ESTRO 2020 Abstract book

S543 ESTRO 2020

PO-0932 Preoperative radiotherapy for early breast cancer: quality assessment of the ROCK trial - NCT03520894 G. Francolini 1 , V. Di Cataldo 2 , S. Cipressi 2 , E. Scoccimarro 1 , V. Maragna 1 , L. Masi 3 , R. Doro 3 , G. Simontacchi 1 , I. Desideri 1 , I. Meattini 4 , L. Livi 4 1 Azienda Ospedaliera Universitaria Careggi, Radiation Oncology Unit, Florence, Italy ; 2 University of Florence - Istituto Fiorentino di Cura ed Assistenza, Radiation Oncology Unit- CyberKnife Center, Florence, Italy ; 3 IFCA, Department of Medical Physics and Radiation Oncology, Florence, Italy ; 4 University of Florence, Department of Biomedical- Experimental- and Clinical Sciences “Mario Serio”, Florence, Italy Purpose or Objective Early stage breast cancer (BC) is currently managed with breast-conserving surgery (BCS) followed by postoperative irradiation. However, preoperative radiation in BC include potential advantages if compared with standard treatment sequence, and some experiences prove the feasibility of this approach. In our Institution, an exploratory study enrolling early BC patients treated with single fraction Cyberknife ® preoperative robotic radiotherapy is currently running (NCT03520894-ROCK trial). Nevertheless, there is no standardised approach for target delineation in this setting. For this reason, we assessed interobserver variability (IOV) in terms of target delineation within the trial and produced a consensus between 3 different operators involved in treatment planning. Material and Methods Overall, 5 patients have currently been included within the trial. Four patients already underwent single fraction 21 Gy preoperative radiotherapy followed by BCS. All patients had planning CT with and without contrast enhancement.Gross Target Volume (GTV) was delineated on contrast enhanced planning CT scan co-registered with Magnetic Resonance Imaging (MRI) obtained in treatment position. Clinical Target Volume is obtained expanding GTV by 1.5cm, excluding 5 mm from skin, thoracic wall, and pectoral muscle. Three different operators involved in the trial (VDC, GF, and SC) independently delineated GTV for all included patients . Simultaneous Truth and Performance Level Estimation (STAPLE) algorithm was used for analysis of contour agreement by obtaining a probabilistic estimate of the“true” contour (GTV-multi) (Figure 1). Dice similarity score (DSC) was calculated between each single operator and GTV-multi and between the 3 different operators. After reviewing the results, a consensus meeting was held in order to establish critical issues to GTV delineation in this setting.

Medicine, Department of Experimental Clinical Oncology and Department of Oncology- Aarhus University Hospital, Aarhus, Denmark ; 3 Department of Clinical Medicine, Research Department of Cardiothoracic and Vascular Surgery- Aarhus University Hospital, Aarhus, Denmark ; 4 Department of Clinical Medicine, Research Department of Cardiothoracic and Vascular Surgery- Rigshospitalet and Aarhus University Hospital, Aarhus, Denmark Purpose or Objective Adjuvant radiation therapy (RT) significantly improves survival in high-risk breast cancer (BC) treated with loco- regional RT, however, it also increases risk of lymphedema in particular in patients operated with axillary lymph node dissection (ALND). ALND and nodal RT may partially obstruct lymph outflow from the ipsilateral arm, chronically raising the afterload of the lymphatic vasculature, which the lymphatic smooth muscle must overcome in order to secure sufficient drainage. Studies have shown a change in contractile function of the lymphatic vessels and demonstrated distinct lymphatic patterns in women diagnosed with breast cancer related lymphedema (BCRL), but no studies have investigated whether these changes occur before clinical edema is detectable, which is the focus of this study. Material and Methods The morphological and functional state of the lymphatic vessels is described using Near-Infrared Fluorescence (NIRF) imaging and plethysmography. The endpoints consist of contraction frequency, velocity, pumping pressure and capillary filtration rate (CFR). The study population is 35 high-risk BC patients included in the DBCG RT Skagen trial 1 and treated with loco-regional RT (50 Gy/25 fr or 40 Gy/15 fr). Patients are invited to participate in the study with an initial examination a few weeks after ended RT and a second examination 6-12 months later, which adds blood sampling for endothelial growth factor and cytokine analysis and a hyperthermia-protocol serving as a stress-test of the lymphatic vasculature to investigate the pumping capacity. Both arms will be examined enabling the patients to serve as their own controls. Results The preliminary results consist of 14 patients. Two patients so far presented with lymphatic abnormalities. A 22% higher pumping pressure was observed in the ipsilateral arm compared to the contralateral when patients with lymphatic abnormalities were excluded (p=0.0105). Patients with lymphatic abnormalities tend to perform a reduced maximal pumping pressure in the ipsilateral arm compared to the contralateral. Currently, no significant difference between the arms was present concerning frequency and CRF. All 35 patients will be accrued and have baseline examination during 2019, and most patients will have their second examination in the early spring of 2020. Conclusion The preliminary results indicate that the lymphatic vessels in the breast cancer related arm are compensating, for the damage inflected by the cancer treatment, by raising the ability to generate pressure. However, at some point the lymphatic vessels begin to develop lymphatic abnormalities and the ability to generate pressure weakens. These lymphatic abnormalities and reduced contractile function could be the initial phase leading to developing BCRL. By using NIRF and plethysmography this study investigates if these changes including biochemical markers can be used to evaluate early BCRL and guide early treatment, which at this point in time has proven to be much more beneficial.

Results Mean DSC between single operators GTV and GTV-multi was 0.9, 0.86 and 0.82 for VDC, GF, and SC, respectively (Table 1). Mean DSC between the single operators coupled was 0.77, 0.73, and 0.71 for VDC and GF, VDC and SC, GF and SC,respectively. After the consensus meeting, all operators agreed that a Window Width and Level of 230 and 40 HU were the best parameters to locate the target