Predicting Breast Cancer-Related Lymphedema (BCRL) following Axillary Lymph Node Dissection (ALND): A Systematic Review of Current Prediction Models

BACKGROUND: Breast cancer-related lymphedema (BCRL) is a debilitating complication affecting one-third of patients undergoing axillary lymph node dissection (ALND). Since BCRL is incurable, early risk identification is essential for prevention. Prediction models have been developed to assess BCRL risk based on demographic, oncologic, and treatment-related factors. However, variability in study populations and model performance has led to inconsistent findings, limiting clinical applicability. This systematic review critically evaluates existing prediction models for BCRL, identifies key risk factors, and synthesizes current evidence to enhance risk stratification. METHODS: A systematic review was conducted on PubMed, Embase, and Web of Science to identify studies reporting prediction models for BCRL in patients undergoing ALND in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines. Outcome data included the cumulative incidence of lymphedema, time from ALND to BCRL diagnosis, model performance (Area Under the Curve), predicted risk factors, odds ratio (OR), and hazard ratio (HR). Studies that include patients who have not had ALND, less than 12 months of follow-up (mean or median), lymphedema before ALND, or lacking incidence of BCRL were excluded. Only statistical models with more than two clinical risk factors were included. RESULTS: Of 1,827 screened articles, 38 met the inclusion criteria. Overall, 21,466 patients with a mean follow-up time of 50.8 months were included. 27.5% of patients developed BCRL on average 16.9 months following ALND. The predictors most commonly associated with BCRL were BMI (19 models), radiation (18 models), chemotherapy (17 models), number of axillary lymph nodes removed (11 models), age (7 models), surgical complications (7 models), level III ALND (5 models), race (3 models), and number of positive lymph nodes (3 models). Higher BMI was significantly associated with an increased risk of BCRL, with BMI >30 showing ORs ranging from 2.36 to 5.27 (p<0.05). Non-white race was linked to a higher BCRL risk (OR 2.34–4.54, p<0.05). Radiation therapy (OR 1.93–15.34, HR 1.38–4.90; p<0.05) and chemotherapy (OR 1.39–5.30, HR 1.45–4.14, p<0.05) significantly increased the risk. The extent of ALND was a key predictor, with the removal of more than 10 nodes associated with ORs of 1.71–2.57 (p<0.05), while the removal of more than 20 nodes had an OR of 5.65 (p<0.05). Level III dissection also elevated risk (OR 2.02–7.89, HR 2.39; p<0.05). Surgical complications within 30 days, including seroma and infections, were significantly associated with BCRL development (OR 1.46–5.27, p<0.05). CONCLUSION: This systematic review is the most comprehensive study that synthesizes existing prediction models for BCRL to date. Patients with higher BMI, racial minorities, and those who receive radiation, chemotherapy, or require more extensive ALNDs are more likely to develop BCRL following surgery. Despite variations in predictive models, these risk factors were consistently identified. However, differences in model validation highlight the need for a standardized, clinically applicable prediction tool to guide prevention strategies.