ESTRO 2020 Abstract book

S369 ESTRO 2020

171 patients with recurrent breast cancer out of 235 patients comprised in the original study were included in the current study. A detailed analysis of the medical records was performed in order to record the treatment prior to diagnosis of the LNM. Furthermore, information about the primary tumor of the patients (e.g. stage, grading, biology) was collected. Reliable and comprehensive data could be collected for 134 of 171 patients. The LNM were allocated to two different groups: ALND vs. no ALND. The LNM in each group were compared with regard to the distribution pattern and size of the LNM and the presence of distant metastases. Color-coded heatmaps marking hotspots of LNM in a CT-template were created (figure 1). Results The ALND group consisted of 131 LNM (63 patients), the no ALND group of 136 LNM (58 patients). Most patients had distant metastases at the time of the recurrence (ALND 63,49%; no ALND 75,86%). The average number of LNM per patient was 2,08 for group ALND and 2,34 for group no ALND. In both groups, LNM occurred mainly in Level I (ALND n=40: 30,53%; no ALND n= 79: 58,09%) and the supraclavicular region (ALND n=27: 20,61%; no ALND n=26: 19,12%). LNM in Level II, III and the internal mammary region occurred more often after ALND (Level II: ALND n=16: 12,21%; no ALND n=10: 7,35% ; Level III: ALND n=22: 16,79%; no ALND n=10: 7,35%; internal mammary region: ALND n=20: 15,27%; no ALND n=10: 7,35%). Figure 1 depicts hotspots of LNM of group ALND (a) compared to LNM of group no ALND (b).

Poster Highlights: Poster highlights 20 PH: Auto- contouring using deep learning

PH-0602 Dosimetric impact of central OAR review on rectal and bladder constraint attainment in PACE-B trial D. Brand 1 , A. Tree 1 , K. Fernandez 2 , O. Naismith 3 , S. Brueningk 2 , E. Hall 4 , S. Gulliford 5 , N. Van As 1 , X. On behalf of the PACE Trial Management Group 4 1 The Royal Marsden NHS Trust & The Institute of Cancer Research, Radiotherapy and Imaging, London, United Kingdom ; 2 The Institute of Cancer Research, Radiotherapy and Imaging, London, United Kingdom ; 3 The Royal Marsden NHS Trust & The National Radiotherapy Trials Quality Assurance RTTQA group, Radiotherapy, London, United Kingdom ; 4 The Institute of Cancer Research, Clinical Trials and Statistics Unit, London, United Kingdom ; 5 The Institute of Cancer Research & University College London Hospital, Radiotherapy, London, United Kingdom Purpose or Objective PACE-B (NCT01584258) is a randomised phase III trial comparing SBRT, 36.25 Gy in 5 fractions (Fr), to conventionally or moderately hypofractionated radiotherapy (CFMHRT), either 78 Gy in 39 Fr or 62 Gy in 20Fr for localised prostate cancer. Most dose volume constraints specified in the trial are based on relative OAR volume. We hypothesised that delineation of OARs by participating centres may differ from per protocol definition and that this may influence dose constraint attainment. Material and Methods 305/874 PACE-B participants with DICOM data available for centralised review at the time of the study were included. The bladder and rectum contours (original OARs) were checked and edited to protocol definition (reviewed OARs) by a single observer [KF], who was quality assured (QA) by two experienced clinicians [DB, AT]. Both organs were solid structures; the rectum was from anus to rectosigmoid junction. DVHs for original and reviewed OARs were calculated. The proportions of patients with any major dose constraint variation (separately for bladder and rectum), were compared for original versus reviewed OARs by exact McNemar test. Major dose constraint variation thresholds were: CFMHRT rectum (Vconventional/Vmoderatehypo): V50/V40Gy>60%, V60/V48Gy>50%, V65/V52Gy>30%, V70/V56Gy>25%, V75/V60Gy>5%. CFMHRT bladder: V50/V40Gy>50%, V60/V48Gy>25%, V74/V59Gy>15%; SBRT rectum: V18.1Gy>50%, V29Gy>20%, V36Gy>2cc; for bladder: V18.1Gy>40%, V37Gy>20cc. OAR volumes were compared by Wilcoxon signed-rank. Differences between original and reviewed OARs across the cumulative dose-volume curve were visualised by: i) rescaling dose so 100% = prescribed PTV dose [x-axis]; ii) subtracting dose bin volumes for original OAR from those of the reviewed OAR [y-axis] (forming the difference between the two cumulative DVH lines). Results For the rectum, significantly fewer original contours (12/305, 3.9%) had major variations compared to reviewed (45/305, 14.8%) (p<0.001). This was not seen for bladder original (12/305, 3.9%) vs reviewed (10/305, 3.3%) contours (p=0.50). Rectal volumes were larger for original (60.2cm 3 IQR 50.4-72.2cm 3 ) vs reviewed (53.2cm 3 , IQR 45.2-64cm 3 ) contours (p<0.001). Bladder volumes were slightly larger for original (272cm3, IQR 180-391cm 3 ) vs

Conclusion LNM occur with different frequencies in all axillary levels. The patients that underwent axillary dissection had less LNM in L I than the patients that did not undergo ALND. Further, the ALND patients had LNM in Level II, III and the internal mammary region more often. The current atlas reveals areas where LNM need to be expected after ALND. These areas need special consideration during regional node irradiation in the primary situation after ALND. PH-0601 Dosimetric Parameters Associated with Esophagitis in Regional Nodal Irradiation for Breast Cancer

Abstract withdrawn