mRNA Vaccines Are Coming for Cancer — The Clinical Trial Results Changing Oncology

The Proof of Concept: Melanoma

The clearest data so far comes from skin cancer. In late 2023, Moderna and Merck published results from KEYNOTE-942, a Phase 2b randomized controlled trial testing mRNA-4157/V940 — a personalized cancer vaccine — combined with Pembrolizumab (Keytruda). The finding: combination therapy reduced the risk of recurrence or death by 49% compared to Pembrolizumab alone in patients with high-risk Stage 3/4 melanoma who had already undergone surgery.

These were patients facing high odds of recurrence after tumor removal. The vaccine is built individually: a biopsy is sequenced to identify the unique mutations in that patient's tumor, an algorithm selects up to 34 of the most immunogenic neoantigens — proteins that appear on cancer cells but not on healthy tissue — and a personalized mRNA vaccine is synthesized and delivered within roughly six weeks. The vaccine teaches the immune system to recognize and attack that specific tumor's molecular fingerprint.

Phase 3 trials for melanoma (KEYNOTE-V940-006) are now enrolling at dozens of centers worldwide. Merck and Moderna expect readout data in 2027.

Pancreatic Cancer: The Hard Test

Pancreatic ductal adenocarcinoma has a five-year survival rate under 12% — one of the worst outcomes of any major cancer. BioNTech, which developed the mRNA COVID vaccine with Pfizer, is running an individualized neoantigen trial (BNT122) for pancreatic cancer in combination with atezolizumab and modified FOLFIRINOX chemotherapy.

Early Phase 1 results published in Nature showed that about half of patients who received the vaccine generated strong T-cell responses targeting their cancer's specific mutations. In the group that responded, the disease-free survival curve separated notably from non-responders at 18 months. The cohort was small and follow-up limited, but these are the first credible data showing durable anti-tumor immunity from individualized mRNA vaccination even in this notoriously treatment-resistant cancer. Phase 2 expansion is underway.

Pancreatic cancer is a harder target than melanoma. It has fewer mutations for the neoantigen selection algorithm to work with, its tumor microenvironment actively suppresses immune activity, and it often metastasizes before diagnosis. The BioNTech data is early. But it exists, which three years ago was not guaranteed.

Non-Small-Cell Lung Cancer and the Pipeline Beyond

Multiple cancer types are now in early trials. Moderna has active studies for non-small-cell lung cancer (mRNA-4359, in combination with checkpoint inhibitors), bladder cancer, and head-and-neck cancers. The common pattern: cancers with higher mutational burden — more mutations available as potential neoantigens — are better targets for individualized vaccination. That is why melanoma and lung cancers driven by smoking or UV damage, which accumulate large numbers of mutations, were natural first choices.

Colorectal cancer with high microsatellite instability (MSI-H subtype) is another active target because of its elevated mutation rate. BioNTech is running mRNA neoantigen trials for this subtype. Roche's Genentech division is in early studies with its own personalized cancer vaccine platform. BioNTech has publicly committed to reaching regulatory submission for at least one cancer indication by 2030.

How the Manufacturing Works — and Why It Is Still a Problem

The core operational challenge is time. Tumor sequencing, neoantigen selection, and mRNA synthesis must happen quickly — delays give the tumor time to evolve or spread. Moderna's current turnaround for personalized vaccine synthesis is approximately six weeks from biopsy to first dose. The company is targeting four weeks by the time Phase 3 reads out.

The neoantigen selection algorithms are a focus of intense ongoing research. The KEYNOTE-942 algorithm selected up to 34 antigens per patient based on predicted immunogenicity and MHC presentation likelihood. Companies are now integrating multi-modal molecular data — proteomics, HLA typing, T-cell receptor profiling from tumor-infiltrating lymphocytes — to sharpen predictions. The hypothesis is that better antigen selection will push response rates above the roughly 50% seen in early pancreatic cancer data.

Cost is the other unresolved question. Personalized cancer vaccines are currently priced at a level that makes them accessible only through clinical trials or well-resourced healthcare systems. Manufacturing happens at specialized mRNA synthesis facilities, and each vaccine is unique to each patient. If Phase 3 succeeds and the FDA grants approval, manufacturing scale-up and pricing will become the dominant challenge — much as they were for CAR-T cell therapy, which has been commercially approved since 2017 but remains difficult to access outside major cancer centers.

What Personalized Cancer Vaccines Are Not

mRNA cancer vaccines as they exist today are adjuvant therapies: given after surgery to reduce recurrence risk, or combined with checkpoint inhibitors to improve response. They are not single-agent cures. They do not work for all cancer types — cancers with low mutational burden have fewer neoantigens to target.

They are also fundamentally different from the COVID vaccine model. The COVID mRNA vaccine trained immune systems against a shared viral antigen that every patient had in common. A personalized cancer vaccine is synthesized from scratch for each patient based on their specific tumor mutations. It cannot be stockpiled, cannot be standardized across patients, and requires per-patient manufacturing infrastructure that does not yet exist at commercial scale.

A separate, related category — preventive cancer vaccines targeting shared tumor antigens that appear across many patients — is in earlier development. Moderna is running a small trial of a prophylactic mRNA vaccine for individuals at high genetic risk for specific cancers. This would resemble a traditional vaccine conceptually, but the science is considerably less mature than the personalized neoantigen approach.

The Most Likely Timeline

The most plausible near-term approval scenario: personalized mRNA cancer vaccines for adjuvant melanoma treatment, contingent on positive Phase 3 data in 2027 and FDA review in 2028–2029. That would make mRNA-4157/V940 the first approved personalized cancer vaccine — and the first major new use case for mRNA technology since COVID-19.

Other indications require separate trials and longer timelines. But the platform infrastructure, once validated for one cancer, can be redeployed to new neoantigen targets faster than traditional drug development allows. The mRNA synthesis process is the same regardless of which antigens are encoded. If the FDA approves for melanoma, the path to trials in other cancer types — and eventually accelerated approval for cancers with strong mechanistic similarity — becomes shorter.

The oncology field is well-acquainted with promising Phase 2 results that fail at Phase 3 — the history of cancer drug development includes many such disappointments. But KEYNOTE-942 was a randomized controlled trial, not an observational study. The 49% reduction in recurrence is a large and statistically robust signal. The question is replication at scale. That answer comes in 2027.