International Journal of Radiation Research

en Radiotherapy-associated alterations in granulomatous mastitis: Clinicopathological and immunohistochemical predictors of recurrence and differential diagnosis from breast tumors Radiation Biology Radiation Biology تحقيق بديع Original Research <div style="text-align: justify;">Background: Granulomatous mastitis (GM) often recurs, and its clinical and pathological overlap with breast malignancies is especially problematic in patients with prior radiotherapy. To identify clinical, pathological, and immunohistochemical predictors of GM recurrence and to evaluate diagnostic challenges in the context of post-radiotherapy alterations. Materials and Methods: A retrospective cohort of 206 GM patients was analyzed (58 with recurrence, 148 without). Baseline clinical factors, pathological features, and immunohistochemical markers were compared. An XGBoost algorithm was applied to construct and validate a recurrence risk model. For patients with a history of breast radiotherapy, additional subgroup analysis was performed, incorporating imaging and pathology review to differentiate GM recurrence from tumor relapse. Results: Recurrence was associated with younger age (<35 years, 41.4% vs. 18.9%, P<0.001), higher prolactin levels (32.5±8.7 vs. 24.8±6.9 ng/mL, P<0.001), and higher BMI (26.4±3.8 vs. 24.1±3.2, P=0.007). Pathological indicators included multinucleated giant cells (31.0% vs. 12.2%, P<0.001), diffuse plasma cell infiltration (82.8% vs. 62.2%, P=0.003), and elevated Ki-67 (65.5% vs. 32.4%, P<0.001). The model achieved strong predictive performance (AUC = 0.89 training, 0.85 validation, 0.83 external). In post-radiotherapy patients, fibrotic distortion and atypical immune marker expression (Ki-67, PD-L1) frequently mimicked malignancy, highlighting the need for integrated clinicopathologic assessment. Conclusion: GM recurrence is driven by hormonal imbalance, immune dysregulation, and tissue injury. Radiotherapy-associated alterations exacerbate diagnostic overlap with breast tumors, underscoring the value of multidimensional prediction models for risk stratification and differential diagnosis.</div> Granulomatous mastitis, breast neoplasms, recurrence, radiotherapy, immunohistochemistry, risk factors, predictive models. 17 24 http://ijrr.com/browse.php?a_code=A-10-1-1449&slc_lang=en&sid=1 X. Zhao 7900319475328460032691 7900319475328460032691 No School for Pregnant Women, The Fourth Hospital of Shijiazhuang Pregnant Women School, Shijiazhuang, Hebei Province, China J.Y. Li lxlong201909@163.com 7900319475328460032692 7900319475328460032692 Yes epartment of Breast Surgery, The Fourth Hospital of Shijiazhuang, Shijiazhuang, Hebei Province, China C.W. Zhang 7900319475328460032693 7900319475328460032693 No Department of Health Management, Zhangjiakou First Hospital, Zhangjiakou, Hebei Province, China L.J. Ma 7900319475328460032694 7900319475328460032694 No epartment of Breast Surgery, The Fourth Hospital of Shijiazhuang, Shijiazhuang, Hebei Province, China Y. Zhao 7900319475328460032695 7900319475328460032695 No epartment of Breast Surgery, The Fourth Hospital of Shijiazhuang, Shijiazhuang, Hebei Province, China X.H. Chen 7900319475328460032696 7900319475328460032696 No epartment of Breast Surgery, The Fourth Hospital of Shijiazhuang, Shijiazhuang, Hebei Province, China X.L. Li xaohecen20XH@163.com 7900319475328460032697 7900319475328460032697 No epartment of Breast Surgery, The Fourth Hospital of Shijiazhuang, Shijiazhuang, Hebei Province, China