ESTRO 38 Abstract book

S1033 ESTRO 38

inclusion of more patients would allow a for better validation of the model. EP-1900 Predictive parameters for long-term cardiac mortality excess related to left breast radiotherapy N. Pasinetti 1 , V. Morandini 1 , S. Berlinghieri 1 , E. Castrezzati 1 , S. Pedretti 2 , L. Spiazzi 3 , P. Frata 1 , L. Grimaldi 1 1 Esine Hospital, Radiation Oncology sevice, Esine, Italy ; 2 Spedali Civili Hospital, Department of Radiation Oncology, Brescia, Italy ; 3 Spedali Civili Hospital, Department of Medical Physics, Brescia, Italy Purpose or Objective As shown by Early Breast Cancer Trialists' Collaborative Group(*) breast cancer radiotherapy is considered mandatory in patients treated with breast conserving surgery or in mastectomy to improve cancer death rate and disease recurs. However, treatment of patients with left breast cancer is burdened by the risk of long-term secondary cardiac effects that may be added to those known induced by chemotherapy (anthracyclines, trastuzumab). Consequently, careful evaluation of the heart-absorbed dose is necessary and, although a risk of cardiac death <1% after breast RT is considered acceptable, the QUANTEC constraint V 25Gy <10% is not sufficient to limit the excess cardiac mortality risk for each patient and for every RT schedule of tangential beams left sided breast RT. Aim of this work is to find a group of patient for which the dose constraint, above mentioned, is respected, but at a further investigation the excess cardiac mortality probability results greater than 1%. Correlation with new “predictive” dosimetric parameters and cardiac mortality risk was investigated and verified on a control group of patients. Material and Methods Analyzing the DVHs on TPS Philips Pinnacle 3 of 240 women, who underwent to left sided breast tangential beams RT, came out that the constraint V 25Gy <10% was always respected. By inspecting the integral DVHs some patients were individuated as “probable false negatives”. For each heart absorbed dose distribution the EQD2 was calculated and the Relative Seriality model (α/β=3Gy, s=1, D 50 =52.4Gy and g=1.28) was used as method to evaluate the probability of late cardiac mortality. Results For 19 patients the V 25Gy <10% was satisfied, but by analyzing the differential DVHs the probability of long- term cardiac mortality was found >1% and up to 6%. The dosimetric heart constraints V 40Gy and D 2% showed a good correlation (R=0.97 and R=0.90 respectively) with the risk of cardiac death. To keep the probability lower than 1%, the cut off levels were determined by the simultaneous occurrence of the conditions: V 40Gy <2%, and D 2% <38Gy. On a control group of 15 other patients (or whom V 25Gy <10 was satisfied) these parameters were tested. What emerged is that if both parameters were satisfied the long term cardiac mortality probability resulted, with an explicit calculation, <1%.

Conclusion Our “predictive” parameters, although they are only a calculation and not an observation of mortality, are closely connected to the irradiation technique used and aimed to specific end-points. Anyway modern TPSs should promote, even more, the use of either radiobiological DVHs or algorithm optimization, especially in the era of hypofractionation. (*) Early breast cancer trialist collaborative group, Lancet 2011; 378: 1707-16 EP-1901 Identifying organs at risk for radiation- induced dysphagia in head and neck cancer patients J. Hedström 1 , L. Tuomi 1 , C. Finizia 1 , C. Olsson 2 1 Inst of Clinical Sciences, Dept of Otorhinolaryngology, Göteborg, Sweden ; 2 Inst of Clinical Sciences, Dept of Radiation Physics, Göteborg, Sweden Purpose or Objective Dysphagia is a common and severe dose-limiting toxicity after oncological treatment of head and neck cancer (HNC). This study aims to investigate the relationship between radiation dose to structures involved in normal swallowing and patient-reported as well as clinically measured swallowing function in HNC patients after curative (chemo)radiotherapy (RT). Material and Methods Patients (n=90) with tumours of the tonsil, base of tongue, hypopharynx, and larynx curatively treated with radiation therapy +/- chemotherapy in 2007-2015 were assessed for dysphagia post-treatment by telephone interview and videofluoroscopy (VFS). A study-specific symptom score, accounting for presence of drinking, eating, swallowing difficulties, and coughing when eating/drinking, was used to determine patient-reported dysphagia (DESdC). The penetration-aspiration scale (PAS) was applied to determine swallowing function by VFS. Anatomical structures involved in normal swallowing were individually delineated on the patients’ original planning CT scans. Radiation dose-volume relationships for these structures were investigated, including ipsi- and contralateral structure dose for the bilateral structures. Univariate logistic regression analysis was performed with structure mean and maximum absorbed doses as predictors for dysphagia. Multivariate regression analysis was subsequently performed to identify the most statistically critical structures associated with swallowing impairment according to DESdC and PAS. Potential effects by relevant clinical factors (comorbidity as scored by ACE-27, age, smoking and BMI) were accounted for. Figure 1. Delineation of organs-at-risk. One representative cross-sectional slice of a pre-treatment planning CT with OARs delineated. A=anterior; R=patient’s