ESTRO 2020 Abstract Book

S546 ESTRO 2020

oncological outcomes in patients with BMs from NSCLC, treated with SRS and TKIs or IO. Material and Methods A total of 195 patients with synchronous and metachronous BMs from NSCLC treated with CyberKnife radiosurgery (CKSRS) and concomitant pre- or post-systemic therapy were analyzed. Exclusion criteria included prior wall brain radiotherapy (WBRT), surgery removal of BMs, prior SRS on the same lesion or insufficient follow-up. The primary endpoint was distant brain failure (DBF). DBFs were defined as any new metastases that developed outside of the previous radiosurgical target volume. Time to DBF was calculated from the start of SRS to DBF. Overall Survival (OS) was the secondary endpoint. We stratified our cohort of patients with Modified Lung Specific Prognostic assessment (GPA) in three prognostic cathegories and subsequently the impact of mutation on OS was tested for each category. Results Median time free from DBF for our patients cohort was 15.3 months (95% CI 11.6-20.4). No statistically significant differences were found between the mutated and not mutated patients (p-value 0.414). For the cohort of patients with the PDL-1 expression (N= 41), no statistically significant differences were found in DBF (p-value 0.371). Considering free from DBF at 12 months (Fig.1), we observed a better trend for patients with expression of PDL-1>1%, 50.6% (95% CI 30.1-68.0), compared to patients without PDL-1 expression (<1%), 31.2% (95% CI 8.1- 58.2). Median OS was 19.9 (95% CI 15.1-27.4) months; no statistical significant differences in OS were noted in patients with targetable mutations (p-value 0.110), even if there was a trend in favor of mutated patients. For patients included in prognostic GPA<2 category we obtained a statistical significative difference in OS (p- value 0.026) in favor of patients with targetable mutations (Fig.2). Conclusion This mono-institutional analysis demonstrates that targetable mutations do not influence the locoregional control of the patients treated with SRS. Otherwise, for patients with altered immune system state, potential benefits in oncological outcomes could be represented by association of SRS with immune-modulatory therapies. For patients with a worse expected prognosis (GPA<2), the presence of targetable mutations seems to improve OS. Of notes we need to confirm these data with a larger and prospective cohort of patients, but our findings represent an encouraging hypothesis generating findings to select patients who benefit for personalized treatments. PO-1024 Fractionated RT is equally effective but less toxic than SBRT for central early-stage NSCLC M. Leenders 1 , S. Peeters 1 , J. Van Loon 1 , A. Van Baardwijk 1 , B. Reymen 1 , K. Verhoeven 1 , M. Öllers 2 , R. Wanders 1 , D. De Ruysscher 1 1 Maastro, Radiotherapy, Maastricht, The Netherlands ; 2 Maastro, Physics, Maastricht, The Netherlands Purpose or Objective Stereotactic body radiotherapy (SBRT) for central early stage non-small cell lung cancer (NSCLC) is reported to lead to 5-25% grade 5 toxicities, mainly bleeding (Tekatli Radiother Oncol 2015, JTO 2016). We hypothesized that a more fractionated accelerated regimen would be as effective as SBRT (van Baardwijk Radiother Oncol 2012), but less toxic. Material and Methods Stage I NSCLC patients were included between July 2012 and January 2019. Pathological confirmation was not mandatory. Systemic anti-cancer treatment was not allowed. All patients were staged with a whole body FDG- PET-CT scan. RTOG definition of a central lung tumor was used: “Tumor within or touching the zone of the proximal bronchial tree, defined as a volume of 2 cm in all

directions around the proximal bronchial tree. Tumors that are immediately adjacent to mediastinal or pericardial pleura (PTV touching the pleura) were also are considered central tumors”. The treatment was delivered in 24 daily fractions of 2.75 Gy (For tumor: EQD2=71 Gy, EQD2,T=69 Gy; for OAR: EQD2=76 Gy) to the PTV. A GTV to CTV margin of 5 mm was used. PTV-margins were individually calculated based on the 4D-CT scan. Daily CBCT was performed for tumor match. Toxicities were reported using the CTCAEv5.0. Follow-up consisted of 3-montly CT- thorax scans and recurrence was evaluated using the RECIST-criteria. Cox-regression analysis and Kaplan-Meier analysis were used to calculate OS and PFS. Results 89 patients were included in the study. Mean age was 75.5 ± 7.6 years (range 55-90 years); 53% were men. The WHO PS was; 0 (7%), 1 (46%), 2 (36%) and 3 (11%). The mean tumor diameter was 2.9 ± 1.2 cm. Pathology was available in 62% of patients (28% squamous , 24% adenocarcinoma and 10% other). The mean lung dose was 7.4 ± 2.8 Gy , the mean heart dose 4.3 ± 5.4 Gy, and the maximum dose on the mediastinal envelope 64.6 ± 5.2 Gy. The 1-, 2-, 3-, 4- years overall survival (OS) and progression-free survival (PFS) were 79%, 60%, 50%, 40% and 71%, 65%, 58%, 51%, respectively. Local tumor recurrence after 1-, 2-, 3-, 4- year was 2.2%, 10.1%, 11.2%, 12.4%, respectively. Six patients (7%) had a toxicity grade 3 or more, classified as dyspnea (2) or radiation pneumonitis (4): 4 patients with grade 3 (5%), 1 with grade 4 (1%), and 1 with grade 5 (1%). No severe bleeding was observed. Conclusion This hypofractionated, accelerated fractionation schedule was as effective as SBRT with 3 year local control rates of 89%, but with less severe toxicity for centrally located early stage NSCLC. PO-1025 Prognostic factors for PFS and OS in radically treated patients with oligometastatic NSCLC M. LeenderS 1 , R. Robeers 1 , L. Hendriks 2 , J. Van Loon 1 , G. Bootsma 3 , R. Wanders 1 , C. Pitz 4 , B. Reymen 1 , R. Houben 5 , A. Van Baardwijk 1 , K. Verhoeven 1 , S. Peeters 1 , D. De Ruysscher 1 1 Maastro, Radiotherapy, Maastricht, The Netherlands ; 2 MUMC+, Pulmonology, Maastricht, The Netherlands ; 3 Zuyderland Medisch Centrum, Pulmonology, Heerlen, The Netherlands ; 4 Laurentius Ziekenhuis, Pulmonology, Roermond, The Netherlands ; 5 Maastro, Data Centre Management, Maastricht, The Netherlands Purpose or Objective Evidence suggest prolonged PFS and OS with oligometastatic non-small cell lung cancer (NSCLC) patients when treated with a radical approach with acceptable side effects. However only 15-25% may experience these beneficial effects. Identification of ‘true’ oligometastatic disease remains a gap in our knowledge. We therefore sought to determine prognostic factors and tried to create a first model. Material and Methods 170 stage IV NSCLC patients were included. All patients were staged with a whole body FDG-PET-CT scan and cerebral MRI. All patients received a radical treatment, surgery or radiotherapy. Systemic anti-cancer treatment was not mandatory. Oligometastatic stage was defined according to the EORTC definition; ≤5 metastatic lesions in ≤3 organs. Follow-up consisted of CT-thorax scans every 3-6 months and a consultation with the pulmonologist. Cox-regression analysis and Kaplan-Meier analysis were used to determine the prognostic factors. Results Mean age 66 ± 8.9 years and 58% was male. The median follow-up time for living patients was for OS 49 months and for PFS 51 months. 83% had a good WHO performance status (0-1) and 95% had only 1 metastatic site. The site of