ESTRO 2020 Abstract Book

S526 ESTRO 2020

Results

(7%) had stable disease; only 4 (14%) patients progressed after CCCR. More than half of the patients were proposed to radical mastectomy (N=15, 64%) and 2 to palliative mastectomy. Subsequently, 8 patients (29%) had locoregional progression, 3 of them together with systemic progression; 19 (68%) patients had distant progression, mainly by hepatic, bone, cutaneous and nodal metastasis. With a median follow-up of 61 months, the OS was 12.7 months (CI 95% 4.6-20.7) and PFS was 4.6 months (CI 95% 2.5-6.6). Conclusion The use of a CCCR scheme for locally advanced disease after primary chemotherapy was associated with satisfactory response rates, with reversal of inoperability in a significant proportion of patients whilst maintaining acceptable toxicity. Prospective randomized studies are needed to validate this treatment. PO-0988 Is there a learning curve for SABR that affects overall survival outcomes in early stage NSCLC? C. Peedell 1 , E. Aynsley 1 , A. Wood 1 , G. Kumar 1 , S. Masinghe 1 , J. Reynolds 1 , C. Huntley 1 , A. Blower 1 , J. Green 1 , J. Bradley 1 , J. Veeratterapillay 1 , A. Hassani 1 , M. Anderson 1 , A. Greenhalgh 1 , J. Daniel 1 , A. Swingler 1 , M. Turnbull 1 , K. Burke 1 1 South Tees Hospitals NHS Foundation Trust, Oncology and Radiotherapy, Near Bedale, United Kingdom Purpose or Objective SABR has become the standard of care for medically inoperable, peripherally located, early stage NSCLC, but the question of whether the excellent long term outcomes published by pioneering academic institutions can be replicated in the “real-world setting” remains a pertinent one. As increasing numbers of institutions set up SABR programmes, it is important to understand whether learning curves that affect patient outcomes exist. Our cancer centre is a non-academic centre located in one of the most socially deprived regions in the UK, serving a population of 1.1million. Our SABR program for NSCLC started in 2009 and we present the long term outcomes of 418 patients treated by 5 different consultants of varying experience. Material and Methods Retrospective data was collected from consecutive patients with medically inoperable early stage (T1a-3 N0) NSCLC treated with SABR between Sept 2009 and May 2019. Patients with previous malignancy within 5yrs, or more than one primary lesion were excluded. Patients without a histological diagnosis required a PET+ve, growing lesion for inclusion. All patients were immobilised using the BodyFix system. The following risk- adapted dose/fractionation schedules were used: 54Gy/3#; 55Gy/5#; 60Gy/8#; 50Gy/10#. Overall Survival was assessed using the Kaplan-Meier method. Five different consultant clinical oncologists (A,B,C,D,E) treated patients during this period and individual consultant outcomes were assessed. OS of the first 25 patients treated versus all subsequent patients was calculated for 3 of the consultant cohorts (A,C,E) Poster: Clinical track: Lung

418 patients were treated as follows: 54Gy/3# (86pts); 55Gy#5 (237pts); 60Gy/8# (61pts); 50Gy/10# (33pts). Median age 76yrs (range 48-93yrs). Histological diagnosis obtained in 57%. 5 year OS of entire cohort was 37.3% (Fig.1). Median OS was 3.47 years. No difference in 5yr OS between the different fractionation schedules, except the 10 fraction schedule, which had poorer OS (p=0.016). No difference in survival between the different treating consultants who varied significantly in experience (Range: 26-195 patients treated, Fig.2). Also no difference in survival for the first 25 patients treated, versus subsequent patients (consultants A,C,E Fig.3) . Toxicity rates were very low with only 4 patients with recorded G3 toxicity (all pneumonitis). Conclusion We have demonstrated that lung SABR can be safely implemented and delivered in the real world, non- academic cancer centre setting, with 5yr Overall Survival rates comparable to the published literature. Toxicity was minimal. There were no differences in OS outcomes between consultants with differing experience, and no differences in OS between the first 25 patients treated and subsequent patients in 3 separate consultant cohorts. This suggests that there may not be a significant learning curve to lung SABR that affects overall survival outcomes, as long as centres follow national guidelines and radiotherapy QA processes (eg UK SABR Consortium and RTTQA guidelines).