HLA Loss of Heterozygosity: Losing Half the Presentation Repertoire
The HLA class I genes are the most polymorphic in the human genome, and most people inherit different alleles from each parent. Tumours can delete one parental haplotype at the HLA locus, a form of loss of heterozygosity that removes up to half the antigen-presentation repertoire without abolishing MHC class I entirely.
Quick Answer
The HLA class I genes are the most polymorphic in the human genome, and most people inherit different alleles from each parent. Tumours can delete one parental haplotype at the HLA locus, a form of loss of heterozygosity that removes up to half the antigen-presentation repertoire without abolishing MHC class I entirely.
Part of a topic cluster
Immuno-Oncology and Tumour Metabolism
Open the complete 15-article guideA Subtler Escape Than Total Loss
Keeping one HLA haplotype lets the tumour cell retain some antigen presentation, avoiding the natural killer cell attack that follows complete MHC class I loss, while dropping the ability to present any neoantigen restricted to the deleted alleles.
In one lung cancer study, HLA loss of heterozygosity was present in a large fraction of tumours and was often subclonal, indicating it is an ongoing adaptation to immune pressure rather than a single early event.
Why It Is Easy to Miss
Standard copy-number tools struggle at the HLA locus because the region is so polymorphic and the sequences of the different genes are similar. Dedicated algorithms that use the patient's germline HLA type are needed to call allele-specific loss reliably.
This means HLA loss of heterozygosity can be present but unreported by a routine tumour panel.
Consequences for Neoantigen Prediction
Neoantigen prediction pipelines assume a tumour can present peptides on all of a person's HLA alleles. If one haplotype is deleted, predictions against those alleles are moot, and the tumour's true presentable repertoire is smaller than it appears.
Accounting for HLA loss improves the correlation between predicted immunogenicity and actual immune pressure, but this refinement is mostly a research technique so far.
Where It Sits in Practice
HLA loss of heterozygosity is not tested for in routine oncology, and no treatment is selected on it. Its main value is explanatory: it is one reason a tumour with a high predicted neoantigen load and visible T-cell infiltration can still evade immunity and fail a checkpoint inhibitor.
For neoantigen-directed trials it matters more, because a vaccine or T-cell product aimed at a peptide restricted to a deleted allele would be inert. Research pipelines increasingly incorporate allele-specific HLA calling for this reason, and the same information helps explain why immune pressure within a tumour is uneven.
Key Takeaways
- ·HLA loss of heterozygosity deletes one parental HLA haplotype, narrowing antigen presentation.
- ·It retains partial MHC class I, avoiding natural killer cell attack, and is often subclonal.
- ·It requires specialised detection and, when present, shrinks the tumour's real neoantigen repertoire.
Put these genes in pathway context
Frequently asked questions
How is HLA loss of heterozygosity different from B2M loss?
B2M loss removes all MHC class I; HLA loss of heterozygosity deletes one parental haplotype, keeping partial presentation, avoiding natural killer cell attack, and dropping only the neoantigens restricted to the lost alleles.
Why might a routine tumour panel miss it?
The HLA region is extremely polymorphic and the genes are similar in sequence, so standard copy-number tools struggle there. Dedicated algorithms that use the patient's germline HLA type are needed to call it reliably.
Does it affect neoantigen prediction?
Yes. Prediction pipelines assume all of a person's HLA alleles are available; if one haplotype is deleted, the tumour's true presentable repertoire is smaller than predicted.
References
- 1Allele-specific HLA loss and immune escape in lung cancer evolution. Cell, 2017. PubMed
- 2Pan-cancer immunogenomic analyses reveal genotype-immunophenotype relationships and predictors of response to checkpoint blockade. Cell Rep, 2017. PubMed
- 3Tumour mutational burden as a biomarker in cancer immunotherapy. Nat Rev Clin Oncol, 2019. PubMed
Continue Reading
B2M and MHC Class I Loss: Hiding From T Cells
3 min read
Neoantigens and Tumour Immunogenicity
3 min read
JAK1/JAK2 Loss and Interferon-Signalling Resistance to Immunotherapy
3 min read
Interpreting Copy-Number Changes: Amplification, Gain, Loss and Deletion
3 min read
PD-L1 Testing Explained: Assays, Scores and Limits
2 min read
Hot Versus Cold Tumours: What T-Cell Infiltration Means
3 min read
Choose your next research step
Move from this explanation into a gene profile, a pathway map, or the next evidence update.
STAT3 has 20+ trials currently recruiting on ClinicalTrials.gov. The GeneAnalyses digest summarises the new and changed ones each day.