A Cancer Vaccine Made for One Person: How Moderna’s mRNA Therapy and Keytruda Work Together

Imagine removing a patient’s melanoma, examining its genetic mutations, and manufacturing a vaccine designed specifically to help that patient’s immune system recognize any cancer cells left behind.

Now combine that vaccine with a drug that prevents cancer from switching off the immune response.

That is the central idea behind Moderna and Merck’s experimental combination of intismeran autogene (V940, formerly mRNA-4157) and Keytruda (pembrolizumab).

The concept is elegant. More importantly, it now has encouraging Phase 3 clinical evidence. But understanding why it works—and where it might fail—is essential before assuming we have found a broadly applicable cancer treatment.

First, why might cancer remain after surgery?

Surgery can remove every tumor the physician can identify without necessarily eliminating every malignant cell. A patient may have microscopic residual disease: cancer cells too few or too dispersed to detect reliably through imaging or standard clinical examinations.

Those cells can remain dormant, proliferate, or eventually establish recurrent tumors. Adjuvant therapy is treatment given after surgery to reduce that risk. The Moderna–Merck approach attempts to make the patient’s immune system better at recognizing and eliminating malignant cells during this vulnerable period.

Step 1: Identify what makes this particular cancer different

Every melanoma contains genetic alterations acquired as its cells evolved. Some mutations change the proteins the cancer produces. Those altered proteins can contain short peptide sequences called neoantigens. A neoantigen is potentially useful because it can distinguish a malignant cell from normal cells.

The first step is to analyze the surgically removed tumor and compare its genetic information with the patient’s normal DNA. Researchers use computational methods to identify candidate neoantigens and predict which ones might be effectively presented to the immune system.

This is considerably more complicated than simply finding mutations. A useful mutation must generate an appropriate peptide, that peptide must be processed and displayed by the patient’s cells, and immune cells must be capable of recognizing it.

The resulting individualized vaccine encodes up to 34 selected neoantigens. Each patient’s vaccine is therefore different.

Step 2: Manufacture an individualized mRNA vaccine

Moderna uses the selected genetic information to produce a synthetic messenger RNA treatment. The mRNA provides temporary instructions for cells to manufacture fragments representing the selected tumor neoantigens. Those fragments can then be presented to the immune system.

This does not involve permanently modifying the patient’s DNA. Nor does it inject live cancer cells into the body. It introduces instructions intended to teach the immune system what features to recognize.

This is an important distinction between a conventional preventive vaccine and a personalized therapeutic cancer vaccine. A measles vaccine trains the immune system to recognize a pathogen before infection. This cancer vaccine is designed after a malignancy has developed, using molecular characteristics of that existing cancer.

Step 3: Train T cells to recognize the target

Immune cells called antigen-presenting cells process the vaccine-produced proteins and display relevant peptide fragments using molecules called the major histocompatibility complex, or MHC. In humans, these presentation molecules are encoded by the HLA system.

Different individuals have different HLA profiles. Consequently, the same tumor mutation may be highly visible to one person’s immune system and much less visible to another’s. When an appropriate neoantigen is presented, it can activate and expand T cells capable of recognizing that target.

CD8-positive T cells are particularly important because they can directly kill cells displaying their target peptides. CD4-positive T cells can help organize and sustain the broader immune response. The hoped-for result is a larger, more durable population of tumor-reactive immune cells.

But recognizing a target in an experiment does not guarantee that immune cells will reach, recognize, and destroy every malignant cell in a patient.

Step 4: Keytruda releases an immune-system brake

Activated T cells express a receptor called PD-1. When PD-1 interacts with its ligands, including PD-L1, signaling pathways can suppress T-cell activity. Tumors can exploit this system to reduce immune attacks.

Keytruda is an antibody that blocks PD-1, limiting this inhibitory signaling. The two treatments are designed to be complementary:

• The personalized vaccine: Helps generate immune recognition of selected tumor neoantigens.
• Keytruda: Helps prevent PD-1-mediated suppression of activated T cells.

One treatment seeks to expand the relevant immune response. The other seeks to preserve its ability to function. The appealing hypothesis is that these mechanisms together can accomplish more than either alone. And we now have randomized evidence that adding the vaccine improves recurrence-related outcomes compared with Keytruda alone.

What treatment actually looks like

This is not an injection that a physician can select from a pharmacy shelf. The process begins with surgery and acquisition of sufficient tumor material. Genetic analysis, neoantigen selection, and individualized manufacturing must occur before vaccination can begin.

In the Phase 3 INTerpath-001 trial, patients received:

• Intismeran or placebo every three weeks for up to nine doses.
• Keytruda 400 mg intravenously every six weeks for up to nine cycles.
• Approximately one year of adjuvant treatment, subject to recurrence, unacceptable toxicity, or the protocol’s treatment limits.

That manufacturing requirement matters clinically. A personalized vaccine must be produced accurately, consistently, and quickly enough to fit the patient’s treatment window. It also matters economically: a commercially successful treatment would require a repeatable manufacturing process for thousands of different patients, rather than mass production of an identical drug.

What the Phase 3 results actually demonstrate

On October 22, 2026, Merck and Moderna announced detailed results from INTerpath-001, a randomized Phase 3 study involving 1,137 patients with completely resected stage IIB–IV cutaneous melanoma. Patients received either personalized vaccine plus Keytruda or placebo plus Keytruda.

At approximately two years:

The primary analysis demonstrated a 42% relative reduction in the hazard of recurrence or death, with a hazard ratio of 0.58. That hazard ratio measures the relative event rate over follow-up. It does not mean 42% more patients were cured.

The two-year recurrence-free survival difference—12.2 percentage points—provides another useful perspective on the magnitude of benefit.

Interestingly, the companies reported little separation between treatment groups during the first six months, followed by a larger difference afterward. That pattern is biologically compatible with an immune response developing over successive vaccine doses. However, the stronger later-period hazard ratios came from a prespecified sensitivity analysis. They should not replace the primary overall trial result.

What we still do not know

We do not yet know whether the combination extends overall survival. The overall-survival data are immature. A favorable trend is encouraging, but it is not proof.

We also do not yet know how long the recurrence benefit persists. Five-year and longer follow-up will matter, particularly if melanoma returns after immune surveillance weakens or resistant cell populations emerge.

And we should avoid assuming the immune response will recognize every remaining cancer cell. Tumors are heterogeneous. Some cells may lack the selected neoantigens. Others may lose antigen-presentation capability, develop mechanisms that exclude T cells, or occupy environments where immune cells cannot function effectively.

The vaccine also produces inflammatory side effects. In Phase 3, fever, injection-site pain, and fatigue were common. Serious immune-related complications remain possible with Keytruda.

The combination is investigational, not an FDA-approved melanoma vaccine as of this evidence review.

Could it work in other cancers?

Potentially. But that is a hypothesis to test, not a conclusion we can draw from melanoma.

Melanoma frequently carries a substantial mutation burden, partly because ultraviolet radiation produces DNA damage. That can create many possible neoantigens for immune recognition. Other cancers can behave differently.

A tumor with relatively few mutations may offer fewer attractive neoantigen targets. A tumor surrounded by an immunosuppressive microenvironment may prevent activated T cells from reaching their targets. Advanced metastatic disease may also be more difficult to eliminate than microscopic disease after surgery.

Merck and Moderna are investigating the platform in non-small cell lung cancer and several other tumor types, with additional studies exploring settings including pancreatic and gastric cancer. Those studies may reveal where the biology translates—and where it does not.

It is also important not to confuse intismeran with autogene cevumeran, the different individualized mRNA neoantigen vaccine developed by BioNTech and Genentech that we discussed in our October 10 pancreatic cancer article. The conceptual approach overlaps, but the products, designs, diseases, and clinical evidence are different.

Our assessment

The attraction of this technology is not simply that scientists can manufacture a different vaccine for every patient. The remarkable possibility is that we can use a tumor’s molecular information to construct a treatment intended to direct the patient’s own immune system toward that particular malignancy.

The Phase 3 melanoma results suggest that this concept can produce meaningful clinical benefit when combined with PD-1 blockade. That is an important achievement.

But three separate propositions must remain separate:

1. We can successfully manufacture individualized neoantigen therapies.
2. Adding one of these therapies to Keytruda can reduce melanoma recurrence following complete resection.
3. Individualized cancer vaccines can broadly improve survival across cancers and treatment settings.

The first is established technologically. The second now has substantial randomized clinical support. The third remains an open question.

That distinction—between an exciting mechanism, a demonstrated clinical effect, and an unproven broader application—is exactly where scientific evaluation becomes useful.

Companion reading: Therapy Comparisons 03 examines how the combination fits alongside existing melanoma treatments and the questions that would determine its place in clinical practice.

Sources and evidence notes

1. Merck and Moderna: INTerpath-001 Phase 3 results, October 22, 2026.
2. Merck and Moderna: INTerpath clinical program backgrounder.
3. National Cancer Institute: Melanoma Treatment (PDQ).

Evidence cutoff: October 24, 2026. The Phase 3 results were sponsor-released on October 22 and announced as accepted for publication in The New England Journal of Medicine.

← Return to Therapeutics