Head and neck cancers rank among the ten most common cancer types worldwide and are frequently diagnosed at advanced stages, when treatment options narrow and prognoses worsen. A defining feature of these tumours is a profoundly immunosuppressive microenvironment, dominated by inhibitory checkpoint molecules such as PD-1 and CTLA-4 that dampen antitumour T-cell responses. Although antibodies targeting PD-1 and CTLA-4 have improved outcomes for some patients, a substantial proportion derive little benefit, a clinical reality that has long suggested the existence of additional immunoregulatory mechanisms operating beyond the classical checkpoint pathways that current drugs address.

The research team, led by investigators at the University Hospital Mannheim and Heidelberg University, isolated small extracellular vesicles from the plasma of 50 head and neck cancer patients, 20 patients with no evident disease after therapy, and 22 healthy donors. Using size-exclusion chromatography, a gentle separation method that preserves vesicle integrity, the researchers obtained preparations that met current international criteria for extracellular vesicle characterisation. Transmission electron microscopy revealed the typical spherical, cup-shaped morphology, while nanoparticle tracking analysis placed the mean diameter at roughly 96 nanometres, with size distributions peaking near 100 nanometres. Western blotting confirmed enrichment of vesicle-associated proteins such as the tetraspanins CD63 and CD81 and the ESCRT component TSG101, while the endoplasmic reticulum protein Grp94 was undetectable, indicating minimal cellular contamination.

Notably, the overall abundance of vesicles did not differ between cancer patients and healthy donors. Mean particle concentrations ranged from about 5 to 9 billion particles per millilitre across the groups, and total protein content was similar as well. This observation is important because it suggests that the pathological relevance of these vesicles lies not in how many are produced but in what they carry and what they do to recipient cells. Confocal microscopy showed that fluorescently labelled vesicles were efficiently internalised by activated CD4-positive and CD8-positive T cells within 18 hours, with activated cells taking up more vesicles than resting ones, confirming that these particles can deliver their cargo directly into the immune cells they target.

The functional consequences were striking. In proliferation assays using a fluorescent cell division tracer, vesicles from cancer patients significantly suppressed the proliferation of both CD4-positive and CD8-positive T cells over four days of co-culture, whereas vesicles from healthy donors or disease-free patients showed little to no suppressive effect. The suppression trended stronger with advanced-stage disease, particularly stage IV. Cancer-derived vesicles also reduced the production of tumour necrosis factor alpha, a key inflammatory cytokine, although interferon gamma levels remained unchanged, hinting that the vesicles preferentially interfere with early activation events and NF-kappa-B-dependent signalling rather than the JAK/STAT pathways driving interferon responses. Expression of CD69, an early activation marker, was also modestly reduced.

Perhaps the most dramatic effect was on cell survival. When CD8-positive Jurkat T cells were exposed to vesicles from late-stage cancer patients, up to 90 percent of the cells underwent apoptosis within 24 hours, as measured by Annexin V staining. The researchers also probed the adenosinergic pathway, an immunosuppressive circuit in which the ectonucleotidases CD39 and CD73 sequentially convert extracellular ATP into adenosine, a molecule that binds the A2A receptor on immune cells and shuts down their activity. Vesicles from advanced-stage patients enhanced the conversion of ATP to ADP and AMP by activated CD4-positive T cells, accompanied by a trend toward increased CD39 expression, although the conversion of AMP to adenosine itself was not altered, and the authors caution that one highly active stage IV sample drove much of this effect.

Immune checkpoint profiling revealed a complex and, in places, counterintuitive picture. On CD4-positive T cells, exposure to cancer-derived vesicles increased the expression of TIGIT and CTLA-4, both associated with immunosuppressive function and frequently found on regulatory T cells, along with the co-stimulatory molecule CD137. At the same time, the inhibitory checkpoints PD-1, LAG-3, and BTLA decreased. On CD8-positive T cells, the vesicles broadly downregulated both inhibitory and stimulatory checkpoints, including PD-1, LAG-3, BTLA, CD137, and OX40. Rather than inducing the canonical exhausted phenotype marked by elevated inhibitory checkpoint expression, the vesicles appear to promote a deactivated state in CD8-positive cells, compounded by apoptosis. Exploratory transcriptomic analysis using the NanoString platform supported these observations at the RNA level: CD4-positive cells showed upregulation of CTLA4 and the immunosuppressive gene SIGIRR, while CD8-positive cells exposed to cancer vesicles uniquely upregulated PDCD1LG2, the gene encoding PD-L2, a ligand for PD-1.

A clinically critical question was whether immune checkpoint blockade could override the vesicle-mediated suppression. Pre-treating T cells with pembrolizumab, an anti-PD-1 antibody, partially attenuated the inhibition of CD4-positive T-cell proliferation and reduced the apoptosis induced by late-stage vesicles in most cases, but it never fully prevented the suppressive effects, even at higher antibody concentrations. ATP metabolism and ectonucleotidase expression were similarly unaffected by PD-1 blockade. The authors note that some of the observed changes, including reduced PD-1 detection, could partly reflect antibody binding to the receptor rather than true biological downregulation, an acknowledged technical caveat. Nevertheless, the incomplete rescue strongly suggests that the vesicles act through pathways beyond the PD-1/PD-L1 axis, potentially implicating TIGIT, LAG-3, or CTLA-4 as co-targets, and raising the possibility that vesicle-associated PD-L1 may even sequester therapeutic antibodies and reduce their availability.

The study is not without limitations, which the authors acknowledge candidly. The gene expression analysis was exploratory, with p-values uncorrected for multiple comparisons, and the high inter-donor variability in checkpoint responses, together with the fact that some effects were also seen with vesicles from healthy donors, means that several findings require confirmation in larger, tightly controlled cohorts. The adenosine metabolism results were heavily influenced by a single patient sample. HPV status, which was positive in 41 percent of the tested patients, did not influence T-cell function in these assays, consistent with earlier observations on tumour-derived vesicles.

Even with these caveats, the work delivers a coherent and consequential message: small extracellular vesicles circulating in the plasma of head and neck cancer patients are functional effectors of systemic immune modulation, capable of reshaping T-cell phenotype, function, and survival in ways that current immunotherapy does not fully counteract. If future studies validate these vesicles as biomarkers of disease activity and immunotherapy response, and if strategies emerge to neutralise their cargo or block their uptake, they could become a linchpin in the next generation of combination treatments for a cancer type that continues to evade the immune system with remarkable sophistication.

Subject of Research: Immunosuppressive effects of plasma-derived small extracellular vesicles on T cells in head and neck cancer

Article Title: Small extracellular vesicles from the plasma of head and neck cancer patients induce dysfunctional T cell phenotypes

Article References: Tengler, L., Jackson, E. K., Bieback, K., Röth, R., Mellein, S., Kern, J., Theodoraki, M.-N., Schütz, J., Huber, L., Scherl, C., Rotter, N., Affolter, A., & Ludwig, S. (2026). Small extracellular vesicles from the plasma of head and neck cancer patients induce dysfunctional T cell phenotypes. British Journal of Cancer. https://doi.org/10.1038/s41416-026-03606-8

Image Credits: AI Generated

DOI: 10.1038/s41416-026-03606-8

Keywords: head and neck cancer, small extracellular vesicles, exosomes, T cells, immune checkpoints, PD-1, immunotherapy resistance, tumour immunology, adenosine pathway, apoptosis, biomarkers, immune evasion