A systematic review of risk prediction models for breast cancer related lymphedema

LU Ningning; HU Fangqi; HUANG Kun; WU Yuqing; SHI Yanyan; CHENG Fang

doi:10.12102/j.issn.2095-8668.2025.07.004

Abstract：Objective:To systematically search and evaluate the performance of risk prediction models for breast cancer related lymphedema.Methods:The relevant studies on risk prediction models of breast cancer related lymphedema in CNKI,WanFang Database,CBM,VIP,PubMed,Web of Science and the Cochrane Library were searched by computer,and the languages were limited to Chinese and English.The retrieval time was from the inception to October 2023.Two researchers independently screened literature and extracted data and analyzed the bias risk and applicability of the included literature.Results:A total of 25 articles were included,including 25 prediction models,with a statistical sample size ranging from 303 to 5 549 cases,and the number of outcome events ranging from 62 to 639 cases.The AUC included models ranged from 0.680 to 0.952.The most common predictors were BMI,axillary lymph node dissection level,radiotherapy,chemotherapy,postoperative complications,etc.All included studies had good applicability,but all had a high risk of bias,which mainly came from failure to select appropriate data sources,unreasonable variable screening,failure to report or use blind methods,and incomplete evaluation of model performance.Conclusion:Current evidence indicates that the research on breast cancer related lymphedema risk prediction model is still in the development stage,showing good differentiation and applicability overall,but the quality of the model needs to be improved,and future prospective cohort studies should be conducted and develop a prediction model applicable to a wider population.