Identification of a Cytokine Biomarker for Prognostic Modeling of Breast Cancer-Related Lymphedema

Authors

Alison J Wu
Neil Lin
Jie Su
Cherie Lin
Madison-Shira Hossack
Wei Shi
Farnoosh Abbas-Aghababazadeh
Wei Xu
Benjamin Haibe-Kains
Simona F Shaitelman
Melissa B Aldrich
Fei-Fei Liu
Jennifer Y Y Kwan

Language

English

Publication Date

2-5-2026

Journal

Cancer Research Communications

DOI

10.1158/2767-9764.CRC-25-0541

PMID

41640157

Abstract

Lymphedema is a chronic complication of breast cancer treatment, and early intervention is crucial to reduce morbidity. This study evaluated the role of blood-based cytokine biomarkers in the prognostication of breast cancer-related lymphedema (BCRL) to improve risk prediction. A secondary analysis of inflammatory biomarkers for BCRL was performed using a previously published cohort of 147 breast cancer patients who had undergone serum cytokine profiling during their treatment at the Princess Margaret Cancer Centre from 2010-2014. Prognostic cytokine variables for lymphedema were selected by regression analysis and independence from known clinical risk factors. Regression-based modeling was employed to integrate prognostic variables for the prediction of lymphedema occurrence. We identified the immunostimulatory cytokine IFN-α2A as a potential biomarker for lymphedema development (OR 3.10, 95% CI 1.05-9.51, p=0.042), independent from known clinical risk factors. Furthermore, Kaplan-Meier analysis demonstrated 3-year lymphedema-free survival of 95% (90-100%) vs. 85% (77-94%) for below vs. above median concentrations of IFN-α2A (p=0.026). In combination with an established clinical risk regression-based model, patients identified as high risk based on clinical factors alone were able to be correctly reclassified as low risk by IFN-α2A in 31% (8/26) of cases. Our combined logistic regression model using both IFN-α2A and clinical risk score achieved an AUC of 0.895 (95% CI: 0.796 - 0.971) and Brier Score of 0.101 (95% CI: 0.061 - 0.149), representing a favourable improvement compared to the logistic regression model using clinical risk factors alone. IFN-α2A in combination with established clinical risk factors may be useful for improving BCRL prognostication.

Published Open-Access

yes

Recommended Citation

Wu, Alison J; Lin, Neil; Su, Jie; et al., "Identification of a Cytokine Biomarker for Prognostic Modeling of Breast Cancer-Related Lymphedema" (2026). Faculty, Staff and Student Publications. 5254.
https://digitalcommons.library.tmc.edu/uthgsbs_docs/5254