Breast reduction and mastopexy procedures on previously irradiated breasts: A retrospective study.

Abstract

Post-radiation fibrosis and vascular compromise significantly complicate elective breast surgeries in previously irradiated patients, despite a high incidence of post-treatment asymmetry and macromastia. This study evaluates complication profiles across surgical techniques to provide evidence-based guidance for breast reduction and mastopexy in irradiated breasts. A retrospective cohort of 41 patients with prior breast irradiation undergoing bilateral breast reduction (n=30), mastopexy (n=7), or mixed procedures (n=4) between 2015-2023 was analyzed. Reduction techniques included superior pedicle (n=4), supero-medial pedicle (n=12), and Thorek free nipple graft (n=14); mastopexies employed Wise-pattern superior pedicle techniques. Outcomes were compared with 1442 non-irradiated breast reductions. Complications were classified as major (reoperation required) or minor (conservatively managed). Statistical analysis utilized Fisher's exact tests and odds ratios (SPSS v26). Irradiated breasts had significantly higher overall complication rates compared to non-irradiated controls (44% vs. 21%; OR 3.017; p=0.0014), with seromas occurring exclusively in irradiated cases (15% vs. 0%; p<0.001). Mastopexies showed higher dehiscence rates than reductions (29% vs. 0%; p=0.01). Among reduction techniques, the supero-medial pedicle had the highest complication rate (67%), while Thorek procedures showed no seromas (p=0.04). No major complications occurred; revisions (15%) addressed asymmetry only. Breast surgery in irradiated fields remains viable when guided by technique-specific risk profiles and conservative tissue handling. Mastopexy carries elevated dehiscence risks, while the Thorek technique may reduce seroma formation in high-risk patients. Because subgroup analyses were under-powered, all technique-related recommendations should be regarded as preliminary and hypothesis-generating, pending validation in larger multicentre cohorts.