Neoantigens and Tumour Immunogenicity

From Mutation to Displayed Peptide

For a mutation to generate a neoantigen, the altered gene must be expressed, the protein must be processed into peptides, at least one peptide must bind the patient's specific HLA molecules, and a T cell with a matching receptor must exist and not be tolerised. Each step removes candidates, so predicted neoantigen counts greatly exceed the number that provoke a real response.

Frameshift insertions and deletions, which create long stretches of novel sequence, tend to be more immunogenic per mutation than single-base substitutions.

Why Mutational Burden Correlates With Response

More mutations generally mean more chances to produce a strong neoantigen, which is the mechanistic reason tumour mutational burden and mismatch-repair deficiency are associated with checkpoint-inhibitor benefit across several cancers.

The relationship is statistical, not deterministic. Some high-burden tumours do not respond, and some low-burden tumours do, because response also depends on antigen presentation, the T-cell repertoire and the tumour microenvironment.

Clonal Versus Subclonal Neoantigens

A neoantigen present in every tumour cell (clonal) is a better target than one present in only a subset (subclonal), because targeting a subclonal antigen leaves antigen-negative cells to regrow. Tumours with many subclonal neoantigens, often from ongoing mutational processes, respond less well.

This is one reason a single biopsy may misrepresent the immunogenic landscape of a heterogeneous tumour.

Therapeutic Use

Personalised cancer vaccines and selected T-cell therapies aim to direct immunity against individual neoantigens, and early-phase trials in melanoma and pancreatic cancer have shown immune responses. These remain investigational.

Neoantigen prediction is also not yet a standardised clinical assay; tumour mutational burden and mismatch-repair status are the validated proxies in routine use.

What This Means for a Report and a Trial

For routine care, the neoantigen concept is applied through its proxies. A high tumour mutational burden or a mismatch-repair-deficient result is used to consider a checkpoint inhibitor, on the reasoning that more mutations mean more chances of a strong neoantigen. Frameshift-rich tumours, such as those with microsatellite instability, are disproportionately immunogenic for their mutation count.

Direct neoantigen targeting is trial territory. Personalised mRNA vaccines encoding a patient's predicted neoantigens, given with a checkpoint inhibitor, have shown relapse-free survival signals in resected melanoma and immune responses in pancreatic cancer. Eligibility for these studies depends on adequate tissue, a manufacturing slot and a minimum predicted neoantigen count.

Key Takeaways

• ·Neoantigens are mutation-derived peptides that T cells can recognise as foreign.
• ·Only a small fraction of predicted neoantigens become effective immune targets.
• ·Clonal neoantigens and higher mutational burden are associated with better checkpoint-inhibitor response.