TOPLINE:

An image-based data mining analysis identified specific regions in the breast where radiation dose was associated with pain, edema, and other complications. Compared with whole-breast dosimetry, region-specific dosimetry was associated with improved prediction.

METHODOLOGY:

• Radiotherapy after breast-conserving surgery reduces the risk for recurrence but can cause acute and late toxicities such as pain, edema, atrophy, and induration that impair quality of life. Traditional dose-volume histogram-based models do not consider organ complexity or interaction with the spatial heterogeneity of dose distributions. Voxel-wise image-based data mining applied to large cohorts may reveal dose-sensitive breast subregions linked to toxicity.
• Researchers analyzed data from 922 patients in the REQUITE study who underwent radiotherapy after breast-conserving surgery. Clinician- and patient-reported pain, edema, atrophy, induration, and nipple retraction were recorded at baseline, at the end of radiotherapy, and 1 and 2 years after radiotherapy and integrated into an area under the curve for toxicity over time (ToxT-AUC) metric.
• The image-based data mining technique required mirroring all left-sided planning CTs and physical dose distributions laterally, so that every patient’s irradiated anatomy was aligned to the right, enabling pooled voxel-wise dose analysis of the ipsilateral breast. Voxel doses were converted to biological doses in terms of equivalent dose in 2-Gy fractions (EQD2) using an alpha/beta ratio of 1.7 Gy (and repeated with 3 Gy); CTs were registered to three reference anatomies with uncertainty blurring.
• Voxel-wise Spearman rank correlation was used to evaluate the association between the EQD2 and ToxT-AUC for each side effect and at each timepoint.

TAKEAWAY:

• Using ToxT-AUC (about 60% clinician- and 40% patient-reported outcomes), all significant regions overlapped with the breast contour: Edema and atrophy clustered in the central and lower quadrants of the breast (atrophy peaked at 1 year); edema regions consistently overlapped axillary lymph node levels 1-4 and matched the postradiotherapy region for pain; and a focal region in the paraspinal muscle near the upper thoracic vertebrae was linked to pain at 1 and 2 years.
• Regions overlapping the breast for pain, atrophy, and induration received higher mean doses than the whole-breast contour at all timepoints. For endpoints with significant differences, patients with toxicity of grade 2 or higher had higher mean EQD2 in the identified regions than those with toxicity of grade 1 or less.
• Identified regions were consistent across all three reference anatomies and with or without EQD2 blurring. Although the volumes of the significant regions varied slightly between anatomies, their locations remained similar even when the alpha/beta ratio was changed.
• In multivariable analyses, only mean EQD2 to the breast region remained significant (P < .05) for most endpoints except induration. Mean EQD2 to the pectoralis, lymph node, and spinal regions was not significant, and breast pain was significant only post-radiotherapy when treatment covariates were accounted for. Older age, larger breast volume, higher BMI, boost dose, supraclavicular radiotherapy, more extensive breast-conserving surgery, and systemic therapy increased the risk for some acute toxicities.

IN PRACTICE:

The study highlights the “importance of local dose to different regions in breast anatomy associated with acute and late breast side effects after breast RT [radiotherapy] without preassuming a region,” the authors of the study wrote. “Identifying these regions may offer a more targeted approach to predicting and managing breast side effects, and outcomes could be improved by reducing high doses to these areas and addressing dose distribution heterogeneity within the irradiation field.”

SOURCE:

The study, led by Tanwiwat Jaikuna, School of Medical Sciences, Faculty of Biology, Medicine and Health, The University of Manchester, Manchester, England, was published online in Radiotherapy and Oncology.

LIMITATIONS:

Half of the patients were treated with less advanced techniques like three-dimensional conformal radiotherapy, while recent breast irradiation methods have shifted to more sophisticated approaches that produce more homogeneous dose distribution. The REQUITE dataset lacked specific details on locations of reported symptoms and data on oncoplastic surgery, which may have influenced outcomes. Patient behaviors, such as bra-wearing post-radiotherapy that may affect fluid drainage, were not considered.

DISCLOSURES:

The study received funding from the European Union’s Seventh Framework Programme, Cancer Research UK RadNet Manchester, the Cancer Research UK Manchester Centre, and other sources. The authors reported no having relevant conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.