# Tumour-associated antigen (TAA)

Source: https://onco.cc/terms/tumor-associated-antigen/  
OnCo record `tumor-associated-antigen` (Term). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

A tumour-associated antigen is a target that is common in a tumour but can also be present in healthy tissue. A treatment aimed at it can affect both, and can also react with a different, look-alike target.

## Summary

Many tumour-associated antigens come from unmutated proteins, so their usefulness depends on higher expression, an unusual expression pattern or accessibility in the cancer. A neoantigen, by contrast, arises from a tumour-specific sequence alteration. For an antibody or a chimeric antigen receptor, the question is accessible surface antigen on tumour and on healthy cells. For a peptide-directed T-cell receptor it is the particular peptide displayed by a particular HLA allele. RNA for the source gene establishes neither recognition surface, and a gene-expression result does not specify accessibility or peptide display.

Two hazards need different controls. On-target, off-tumour recognition means the receptor recognises its intended target in the wrong tissue. Off-target cross-reactivity means the receptor also recognises a different molecular target, so screening the intended antigen alone cannot rule it out. A published example of the second is an affinity-enhanced T-cell receptor directed at HLA-A*01-restricted MAGE-A3: in a clinical study the first two treated patients developed cardiogenic shock and died within days, no MAGE-A3 expression was detected in heart tissue at autopsy, and the T cells recognised an unrelated peptide from the striated-muscle protein titin (Linette et al., 2013; Cameron et al., 2013, who describe the preclinical analysis as having found no off-target concern). That illustrates the hazard for one engineered receptor and is not a toxicity rate for tumour-associated targets. Immune tolerance also shapes the available response to self antigens, and a stronger receptor does not remove the need to test healthy tissues and alternative targets.

A shared antigen can support a treatment intended for more than one person's cancer, with a trade-off between tumour coverage and healthy-tissue exposure; expression prevalence in a cohort is not a safe recognition threshold for a product. For vaccine-elicited peptide-specific T cells, a shared antigen offers another candidate when credible private targets are few, which is a coverage hypothesis, not demonstrated benefit. A peptide from a shared antigen and one from a neoantigen may depend on the same HLA allele, so changing the antigen does not restore a missing display route: Tran et al. (2016) documented escape after loss of the restricting HLA allele in one colorectal cancer patient. Antibodies and CARs have different recognition requirements, and epitope spreading is a separate widening of recognition that adding a shared antigen does not guarantee.

## Fields

- Kind: Term
- Last checked: 2026-11-13
- Also known as: TAA; TAAs; shared tumour antigen; on-target off-tumour recognition; off-target cross-reactivity
- Tags: education-immunology-2026-11

## Notes

- The existing antigen record already lists tumour-associated antigen as an alias; this record carries the safety distinctions and uses different aliases.
- Fictional X/Y examples in the teaching source were not carried over. Abstract-depth reads on 9 November 2026.
- NY-ESO-1 illustrates a shared cancer-testis antigen, rather than a private mutation. Ademuyiwa and colleagues reported immunohistochemical expression in 16% of 168 triple-negative breast cancers and 2% of 47 ER-positive/HER2-negative comparators; the comparison groups had different sampling characteristics, and staining was heterogeneous. Their background describes normal expression restricted to testis. Expression prevalence is neither proof of a displayed peptide on every tumour cell nor evidence that a receptor or vaccine is effective or safe. A shared peptide-directed product still requires the appropriate HLA and functional testing.
- Viral proteins provide another possible source of non-self targets, distinct from an altered human protein. NCI identifies HPV as a cause of cervical and oropharyngeal cancers, among other sites. For a proposed viral peptide target, establish viral-protein expression in the relevant cells, compatible peptide-HLA display and receptor recognition. A breast-cancer subtype label alone supplies none of that evidence; it does not establish a viral target or predict benefit from targeting it.
- The germ-cell HLA qualification is about a measured recognition surface, not all cells in the testis. Guillaudeux and colleagues purified human pachytene spermatocytes and round spermatids and found low levels of class I HLA transcripts, but detected neither surface nor intracellular HLA-I heavy-chain protein with the antibodies used. This supports limited classical HLA-I display in those examined germ-cell populations, despite transcription. It does not establish absence of every HLA molecule in every testicular cell or condition, and it does not by itself prove that a peptide-directed engineered receptor is safe.

## Sources

- Linette et al., 2013, Blood (PMID 23770775): https://doi.org/10.1182/blood-2013-03-490565
- Cameron et al., 2013, Science Translational Medicine (PMID 23926201): https://doi.org/10.1126/scitranslmed.3006034
- Tran et al., 2016, New England Journal of Medicine (PMID 27959684): https://doi.org/10.1056/nejmoa1609279
- Ademuyiwa et al.: NY-ESO-1 expression in the studied breast-cancer cohorts (2012): https://pubmed.ncbi.nlm.nih.gov/22761704/
- Ademuyiwa et al.: correction adding authors and contributions (2012): https://doi.org/10.1371/annotation/5cdf6105-2a52-497a-86b3-db8f4a4e439c
- NCI: HPV and Cancer, established cancer sites: https://www.cancer.gov/about-cancer/causes-prevention/risk/infectious-agents/hpv-and-cancer
- Guillaudeux et al., expression of HLA class I genes in meiotic and post-meiotic human spermatogenic cells (Biology of Reproduction 1996): https://pubmed.ncbi.nlm.nih.gov/8793064/

## Connected records

- terms: [Antigen](https://onco.cc/terms/antigen/), [Antigen escape (antigen loss, lineage switch)](https://onco.cc/terms/antigen-escape/), [HLA (human leukocyte antigen) and MHC](https://onco.cc/terms/hla/), [HLA allele names and typing reports](https://onco.cc/terms/hla-typing-and-nomenclature/), [Neoantigen](https://onco.cc/terms/neoantigen/)
- targets: [MUC1 (mucin-1)](https://onco.cc/targets/muc1/)

---
JSON: https://onco.cc/api/v1/entities/tumor-associated-antigen.json