ESTRO 38 Abstract book

S765 ESTRO 38

superior slice was established. The median and range of the final 13 differences in position in cm is documented. The same was done for the inferior slices. As the 13 tumours varied in size looking at cm 3 was less useful for difference in volume, instead percentage difference between the largest and smallest volume was used. E.g. if largest volume is 16cm 3 and smallest is 8cm 3 then there is a 50% difference in the volumes.

Conclusion Inter and intra-observer variability was significant despite markers; small variations at the inferior and superior border still contributed to considerable variation in the final volume. Fiducial placement and retention at the inferior border was lower than would be anticipated, however inter- observer variability was similar to the superior slice. In the context of multi-modality imaging techniques to identify GTV, it is likely marker contribution has been restricted. Cases where they are most likely to contribute are tumours difficult to appreciate on imaging. Given the rare but recognised complication of death following perforation at EUS, identifying patients most likely to benefit is key. EP-1406 Mapping Pattern of LNMs for Postoperative Radiotherapy in TESCC: Defining the Clinical Target Volume J. Yu 1 , C. Li 1 , W. Ouyang 1 , Y. Xu 1 , J. Zhang 1 , C. Xie 1 1 Zhongnan Hospital of Wuhan University, Radiation and Medical Oncology, Wuhan, China Purpose or Objective The clinical target volume (CTV) for postoperative radiotherapy for thoracic esophageal squamous cell carcinoma (TESCC) needs to be defined. The study aim was to map the first metastatic lymph nodes (LNMs) in a computed tomography (CT)-based atlas and to guide CTV delineation in postoperative radiotherapy for TESCC. Material and Methods Sixty-nine patients with primary regional LNMs after esophagectomy were included. The LNM epicenters were registered onto corresponding anatomic axial CT images of a standard patient in the treatment position, with reference to the surrounding vascular and bony structures. The LNM sites were based on lymph node map of esophageal cancer, AJCC 8 th . Accordingly, regional lymph node stations 1 to 8M were defined as the upper-middle mediastinum region (UMMR); stations 8Lo, 9, and 15 were defined as the inferior mediastinum region (IMR); and stations 16 to 20 were defined as the upper abdominal lymph node region (UAR). The lymph metastasis risk for different segments was assessed. Results One hundred and seventy-nine LNMs were mapped onto standard axial CT images. The upper-middle mediastinum region (station 1 to 8M) was the most common site for LNMs. It contained 97% of metastases in the upper segment, 90% in the middle segment, and 66% in the lower one. Advanced pathological stage (≥IIIB) might be a predictive factor for upper abdominal region (UAR) relapse in lower TESCC. Lower cervical para-tracheal LNMs were within a 4.3-cm bilaterally expanded area from the midline and a 2.2-cm expanded area from the anterior, from the superior border of the C7, to the inferior border of the first thoracic vertebra.

Tolerance to the procedure was generally good. Ten patients presented acute toxicity, being asthenia Grade 1 or 2 the most frequent in seven patients. Chronic toxicity was reported in two patients (6.25%), one Grade 3 enteritis and one Grade 3 abdominal abscess. No Grade 4- 5 acute or chronic toxicity was reported. Conclusion In our experience SBRT for adrenal gland metastases is safe and effective, achieving good local control rates, which are directly related with the total dose administered. Further studies are needed to consolidate these results and evaluate the influence of SBRT in patient survival and quality of life. EP-1405 Feasibility study of fiducial markers in oesophageal cancer radiotherapy A. Morton 1 , D. McIntosh 1 , S. Currie 2 , A.J. Stanley 3 , S. Paterson 4 , D. Grose 1 , H. Marashi 1 , V. MacLaren 1 1 Beatson West of Scotland Cancer Centre, Clinical Oncology, Glasgow, United Kingdom ; 2 Beatson West of Scotland Cancer Centre, Physics, Glasgow, United Kingdom ; 3 Glasgow Royal Infirmary, Gastroenterology, Purpose or Objective To determine whether insertion of 5mm radio-opaque gold fiducials (Cook Medical USA) during EUS impacts on inter- and intra-observer variability when defining oesophageal cancer GTV. Material and Methods 14 patients with oesophageal cancer planned to undergo radical radiotherapy, with or without chemotherapy, underwent EUS to place 1-2 fiducials at the proximal and distal end of tumour unless stricturing prevented this. GTV of primary tumour was defined on a planning CT retrospectively 3 times by 3 separate clinicians using fiducials with CT-PET and EUS report. Results 13 patients were imaged with 4 having superior/ inferior markers visible. 8 had only superior and 1 had no visible markers (4 due to stricturing tumours preventing placement, 5 due to marker migration). Unfortunately 1 patient died of mediastinitis secondary to oesophageal perforation following EUS at which stage the trial stopped early. Intra-observer variability is shown in the attached table for clinicians 1, 2 and 3. Each clinician completed 3 separate contours and these were compared for all 13 cases. Inter-observer variability was assessed by comparing the first contours from each clinician, then the second contours separately followed by the third as shown in table as contour 1, 2 and 3. For example in ‘Contour 2’ column, there were 3 contours for all 13 patients representing the second volume done by each of the 3 clinicians. For each patient the maximum difference between the 3 contours in the position of the Glasgow, United Kingdom ; 4 NHS Forth Valley, Gastroenterology, Larbert, United Kingdom