Comprehensive Longitudinal Linear Mixed Modelling of CTCs Illuminates the Role of Trop2, EpCAM, and CD45 in CTC Clustering and Metastasis
Cancers. 2025;17(16):2717. DOI 10.3390/cancers17162717.
How to cite
AMA
Merkley SD, Kang H, Brown-Glaberman U, Marchetti D. Comprehensive Longitudinal Linear Mixed Modelling of CTCs Illuminates the Role of Trop2, EpCAM, and CD45 in CTC Clustering and Metastasis. Cancers. 2025;17(16):2717. doi:10.3390/cancers17162717APA
Merkley, S. D., Kang, H., Brown-Glaberman, U., & Marchetti, D. (2025). Comprehensive Longitudinal Linear Mixed Modelling of CTCs Illuminates the Role of Trop2, EpCAM, and CD45 in CTC Clustering and Metastasis. Cancers, 17(16), 2717. https://doi.org/10.3390/cancers17162717BibTeX
@article{merkley2025comprehensive,
title = {Comprehensive Longitudinal Linear Mixed Modelling of CTCs Illuminates the Role of Trop2, EpCAM, and CD45 in CTC Clustering and Metastasis},
author = {Merkley, Seth D. and Kang, Huining and Brown-Glaberman, Ursa and Marchetti, Dario},
journal = {Cancers},
volume = {17},
number = {16},
pages = {2717},
year = {2025},
doi = {10.3390/cancers17162717}
}Circulating tumor cells are the cells that carry a cancer from its primary site into the bloodstream and on to distant organs, and their presence signals a poorer prognosis. But they vary widely between patients and even within one patient, and the classical way of spotting them — EpCAM-positive, cytokeratin-positive, CD45-negative — was built around epithelial tumors, so it can miss cells that have drifted toward a mesenchymal state.
This team followed 51 metastatic breast cancer patients over 32 months, counting both classical and Trop2-positive circulating tumor cells across 305 blood samples and relating each marker to how the cells cluster. EpCAM, Trop2, and the presence of CD45-positive cells all predicted cluster size, while pan-cytokeratin did not, and EpCAM and larger clusters rose over time in HER2-positive but not hormone-receptor-positive cancers after metastasis was diagnosed.
Key findings
- Two circulating tumor cell populations were tracked across a 32-month longitudinal series. 305 blood samples from 51 metastatic breast cancer patients were scored — 205 samples from 51 patients for classical CTCs and 100 samples from 26 patients for Trop2+ CTCs — yielding more than 6,000 images for review.
- EpCAM, Trop2, and CD45+ cells predicted CTC cluster size; pan-cytokeratin did not. Among the strong correlations (r ≥ 0.7), EpCAM+ CTCs and the presence of CD45+ cells tracked cluster presence and size — with a correlation of 0.78 between EpCAM+ CTCs and CD45+ cells — while pan-cytokeratin reached no strong correlation with any biomarker.
- EpCAM and larger CTC clusters increased over time in HER2+ but not HR+ cancers after metastatic diagnosis. In the longitudinal model, slopes differed significantly between receptor subtypes for EpCAM (p = 0.02183) and for clusters of more than two cells (p = 0.00691), both rising in HER2+ disease while HR+ showed no significant change.
CyteFinder II in the methods
“All slides were scanned using the Cytefinder II® imaging platform and analyzed with CyteHub® software. Images output by CytHub were then quality-checked by trained technicians to confirm classification as individual CTCs and/or clusters.”
— Merkley et al., Cancers (2025), Methods, “Sample Processing, Scanning, and Analysis”
Why it matters for CyteFinder II users
If you are weighing CyteFinder II for rare-cell work, look at what it carried here. Every slide in this 32-month study was scanned on the CyteFinder II platform, and the images it produced were analyzed in CyteHub to enumerate single CTCs and to classify each as a singlet or as a two-, three-, or larger-cell cluster. That cluster call is the unit the whole analysis rests on: the correlations with EpCAM, Trop2, and CD45, and the receptor-subtype trajectories over time, are all built from cluster presence and size read off these images. Two things follow for your own workflow. Because the entire nucleated-cell layer is laid onto the slide before scanning rather than captured on EpCAM first, the imaging step sees Trop2-positive and EpCAM-low cells that an enrichment step would have discarded. And because clusters are scored image by image with technician review, heterotypic CD45-positive configurations are counted rather than dissociated away. For a rare, heterogeneous population, that whole-slide readout is the point.
