Rethinking Neoantigen Selection in High–Mutational Burden Cancers

Neoantigen-based cancer vaccines are designed to help the immune system recognize tumors by targeting mutations found only in cancer cells. Although this approach has shown promise in several cancers, its success has been inconsistent. A major challenge is that only a small fraction of predicted neoantigens actually generate immune responses capable of slowing or eliminating tumors.¹ A recent study by Adams and colleagues, published in the Journal for ImmunoTherapy of Cancer, examines this problem in cutaneous squamous cell carcinoma (cSCC) and offers new insight into why some neoantigens succeed where others fail.²

The Challenge of Neoantigen Selection

Most current methods for selecting neoantigens rely on predicting how well a mutated peptide binds to major histocompatibility complex (MHC) molecules and whether the mutated gene is expressed by the tumor. Although these factors are necessary for immune recognition, they have proven insufficient for reliably identifying neoantigens that lead to tumor control. This limitation is particularly evident in cancers like cSCC, which carry thousands of mutations due to chronic ultraviolet (UV) exposure.

Using genomic data from nearly 150 human cSCC tumors, the investigators found that most mutations were unique to individual patients. Only a small number of mutations were shared across tumors, and even fewer were predicted to bind MHC molecules well enough to be seen by T cells. These findings suggest that broadly applicable, “shared” neoantigen vaccines are unlikely to benefit most patients with cSCC and that personalized approaches are likely necessary.

A Mouse Model That Reflects Human Disease

To better understand which neoantigens actually matter for immune control of tumors, the researchers developed a transplantable mouse model of cSCC using UV-induced tumors. Importantly, these tumors closely resembled human cSCC at the genetic level, including the characteristic UV-related mutation pattern and changes in key cancer-related genes such as Trp53.

When these tumors were implanted into mice, their growth was significantly slowed by the immune system. Experiments showed that CD8⁺ T cells—immune cells responsible for directly killing cancer cells—played a central role in controlling tumor growth. This confirmed that the model was suitable for studying immune responses to tumor-specific mutations.

Identifying Tumor-Rejecting Neoantigens

From thousands of mutations present in the tumors, the investigators used standard prediction methods to narrow the list to 30 candidate neoantigens. These candidates were then tested in immune assays and vaccination experiments to determine whether they could stimulate meaningful T-cell responses.

Only 2 neoantigens stood out, according to investigators. Both triggered strong CD8⁺ T-cell responses and were able to slow tumor growth when used as vaccines. Notably, these effects occurred without the use of immune checkpoint inhibitors, highlighting the intrinsic immune activity of these neoantigens.

Why These Neoantigens Were Different

Closer examination revealed that the 2 successful neoantigens worked for different reasons. One bound more strongly to its MHC molecule than the corresponding normal (nonmutated) peptide, fitting well with traditional expectations of neoantigen immunogenicity.

The second neoantigen, however, did not show improved MHC binding. Instead, structural modeling revealed that the mutation caused part of the peptide to protrude more prominently from the MHC surface. This change made the mutated region more visible to T-cell receptors, likely improving immune recognition despite similar MHC binding.

When the researchers expanded their analysis to include tumor-rejecting neoantigens from other cancer models, a consistent pattern emerged. Among neoantigens that did not differ much in MHC binding, those that changed how the peptide was displayed to T cells, particularly by increasing exposure of the mutated region, were more likely to mediate tumor rejection.

Implications for Clinical Development

This study highlights an important limitation of current neoantigen selection strategies and suggests a potential way forward. Although MHC binding remains important, it does not fully explain which neoantigens will lead to tumor control. Structural features that affect how T cells “see” the mutation may be just as critical.

For clinicians and researchers developing personalized cancer vaccines, these findings support incorporating structural modeling into neoantigen selection pipelines. Although further validation in human studies is needed, this approach may improve the likelihood that selected neoantigens translate into meaningful clinical benefit.

In high–mutational burden cancers such as cSCC, where most mutations are unique to each tumor, refining how neoantigens are chosen may be as important as identifying them in the first place.

References

1. Naffaa MM, Al-Ewaidat OA, Gogia S, Begiashvili V. Neoantigen-based immunotherapy: advancing precision medicine in cancer and glioblastoma treatment through discovery and innovation. Explor Target Antitumor Ther. 2025;6:1002313. doi:10.37349/etat.2025.1002313
2. Adams AC, Macy AM, Borden ES, et al. Structural changes from wild-type define tumor-rejecting neoantigens. J Immunother Cancer. 2025;13(10):e013148. doi:10.1136/jitc-2025-013148