MechanoAge, a machine learning platform to identify individuals susceptible to breast cancer based on mechanical properties of single cells

作者：Stefan Hinz, Sturla Magnus Grøndal, Masaru Miyano, Jennifer C. Lopez, Kristen L. Cotner, Taylor Thomsen, Chang Chen, Edward J Hester, Lisa D. Yee, Victoria L. Seewaldt, James B. Lorens, Lydia L. Sohn, Mark A. LaBarge · 发表于：EBioMedicine · 年份：2026 · DOI：10.1016/j.ebiom.2026.106241 · 研究领域：Cellular Mechanics and Interactions、Cancer Cells and Metastasis、AI in cancer detection

BACKGROUND: Emerging evidence links cellular ageing and biophysical alterations with cancer susceptibility. Existing breast cancer risk models inadequately identify individuals at latent risk, particularly among women without known genetic mutations or family history. Risk is often underestimated or overestimated due to reliance on population-level data and absence of individualised tissue-based markers of breast cancer risk. METHODS: We profiled primary human mammary epithelial cells (HMECs) from women of varying ages and risk backgrounds using mechano-node-pore sensing (mechano-NPS), a high-throughput microfluidic platform that measures single-cell physical and mechanical properties. We developed a machine learning classifier, MechanoAge, to estimate chronological age based on mechanical phenotypes, and a biological age-based risk index, Mechano-RISQ. We further assessed cytoskeletal protein keratin 14 (KRT14) as a key mediator of underlying mechanical states through overexpression and knockdown experiments. FINDINGS: Epithelial cells from normal tissue of young BRCA1/2 mutation carriers (n = 4), women with family history of breast cancer (n = 3), and tissue contralateral to a tumour-bearing breast (n = 9) exhibited elevated Mechano-RISQ scores, which reflects accelerated biological ageing compared to age-matched controls (n = 18). KRT14 overexpression induced a biologically aged phenotype in cells obtained from younger women, whereas knockdown partially reversed this state...