ESTRO 36 Abstract Book

S376 ESTRO 36 _______________________________________________________________________________________________

Results Three hundred forty-one patients were treated by IMRT for gynecological or anal cancer between 2007 and 2014. Fifteen patients had at least one pelvic fracture occurred after external radiotherapy, an overall incidence of 4.4%. The age and menopausal status were linked to an increased fracture risk (p = 0.0274 and p <0.0001, respectively). The site of the primary tumor (gynecological or anal canal) was not related to an excess risk of fracture. The median maximum dose received at the fracture site was 50.3 Gy (range, 40.8-68.4). Conclusion The incidence of pelvic fracture after IMRT was low but more important after age 50 and postmenopausal patients. The pre-therapeutic assessment of bone density by densitometry could be relevant for these patients. PO-0717 Role of Radiation Therapy in Vulvar Cancer Patients with One or More Positive Inguinal Lymph Nodes M.J. Kanis 1 , N. Rydzewski 2 , E.D. Donnelly 2 , J.R. Lurain 1 , J. Strauss 2 1 Northwestern University, Gynecologic Oncology, Chicago, USA 2 Northwestern University, Radiation Oncology, Chicago, USA Purpose or Objective Using a large national cancer database, we aimed to investigate outcomes for women with one or more than one positive inguinal lymph nodes in squamous cell vulvar cancer who were treated with external beam radiation therapy (EBRT) compared to patients that received no adjuvant radiation. Material and Methods The National Cancer Database (NCDB) was queried to identify women with vulvar cancer that had one or more positive lymph nodes diagnosed between 2004 and 2012. All patients were surgically staged. Chi-square tests and multivariate logistic regression were performed to analyze factors associated with receipt of radiation. Survival analysis was performed using overall survival confidence intervals (CI), log-rank test, Kaplan-Meier estimates, and Cox proportional hazards regression. Results

Significant predictors of receiving EBRT over no radiation included age younger than 70 and a closer distance to the hospital. 5-year overall survival (OS) for the entire cohort was 40% (95% CI, 38%-42%). 5-year OS for 1 node positive was 47% (95% CI, 42%-52%) for no EBRT and 53% (95% CI, 49%-57%) in those receiving EBRT. 5-year OS for 2 or more nodes positive was 19% (95% CI, 15%-23%) for no EBRT and 35% (95% CI, 32%-38%) for EBRT. Kaplan-Meier estimates with log-rank test for equality of survivor functions showed improved survival with EBRT in 1 node positive (p=0.0129) and 2 or more nodes positive (p<0.0001). Cox survival multivariate regression model controlling for confounding variables observed a Hazard Ratio (HR) for EBRT compared to no radiation of 0.88 (p=0.122) for 1 node positive, and 0.58 (p<0.001) for 2 or more nodes positive. HR without covariates for 1 node positive was 0.82 (p=0.008). Conclusion In a large national cancer database, receipt of EBRT was associated with improved survival in woman with vulvar cancer and two or more positive inguinal nodes. In patients with one positive node, a survival advantage was seen using Kaplan-Meier estimates and Cox survival model, but lost statistical significance in a multivariate regression model. These data lend support to the role of EBRT in the management of node positive vulvar cancer. Further research with more detail of other known prognostic factors are needed to validate these findings. PO-0718 18-FDG PET/CT parameters to predict survival and recurrence in cervical cancer N. Scher 1 , F. Herrera 1 , A. Depeursinge 2 , T. Breuneval 1 , J. Bourhis 1 , J. Prior 1 , M. Ozsahin 1 , J. Castelli 1 1 Centre Hospitalier Universitaire Vaudois, Radiation Oncology department, Lausanne Vaud, Switzerland 2 Ecole Polytechnique Fédérale, imaging, Lausanne Vaud, Switzerland Purpose or Objective In the context of locally advanced cervical cancer (LACC) treated with chemo-radiotherapy, the aim of this study was to identify the best predictive 18-FDG PET-based parameters for local-control, disease free- and overall survival, testing different threshold to compute MTV and TLG. Material and Methods Thirty-seven patients treated with standard chemo- radiotherapy followed by brachytherapy underwent a pre- therapy 18-FDG PET/CT. Using different thresholds (from 2.5 to 8 mg/mL and from 30% to 70% of SUVMax), the following PET parameters were computed: maximum standardized uptake value (SUVmax), mean standardized uptake value (SUVmean), metabolic tumor volume (MTV) for primary tumor and lymph nodes, total lesion glycolysis (TLG), and a new parameter combining the MTV and the Euclidian distance between lymph nodes and the primary tumor, namely metabolic nodes distance (MND). Correlation between PET and clinical parameters with clinical outcome (OS, DFS and LRC) was assessed using univariate and multivariate Cox-model. An internal validation of the final model was performed using a cross validation with 5 folds. Results The median follow-up was 52 months (range: 7 - 128). The 3-year OS, DFS and LRC were 71.2 % (95% Confidence Interval (95CI): 56%–86%), 64.1 % (95CI: 48–80) and 69.4 % (95CI: 53 –84) respectively. In univariate analysis, PET/CT parameters associated with OS and DFS were: MTV Tumor, TLG Tumor, TLG Lymph Nodes, and MND. The most predictive threshold segmentation for MTV and TLG was 48 % of SUV max for the primary tumor and 30% for the lymph nodes. In multivariate Cox analysis, the TLG T 48% and MND were the two independent risk factor for OS (p<0.01), DFS (p<0.01) and LRC (p=0.046). The c-index of the model for OS, DFS and LRC (adjusted after cross

Of 2,859 patients identified, 1,024 (36%) received no adjuvant radiation compared to 1,835 (64%) who received EBRT. Mean number of regional nodes examined was 12.7 for the no radiation group and 12.4 for the EBRT group.