ESTRO 2023 - Abstract Book

S1136

Digital Posters

ESTRO 2023

Purpose or Objective Radical resection (R0) represents the best curative treatment for local recurrence (LR) of rectal cancer and represents the only therapeutic option that can provide long-term survival. Re-irradiation (re-RT) can increase the rate of R0 resection. Currently, guidelines on Re-RT for LR rectal cancer are lacking. The Italian Association of Radiation and clinical Oncology study group for gastrointestinal tumors (AIRO-GI) disseminated a national survey to investigate the current clinical practice of external beam radiation therapy (EBRT) in this group of patients. Materials and Methods In February 2021, the survey was designed and then distributed to members of the GI working group. The questionnaire consisted of 40 questions regarding center characteristics, clinical indications, doses, and treatment techniques of re-RT for LR rectal cancer. Results Thirty-seven questionnaires were collected. Re-RT was reported as an option for neoadjuvant treatment in resectable and nonresectable disease by 55% and 75% of respondents, respectively. In most centers, long-course treatment with 30-40 Gy (1.8-2 Gy/die, 1.2 Gy bid) and hypo-fractionated regimen of 30-35 Gy in 5 fractions were used. A total dose of 90-100 Gy as EqD2 dose (alfa/beta=5 Gy), taking into account previous treatment, was delivered by 46% of respondents. IMRT/VMAT and daily IGRT protocols were used in 94% of centers. Conclusion Our survey showed high quality of treatment and management of rectal cancer LR. Considerable variation was observed in terms of dose and fractionation, highlighting the need for consensus on a common treatment strategy that could be validated in prospective studies. 1 Ajou University School of Medicine, Radiation Oncology, Suwon, Korea Republic of; 2 DR Lab, Data Science Team, Seoul, Korea Republic of Purpose or Objective We developed a deep learning model to predict pathological response based on magnetic resonance imaging in rectal cancer. Materials and Methods A total of 242 patients with locally advanced rectal cancer who received preoperative chemoradiotherapy followed by surgical resection were collected from single center. Surgical resection was performed at 6 to 8 weeks after the completion of preoperative CRT. Achieving pathogic complete response (pCR) was defined as complete absence of any tumour cells, in both the primary site and the dissected lymph node in surgical specimens. Based on pre-chemoradiotherapy T1-weighted axial 3D MR images, deep learning models were developed to predict pCR ,respectively. Results To calculate the probability of pCR using various convolutional neural network (CNN) architectures, several deep learning models were developed. The data input to the deep learning model was 3D MRI image, clinical data, and MRI metadata. Among 242 patients, 50 (20.7%) had evidence of pCR. The best model was developed based on SeResNet using transfer learning by MONA framework. The deep learning model showed an area under the receiver operating characteristic curve, PO-1407 AI model for predicting tumor response in rectal cancer using magnetic resonance imaging. J. HEO 1 , Y. Oh 1 , O.K. Noh 1 , M. Chun 1 , H. Jang 2

sensitivity, specificity, and accuracy of 0.926, 0.831, 0.921, and 90.2% for predicting pCR. Figure 1. Overview of the development and evaluation of the tumor response algorithm.