ESTRO 38 Abstract book

S873 ESTRO 38

provide heterogeneous outcomes due to group survival estimation. The purpose of this study was to develop and internally validate an individualized predictive model for patients with brain metastasis who underwent WBRT. Material and Methods Based on a series of 178 retrospectively analyzed patients who underwent WBRT, a multivariable piecewise Cox regression model was developed. Individualized survival estimates at 12-month, 6-month, and 1-month were generated using the most parsimonious model. Model's predictive ability was estimated by the harrel's C statistics corrected for model optimism following a bootstrap strategy. Calibration was evaluated by comparing predicted and observed (Kaplan Meier) survival estimates obtained from bootstrap resamples. Results The majorities of patients had lung cancer, ECOG-PS score ≥ 2, metastatic spread, and multiple BMs and were in Recursive Partitioning Analysis (RPA) score II or had a Diagnosis-Specific Graded Prognostic Assessment (DS-GPA) score of 0.5–2. Median overall survival was 6.9 month in patients with a breast cancer, 2.9 month in patients with lung cancer and 3.8 month in patients with other primary site. The final multivariable regression model was based on: primary site, metastatic spread, age at diagnosis, ECOG-PS and intracranial hypertension signs. Comparison with DS-GPA and RPA scores showed that individualized survival estimates outperformed RPA and DS-GPA scores at 1-month, 6-months and 12- months. Conclusion Our model provides accurate individualized estimates of survival, outperforming actual grouped estimates. A nomogram and shiny application were developed for individualized prediction with our model, to provide easy- to-use tools for clinical practice. Our model could be useful for assisting rational prescription of WBRT, especially for survival limited patients. External validation of our model is recommended before use in clinical practice. EP-1621 First results of the first cohort of a phase I dose-escalation trial on SABR for oligometastases C. Mercier 1 , P. Dirix 1 , C. Billiet 1 , P. Meijnders 1 , P. Vermeulen 2 , C. Rypens 2 , P. Huget 1 , D. Verellen 1 1 Iridium Cancer Network, Department of Radiotherapy Oncology, Antwerp, Belgium ; 2 Oncological Centre GZA, Translational Cancer Research Unit TCRU, Antwerp, Belgium Purpose or Objective Stereotactic ablative body radiotherapy (SABR) is an emerging treatment option for oligometastatic cancer. However, limited prospective evidence is available. We report on the primary and secondary endpoints of the first cohort of a phase I dose-escalation trial on SABR for non- spine bone and lymph node oligometastases. Material and Methods In a prospective clinical trial, patients with 3 or less metastases on functional imaging received SABR in 5 fractions on all lesions. Primary endpoint was toxicity at 6 months after SABR. Secondary endpoints were local control (LC, defined through repeated functional imaging at 6 months), progression-free survival (PFS, defined as clinical or biochemical progression or death from any cause), and quality of life (QoL). Kaplan-Meier methods were used to determine LC and PFS. Toxicity was graded using Common Terminology Criteria for Adverse Event version 4.0. QoL was assessed using European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 at inclusion, as well as at 3, 6, 9 and 12 months after SABR. Results From July 2017 to December 2017, 30 consecutive patients received SABR to a total of 32 oligometastases. One extra patient was excluded in the screening period because of a technical unsuitable location (mediastinal lymph node in

on the left and 3 on the right. Two patients were irradiated bilaterally and one pulmonary metastasis was simultaneously irradiated in one patient. 75% of patients have lung cancer (50% of them - squamous cell carcinoma). From 4D CT gross tumor volume (GTV) is delineated in different breathing phases. From the combination of GTVs the internal target volume (ITV) was determined. The planning target volume (PTV) includes an additional margin of 3-5mm. PTV volumes range from 14.6 to 59.7cc. The treatment plans are performed in one or in five fractions (4 patients for both groups). The prescribed doses at 92% are – for one fraction 18 to 21 Gy and for five 30 to 40Gy. Conformity index (CI), non-homogeneity index (NHI) and gradient index outside the PTV (R50%) are defined. The organs at risk such as spinal cord, adjacent and contralateral kidney, intestine, aorta, liver and pancreas are evaluated from dose-volume histograms. Radiotherapy is done on tomotherapy machine and in one of the cases оn conventional linear accelerator. Patients who are treated in one fraction on tomotherapy are made on two pass to make verification of positioning after each exposure of 7 to 10 minutes. Results The results of dosimetry plans for all patients are shown in Table 1. It is obvious that the CI and NHI are independent of the PTV volume and the mean values are as follows: CI - 1.25 (SD=0.09); NHI – 1.35 (SD=0.07). The gradient index outside the PTV varies from 3 to 4.6 for all plans with the exception of the one of the patient with bilateral adrenal glands and pulmonary metastasis where R50% increases to 5,8. Regarding the organs at risk only the values for pancreas and intestine in the plans of patients irradiated bilaterally exceed our criteria. The patients are followed up for 2-12 months (average 5.5 months). One of them died of progression of the primary tumor and the remaining 7 (87.5%) are alive. In 5 of them a control study was performed 3 months after irradiation - CT with a contrast of chest and abdomen or PET / CT. The control study indicates that there is a complete local control after the treatment in 2 patients, in 1 patient there is a partial response of 30% and in the other one there is no dynamic of the lesion in the adrenal gland. There are no side effects and progression in all patients.

Conclusion Our experience has shown promising results for SBRT treatment of adrenal gland metastasis with the given criteria for evaluation of the dosimetry plan (for tumor and organs at risk). The SBRT treatment provides good tumor control and a lack of acute and late toxicities. EP-1620 A model for individualized estimation of survival in patients who underwent whole-brain radiotherapy C. Marchand-Créty 1 , J. Riverain 2 , Y. Drouet 3 , J. Thariat 2 , S. Servagi-Vernat 1 1 Institut Jean Godinot, Radiation therapy, Reims, France ; 2 Centre François Baclesse, Radiation Therapy, Caen, France ; 3 Centre Léon Bérard, Statistics, Lyon, France Purpose or Objective Historic palliative role of whole brain radiation therapy (WBRT) is questioned in management of survival limited patients with multiple brain metastases (BMs). Most scores