Nomogram for Predicting Risk of Arm Lymphedema Following Axillary Lymph Node Dissection in Breast Cancer Patients

Miaomiao Jia¹Lihui Pan¹Haibo Yang¹Jinnan Gao¹Wenzhuang Shen^2*Xiaojun Zhang^1*

• ¹Department of Breast Surgery, Shanxi Bethune Hospital, Taiyuan, China
• ²Huazhong University of Science and Technology Tongji Medical College Tongji Hospital, Wuhan, China

Purpose: Breast cancer-related arm lymphedema (BCRaL) is a prevalent and severe complication post-breast cancer treatment, especially following axillary lymph node dissection (ALND). This study aimed to develop a nomogram for BCRaL risk prediction by identifying and integrating key risk factors, including chemotherapy type (neoadjuvant vs. adjuvant), to enhance individualized patient monitoring and prevention strategies. Patients and Methods: We conducted a retrospective analysis of clinical data from 535 breast cancer patients who received ALND and chemotherapy. Patients were divided into a training cohort (70%) and a validation cohort (30%). Univariate and multivariate Cox regression analyses identified independent risk factors for BCRaL, which were subsequently used to construct a nomogram. The model's performance was assessed through calibration curves, ROC curves, and clinical decision curve analysis (DCA). Results: The incidence of BCRaL in our cohort was 20.6%. Multivariate analysis identified several independent risk factors for BCRaL, including elevated body mass index (BMI), increased number of positive axillary lymph nodes, neoadjuvant chemotherapy (NAC), HER2-targeted therapy, and supraclavicular radiotherapy (SCRT). The nomogram developed based on these factors demonstrated strong predictive accuracy, with C-index values of 0.692 in the training cohort and 0.719 in the validation cohort. ROC curve analysis revealed AUC values reaching 0.760, indicating good discriminative ability. Time-dependent ROC curves further confirmed the model's consistency across different follow-up periods. DCA validated the clinical utility of the nomogram, while survival analysis clearly distinguished between high-risk and low-risk BCRaL groups. Conclusion: This study developed and internally validated a predictive model that integrates modern treatment modalities (NAC, HER2-targeted therapy, SCRT) with traditional risk factors to identify high-risk BCRaL patients undergoing ALND and chemotherapy. The model requires external validation in future studies. Consequently, the nomogram presents a potential tool for strategizing precision prevention, necessitating further evaluation before its broader adoption in clinical practice.