ESTRO 37 Abstract book

S791

ESTRO 37

EP-1453 Defining patterns of failure in neoadjuvant radiotherapy for borderline resectable pancreatic cancer G. Murphy 1 , L. Keenan 1 , R. Rahman 2 , G. McVey 1 1 St Luke's Radiation Oncology Network- Dublin, Radiation Oncology, Dublin, Ireland 2 St Vincent's University Hospital, Medical Oncology, Dublin, Ireland Purpose or Objective To evaluate patterns of failure and survival in patients following neoadjuvant radiotherapy for borderline resectable pancreatic cancer (BRPC). Material and Methods We performed a retrospective analysis of all patients with BRPC who were treated at our institution with radiotherapy from February 2012 to July 2016. Recurrence and survival data were analysed. Recurrences were mapped onto each patients planning CT scan. Progression of disease (POD) was classified as in-field, field edge, out of field or distant recurrence. Results Eighty-four patients were included in the analysis. Patients either received short course (39%, n=33) or long course radiotherapy. The majority of patients received induction chemotherapy (93%). The majority of patients receiving short course radiotherapy proceeded to surgery (51%, n=17), while fewer who received long course radiotherapy had surgery (23%, n=12). Median follow-up was 11.8 months. Disease recurrence data was available in 82% (n=69) of patients. There was POD in 91% (n=58) of patients. Distant disease was the most frequent first site of treatment failure (47% n=33). 36%, (n=25), of patients had in-field failure. Four patients, (5.8%), had synchronous distant and locoregional failure. Of those patients who underwent surgical resection post neoadjuvant RT, 48% (n=12) had distant failure, 24% (n=6) had in field failure, 8% (n=2) had synchronous distant and in field failure and 20% (n=5) had no evidence of treatment failure at last follow-up. Median progression free survival was 6.2months (95% CI: 5.0-7.3). Median Overall Survival was 13.3 months (95% CI: 11.4–15.2). Conclusion Regional recurrence rate remains high and it is therefore important to accurately define a clinical target volume. Defining the rate of in-field, field edge and out of field recurrences will aid in improving target delineation. EP-1454 Is MRI necessary before performing liver metastases SBRT? D. Romanov 1 1 FSBI "N.N. Blokhin's NMRC of oncology" Russian Ministry of Heath, Radiation Oncology, Moscow, Russian Federation Purpose or Objective The aim of the study was to evaluate the effect of detecting new metastases in the liver after SBRT on the life expectancy of patients. Material and Methods We analyzed the overall survival rates of 80 patients after SBRT of metastases in the liver and their dependence on various factors. In 41 patients after SBRT, new liver metastases were detected. However, the detection of new metastases alone did not significantly affect life expectancy: the median in these patients was 27 months compared to 31 months in patients without the appearance of new metastases in the liver during the entire follow-up period (p=0,397). The amount of time

between the end of SBRT and the appearance of such metastases had a significant effect on life expectancy: median for patients, new metastases in the liver appeared no later than 3 months after SBRT - 7 months, later - 31 months (p = 0.01333); similarly for 6 months: medians: 10 and 35 months (p = 0.04688). Results The probability of occurrence of new metastases in the liver for 3 months was significantly associated with non- CRC histological types of the tumor (9.4% among patients with CRC and 28.6% among patients with non-CRC, p<0.04). The increase in the likelihood of new metastases in the liver within 6 months after SBRT was also significantly associated with the presence of tumor deposits outside the liver (primary tumor or distant metastases) at the time of the SBRT (20% in the group of patients, the spread of the tumor process at the time of treatment limited to the liver and 41.2% in the group of patients who at the time of the treatment had already detected extrahepatic deposits of the disease, p <0.05). In CRC patients who had not performed such diagnostic studies as MRI and PET-CT during the month of the month when the SBRT started, the probability of detecting new liver metastases in the near future after SBRT was significantly higher. Thus, within 3 months after SBRT of metastasis of CRC in the liver, the risk of detecting new metastases in the liver was significantly lower among patients who were undergoing MRI (p = 0.017) or MRI / PET-CT (p = 0.04): 0% and 0% in comparison with 50% among patients who had only CT scan at the stage of the examination. Within 6 months after SBRT, the risk of detecting new metastases in the liver was also significantly lower among patients who were undergoing MRI (p = 0.01) or MRI / PET-CT (p = 0.012): 9% and 12% with 67% among patients who had only CT scan. Conclusion We found that the detection of new metastases in the liver for 6, and especially 3 months after SBRT, significantly decreases the life expectancy of patients. Significantly more often at these times of observation, new liver metastases were detected in those patients who had not performed an abdominal MRI with contrast for a month before SBRT. It should be assumed that in such patients the volume of the examination, including CT with contrast, did not allow to detect small metastases that became visible in the control examination after SBRT. So, the abdominal MRI with contrast is necessary before deciding to conduct a metastasis in the liver SBRT. EP-1455 Interobserver and interfraction variability using free-breathing cbCT for SBRT of the upper abdomen L. Nicosia 1 , T. Brunner 2 , I. Popp 2 , E. Gkika 2 , F. Capilli 3 , H.C. Rischke 2 , N. Volegova-Neher 2 , A. Thomsen 2 , I. Strouthos 2 , A.L. Grosu 2 , S. Kirste 2 1 University of Rome "Sapienza", Radiation Oncology, Roma, Italy 2 University of Freiburg, Radiotherapy, Freiburg, Germany 3 University of Freiburg, Radiology, Freiburg, Germany Purpose or Objective To evaluate the interobserver and interfraction setup error with the use of cbCT in stereotactic treatment of tumors of the upper abdomen. Material and Methods cbCT data sets from 10 patients who completed SBRT for tumors of the upper abdomen were chosen. Matching data of each cbCT from 11 observers (7 radiation