observer variability analysis. At the moment we reported data of 11 and 16 radiotherapy centers who submitted targets delineated on both CT scans and MRI images, for Case 1 and 2, respectively. The mean volumes and L-R, A-P and C-C deviations are reported in Table 1 and showed in Figure 1. Mean DICE index for GTV and duodenum delineation for Case1 were 0.56 and 0.56 for CT, and 0.45 and 0.46 for MRI, respectively. Mean DICE index for GTV and duodenum delineation for Case2 were 0.36 and 0.41 for CT, and 0.33 and 0.37 for MRI, respectively.

Conclusion The large number of centers that joined to the study showed a great interest in MRI for GTV delineation in pancreatic cancer. At the current analysis, MRI resulted comparable to CT concerning the GTV delineation of borderline resectable case. Preliminary DICE index values confirmed the great criticism of GTV delineation in pancreatic cancer, especially when vascular structures are involved, as in unresectable case. Final results are expected to define the real impact of MRI inter-observer variability in contouring compared to CT scan. PO-0776 Induction chemotherapy with TPF followed by chemoradiotherapy for esophageal squamous cell carcinoma H. Ariga 1 , K. Kikuchi 1 , T. Iwaya 2 , H. Oikawa 1 , H. Kakuhara 1 , T. Segawa 1 , S. Yamaguchi 1 , R. Nakamura 1 1 Iwate Medical University, Radiation Oncology, Uchimaru- Morioka, Japan 2 Iwate Medical University, Surgery, Uchimaru- Morioka, Japan Purpose or Objective Induction chemotherapy with docetaxel, cisplatin, and fluorouracil (TPF) is emerging as more effective regimen than cisplatin and fluorouracil (FP) for locally advanced squamous cell carcinomas. The aim of this study is to retrospectively evaluate the efficacy and safety of induction chemotherapy with TPF followed by concurrent chemoradiotherapy (CCRT) compared with CCRT alone in patients with locally advanced esophageal squamous cell carcinoma. Material and Methods Between 2011 and 2016, 77 patients with localy advanced esophageal squamous cell carcinoma underwent definitive CCRT at one institution. CCRT basically consisted of two cycles of FP with concurrent 3D conformal irradiation of 60 Gy at 2 Gy per once-daily fraction. Of the 77, 33 patients received induction chemotherapy, which was TPF regimen (docetaxel 30 mg/m² on day 1 and 8, cisplatin 80 mg/m² on day 1, and continuous fluorouracil 800 mg/m² per day from day 1 to day 5; n=31) or FP regimen (cisplatin 80 mg/m², fluorouracil 800 mg/m² day 1 to 5; n=2). Overall survival was calculated from the treatment start date (induction chemotherapy or CCRT) to death. The clinical response to induction chemotherapy was evaluated on the basis of the RECIST version 1.1 and the evaluation of toxicity was based CTCAE version 4.0. Results Induction chemotherapy was performed on patients with more advanced stage (p=0.026). After a median follow-up of 25.4 months (IQR 11.8-36.6), 2-year overall survival of patients with/without induction chemotherapy was 49.0% and 62.2%, respectively (ns). Of the 33 patients with induction chemotherapy, 19 experienced complete to partial response (CR/PR), and 14 had stable to progressive disease (SD/PD). 2-year overall survival of patients evaluated CR/PR and SD/PD were 72% and 23.4%, respectively (p=0.0036). This overall survival of