Circulating tumor cells predict relapse in head and neck mucosal squamous cell carcinoma
European Journal of Surgical Oncology. 2026;52(2):111357. DOI 10.1016/j.ejso.2025.111357.
How to cite
AMA
Yeo D, Liang X, Bastian A, Strauss H, Lim G, Mahfouz FM, et al. Circulating tumor cells predict relapse in head and neck mucosal squamous cell carcinoma. Eur J Surg Oncol. 2026;52(2):111357. doi:10.1016/j.ejso.2025.111357APA
Yeo, D., Liang, X., Bastian, A., Strauss, H., Lim, G., Mahfouz, F. M., et al. (2026). Circulating tumor cells predict relapse in head and neck mucosal squamous cell carcinoma. European Journal of Surgical Oncology, 52(2), 111357. https://doi.org/10.1016/j.ejso.2025.111357BibTeX
@article{yeo2026circulating,
title = {Circulating tumor cells predict relapse in head and neck mucosal squamous cell carcinoma},
author = {Yeo, Dannel and Liang, Xiaoqi and Bastian, Althea and Strauss, Heidi and Lim, Grace and Mahfouz, Fawaz M. and others},
journal = {European Journal of Surgical Oncology},
volume = {52},
number = {2},
pages = {111357},
year = {2026},
doi = {10.1016/j.ejso.2025.111357}
}Head and neck cancers relapse in more than half of patients within two years of surgery, yet no blood test flags who is at highest risk. Circulating tumor cells are a candidate marker, but they are rare and easily missed, and methods that capture them by an epithelial antigen can overlook cells that express that antigen weakly.
This study drew blood from patients before and after head and neck cancer surgery and processed it with an enrichment-free, slide-imaging method that isolates cells by density rather than by antigen, then asked whether the counts predicted relapse.
Circulating tumor cells were found in most mucosal squamous cell carcinoma patients, and a high preoperative count was independently associated with relapse. In an exploratory set of post-surgery samples the association was stronger still, though based on few events.
Key findings
- Circulating tumor cells were detected in 81% of mucosal squamous cell carcinoma patients. The workflow found CTCs in 46 of 57 HNmSCC patients (mean 19 per 7.5 mL of blood), at 89% specificity and 98% positive predictive value across epithelial head and neck cancers, benchmarked against 9 non-epithelial controls.
- A high preoperative CTC count predicted relapse independently of standard pathology. Patients with 5 or more CTCs had a 9.5-fold higher odds of relapse (odds ratio 9.52, p=0.022), and across the individual clinicopathologic factors the adjusted odds ratio ran from 4.44 to 10.09, near 8 to 10 for most but attenuated when adjusting for tumor stage or size.
- CTC status and surgical margin involvement carried complementary, independent prognostic value. In one adjusted model both CTC-high status (adjusted odds ratio 10.09, p=0.043) and positive margins (adjusted odds ratio 6.17, p=0.024) remained significant, and 36% (14 of 39) of CTC-high patients relapsed versus 6% (1 of 18) of CTC-low patients.
The AccuCyte–CyteFinder workflow in the methods
“7.5 mL of whole blood were processed using the AccuCyte-CyteFinder platform (RareCyte)”
— Yeo et al., European Journal of Surgical Oncology (2026), Methods, “CTC detection using the AccuCyte-CyteFinder platform”
Disclosure: RareCyte is named in the competing-interests statement of the publication cited above.
Why it matters for The AccuCyte–CyteFinder workflow users
If you are weighing circulating tumor cells as a prognostic readout, look at what this study needed from its detection step. The finding rests on counting rare tumor cells reliably enough to set a threshold — here, five cells — and then showing that threshold separates the patients who relapse from those who do not. That only works if you do not lose cells before you count them. The AccuCyte–CyteFinder workflow isolates every nucleated cell by density instead of selecting for an epithelial antigen first, so cells that carry little or no EpCAM are still on the slide when imaging begins. The authors stained two slides per patient with a fixed four-marker readout — a nuclear dye, cytokeratin, EpCAM and CD45 — and called a cell a CTC when it was nucleated, CD45-negative, and above a set epithelial-marker intensity. For your own work the trade is explicit: you image a large nucleated-cell population rather than a pre-selected subset, and that is what let this study detect CTCs in 81% of patients and hold a prognostic signal after adjustment. When the endpoint is a count you will threshold on, keeping every candidate cell in view is the property that carries it.
