Abstract
OBJECTIVE: Breast comprises different tissues with potentially different dose responses to radiation therapy (RT). This study investigates the correlation between RT dose, breast composition, and side effects from breast RT. MATERIAL/METHODS: Data from 922 early-stage breast cancer patients who underwent breast-conserving surgery and RT from the REQUITE study were included. Breast pain, oedema, atrophy, and induration were assessed immediately post-RT, one-year, and two-years post-RT. Maximum severity scores for each toxicity were used for analysis. Breast tissue was divided into "fat" and "fibroglandular" substructures from computed tomography (CT) using a Gaussian Mixture Model. The correlation between breast characteristics, toxicity, dosimetric parameters, and patient and clinical variables was investigated using ordinal regression. The model's fit was evaluated using the Akaike Information Criterion in SPSS v.29. RESULTS: Breast volume and breast density were associated with increased risk of breast oedema, atrophy, and induration in multivariable analysis (p<0.05). Higher mean dose and dose uniformity were observed for fibroglandular compared to fatty tissue at all severity levels, while there was no significant difference in the maximum dose to either substructure. Higher dose deposit to fat was associated with breast pain and oedema, while breast atrophy and induration were associated with dose to fibroglandular tissue. All best-performing toxicity models included dosimetric parameters derived from breast composition. CONCLUSION: Breast characterisation offers new insight into the link between dose and toxicity. Breast density and dose parameters from different substructures were associated with different breast toxicity. These findings further support the importance of dose homogeneity of breast RT planning.