Topological analysis of the human lymph node reticular network predicts outcome in breast cancer.

Axillary LNs (ALNs) initiate immune responses in breast cancer (BC) but how and when ALNs become dysfunctional, facilitating metastasis, is unclear. The fibroblastic reticular cell (FRC) network within ALNs provides structural support and mediates immune homeostasis, but we have yet to elucidate whether this network changes during BC progression. An unbiased computational approach was used to quantify features of the immunolabelled FRC network in ALNs derived from patients with BC. Platelet-derived growth factor receptor β (PDGFRβ) was identified as a robust immunomarker for human FRC and used to quantify how FRC network topology changes during BC progression and after treatment. Formalin-fixed paraffin-embedded ALNs (n = 331) from 179 patients with BC and 23 benign reactive controls were assessed for FRC network metrics, including lacunarity and branchpoints, alongside de-identified clinico-pathological data. These data were then integrated using multivariate, principal component and survival analyses. In node-negative, treatment-naïve triple-negative BC, denser FRC networks in uninvolved nodes significantly improved survival (p = 0.0365). Conversely, similar changes seen in node-positive BC significantly worsened survival (p = 0.0407), regardless of BC subtype or treatment. In metastatic ALNs, FRC network disruption grew proportionately to axillary tumour burden, and this significantly correlated with poorer outcomes (p = 0.043). Interestingly, increased FRC alignment within these metastases significantly improved survival (p = 0.0205). This study showed that changes in human ALN FRC network topology predicts BC prognosis. This could improve how we risk stratify patients in future, and provide a new avenue for mechanistic, translational research. © 2026 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.