Evolution of delayed resistance to immunotherapy in a melanoma responder
Nature Medicine. 2021;27(6):985-992. DOI 10.1038/s41591-021-01331-8. PMID 33941922. PMCID PMC8474080.
How to cite
AMA
Liu D, Lin JR, Robitschek EJ, Kasumova GG, Heyde A, Shi A, et al. Evolution of delayed resistance to immunotherapy in a melanoma responder. Nat Med. 2021;27(6):985-992. doi:10.1038/s41591-021-01331-8APA
Liu, D., Lin, J. R., Robitschek, E. J., Kasumova, G. G., Heyde, A., Shi, A., et al. (2021). Evolution of delayed resistance to immunotherapy in a melanoma responder. Nature Medicine, 27(6), 985-992. https://doi.org/10.1038/s41591-021-01331-8BibTeX
@article{liu2021evolution,
title = {Evolution of delayed resistance to immunotherapy in a melanoma responder},
author = {Liu, David and Lin, Jia-Ren and Robitschek, Emily J. and Kasumova, Gyulnara G. and Heyde, Alex and Shi, Alvin and others},
journal = {Nature Medicine},
volume = {27},
number = {6},
pages = {985--992},
year = {2021},
doi = {10.1038/s41591-021-01331-8},
pmid = {33941922}
}Some melanomas respond well to immune checkpoint blockade and then, years later, return. Following that slow relapse inside one patient is hard, because it means tracking how the tumor and the immune cells around it change across dozens of samples and many years, one cell at a time.
Over 9 years and 37 tumor samples from a single patient who responded to immunotherapy and later died of disease, the authors combined whole-exome and RNA sequencing with whole-slide multiplexed tissue imaging. They reconstructed seven co-evolving tumor lineages and traced a de-differentiated, neural-crest-like tumor population that was PD-L1-high and sat close to immune cells, with different spatial patterns in subcutaneous versus lung metastases.
Key findings
- A single melanoma patient was followed across 9 years and 37 longitudinal tumor samples. Phylogenetic reconstruction resolved 7 co-evolving lineages, each carrying convergent but independent resistance-associated alterations.
- All recurrent tumors arose from one lineage marked by loss of chromosome 15q. Post-treatment clones then acquired additional genomic driver events as resistance emerged.
- A PD-L1-high, NGFR-high neural-crest-like tumor population sat close to immune cells. Rapid autopsy resolved 2 distinct spatial patterns: high immune proximity in subcutaneous tumors versus a diffuse arrangement in lung metastases.
CyteFinder in the methods
“Imaging was performed on a CyteFinder slide scanning fluorescence microscope (RareCyte Inc. Seattle WA) using a 10X objective (for batch 1) and 20X objective (for batch 2&3).”
— Liu et al., Nature Medicine (2021), Methods, “MULTIPLEXED IMMUNOFLUORESCENCE”
Disclosure: RareCyte is named in the competing-interests statement of the publication cited above.
Why it matters for CyteFinder users
If you are weighing the CyteFinder for a tissue-imaging program, this study shows what it supports across a demanding longitudinal cohort. The biology here is spatial: whether a PD-L1-high, NGFR-high tumor population sits next to immune cells, and how that arrangement differs between subcutaneous and lung metastases, resolves only when you image whole FFPE sections at single-cell resolution instead of sampling small fields. The authors ran tissue-based cyclic immunofluorescence (t-CyCIF), acquiring each staining and imaging cycle on the CyteFinder slide-scanning microscope and stacking the registered cycles into single-cell intensity values for markers including S100, MITF, CD3, and NGFR. Those single-cell coordinates then fed the gating, clustering, and spatial-proximity analysis that carried the paper's imaging conclusions. For your own work the takeaway is concrete: when the signal lives in where cells sit relative to one another across a whole specimen, the CyteFinder is the instrument that captures it, cycle after cycle, on a section you can return to.
