ESTRO 36 Abstract Book

S366 ESTRO 36 _______________________________________________________________________________________________

follow-up of 20 months, the median PFS was 20 months. One-year, two-year PFS and three-year PFS were 70%, 53% and 42%, respectively. Two-year and three-year LC were 95% and 79%, respectively. Two-year and three-year MFS were 53% and 42%, respectively (median, 20 months). Median OS was 19.2 months.One-year OS, two-year OS and three-year OS were 77%, 47% and 37%, respectively. The median OS for borderline resectable patients was 21.5 months compared with 14 months for unresectable patients (p=0.3). Patients who underwent resection had a significantly longer median OS compared with non- resected patients (37.6 months vs 13 months, p=0.03). Conclusion This protocol treatment represents a well-tolerated promising approach for patients with borderline resectable and unresectable pancreatic cancer. Continued optimization in multimodality therapy and an accurate patient selection are crucial for the appropriate treatment of patients. PO-0699 Is stereotactic radiotherapy following radiochemotherapy useful in local advanced pancreatic cancer? G. Mattiucci 1 , A. Nardangeli 1 , L. Boldrini 1 , M. Balducci 1 , F. Cellini 1 , S. Chiesa 1 , G. Chiloiro 1 , F. Deodato 2 , N. Dinapoli 1 , V. Frascino 1 , M. Gambacorta 1 , G. Macchia 2 , A. Morganti 3 , V. Valentini 1 1 Università Cattolica del Sacro Cuore- Gemelli ART, Radiation Oncology, Rome, Italy 2 Fondazione “Giovanni Paolo II”- Università Cattolica del Sacro Cuore, Radiotherapy Unit, Campobasso, Italy 3 Università di Bologna- Ospedale S. Orsola-Malpighi, Radiation Oncology Center- Department of Experimental- Diagnostic and Specialty Medicine – DIMES, Bologna, Italy Purpose or Objective To evaluate the feasibility and efficacy of stereobody radiotherapy (SBRT), following radiochemotherapy (RTCT) and chemotherapy (CT), in patients (pts) with unresectable, locally advanced pancreatic carcinoma (LAPC). Primary endpoints were toxicity, local control (LC) and pain-free progression (PFP); secondary endpoints were overall survival (OS) and disease-free survival (DFS). Material and Methods Patients affected by unresectable LAPC already treated with RTCT (50.4 Gy in 28 fractions (fr) to visible pancreatic tumour and 39.6 Gy in 22 fr to nodal drainage area with concurrent gemcitabine) and chemotherapy (Gemcitabine or Folfirinox), with no evidence of metastatic disease at the restaging imaging, were selected for a SBRT boost on the primary lesion. The pain assessment was defined by Kersh-Hazra scale. Results From 2003 to 2015, 26 consecutive pts (16 male, 10 female), with a median age of 65,5 years (range 47-80), were enrolled in this study. The 53,8% was a cT4 and the 50% was a N1. The tumor was localized in the head of the pancreas in 53.8% of pts. The SBRT boost was delivered to the primary lesion with a total dose depending on the dose received by the duodenum (maximum dose to the duodenum 90Gy in EQD2 α/β 2 summing the dose received during RT-CT and SBRT): 20Gy in 5 fr for 4 pts, 20Gy in 4 fr for 5 pts, 25Gy in 5fr for 16 pts and 30Gy in 6fr for 1pt. The median follow up was 25 months (range 15-154). The median interval between RTCT and SBRT was 8 months (range 3-16 months). None Grade 3 or 4 acute or late gastrointestinal toxicities were developed among all pts after SBRT. Pain control was achieved in 19 pts (73,1%) after SBRT boost. After SBRT: 2-years and 3-years LC were 62.4% and 41.6%, with a median of 36 months; 2-years and 3-years PFP were 64.3% and 32.1%, with a median of 25 months; 2-years and 3-years DFS were 27.8% and 18.5%, with a median of 7 months. The 2-years and 3-years OS after SBRT were 57% and 42.7%, with a median of 29

months. Since diagnosis: 2-years and 3-years LC rate were 71.1% and 64.6%, with a median of 45 months; the 2-years and 3-years PFP were 86.3% and 49.3%, with a median of 35 months; 2-years and 3-years DFS were 45.8% and 29.7%, with a median of 23 months; 2-years and 3-years OS was 77.3% and 58%, with a median of 41 months. Conclusion A SBRT boost on primary lesion after RTCT and CT could be delivered safely and effectively in pts with non- metastatic, unresectable LAPC with acceptable side effects and with promising local and pain control. PO-0700 Significant heart dose reduction by deep inspiration breath hold for RT of esophageal cancer M. Dieters 1 , J.C. Beukema 1 , A.C.M. Van den Bergh 1 , E.W. Korevaar 1 , N.M. Sijtsema 1 , J.A. Langendijk 1 , C.T. Muijs 1 1 UMCG University Medical Center Groningen, Radiation oncology, Groningen, The Netherlands Purpose or Objective As survival for esophageal cancer (EC) patients improves 1 , reduction of long term radiation-induced toxicity will become increasingly important. For radiotherapy of left- sided breast cancer patients, deep inspiration breath-hold is used routinely to minimize the radiation dose to the heart 2 . For EC patients, the expiratory phase might be more beneficial to reduce the heart dose, while the inspiration phase might be better for the dose to the lungs, consequently allowing for cardiac dose reduction. Therefore, the main objective of this study was if breath hold in photon radiotherapy of esophageal cancer minimized the dose to the heart, without compromising the dose to the lungs and the target. References: 1. Shapiro J, van Lanschot JJB, Hulshof MCCM, et al. Lancet Oncol. 2015;16:1090–1098. 2. Sixel KE, Aznar MC, Ung YC. Int J Radiat Oncol Biol Phys 2001;49:199–204. Material and Methods Ten EC patients were included in this in prospective cohort study. All patients received a free breathing (FB) 4D- plannings-CT and additionally a CT-scan in deep inspiration (DIBH) and expiration breath-hold (EBH) using Active Breathing Control (ABC). Treatment volumes and organs at risk were delineated. No ITV margin was used in the breath-hold CTs assuming absence of respiratory movement. Full VMAT treatment plans (3 arcs, per arc <20sec delivery time) were constructed on all 3 planning CTs (FB, DIBH and EBH) with the aim to cover the target with a prescribed dose of 41.4Gy or 50.4Gy, while reducing the cardiac dose, without compromising the dose to the lungs. These plans were compared to the clinically robust partial VMAT /IMRT treatment plans (clinFB), for volume and DVH differences using one way ANOVA with Bonferroni correction. Results Breath-hold, both DIBH and EBH, was feasible in all patients. The GTVs were similar on all 3 CT-scans(p=0.99). With DIBH, lung volumes were significantly larger (on average 2800 cc) than with FB andEBH (p<0.01). The mean heart dose (MHD) and V30 heart were also significantly different among the4 types of treatment plans (p=0.02 and p<0.01) (table 1). The reduction of the MHD and V30 heartwere most pronounced using DIBH and was consistent over the entire range of MHD/MLD, asillustrated by the example in figure 1.b. On average, the MHD and V30 heart reduced from 22Gy inthe clinFB plans to 13.8Gy in the DIBH plans (mean difference 8.2Gy, 95%CI 1.2-15.3) (figure 1.a). Thereduction in cardiac dose can be explained by: 1) Use of full VMAT instead of partial VMAT/IMRT, 2)the absence of ITV margin when using BH, 3), an increase of lung volume, which allows cardiac dosereduction, corresponding to average MHD reductions of 2.8 Gy (1), 2.2 Gy (2) and 3.2 Gy (3),respectively (figure 1.c).