ESTRO 38 Abstract book

S420 ESTRO 38

escalation in definitive radiotherapy for esophageal and GEJ cancer is feasible with similar normal tissue doses compared to standard treatment. A randomized dose escalation trial will be designed based on this knowledge. PO-0808 Definitive involved-field radiotherapy for esophageal cancer: are we missing the target? C. Viveiros 1 , G. Fernandez 1 , C. Pedro 1 , A. Pimenta 1 , L. Mirones 1 , M. Fortunato 1 , F. Santos 1 1 Instituto Português de Oncologia de Lisboa Francisco Gentil- EPE, Radiotherapy, Lisboa, Portugal Purpose or Objective Definitive concurrent chemoradiotherapy became the standard of care for patients with esophageal cancer not candidates to surgery. However, it is still controversial whether radiotherapy requires elective nodal irradiation (ENI) or only involved-field irradiation (IFI). The purpose of this study was to analyze failure patterns and survival Retrospective study of patients with pathologic proven esophageal cancer, not candidates to surgery, who have received IFI with curative intent, with or without concurrent chemotherapy, between January 2013 and December 2017. Comparing the target volume with radiological response, patterns of failure at the first recurrence were defined as in-field, regional out-of-field and distant failure. Response to treatment was also assessed and persistence disease was noted. Demographic data, tumor characteristics, imaging and treatment factors were recorded. Survival and disease-control outcomes were calculated from the first day of treatment with Kaplan-Meier method. Results 105 patients were included, 95% with esophageal squamous cell carcinoma and 71.4% with lymph node involvement. All radiation treatments were delivered as three-dimensional conformal radiation therapy (3D-CRT) with standard fractionation (1.8Gy fraction). 93.3% with a total dose of 50.4Gy and a median overall treatment time of 39 days. 94.3% patients received concurrent chemotherapy. With a median follow-up of 13.2 months, 33 patients (31.4%) had persistent disease, most of whom evolved with distant metastasis (11) and out-of-field recurrences (4). Amongst the patients with complete response (72), there were 5 cases of out-of-field recurrences (only 2 of which were isolated out-of-field relapses, and both were in regions that would not have been included in ENI), 9 with in-field recurrences and 22 with distant metastasis. The median dose at sites where out-of-field relapse occurred was 13.8Gy (1–44 Gy). 21 patients (20%) had mixed failure patterns on the first evidence of disease failure (persistence and/or recurrence). Median overall survival (OS) was 11.6 months (95% CI 8.2– 14.9 months). There were significant differences in OS for patients with and without disease persistence (median OS 10.2 months vs 13 months, p < 0.05). Median time to in- field, out-of-field and distant recurrences was 17.3 months (CI 95% 0–38 months), 9.7 months (CI 95% 0- 25months) and 8.2 months (CI 95% 6.8–9.7 months), respectively. in patients receiving IFI. Material and Methods

A national randomized phase II study to investigate FDG- guided dose escalation in order to improve locoregional control and survival in this patient cohort is planned. Prior to final design of the randomized study, all centres participated in a delineation workshop and a treatment planning study; data from these are presented here. Material and Methods In the delineation workshop 5 patients were delineated by participants from all enrolment centres. For the treatment planning study, a common delineation was used. A homogenous dose of 50 Gy/25 fractions (fx) prescribed to the PTV is planned as the standard arm. In the experimental arm, the dose is escalated heterogeneously to the FDG-PET avid tumour volume (defined as 50% of SUV peak ), with mean dose up to 63 Gy/25 fx for the most PET avid volumes of the primary tumour, and 60 Gy/25 fx for all malignant lymph nodes. The escalation dose is limited in favour of OAR constraints. A standard and an experimental treatment plan will be optimized for each patient prior to randomization. Mean dose to lungs and heart in the experimental plan must be kept within ±1Gy corresponding doses in the standard plan. Results Data from the target delineation workshop is presented in Table 1. Consistency of the delineated GTV-T, GTV-N and CTV volumes were acceptable. The two patients with the smallest target volumes had the largest relative range. In the treatment planning study, the dose escalated FDG- PET avid part of tumour (PET GTV-T ) constituted in median 45% (range 32% - 57%) of the GTV-T and received an average mean dose of 62.3 Gy. This resulted in an average mean dose to the GTV-T of 59.9 Gy. For the GTV-N, the average mean dose was 59.6 Gy. The combined clinical target volume (CTV-total) received an average mean dose of 55.3 Gy. Normal tissue doses were similar for the experimental and standard plan, with a difference in mean lung dose and mean heart dose in the two arms of 0.1 Gy and 0.2 Gy, respectively. The maximum dose for the standard plans was 55 Gy (110%). Higher doses up to 67 Gy were applied for the experimental plans, but only to small volumes respecting the strict normal tissue constraints (see figure)

.

Conclusion Data from a target delineation workshop and a pilot treatment planning study suggest that FDG-guided dose