ESTRO 37 Abstract book

S1228

ESTRO 37

macroscopic lymph nodes) with concurrent platine-based chemotherapy (if no contraindication because of general health status or renal impairment) followed by an IGABT boost. The impact of BMI was examined by considering different thresholds and examining BMI at the time of IGABT as a quantitative parameter. Results 260 pts were identified. Median age was 48.0 years, (Interquartile [IQ] 25-75: 41.0-55.3). Median follow-up was 5.0 years (IQ25-75: 4.5–5.4). Estimated 5-year overall survival and disease-free survival (DFS) rates were 72.0% (95% confidence interval [CI]: 66.1–79.0) and 63.3% (95%CI: 57.0–70.0), respectively. In univariate analysis, pts with a BMI > 25 had a poorer DFS (p=0.021) and poorer survival without pelvic nodal failure (p=0.025), as compared with normal weighted patients. Probability of pelvic nodal failure was 11.7% in pts with BMI ≤ 25 vs 19% in pts with BMI > 25 (p=0.019). A difference was seen whatever initial pelvic nodal status, but reached significance only for pts who had no pelvic nodal involvement at diagnosis (if N0: 6.6% vs 19%; p=0.037; if N+: 19% vs 31% p =0.205). No significant impact was seen for survival without local failure (p=0.382), and there was only a trend for distant failure free survival (DFFS) (p=0.059). Other BMI thresholds or quantitative analyses were not significant. In a multivariate analysis including other univariate variables with p value < 0.1 (nodal involvement, differentiation, anemia and neutrophilia at diagnosis, use of concurrent chemotherapy, and high-risk clinical target volume [CTV]), the prognostic value of a BMI > 25 remained significant in terms of DFS, with HR = 1.922 (95%CI: 1.152– 3.393), as well as for survival without pelvic nodal failure, with HR = 1.968 (95%CI 1.088 – 3.561). Conclusion BMI seems to be independent prognostic factor for DFS and survival without pelvic lymph node failure. Further studies are needed to better understand the impact of overweight and the underlying mechanisms which might be biological, as well as dosimetric, including dose distribution uncertainty in these patients, insufficient margins around nodal CTV, or loss-of-weight during treatment yielding to lymph node underdosages. EP-2222 Intrafractional applicator shifts in MRI adaptive high dose rate brachytherapy for cervical cancer C. Lapuz 1 , S. Hanna 1 , W.L. Ong 1 , A. Lim 1 1 Austin Health, Radiation Oncology, Heidelberg, Australia Purpose or Objective In some centres due to resource availability, image guided adaptive high dose rate (HDR) cervical brachytherapy requires multiple patient transfers between applicator insertion and treatment delivery, raising concerns about applicator shifts and dosimetric uncertainties in the high risk clinical target volume and organs at risk. The aim of this study was to determine the intrafractional applicator displacements during the brachytherapy process at our institution. Material and Methods Thirteen patients treated with HDR brachytherapy for cervical cancer at our institution between October 2016 and September 2017 were included in this study. All patients were treated with definitive external beam radiotherapy and 3-4 fractions of HDR brachytherapy. Applicator insertion occurred in theatre under anaesthesia, with transfer to the radiology department for magnetic resonance imaging (MRI) planning scan and final transfer to the radiation oncology department for treatment delivery. Transabdominal ultrasound (US) was performed at insertion and immediately prior to treatment to verify applicator position. Mid-sagittal and transaxial US images were obtained and distances

recorded from the central tandem to reference points on the anterior and posterior uterine surface (0, 2 and 4cm from flange and tandem tip) and from tandem tip to fundal surface. Corresponding distances were measured on the MRI planning scan. Differences between insertion US, pre-treatment US and planning MRI were analysed using paired t-test for each fraction and correlation assessed using Pearson's correlation coefficient test. Results Forty-four brachytherapy applications were evaluable. Average time from insertion to treatment delivery was 6 hours. Mean (standard deviation (sd)) differences between insertion and pre-treatment US measurements were 0.05 (3.3), 0.25 (4.02), 0.92 (3.41) and 0.11 (4.6) mm anteriorly, 0.81 (1.99), 0.05 (2.69), 0.77 (2.66) and 1.4 (4.33) mm posteriorly and 0.83 (5.0) mm tip to fundus. The difference in posterior distance at the flange was the only statistically significant value (p=0.01). Mean (sd) differences between planning MRI and pre-treatment US measurements were 2.26 (2.99), 3.15 (2.37), 3.81 (2.47) and 1.39 (2.70) mm anteriorly, 0.84 (3.04), 0.05 (2.55), 0.26 (3.52) and 1.61 (2.99) mm posteriorly and 1.06 (3.86) mm tip to fundus. Differences in anterior measurements at 0, 2 and 4 cm from the flange were statistically significant (p<0.001). There was strong correlation between MRI and pre-treatment US for all measurements, except for posterior distance at the flange where correlation was moderate (R=0.62). Conclusion Intrafractional applicator shifts from insertion to treatment delivery were minimal based on US measurements. However, the ambiguity in identifying the uterine surface on US resulted in unacceptable differences with MRI measurements. Whilst US is useful in applicator position verification, it is highly user- dependent and MRI remains the gold standard for brachytherapy planning. EP-2223 Is it necessary to repeat planningCT after first vaginal brachytherapy? Outcomes and Quality of life S. Cordoba Largo 1 , M.J. Expósito Casas 2 , V. García Jarabo 3 , D. Martínez Barrio 1 , J. Corona Sánchez 1 , A. Doval González 1 , M. Gaztañaga Boronat 1 , P. Alcántara Carrió 1 , M. Vázquez Masedo 1 , M. De las Heras 1 1 Hospital Clinico San Carlos, Radiation Oncology, Madrid, Spain 2 H. Fuenlabrada, Radiation Oncology, Madrid, Spain 3 Hospital Universitario Puerta de Hierro Majadahonda, Radiation Oncology, Madrid, Spain Purpose or Objective To evaluate clinical outcomes and quality of life (QoL) of patients receiving adjuvant vaginal brachytherapy (VB) for stage I endometrial carcinoma (EC-I) and to evaluate dosimetric differences of using either replanning prior to each fraction or a single plan, in vaginal brachytherapy (VB). Material and Methods Between 2011-2016, 65 p. with pathological EC-I were treated by VB in our department. Mean age: 62y. Stage/grade: IA/G1-2: 24p; IA/G3: 11p; IB/G1-2: 23p; IBG3: 3p. All patients were treated with 3D-VB to a total dose of 21 Gy at 5mm depth from applicator surface, in 3 weekly fractions. a) Outcomes and QoL: The effects of pathologic stage and grading on loco-regional relapse (LRR), metastases, and tumor-related death were investigated. EORTC QLQ-C30 and QLQ-EN24 questionnaires were used to evaluate vaginal toxicity and QoL. Statistics: Kaplan-Meier method, Logrank test and Chi-square test. b) Planning: A CT simulation was performed after which a treatment plan was generated. We calculated the real dose (RD) to OAR for each fraction. We compared this with the estimated dose if imaging and planning were done only on the first fraction (single plan based on a