Immune organization in sentinel lymph nodes of melanoma patients is prognostic of distant metastases
Amitay, Y.; Milo, I.; Keidar-Haran, T.; Deis, S.; Truzman, G.; Elhanani, O.; Salame, T. M.; Azimov, M.; Stein, I.; Cohen, J.; Lotem, M.; Pikarsky, E.; Keren, L.
Show abstract
Sentinel lymph node (sLN) biopsy is part of melanoma staging, as involved LNs indicate a higher risk of recurrence. However, how the sLN is shaped by the tumor and reciprocally affects metastatic progression is poorly understood. Here, we mapped immune organization in involved and non-involved sLNs of 69 melanoma patients using high-resolution spatial proteomics, spatial transcriptomics and deep learning, leveraging the data for prognostic evaluation. In patients with involved LNs, a robust T cell response correlated with absence of recurrence. In non-involved LNs, protection from metastases was linked to expanded sinuses colocalized with plasmablasts whereas CCR7+ cells and Tregs correlated with future development of distant metastases. We trained a model that predicts development of distant metastases with 79% and 93% AUC for non-metastatic and metastatic LNs, respectively. Our findings reveal conserved immune patterns in sLNs that prospectively identify patients at risk for metastatic disease and may aid in therapeutic decisions.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Cancer-associated fibroblast compositions change with breast cancer progression linking S100A4 and PDPN ratios with clinical outcome 97%
- Metastasis founder cells activate immunosuppression early in human melanoma metastatic colonization 96%
- Cell lineage as a predictor of immune response in neuroblastoma 96%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.