Pre-treatment clinico-dosimetric risk stratification to inform tumor bed boost–associated skin toxicity in breast cancer radiotherapy: a propensity score–matched retrospective study
Year: 2026
Type: article
Study design: Observational Study
Abstract: Background Radiation-induced skin injury (RISI) is a frequent and clinically relevant toxicity in breast cancer patients receiving tumor bed boost following breast-conserving surgery. Although boost irradiation improves local control, the associated increase in skin dose may adversely affect treatment tolerance and cosmetic outcomes. In real-world practice, the risk of RISI is determined by a combination of dosimetric and patient-related factors; however, pre-treatment, decision-oriented risk stratification remains limited. Methods This retrospective single-center study included 140 consecutive breast cancer patients treated with breast-conserving surgery and adjuvant radiotherapy between 2020 and 2025. Seventeen patients were excluded because of incomplete clinical or dosimetric records. Propensity score matching (1:1) was performed based on age, body mass index (BMI), and concurrent chemotherapy, yielding 58 matched pairs (N = 116). Acute RISI was graded according to the RTOG acute radiation morbidity scoring criteria, with Grade ≥2 defined as clinically significant toxicity. Multivariable logistic regression was used to identify independent risk factors. Model performance was assessed using receiver operating characteristic (ROC) analysis, calibration, and decision curve analysis (DCA). Results In the matched cohort, Grade ≥2 RISI occurred in 27 of 58 patients (46.6%) receiving tumor bed boost and in 20 of 58 patients (34.5%) without boost. Skin maximum dose (D max ) (OR, 1.11 per Gy; 95% CI, 1.01–1.21; P = .027) and concurrent chemotherapy (OR, 2.39; 95% CI, 1.06–5.39; P = .036) were independently associated with RISI. The model demonstrated moderate discriminative performance (AUC, 0.74; 95% CI, 0.66–0.82) with good calibration. DCA suggested a modest net benefit within a limited range of intermediate threshold probabilities. Conclusion Tumor bed boost–associated RISI is predictable based on pre-treatment clinico-dosimetric factors. This risk stratification model may assist clinicians in identifying patients at elevated risk before treatment planning and support individualized skin-sparing strategies. As a single-center analysis, the model requires external validation in independent, multi-institutional cohorts before routine clinical use.
Source: Frontiers in Oncology
Authors: Wei Ye(Conceptualization, Formal analysis, Methodology, Writing – original draft), Xiaofeng Wan(Data curation, Investigation, Writing – review & editing), Ling Gui(Data curation, Investigation, Writing – review & editing), Lu Wang(Supervision, Writing – review & editing), Hongbo Zuo(Supervision, Writing – review & editing)
Institution: Jiujiang First People's Hospital
Language: English
Subfield: Radiation
Field: Physics and Astronomy
Domain: Physical Sciences
Keywords: radiation dermatitis, breast cancer, tumor bed boost, breast-conserving surgery, adjuvant radiotherapy, risk stratification, skin dose, concurrent chemotherapy, acute radiation toxicity, risk prediction model, decision curve analysis, propensity score matching, body mass index, treatment tolerance, cosmetic outcome, local control
Open Access status: gold
Estimated APC paid: $3,801
OpenAlex ID: W7235250683
