ESTRO 2020 Abstract Book

S511 ESTRO 2020

PO-0957 Why does deep inspiration work for radiotherapy of the internal mammary nodes? F. Voelk 1 , M. Düsberg 1 , M. Oechsner 1 , S.E. Combs 1 , K.J. Borm 1 , N.D. Marciana 2 1 Klinikum rechts der Isar Technische Universität München, Klinik und Poliklinik für Radioonkologie und Strahlentherapie, München, Germany ; 2 Universitätsklinikum Jena, Klinik für Strahlentherapie und Radioonkologie, Jena, Germany Purpose or Objective Cardiac toxicity is a limiting factor during irradiation of the internal mammary nodes (IMN) in breast cancer patients. Deep inspiration breath-hold (DIBH) can help to reduce the heart dose. So far, the impact of DIBH during regional lymph node irradiation including the IMN is not fully understood. The goal of the present study was to investigate the movement of the IMN during DIBH in relation to the movement of the heart. Material and Methods We randomly chose 32 breast cancer patients from our database who underwent both DIBH and free-breathing (FB) treatment planning. Contouring of the axillary lymph node levels (LI, LII, and LIII and IMN) was performed retrospectively according to the ESTRO contouring guidelines. In addition, we contoured organs at risk (OARs) including the heart and the left anterior descending artery (LAD) according to Duane et al. We then analysed the cranio-caudal overlap (CCO) of the left IMN CTV (L_CTV_IMN) and heart as well as the CCO of L_CTV_IMN and the LAD in the FB-CT and DIBH-CT. In addition to this, we used deformable registration algorithms and MATLAB R2019a (The Mathworks Inc., Natick, MA) with image processing toolbox to delineate and analyse the changes of the IMN position with regard to the heart structures. Results The mean CCO of L_CTV_IMN and heart was significantly smaller (p<0.01) in DIBH-CTs (2.55cm) in comparison with the FB-CTs (5.73cm). Further it was shown that CCO of the CTV IMN and the heart was smaller in all patients in the DIBH-CT. Regarding the CCO of LAD and L_CTV_IMN, we found that in 19 patients there was no overlap in the DIBH- CT. In 5 patients the overlap in the DIBH-CT was <15% of the total volume of the LAD. Only 4 patients showed an overlap of >30% in DIBH-CT. In general, the mean volume of the LAD that was cranio-caudal overlapped by the L_CTV_IMN was significantly smaller (p<0.01) in the DIBH- CTs (0.14cm³ ≙ 9.5% of total volume) in comparison with the FB-CTs (0.59 cm³ ≙ 43.0% of total volume). The differences found are also graphically displayed and analysed (Figure 1).

Radiotherapy, Paris, France ; 4 CHRU Jean Minjoz, Oncology Radiotherapy, Besançon, France ; 5 Centre Oscal Lambret, Oncology Radiotherapy, Lille, France ; 6 Radboudumc, Oncology Radiotherapy, Nijmegen, The Netherlands ; 7 University-Cerrahpaşa, Oncology Radiotherapy, Istanbul, Turkey ; 8 Centre Georges François Leclerc, Oncology Radiotherapy, Dijon, France ; 9 Hopital Tenon, Oncology Radiotherapy, Paris, France ; 10 Institut Jules Bordet, Oncology Radiotherapy, Bruxelles, Belgium ; 11 Hôpital du Valais, Oncology Radiotherapy, Sion, Switzerland ; 12 Instituto de radiomedicina, Oncology Radiotherapy, Santiago, Chile ; 13 Centre François Baclesse, Oncology Radiotherapy, Caen, France ; 14 West Virginia University, Oncology Radiotherapy, Morgantown, USA ; 15 Centre Hospitalier universitaire Vaudois, Oncology Radiotherapy, Lausanne, Switzerland Purpose or Objective Primary neuroendocrine tumors (NET) of the breast are rare breast tumor, representing less than 1% of breast cancers.Their prognosis appears to be poorer than that of classical invasive breast carcinomas, with shorter OS and DFS at equal stage. There are no specific therapeutic guidelines for those tumors. The aim of this study was to analyze the histological, prognostic and therapeutic specificities of this rare disease. Material and Methods All women with primary NET of the breast from 14 different centers and 8 countries between 1995 and 2015 were included. Clinical, pathological and therapeutic data were collected retrospectively. Statistical analyses were performed using R v3.4.3 . Survival analysis was determined using the Kaplan-Meier method. Factors influencing overall survival and disease- free survival were first evaluated with a univariate Cox analysis with a significant cutoff of p<0.2. Multicollinearity was also checked using a VIF threshold of 4. A stepwise backward procedure was used to build the multivariate Cox model to evaluate the impact of potential variables on survival. Results Among the 97 women included, median age was 66 (33- 93). At diagnosis, 36% were stage I, 41 II, 14% III and 8% IV. Lymphovascular invasion was found in 39% of tumors, perineural invasion in 8%, fibrosis in 18% and necrosis in 19%. NETs were well differentiated in 40% of cases, poorly differentiated in 12%, and it was carcinoma with neuroendocrine differentiation in 47% of cases. Expression of synaptophysine was found in 84% of cases, chromogranine in 66% and HR in 88%. The majority of the patients (95%) had breast surgery (57% of partial mastectomy, 43% of total mastectomy), with axillary dissection (56%). Chemotherapy, endocrine therapy and radiotherapy were administrated before surgery in 14%, 10% and 2% of cases respectively, and after surgery in 28%, 73% and 64% of cases. Chemotherapy regimens were heterogeneous: most of patients received taxans and/or anthracyclins (86% in neoadjuvant situation, 69% in adjuvant situation), but several patients had platinum, etoposide, bevacizumab or somatostatin. With a median follow up of 89 months (95CI 68-110 months), OS and DFS at 5 years were respectively 66.5% (95CI 57.0-77.6) and 75.0% (95CI 66.2-85.1) (Figures 1 and 2). Age, stage and Ki67 were significantly correlated with OS in multivariate analysis. Conclusion The management of primary TNE of the breast is heterogeneous, especially concerning chemotherapy regimens. TNM stage and Ki67 are important prognostic factors. Prospective studies are needed to determine specific therapeutic guidelines.

Conclusion Our findings show that for the majority of patients in DIBH there is no relevant overlap between the LAD and the lower edge of the L_CTV_IMN. In addition the overlap of heart (and the LAD) and L_CTV_IMN decreases during DIBH. The observed mechanism might account to a relevant part to the dose reduction during DIBH in patients that receive regional nodal irradiation.