Abstract Book

S727

ESTRO 37

physicians and patients using a four-point scale according to Harvard criteria. Results Median age was 53 years (26-85). Median breast and tumor bed volume were 523 cc (85-2161) and 26 cc (2- 142) respectively. Most patients received quadrantectomy + sentinel node biopsy (70%). Median follow-up was 23.91 months (interquartile range: 17.66- 26.09). Planning target volume (PTV) coverage and healthy tissues sparing were satisfactory. Median value of V95% was 100% and 99% for PTV WB and PTV TB , respectively. Median maximum dose to PTV WB and PTV TB were 113% and 103%, respectively. The table presents the highest grade of toxicities reported at the last follow-up. Neither grade 4 skin ulceration nor soft tissue necrosis was observed. Grade 3 toxicity was very limited. At univariate analysis, predictor for any severe toxicity (LENT/SOMA grade ≥2) was breast volume only ( p<0.001), while tumor bed volume showed a trend ( p= 0.06). Statistically significant relationship between age and any grade late toxicity was also found. Cosmetic outcome was categorized into excellent/good (89% for physicians and 83% for patients) and fair/poor. However, patient and physician-rated cosmesis showed poor agreement [weighted Kappa of agreement =0.50 (95% CI,0.41-0.59)].

Gallen. Consensus Criteria. Kaplan Meier log-rank and Cox-regression multivariate analysis were performed to evaluate the relationship between molecular subtypes and locoregional recurrence. Results Median follow-up was 4.2 years (range 1-12 yrs). Molecular subtypes were distributed as follows: 57.6% LA, 24.9% LB, 8.7% luminal Her-2, 1.9% Her- 2, 6.8% TN. Two hundred seventy (83.1%) patients underwent breast conserving surgery and fifty-five (16.9%) received mastectomy. All patients received adjuvant radiotherapy to whole breast or chest wall. Sixty-four (19.7%) patients received also radiotherapy to supra-infraclavicular nodes. Luminal A subtype had lowest histologic grades (p=<0.001), lowest extracapsular spread rate (p=0.03). Four (1.24%) loco-regional relapse occurred, 2 (50%) in breast and 2 (50%) in the chest wall area. No difference in local recurrence rate was observed according to molecular subtypes, irradiation of supra-infraclavicular nodes, age, tumor diameter (p=NS). Only the number of positive nodes (1 vs 2-3) was associated with local recurrence risk (p=0.017). Luminal A subtypes had a 10 years metastasis free survival rate higher compared to other subtypes (86.9% vs 64.6%; p=<0.001). Conclusion The loco-regional recurrence rate observed was very low. Postoperative radiotherapy even if not associated with supra-infraclavicular nodal irradiation seem to be effective in lowering loco-regional recurrence rate regardless of molecular subtype. EP-1328 Impact of deep inspiration breath hold on heart and lung dose constraints in a large breast series G. Loganadane 1,2 , B.H. Kann 1 , M.R. Young 1 , C.A. Knowlton 1 , S.A. Higgins 1 , Y. Belkacemi 2 , L.D. Wilson 1 , M.S. Moran 1 1 Yale University School of Medicine, Department of Therapeutic Radiology, New Haven- Connecticut, USA 2 APHP. University of Paris-Est Creteil UPEC, Department of radiation oncology & Henri Mondor Breast Center, Créteil, France Purpose or Objective The purpose of this study was to evaluate the dosimetric impact of deep inspiration breath hold (DIBH) on the heart and the lungs in a large cohort of breast cancer (BC) patients Material and Methods This study included 360 patients undergoing three dimensional conformal (3D) tangential radiation using a field-in-field technique treated in our department between January 2015 and July 2017. Patients were divided into 3 cohorts for comparison: right-sided tumors (n=180), left tumors with DIBH (n=120) and left sided tumors with free breathing (FB) (n=60). The median values of various parameters were compared between groups (lung V20, lung V30, mean lung dose, heart V20, heart V30 and heart mean dose). For patients treated with hypofractionated radiation, we used V17, V28 and V38 instead of V20, V30 and V40 respectively to facilitate comparisons. Results There were significant differences in the 3 cardiac dosimetric parameters evaluated (V20, V40 and mean heart dose) in equivalent 2 Gy per fraction (EQD2), but differences in lung volume parameters were not significant. For left sided tumors, DIBH improved cardiac dosimetric constraints compared to FB: V17/V20 : 0.45% vs 1.6% (-71.9%) (p<0.0001), V38/V40: 0% vs 0.35% (-100%) (p=0.003) and mean heart dose (EQD2): 18Gy vs 26.4 Gy (-31.6%) (p<0.0001). Regarding the ipsilateral lung, the comparison of the 3 cohorts by the Kruskall-Wallis test (right sided tumors, left sided tumors with DIBH and left sided tumors with FB) did not show significant differences for any of the parameters: V17/V20: 16.6% vs 17% vs

Conclusion although larger breast volume correlated with increased severe toxicity, Grade ≥ 2 toxicity was mild and acceptable. Cosmesis was mostly rated as excellent/good. These promising results must be confirmed with a longer follow-up. EP-1327 Risk of relapse according to molecular subtype in patients with 1-3 positive nodes after adjuvant RT A. Di Donato 1 , S. Silipigni 1 , E. Ippolito 1 , G.M. Petrianni 1 , C.G. Rinaldi 1 , A. Iurato 1 , B. Santo 1 , P. Trecca 1 , R.M. D'Angelillo 1 , L. Trodella 1 , S. Ramella 1 1 Campus Biomedico University, Radiotherapy, Roma, Italy Purpose or Objective To evaluate the impact of molecular subtype on locoregional failure in breast cancer patients with 1-3 positive nodes after postoperative radiotherapy. Material and Methods From an institutional database we identified 321 patients with 1-3 positive nodes treated with postoperative radiotherapy from 2005 to 2016. Molecular subtypes were defined as luminal A (LA), luminal B (LB), luminal Her 2, Her 2 and triple negative (TN) based on the 2015 St.