Risk factors for granulomatous mastitis and establishment and validation of a clinical prediction model (Nomogram)

Abstract

Abstract Purpose: This study aimed to explore the risk factors and clinical characteristics of granulomatous mastitis (GM) through a case-control study and establish and validate a clinical prediction model (Nomogram). Method: This retrospective study was conducted at Beijing Hospital of Traditional Chinese Medicine, Capital Medical University, Xiyuan Hospital of China Academy of Chinese Medical Sciences and Guang’ anmen Hospital, China Academy of Chinese Medical Sciences from June 2017 to December 2021. In the design of the case-control study, a total of 1634 GM patients and 186 healthy women during the same time period were included and randomly divided into the modeling group and validation group with a 7:3 ratio. To identify the independent risk factors of GM, univariate and multivariate logistic analyses were conducted and used to develop a Nomogram . The prediction model was internally and externally validated using the Bootstrap technique and validation cohort. The receiver operating characteristic (ROC) curve and the calibration curve were used to evaluate the discrimination and calibration of the prediction model. Decision curve analysis (DCA) and clinical impact curve (CIC) were utilized to evaluate the clinical significance of the model. Result: The average age of GM patients was 33.14 (mainly 20 to 40). The incidence was high within five years after delivery. It mainly occurs in the unilateral breast. Majority of the patients exhibited local skin alterations, while some also presented with systemic symptoms. Univariate analysis showed GM was relevant to gestation history, menopause, nipple discharge and invagination, high prolactin level, sex hormone intake, thyroid function, SDS score, breast trauma and diet preference (P < 0.05). Multivariate logistic analysis showed ages (20-40 years old), high prolactin level, sex hormone intake, breast trauma, nipple discharge or invagination and high SDS score were independent risk factors for GM. The mean area under the curve (AUC) in the modeling group was 0.899, and the AUC in the validation group was 0.889. The internal and external validation demonstrated the model's predictive ability and clinical value. Conclusion: The lactation-related factors are the main risk factors of GM, which could lead to milk siltation or ductal secretion increasing. Meanwhile, hormone disorders could affect the secretion and the expansion of mammary ducts. They all can obstruct or injure the duct, inducing inflammatory reactions and immune responses. Blunt trauma, depressed mood and diet preference can accelerate the process. The Nomogram can effectively predict the risk of GM's occurrence.