Upper Extremity Lymphedema​​​: Performance of a Risk Prediction Model in Women Undergoing Surgery for Breast Cancer

Area of Science:

• Oncology
• Lymphedema Research
• Clinical Prediction Modeling

Background:

• Upper extremity lymphedema is a frequent complication after breast cancer treatment.
• Early detection and intervention tools are crucial for managing lymphedema.
• A predictive formula for lymphedema risk was proposed by Kwan et al. in 2020.

Purpose of the Study:

• To assess the predictive performance of the Kwan lymphedema risk model.
• To evaluate the model's classification of lymphedema risk (low, moderate, high) in Brazilian women.
• To determine the model's accuracy, sensitivity, specificity, and predictive values.

Main Methods:

• Retrospective analysis of 190 breast cancer surgery patients with at least 6 months follow-up.
• Application of the Kwan model using variables: age, BMI, mammographic density, lymph node status, and axillary surgery type.
• Comparison of predicted lymphedema risk with observed outcomes via arm volumetry.

Main Results:

• The Kwan model exhibited low predictive accuracy (RMSE=223.54, R2=-0.18).
• Overall accuracy was 76%, but sensitivity was low (48%) with a positive predictive value of 22%.
• Specificity was 79% and negative predictive value was high (92%), influenced by low lymphedema incidence.

Conclusions:

• The Kwan model demonstrated limited effectiveness in predicting positive lymphedema cases.
• The model may serve as a screening tool for ruling out lymphedema risk.
• Future models incorporating machine learning and broader variables could enhance positive predictive accuracy.