ESTRO 37 Abstract book

S198

ESTRO 37

Purpose or Objective To evaluate the evolution of external beam radiotherapy (EBRT) practice between EMBRACE I and the in itial phase of EMBRACE II and the consequences on irradiated volumes. Material and Methods The international prospective EMBRACE I study enrolled 1416 locally advanced cervical cancer (LACC) patients from 23 institutions from 2008 to 2015. EBRT was delivered as IMRT or conformal radiotherapy (CRT), and dose prescription ranged from 45-51Gy/25-30fx to the elective target. PTV margins were according to institutional practice. The elective target included pelvic, ±para-aortic (PAN), and ±inguinal Lymph Nodes (LN). In 1270 patients treated with pelvic±PAN irradiation (patients with inguinal irradiation were excluded), PTV volume and V43Gy were available. Among them, 448 patients with LN boosting had nodal PTV (PTV-N) volume and/or V57Gy available. Differences in PTV volume across centres were compared for centres which enrolled >15 patients: 18 centers for pelvic and 5 for PAN. The EMBRACE II study was initiated in 2016 and 3 centres enrolled 68 patients by 10-2017. EMBRACE II implies common contouring guidelines, IMRT, daily IGRT and a joint dose prescription protocol including 45Gy/25fx for the elective volume (pelvic±PAN) and 55/57.5Gy for simultaneously integrated LN boost with coverage probability planning. A 5mm PTV margin and a PTV constraint of V95%>95% were required. PTV and V43Gy were reported in 50 patients, V50Gy and PTV-N in 29 boosted patients. Results EMBRACE I: CRT was used in 60% of patients and IMRT in 40%. For pelvic irradiation (84%), median PTV (first - third quartile) was 1549 (1312-1807) cm 3 and V43Gy 2390 (2028-2799) cm 3 and for PAN patients (16%), 1924 (1663- 2175)cm 3 and 2909 (2528-3462)cm 3 , resulting in V43/PTV of 1.54 and 1.51, respectively. For pelvic patients treated with IMRT, V43/PTV was 1.31. Median institutional PTV varied at most by 745 for pelvic and 588cm 3 for PAN patients. In pelvic patients, median V43Gy was 513cm 3 larger with CRT as compared to IMRT, and 357cm 3 larger with ≥48Gy as compared to 45Gy, as showed in Figure 1. Median V57Gy for boosted patients was 98cm 3 and PTV-N 57cm 3 . EMBRACE II: For pelvic irradiation (66%), median PTV was 1350 (1206-1424) cm 3 and V43Gy 1409 (1278-1544) cm 3 and for PAN patients (34%), median PTV was 1803 (1735- 2038)cm 3 and V43Gy 1935 (1794-2109)cm 3 , resulting in V43/PTV of 1.04 and 1.07, respectively. Median institutional PTV varied at most by 217cm 3 for both pelvic and PAN patients. For boosted patients, median V50Gy was 67cm 3 and PTV-N 43cm 3 .

Conclusion Adherence to the EMBRACE II protocol improved inter- institutional consistency, decreased irradiated volumes and improved conformality considerably. By application of joint contouring guidelines, IMRT and daily IGRT, V43Gy was reduced by >950cm 3 for pelvic and PAN patients. These encouraging results need to be confirmed through future patient accrual. The overall decrease in irradiated volumes is expected to lower the incidence and intensity of EBRT related complications. OC-0392 Impact of image-guided brachytherapy on pattern of relapse in the RetroEMBRACE cervical cancer study L.T. Tan 1 , K. Kirchheiner 2 , A. Sturdza 2 , L. Fokdal 3 , C. Haie-Meder 4 , P. Hoskin 5 , I.M. Jürgenliemk-Schulz 6 , J. Lindegaard 3 , C. Kirisits 2 , K. Tanderup 3 , R. Pötter 2 1 Cambridge University Hospitals, Oncology Centre, Cambridge, United Kingdom 2 Medical University of Vienna, Department of Radiation Oncology, Vienna, Austria 3 Aarhus University Hospital, Department of Oncology, Aarhus, Denmark 4 Institut Gustave Roussy, Department of Radiation Oncology, Villejuif, France 5 Mount Vernon Cancer Centre, Department of Oncology, Northwood, United Kingdom 6 University Medical Center Utrecht, Department of Radiation Oncology, Utrecht, The Netherlands Purpose or Objective The RetroEMBRACE study has previously shown that IGBT for cervical cancer improves pelvic control to 87% and overall survival by ~10%. This present analysis investigates the impact of IGBT on patterns of failure. Material and Methods Data from 731 patients treated in 12 institutions were analysed. All patients received initial EBRT - 84% to the pelvis alone and 16% to the pelvis and paraaortic nodes (PAN). 77% patients received concomitant chemotherapy. MRI-based IGBT was used in at least one application in 81% of patients, and CT alone in 19%. In 23% of patients, combined intracavitary/interstitial BT was used in at least one fraction. Treatment failures were classified as local (cervix, upper vagina and/or parametria), regional (pelvic nodes), pelvic (local and/or regional), systemic (excluding PAN relapse) or distant (including PAN relapse). The pattern of failure at time of first relapse was analysed and compared with two large published series treated with RT + conventional BT without