How Tumors Evade Peptide Vaccines: Immune Escape

Peptide Vaccine Challenges

3 Layers of Tumor Evasion

Tumors escape peptide vaccines at three levels: they hide the target antigen, suppress arriving T cells, and create hostile microenvironments that neutralize immune responses.

Wells et al., Cell, 2020

Wells et al., Cell, 2020

View as image

A peptide vaccine can generate potent T cells in the blood that never kill a single cancer cell. The immune system may recognize the target, expand the right T cell clones, and still fail because the tumor has evolved defenses against immune attack. Wells et al. (2020) assembled a global consortium to study why predicted neoantigens so often fail to produce clinical responses, identifying key parameters that govern whether a peptide-MHC complex will actually trigger productive immunity versus being ignored or suppressed.^[1] For context on why peptide vaccines are challenging to develop, see why peptide vaccines are hard: HLA restriction, evasion, and immunogenicity.

Escape Mechanism 1: Antigen Loss and MHC Downregulation

The Antigen Disappears

The most direct evasion strategy: the tumor stops expressing the protein that the vaccine targets. If a peptide vaccine generates T cells against a specific neoantigen derived from a somatic mutation, tumor cells that lose that mutation (or stop expressing the mutated gene) become invisible to the vaccine-induced T cells. These antigen-loss variants survive while antigen-expressing cells are killed, a process called immunoediting.

This is why targeting a single antigen is risky. Tumors are genetically heterogeneous: not every cell in a tumor carries the same mutations. A vaccine targeting one neoantigen applies selective pressure that favors outgrowth of cells lacking that neoantigen. Peterson et al. (2020) compared personal neoantigen vaccines versus shared frameshift neoantigen vaccines in a mouse mammary cancer model, finding that multi-target approaches provided broader coverage against antigen loss than single-epitope strategies.^[7]

The Display System Fails

Even when the antigen is present, the tumor can prevent it from being displayed to T cells. MHC class I molecules are the display platform: they present intracellular peptide fragments on the cell surface where CD8+ T cells can inspect them. Tumors frequently downregulate MHC class I expression through:

• Beta-2-microglobulin (B2M) mutations: B2M is essential for MHC class I surface expression. Loss-of-function mutations in B2M eliminate all MHC class I display.
• Transcriptional silencing: epigenetic changes can silence MHC genes without deleting them.
• Defective antigen processing: mutations in TAP (transporter associated with antigen processing) or proteasome subunits prevent peptides from being loaded onto MHC molecules.

Pang et al. (2018) identified a structural mechanism underlying poor T cell recognition: peptide-binding groove contraction. When certain peptide-MHC complexes form, the binding groove adopts a conformation that is structurally incompatible with T cell receptor engagement, even though the peptide binds MHC with high affinity. This means the standard prediction pipeline (predict MHC binding, assume T cell recognition follows) can fail at the structural level.^[4]

Escape Mechanism 2: The Immunosuppressive Tumor Microenvironment

Checkpoint Molecules

Even when vaccine-induced T cells reach the tumor and recognize their target, the tumor can shut them down through checkpoint ligand expression. PD-L1 (programmed death ligand 1) on tumor cells binds PD-1 on T cells, transmitting an inhibitory signal that exhausts the T cell and prevents killing.

Hailemichael et al. (2018) conducted a study that revealed a deep problem with peptide vaccine-checkpoint inhibitor combinations. In their model, gp100 peptide vaccination combined with anti-CTLA-4 (ipilimumab) showed no benefit, mirroring the landmark clinical trial result. The reason: the vaccine formulation (incomplete Freund's adjuvant) created a persistent antigen depot at the injection site that sequestered vaccine-specific T cells away from the tumor. These trapped T cells became exhausted and apoptotic at the vaccine site rather than trafficking to the tumor. Changing the vaccine formulation to a short-lived adjuvant restored synergy with both CTLA-4 and PD-L1 blockade.^[2]

This finding was pivotal: it showed that the failure of the ipilimumab + gp100 vaccine clinical trial was due to formulation, not the fundamental concept of combining vaccines with checkpoint inhibitors. For the gp100 clinical trial history, see gp100 peptide vaccine for melanoma.

Regulatory T Cells

Tumors actively recruit and expand regulatory T cells (Tregs) that suppress anti-tumor immunity. Tregs express FOXP3, a transcription factor that drives their immunosuppressive program. In the tumor microenvironment, Tregs produce IL-10 and TGF-beta, which suppress CD8+ cytotoxic T cells and prevent them from killing tumor cells.

Hawley et al. (2022) took a peptide-based approach to this problem. They developed stapled alpha-helical peptides that directly inhibit FOXP3, the master transcription factor of Tregs. By dampening Treg function within the tumor microenvironment, these FOXP3-inhibiting peptides could potentially restore the efficacy of vaccine-induced anti-tumor T cells. The approach overcomes the dosing limitations and off-target effects of antibody-based Treg depletion strategies.^[5]

Immunosuppressive Cytokines and Cells

Beyond checkpoints and Tregs, tumors create a hostile microenvironment through:

• Myeloid-derived suppressor cells (MDSCs): immature myeloid cells recruited by tumor-secreted factors that suppress T cell function
• Tumor-associated macrophages (TAMs): M2-polarized macrophages that promote tumor growth and suppress immunity
• Metabolic deprivation: tumors consume glucose and amino acids (especially tryptophan via IDO) that T cells need for effector function
• Hypoxia: low oxygen tension in the tumor center impairs T cell metabolism and promotes immunosuppressive gene expression

Escape Mechanism 3: Immunogenicity Failures

The Wrong Epitope

Not all tumor mutations create immunogenic epitopes. Capietto et al. (2020) demonstrated that where a mutation falls within the peptide sequence determines whether T cells recognize it as foreign. Mutations at anchor residues (positions that contact MHC rather than the T cell receptor) change binding affinity but not the surface presented to T cells. Mutations at TCR-contact residues are more likely to generate neoantigens that T cells recognize as non-self. Standard prediction pipelines do not adequately distinguish between these positions.^[3]

Richters et al. (2019) published best practices for bioinformatic neoantigen characterization, emphasizing that the field needs to move beyond simple MHC binding prediction to incorporate gene expression, clonality, and TCR-facing mutation position into the prediction pipeline.^[8]

Poor Immunogenicity of Peptides

Short synthetic peptides are intrinsically poor immunogens. Without the danger signals that accompany natural infection (pathogen-associated molecular patterns), the immune system often tolerizes against vaccine peptides rather than mounting an attack. Shae et al. (2020) addressed this by co-delivering peptide neoantigens with STING (stimulator of interferon genes) agonists in nanoparticle formulations. The STING agonist provided the innate immune activation signal that the peptide alone lacked, enhancing CD8+ T cell responses and improving tumor control in mouse models.^[6]

Chen et al. (2020) reviewed advances in personalized neoantigen vaccination with synthetic long peptides (SLPs), which improve immunogenicity by requiring intracellular processing by dendritic cells rather than directly loading onto MHC. SLPs generate both CD4+ and CD8+ T cell responses, and the dendritic cell processing step provides activation signals that short peptides skip.^[9]

Strategies to Overcome Evasion

Multi-Epitope Vaccines

Targeting multiple antigens simultaneously makes antigen loss harder. If a vaccine targets 10 different neoantigens, the tumor must lose all 10 to fully escape. Shimizu (2018) described a combination approach targeting both shared antigens (glypican-3) and personalized neoantigens, providing redundant targeting that resists single-antigen escape.^[10]

Combination with Checkpoint Inhibitors

The most clinically advanced strategy. Peptide vaccines generate tumor-specific T cells; checkpoint inhibitors remove the molecular brakes that tumors use to suppress them. The Hailemichael formulation study shows this combination works when the vaccine is properly designed, and multiple clinical trials are testing peptide vaccine + anti-PD-1/PD-L1 combinations. For a comprehensive look at how peptide vaccines are designed and formulated, see how peptide vaccines are designed: from epitope to injection.

Treg Depletion or Inhibition

Removing regulatory T cells from the tumor microenvironment restores vaccine-induced T cell function. The FOXP3-inhibiting stapled peptides (Hawley et al., 2022) represent a targeted approach to this problem that uses peptide chemistry against a peptide vaccine barrier. Existing Treg depletion strategies (anti-CD25 antibodies, low-dose cyclophosphamide) are blunt tools that can also deplete effector T cells. The stapled peptide approach offers selectivity by targeting the transcription factor specific to Treg suppressive function rather than a surface marker shared with activated effector cells.

Improved Adjuvants and Delivery

The innate immune activation gap is addressable through better adjuvant design. The Shae et al. STING agonist nanoparticle approach is one example. Others include TLR agonists (poly-ICLC, CpG oligonucleotides), saponin-based adjuvants (ISCOMATRIX), and self-assembling peptide nanoparticles that combine delivery with immune activation. Chen et al. (2020) emphasized that synthetic long peptides (15-30 amino acids) are inherently better immunogens than short minimal epitopes because they require processing by dendritic cells, which provides the activation context that MHC class I-restricted 8-10mer peptides lack.^[9] For the broader HLA restriction problem, see the HLA problem: why one peptide vaccine does not fit everyone.

Limitations

Most studies of tumor immune evasion from peptide vaccines use mouse tumor models, which have significant immunological differences from human cancers. Mouse tumors are typically faster-growing, less heterogeneous, and implanted into hosts with intact immune systems, which does not fully model the gradual immune evasion that occurs during human cancer progression.

The relative contribution of each evasion mechanism varies by tumor type, stage, and anatomical location. Melanoma and lung cancer (high mutation burden) may respond differently to neoantigen vaccines than pancreatic cancer or glioblastoma (lower mutation burden, stronger immunosuppressive microenvironments).

Clinical trial designs for peptide vaccine-checkpoint combinations are still being optimized. The Hailemichael finding about vaccine formulation affecting checkpoint synergy suggests that past negative trials may have failed due to formulation rather than concept, but this hypothesis requires prospective validation in randomized human trials.

The temporal dynamics of immune evasion also complicate vaccine strategy. Early-stage tumors may be more susceptible to peptide vaccines because they have had less time to evolve escape mechanisms. Advanced tumors have undergone extensive immunoediting and typically exhibit multiple layers of evasion simultaneously. Whether peptide vaccines should be deployed in adjuvant settings (after surgical removal of the primary tumor, targeting micrometastases) or in advanced disease remains an open strategic question with different evasion profiles at each stage.

The Bottom Line

Tumors evade peptide vaccines through three main mechanisms: antigen loss and MHC downregulation (hiding the target), immunosuppressive microenvironments (Tregs, checkpoints, MDSCs), and immunogenicity failures (wrong epitope position, poor innate activation). Understanding each mechanism has led to countermeasures: multi-epitope vaccines resist antigen loss, checkpoint combination restores T cell function (when formulation is correct), and novel adjuvants like STING agonists overcome peptide immunogenicity barriers. The Hailemichael finding that vaccine formulation determines checkpoint synergy was a breakthrough that explains past clinical trial failures.

Frequently Asked Questions

Why do tumors escape peptide vaccines?

Tumors evade peptide vaccines through three mechanisms: they stop expressing the target antigen (antigen loss), they downregulate MHC class I molecules so T cells cannot see the antigen, and they create immunosuppressive microenvironments using regulatory T cells, PD-L1, and metabolic deprivation that neutralize vaccine-induced T cells even when they reach the tumor.

What is antigen loss in cancer?

Antigen loss occurs when tumor cells stop expressing the protein that a vaccine targets. Tumors are genetically heterogeneous, with different cells carrying different mutations. A vaccine targeting one neoantigen kills cells expressing it but spares cells that lack it, allowing antigen-negative cells to grow. This is why multi-epitope vaccines that target many antigens simultaneously are preferred.

Why don't peptide vaccines work with checkpoint inhibitors?

They can work together, but only with the right vaccine formulation. Hailemichael et al. (2018) showed that incomplete Freund's adjuvant created a persistent antigen depot that trapped T cells at the injection site, preventing them from reaching the tumor. When formulation was changed to a short-lived adjuvant, the vaccine synergized with both CTLA-4 and PD-L1 blockade. Past clinical trial failures may have been formulation problems.

What is the tumor microenvironment?

The tumor microenvironment is the cellular neighborhood surrounding and infiltrating a tumor. It includes immune cells (regulatory T cells, macrophages, MDSCs), blood vessels, fibroblasts, and the extracellular matrix. Tumors actively shape this environment to suppress anti-tumor immunity through checkpoint ligands (PD-L1), immunosuppressive cytokines (IL-10, TGF-beta), and metabolic deprivation.

How can peptide vaccines overcome immune escape?

Three main strategies: target multiple antigens simultaneously so the tumor cannot escape by losing one (multi-epitope vaccines), combine with checkpoint inhibitors to remove the molecular brakes on T cells (anti-PD-1/PD-L1), and use strong adjuvants or delivery systems (STING agonists, nanoparticles) that provide the innate immune activation signals that short peptides lack on their own.

Does mutation position affect peptide vaccine success?

Yes. Capietto et al. (2020) showed that where a mutation falls within the peptide determines immunogenicity. Mutations at MHC anchor residues change binding but not the surface that T cells see. Mutations at TCR-contact residues are more likely to create neoantigens that T cells recognize as foreign. Standard prediction tools do not adequately distinguish between these positions.