mRNA Cancer Vaccines and Personalized Neoantigen Therapy: How They Work and What Cancers Are Being Tested

Research-informed explainer — last updated April 12, 2026

Personalized mRNA cancer vaccines — which use the same mRNA platform that enabled COVID-19 vaccines to teach a patient's immune system to recognize their tumor's unique mutations — reduced the risk of melanoma recurrence or death by 49% when combined with pembrolizumab compared with pembrolizumab alone in a 2024 randomized trial. These vaccines are not a single product; each one is manufactured individually for one patient, based on sequencing of that specific patient's tumor.

The research behind neoantigen vaccine development has been shaped by investigators at Boston-area cancer centers. Elizabeth Buchbinder, Assistant Professor of Medicine at Harvard Medical School, is the first author of both the 2017 Nature proof-of-concept neoantigen vaccine trial in melanoma (2,838 citations) and the 2021 Nature Medicine study showing that neoantigen vaccines induce persistent memory T cells and immunological spreading beyond the targeted epitopes (491 citations). Buchbinder also contributed to the 2024 KEYNOTE-942 randomized trial of mRNA-4157 plus pembrolizumab in resected melanoma (642 citations). Patrick Ott, Professor of Medicine at Harvard and Clinical Director of the Melanoma Disease Center at Dana-Farber, published the 2021 Nature Reviews Clinical Oncology review of personalized neoantigen-based therapeutic cancer vaccines (978 citations) and led the phase Ib trial of personalized neoantigen therapy plus anti-PD-1 across melanoma, non-small cell lung cancer, and bladder cancer (717 citations). Andrew Aguirre, Associate Professor at Harvard Medical School and Dana-Farber, contributed to the 2023 Science paper demonstrating that mRNA neoantigen vaccines stimulate T cells in pancreatic cancer — extending the concept beyond melanoma to one of the hardest-to-treat tumor types.

What a neoantigen is and why it matters

When a normal cell becomes a cancer cell, it accumulates mutations. Some of those mutations produce abnormal protein fragments — called neoantigens — that are displayed on the surface of tumor cells and recognized by T cells as "foreign." In principle, the immune system should attack cells displaying these neoantigens. In practice, tumors develop mechanisms to suppress this recognition, including checkpoint pathways.

The neoantigen vaccine concept is to identify a patient's specific neoantigens through tumor sequencing, manufacture a vaccine encoding those neoantigens, and deliver it to prime the immune system to mount a specific, sustained attack against cells displaying those mutations.

How the manufacturing process works

Personalized neoantigen vaccine manufacturing is a multi-step process:

1. Tumor biopsy and germline sampling — A fresh tumor sample is sequenced alongside normal (germline) DNA from the same patient.
2. Neoantigen prediction — Bioinformatics pipelines identify mutations unique to the tumor and predict which peptide fragments are likely to be presented by that patient's HLA (immune recognition) molecules and generate an immune response.
3. Vaccine design and manufacturing — The identified neoantigens are encoded in mRNA (for the mRNA-4157/V940 platform) or synthesized as peptides (for earlier platforms). Manufacturing takes approximately 6–8 weeks.
4. Vaccination series — Patients receive a series of injections, typically combined with a checkpoint inhibitor such as pembrolizumab.

The KEYNOTE-942 trial: the pivotal evidence

KEYNOTE-942 was a randomized phase 2b trial in which 157 patients with resected high-risk melanoma (stage IIB–IV) received either mRNA-4157 plus pembrolizumab or pembrolizumab alone after surgical resection. At 18 months, recurrence-free survival was 78.6% with the combination versus 62.2% with pembrolizumab alone — a 49% reduction in the risk of recurrence or death. Distant metastasis-free survival also favored the combination.

The vaccine was manufactured for each individual patient based on their tumor's unique mutation profile. Median manufacturing time was about 7 weeks.

What Patrick Ott's phase Ib trial added

Patrick Ott's 2020 Cell paper enrolled 10 patients across melanoma, non-small cell lung cancer, and bladder cancer — the first prospective trial of personalized neoantigen therapy combined with anti-PD-1 (nivolumab) in a multi-tumor setting. All 10 patients received personalized vaccines based on tumor sequencing, and the trial demonstrated that neoantigen-specific T cell responses were induced in all patients. Responses were durable and broadened over time through epitope spreading — meaning the immune system began attacking additional cancer antigens beyond those encoded in the vaccine.

This multi-tumor result is significant. It suggests the neoantigen vaccine platform is not limited to melanoma, which has one of the highest mutation burdens of any cancer type.

Extension to pancreatic cancer: a 2023 Science paper

Andrew Aguirre's lab contributed to a landmark 2023 Science paper demonstrating that mRNA neoantigen vaccines generate detectable, functional neoantigen-specific T cell responses in pancreatic ductal adenocarcinoma — a cancer historically resistant to immunotherapy due to its immunosuppressive microenvironment. In 8 of 16 patients, vaccines induced T cell responses to at least one neoantigen, and these patients showed longer recurrence-free survival than non-responders. This is early but scientifically compelling evidence that neoantigen vaccines could be active in cold tumors, not just highly mutated ones.

Who are these vaccines being developed for?

As of April 2026, personalized mRNA cancer vaccines are being studied in multiple tumor types including:

• Melanoma (KEYNOTE-942, the furthest advanced clinical trial)
• Pancreatic cancer (early-phase trials showing T cell induction)
• Non-small cell lung cancer (phase Ib data, larger trials ongoing)
• Bladder cancer (phase Ib data)
• Colorectal cancer (early-phase)
• Head and neck squamous cell carcinoma (early-phase)

Higher tumor mutation burden generally correlates with more neoantigens and potentially more robust vaccine responses, which is why highly mutated tumors like melanoma, lung cancer, and MSI-H colorectal cancer are priorities. However, Aguirre's pancreatic cancer data suggests the approach may work in some low-mutation-burden cancers too.

The checkpoint blockade connection

Jedd Wolchok's research on the biology of PD-L1 and PD-1 pathway blockade helps explain why neoantigen vaccines are typically combined with checkpoint inhibitors. Vaccines prime T cells to recognize neoantigens, but tumor immune suppression — mediated by PD-1/PD-L1 signaling — can still inactivate those T cells when they reach the tumor. Combining a vaccine with a checkpoint inhibitor simultaneously generates trained T cells and removes the brakes on their activity, achieving synergistic immune activation.

Questions to ask your doctor

• Is an mRNA neoantigen vaccine trial open at my cancer center for my tumor type?
• Has my tumor been genomically sequenced, and would that data be sufficient to identify potential neoantigens?
• If a trial is available, how does the manufacturing process work and how long will it take?
• What are the known side effects of mRNA neoantigen vaccines when combined with pembrolizumab?
• Given my cancer type and stage, is a personalized vaccine approach more likely to benefit me as adjuvant therapy (after surgery) or in the metastatic setting?

The bottom line

Personalized mRNA cancer vaccines represent the most individualized approach to cancer immunotherapy developed so far — each vaccine is designed for a single patient's tumor. The KEYNOTE-942 trial produced a striking reduction in melanoma recurrence when the vaccine was combined with pembrolizumab, and early data in pancreatic cancer suggests the concept may extend well beyond melanoma. These vaccines are not yet standard of care for any indication, but clinical trials are actively enrolling and patients with high-risk resected or advanced disease should ask their oncologist whether they are eligible.