Abstract Book

ESTRO 37

S422

Purpose or Objective EMBRACE-II is a prospective interventional and observational multicentre study of IMRT and MRI-based IGBT in cervix cancer. The study involves several interventions which imply a change of practice for many centres. An online continuing medical education (CME) programme was therefore implemented to support radiotherapy quality assurance (RTQA). This abstract evaluates the CME programme for IMRT target volume delineation (TVD). Material and Methods The RTQA and CME programmes were hosted on an open- source learning management system which linked to a bespoke online contouring tool. The RTQA programme comprised a compliance questionnaire, evaluation of IMRT target/OAR contouring, IMRT dose planning evaluation and IGBT contouring evaluation for selected centres. CME content included questionnaires on previous experience, practice cases for contouring and dose planning, quizzes on key aspects of the protocol and an IMRT contouring atlas. For IMRT TVD accreditation, the PI from each centre was required to contour on one case. The ROIs assessed were GTV, CTV-HR, CTV-LR, CTV-E, CTV-N and ITV45. ROIs were scored as 2 (excellent), 1 (fair) or 0 (revision required). A total score of ≥9 (out of 12) was needed to pass. Results To date, 114 participants (all staff groups) from 49 centres in 19 countries have registered for the CME programme. 18 oncologists completed the questionnaire on previous experience; all were using IMRT in clinical practice, but only 7 (39%) had experience with the ITV concept. The most commonly reported areas of TVD difficulty were vagina (6=33%), ITV (6=33%) and parametrium (4=22%). 5 (28%) oncologists reported ‘no particular difficulty’ with TVD. 45 oncologists (including 6 non-PIs) from 39 centres submitted contours for evaluation. 21 (47%) passed at the first attempt and a further 14 (31%) after individualised feedback. The ROIs that received the lowest scores were CTV-E (0.9) and ITV45 (1.0) [Figure 1].

guide. Further work is underway to identify effective learning activities which are time-efficient. PO-0810 Multimodality functional imaging and chemoradiotherapy outcome in cervical cancer K. Skipar 1 , T. Hompland 1 , K. Vassmo Lund 2 , E. Nakken 3 , K. Bruheim 3 , H. Lyng 1 1 Oslo University Hospital, Department of radiation biology, Oslo, Norway 2 Oslo University Hospital, Department of radiology, Oslo, Norway 3 Oslo University Hospital, Department of oncology, Oslo, Norway Purpose or Objective Cervical cancer patients undergo multimodality imaging, including diffusion weighted magnetic resonance imaging (DW-MRI) and fluorine 18-fluorodeoxyglucose positron emission tomography (18F-FDG-PET) for appropriate staging and treatment decision in clinical practice. The prognostic potential and biological information available from these imaging modalities have not been fully explored. In this study we investigated the correlation between the apparent diffusion coefficient (ADC) derived from DW-MRI and the standardized uptake volume (SUV) from 18F-FDG-PET relative to patient outcome. Material and Methods We identified fifty-six patients with cervical c ancer FIGO IB-IIIB, who had received definitive chemoradio- therapy in 2014-15. All patients underwent pretreatment DW-MRI and 18F-FDG-PET as part of the diagnostic work- up. The primary tumour, defined as the region of interest (ROI), was contoured manually on both imaging modalities. ADC (reflecting cellular density) and SUV (reflecting metabolic profile) were calculated for each tumour voxel and presented in histograms. The tumour voxel with the highest SUV value was referred to as SUVmax. Explorative correlation with PFS was searched for by percentile screening of the entire histogram, using the log-rank test. The clinical endpoint was progression free survival (PFS). Results Eight patients relapsed. Median follow up time for the entire cohort and the surviving patients was 32.7 months and 35.7 months, respectively. SUV correlated significantly with PFS. For the highest SUV percentile (SUV Max ) a difference of 26 percent in PFS probablility was found (p=0.012). Twenty-five patients had PET positive lymph nodes in the iliac and/or paraaortic region, and we found a significant higher SUV in the primary tumour in this patient group (p= 0.016 for SUV Max ; Student t-test). When the same analyses were performed for ADC, no correlation with outcome or presence of PET positive lymph nodes was found. However, we found a significant negative correlation between ADC and SUV (r=-0.28, p=0.03 for SUV Max versus ADC Median , Spearman's correlation analysis). Conclusion The 18F-FDG-PET parameter SUV may provide prognostic information in cervical cancer patients. The DW-MRI parameter ADC showed no prognostic value when analysed separately, but may be important in combi- nation with other imaging parameters. Our findings indicate that the metabolic activity of the tumour is of greater importance than cell density. PO-0811 Patient-reported quality of life in cervical cancer patients treated with definitive chemoradiation J.L. Conway 1 , S. Felder 1 , J. Tang 1 , A. Fyles 1 , M. Milosevic 1 , J. Lukovic 1 , K. Han 1 , J. Croke 1 1 Princess Margaret Cancer Centre, Radiation Oncoogy, Toronto, Canada

An analysis of user engagement with CME content was carried out. All 45 oncologists accessed the contouring atlas multiple times. 28/45 (62%) accessed the practice cases, and 16/28 (57%) submitted contours. 14/45 (31%) accessed the quizzes, and 8/14 (57%) completed them. The quiz question receiving the lowest score (42%) was also on the CTV-E. Conclusion Our study has shown the potential of a CME programme to support clinical trial participants (PIs and non-PIs). Difficulty with the ITV was anticipated as it was an unfamiliar concept. Difficulty with the CTV-E was more surprising, and change in practice to a risk-adapted volume may have contributed. Uptake of practice cases was moderate, probably because of the time necessary to contour. Quizzes may be a quick way of identifying areas of difficulty and reinforcing key aspects of the protocol. Analysis of participant responses has led to development of targeted content, including a detailed ITV step-by-step