ESTRO 2021 Abstract Book

S443

ESTRO 2021

We found that for both metrics (RECIST and tumor volume) the automatically generated segmentations have more prognostic power (Fig. 2C). Conclusion Our proposed pipeline can potentially provide a low-cost, observer-independent and reproducible method for the detection and segmentation of lung cancers on CT images, facilitating diagnosis, robust contouring and allowing for the automatic RECIST and volumetric RECIST measurements. OC-0558 No correlation between mean heart dose and coronary stenosis in 17,088 DBCG breast cancer patients M.L. Milo 1 , D.S. Møller 2 , T.B. Nyeng 2 , L. Hoffmann 2 , I. Jensen 3 , H.D. Nissen 4 , E.L. Lorenzen 5 , K.M. Nielsen 6 , L.B.J. Thorsen 7 , S.P. Johnsen 8 , J. Alsner 1 , B.V. Offersen 9 1 Aarhus University Hospital, Department of Experimental Clinical Oncology, Aarhus, Denmark; 2 Aarhus University Hospital, Department of Medical Physics, Aarhus, Denmark; 3 Aalborg University Hospital, Department of Medical Physics, Aalborg, Denmark; 4 University Hospital of Southern Denmark, Department of Medical Physics, Vejle, Denmark; 5 Odense University Hospital, Laboratory of Radiation Physics, Odense, Denmark; 6 Aarhus University Hospital , Department of Cardiology, Aarhus, Denmark; 7 Aarhus University Hospital, Department of Oncology, Aarhus, Denmark; 8 Danish Center of Clinical Health Services Research , Department of Clinical Medicine , Aalborg, Denmark; 9 Aarhus University Hopital, Department of Experimental Clinical Oncology, Aarhus, Denmark Purpose or Objective Radiation therapy (RT) is widely used in the adjuvant therapy of early breast cancer (BC) to reduce the risk of recurrence and improve overall survival. Coronary artery disease (CAD) has been reported as a long-term side effect. A dose-response relationship between mean heart dose (MHD) and CAD has been demonstrated for patients receiving outdated RT regimens pre-dating 3D-based RT. This study aims to investigate the incidence of CAD in early BC patients treated with 3D-CRT (conformal RT) and to report RT doses to the heart and cardiac substructures in patients with subsequent CAD. Materials and Methods BC patients registered in the Danish Breast Cancer Group database from 2005 to 2016 were included. Standard 3D-CRT was tangential fields using 48 Gy in 24 fractions (fr) until 2009, whereafter it was 50 Gy in 25 fr. From 2014, 40 Gy in 15 fr became standard for whole breast irradiation. If boost was applied, it was 10 Gy in 5 fr or 16 Gy in 8 fr. The endpoint CAD was obtained from the Western Denmark Heart Registry and was defined as a stenosis of minimum 50% of the lumen. For all patients with a subsequent CAD, the planning CT scans and dose plans were retrieved, if available. The heart and 24 cardiac substructures were delineated by an automatic atlas segmentation workflow developed in the MIM software system, version 7.0.4. Mean doses to the heart and cardiac substructures were extracted from the dose volume histograms. The hypothesis of no difference in