ESTRO 2020 Abstract book

S573 ESTRO 2020

PO-0990 Prevention of oncologic pulmonary death by control for pulmonary oligometastases treated with SBRT T. Yamamoto 1 , N. Yuzuru 2 , K. Yamada 3 , M. Aoki 4 , H. Onishi 5 , K. Katsui 6 , Y. Dekura 7 , A. Nishikawa 8 , Y. Manabe 9 , S. Kubota 10 , H. Yamashita 11 , K. Jingu 1 1 Tohoku University Hospital, Radiation Oncology, Sendai, Japan ; 2 Toho University Omori Medical Center, Radiology, Tokyo, Japan ; 3 Seirei Mikatahara General Hospital, Radiation Oncology, Hamamatsu, Japan ; 4 Hirosaki University, Radiology, Hirosaki, Japan ; 5 Yamanashi University, Radiology, Yamanashi, Japan ; 6 Okayama University, Proton Beam Therapy, Okayama, Japan ; 7 Keiyu-kai Sapporo Hospital, Radiation Oncology, Sapporo, Japan ; 8 Shikoku Cancer Center, Radiation Oncology, Ehime, Japan ; 9 Nagoya City University, Radiology, Nagoya, Japan ; 10 Nagoya University Hospital, Radiology, Nagoya, Japan ; 11 University of Tokyo, Radiology, Tokyo, Japan Purpose or Objective The purpose of current study is to investigate freedom from oncologic pulmonary death (FOPD) rate and to analyze affecting factors for FOPD. Material and Methods Eligibility of this retrospective survey needed that SBRT was performed between 2004 and 2015, number of metastases was 5 or fewer and all metastases located in lung (i.e., pulmonary oligometastases), primary lesion and extrathoracic metastases needed to be controlled before SBRT and biological effective dose needed 75 Gy or more. FOPD was defined as freedom from oncologic pulmonary death which was caused by respiratory failure or hemoptysis following lung metastases, pleural effusion or intrathoracic dissemination, but fatal lung toxicity of SBRT was not included. Primary disease death from non- oncologic pulmonary cause was censored in model 1 and excluded in model 2, and non-primary disease death was censored in both models. Kaplan-Meier estimator and a log-rank test was used to calculate and compare the stratified FOPD. Cox proportional hazards model was applied to multivariate analyses (MVA). Results A total of 1172 patients with 1315 tumors were enrolled. During a median follow-up of 24.5 months, oncologic pulmonary deaths occurred in 101 out of 221 primary disease deaths. The 3-year local control rate, cause- specific survival (CSS) rate, OS rate were 82.9%, 78.5% and 69.6%, respectively. The 1-, 3-, 5-year FOPD rate was 98.2%, 89.4% and 84.0%, respectively. Median interval to pulmonary death was 23.0 months. MVA for FOPD revealed that local failure of irradiated tumor, squamous cell pathology and chemotherapy after SBRT had significant relationship with worse FOPD rates in both models 1 and 2.

Conclusion Successful local control for pulmonary oligometastases by SBRT was contributed to higher FOPD rate. PO-0991 Impact of residual setup errors after image guidance on heart dose in NSCLC patients C. Johnson-Hart 1 , G. Price 2 , A. McWilliam 1 , A. Abravan 1 , C. Faivre-Finn 1 , M. Van Herk 1 1 The University of Manchester, Radiotherapy Related Research, Manchester, United Kingdom ; 2 The Christie NHS Foundation Trust, Radiotherapy Related Research, Manchester, United Kingdom Purpose or Objective A recent study of NSCLC patients showed small residual setup errors (shifts) in the direction of the heart following image-guidance were significantly related to overall survival. This study of the dosimetric effects of these residual shifts investigates the hypothesis that observed survival differences were related to a change in heart dose. Material and Methods Accumulated doses including shifts for each fraction were determined for 546 NSCLC patients treated with 55Gy in 20 fractions, assuming shift invariance (validated in a subset). Planning CTs and corresponding dose distributions were deformed to a reference case. Image-based data- mining and permutation testing techniques were then applied to the difference between the planned and accumulated dose (Δdose) to determine where Δdose relates to 1-year survival. The significance of Δdose in the identified region was assessed using multivariable Cox analysis, correcting for age, sex, GTV, N stage and ECOG- PS. The cohort was then split into octiles, based upon planned dose to the region, and multivariable Cox analysis performed for each sub-cohort to explore the dose response relationship. The identified dose threshold for damage was then tested in an independent validation cohort of 1482 NSCLC patients (treated with 55Gy in 20 fractions) from the same institution, based upon the planned dose to the region (as shifts were unavailable for this cohort). Results Permutation testing identified a small region in the heart base, near the right coronary artery, where Δdose correlated with 1-year survival (Figure 1). Δdose in this region showed no correlation with tested common clinical variables – meaning it is likely not confounded and thus is a sensitive test metric. Δdose was significant in multivariable Cox regression (p<0.001, HR 1.221/Gy), with increasing change in dose from plan resulting in greater risk of death. Octile analysis revealed Δdose to be significant only in the 7 th octile, planning dose range 16.2– 23.4Gy, suggesting a steep dose-effect relation for heart damage in this range. Taking 16.2Gy as a conservative threshold dose, this result was successfully validated.