Neoantigen Peptide Vaccines 2026: Clinical Pipeline Update

Peptide-based cancer vaccines were an early focus of tumor immunotherapy in the 1990s and 2000s, but most first-generation programs produced modest clinical signal. Two technical shifts have re-opened the field: routine patient-specific neoantigen identification by tumor-normal sequencing, and improved understanding of delivery platforms that reliably drive HLA class I presentation. The 2024 mRNA-4157 readout in melanoma served as broader validation of the neoantigen concept, and peptide formats have re-emerged as complementary rather than replaced approaches [1].

This article reviews the 2026 pipeline across synthetic long peptide (SLP) vaccines, peptide-pulsed autologous dendritic cell (DC) vaccines, and short-peptide HLA-I-targeted formats — with particular attention to the 2025 Mount Sinai-led peer-reviewed analysis that repositioned peptide-pulsed DCs in the therapeutic landscape.

A 2025 review led by investigators at Mount Sinai and collaborating institutions synthesized the clinical and translational data for peptide-based cancer vaccines and compared per-epitope CD8+ T-cell response rates across delivery formats. The review highlighted three conclusions relevant to the 2026 pipeline [2]:

• Peptide-pulsed DCs deliver HLA class I presentation most directly among the peptide-format options, because exogenous loading of short peptides onto MHC-I can occur at the cell surface or via cross-presentation.
• Synthetic long peptides (SLPs) rely on DC uptake, processing, and presentation in vivo, which is variable across patients and adjuvants.
• Per-epitope induction of detectable CD8+ T-cell responses is comparable between well-executed peptide-pulsed DC and mRNA neoantigen platforms in the studies compared, with different manufacturing and scalability profiles.

The review is notable not for claiming peptide formats are superior to mRNA, but for formally re-centering peptide-pulsed DCs as a clinically viable alternative with specific mechanistic advantages for HLA class I priming.

CD8+ T-cell responses against tumor neoantigens require presentation of 8- to 11-amino-acid peptides in the groove of HLA class I molecules on the surface of antigen-presenting cells (APCs). Three routes can deliver a neoantigen into the HLA-I pathway [3]:

• Endogenous synthesis (mRNA or DNA vaccines) — the antigen is translated inside the APC and enters the proteasome-TAP-MHC-I pipeline.
• Cross-presentation (SLP vaccines, protein vaccines) — exogenous long peptide is taken up, routed through the cross-presentation pathway, and loaded onto MHC-I.
• Direct exogenous loading (peptide-pulsed DCs with short epitopes) — minimal HLA-binding peptides are loaded onto surface MHC-I without requiring internal processing.

Peptide-pulsed DCs with short epitope-length peptides bypass the cross-presentation bottleneck that limits SLP efficiency in some patients, which is the core mechanistic argument behind the Mount Sinai position [4].

Bladder cancer has one of the highest tumor mutational burdens among solid tumors, which translates to a richer neoantigen repertoire. Active 2026 peptide-vaccine programs include personalized long peptide constructs combined with anti-PD-1 in muscle-invasive bladder cancer, and BCG-combination peptide programs in high-risk non-muscle-invasive bladder cancer (NMIBC) [5]. Early data from combination trials have reported immune responses against predicted neoantigens and acceptable safety profiles, with efficacy endpoints under evaluation in ongoing Phase 2 studies.

Pancreatic ductal adenocarcinoma (PDAC) has been resistant to most immunotherapy approaches, but peptide programs targeting recurrent KRAS mutations (G12D, G12V, G12C, G12R) and personalized neoantigen sets have shown signals. A landmark mRNA neoantigen program in resected PDAC reported T-cell responses in approximately half of treated patients and an association with delayed recurrence in immunological responders [6]. Peptide-format KRAS vaccine programs are advancing in parallel, leveraging the same shared-driver rationale.

Glioblastoma (GBM) and IDH-mutant glioma programs have produced some of the most mechanistically informative peptide-vaccine datasets. An IDH1 R132H-targeted long peptide vaccine has demonstrated T-cell responses against the shared driver neoantigen in IDH1-mutant glioma patients in a randomized Phase 1 trial [7]. Personalized peptide approaches in GBM combining patient-specific neoantigens with non-mutated glioma-associated antigens are also in clinical evaluation.

• mRNA platforms: 6-8 week manufacturing, scalable once pipeline is established, established with the mRNA-4157 melanoma readout
• Peptide-pulsed DCs: direct HLA-I presentation, strong co-stimulation, but require individualized cell therapy manufacturing
• Synthetic long peptides: simpler chemistry, off-the-shelf shared-antigen formats possible, dependent on cross-presentation
• Short peptide HLA-I formats: best epitope control, HLA-restricted patient selection required

In 2026 the field is converging on a view that peptide and mRNA modalities are likely to coexist, with selection driven by indication biology, HLA restriction, manufacturing infrastructure, and the relative weight of CD8 versus CD4 responses for the target disease.

Expected milestones through 2028 include randomized Phase 2 readouts for personalized peptide vaccines in muscle-invasive bladder cancer, further KRAS-targeted peptide data in PDAC, long-term follow-up of IDH1 R132H glioma peptide vaccine programs, and ongoing head-to-head methodological comparisons against mRNA platforms. Regulatory frameworks for individualized biologic products are evolving in parallel, with FDA and EMA guidance documents under active revision [8].