ESTRO 2024 - Abstract Book

S648

Clinical - Breast

ESTRO 2024

Thirty- three/88 patients exhibited grade≥2 erythema (37.5%, grade2=26.3%, grade3=11.3%, CTCAE v.5.0). Table 1 reports the toxicity rates stratified by treatment characteristics, MVHS and hemoglobin concentration before radiotherapy. We combined the two quantitative features, defining a new composite local-systemic score LSS = VHS*[Hb]. The mean values for LLS are 17.7 vs 28.7 in patients with/without toxicity, respectively (t-test p value=0.03). We developed a two- variable logistic regression model predicting grade≥2 erythema (AUC=0.74) and including LSS (Odds Ratio=0.955 for 1 point increase, range for LSS 2.4-91.4, p=0.007, highest score entails higher microvascular health) and the use of sequential boost (yes/no) (Odds Ratio=3.4, p=0.02) (Figure 2). We could identify two cut-offs for clinical application: LSS<15 and LSS>40, identifying patients at a very high (>60% patients for patients with sequential boost and >40% for patients without sequential boost) and very low (<30% patients for patients with sequential boost and <10% for patients without sequential boost) risks of toxicity, respectively.