Updated on 24 October 2026, following the presentation of INTerpath-001 at the European Society for Medical Oncology (ESMO) Congress in Madrid, Spain.
A good tailor does more than choose a size. They take measurements, observe how their client moves, and cut the cloth with one particular person in mind. A similar principle underlies INTerpath, the clinical trial programme through which Merck (known as MSD outside the United States and Canada) and Moderna are investigating a cancer treatment designed for each patient.
For people whose melanoma has been completely removed by surgery, this tailor-made approach has met its primary endpoint in a pivotal phase 3 trial. Presented by Georgina Long of Melanoma Institute Australia on 24 October 2026 at the ESMO Congress in Madrid, the findings show statistically significant improvements in recurrence-free survival and distant metastasis-free survival when the personalised therapy is added to pembrolizumab. They also provide a clearer picture of its limitations, risks and practical challenges.
The Tumour’s Flags: To understand the idea, it helps to consider a characteristic feature of cancer. Tumour cells accumulate changes in their genetic material, known as mutations. Some give rise to altered proteins whose fragments may be recognised as foreign by the immune system. These fragments are called neoantigens. Think of them as small flags that can help distinguish certain cancer cells from healthy ones. The challenge is to get the body’s defences to recognise those signals and respond to them.
The treatment is designed by analysing genetic material from the patient’s tumour and a sample of normal tissue. Comparing the two allows researchers to identify changes specific to the cancer and distinguish them from inherited genetic variants. Using computational algorithms, they select up to 34 neoantigens and manufacture synthetic messenger RNA (mRNA) encoding their sequences. Once administered, the mRNA provides instructions for cells to produce proteins containing the selected neoantigen sequences. These proteins are processed into fragments that can be presented to T cells. The aim is to activate T cells capable of recognising those signals and attacking tumour cells that display them.
Messenger RNA carries instructions that cells use to make proteins. The technology is also used in some COVID-19 vaccines, but here the instructions are selected to trigger an immune response against one person’s cancer.
The treatment is called intismeran autogene, usually shortened to intismeran, and is also known as V940 or mRNA-4157. It is described as a therapeutic vaccine because it seeks to induce an immune response against tumour cells. Rather than preventing a tumour from developing in the first place, it aims to help the immune system recognise and destroy any remaining cancer cells, thereby reducing the risk of recurrence after surgery. It remains investigational.
Two Treatments with Complementary Roles: In the melanoma trials, intismeran is combined with pembrolizumab, an antibody that blocks PD-1, an immune checkpoint receptor that can restrain T-cell activity and limit the response against a tumour. The treatments have complementary roles. Intismeran is intended to prime or expand T cells against selected tumour-specific signals, while pembrolizumab blocks an inhibitory mechanism that can limit their activity.
The combination is given after all visible tumour has been removed, when tests show no evidence of disease but microscopic cancer cells may still remain. Treatment given after surgery to reduce the risk of recurrence is known as adjuvant therapy.
In INTerpath-001, the planned regimen consisted of up to nine doses of intismeran, given every three weeks, alongside pembrolizumab at 400 mg every six weeks for up to nine cycles—approximately one year. The comparison group received placebo instead of intismeran, together with the same pembrolizumab regimen.
What Phase 2 Showed: Clinical trials progress through phases that address different questions. Phase 2 assesses a treatment’s activity and further characterises its safety. Phase 3 seeks more robust evidence of benefit, usually in a larger population and against a defined comparator.
The phase 2b KEYNOTE-942 trial enrolled 157 patients with completely resected, high-risk stage IIIB–IV cutaneous melanoma. A five-year update, based on a data cut-off of 15 December 2025, showed that the combination maintained its advantage over pembrolizumab alone.
The hazard ratio for recurrence or death was 0.51, corresponding to a 49% relative reduction in the instantaneous rate of these events. For distant metastasis or death, the hazard ratio was 0.41, corresponding to a 59% relative reduction. The estimated five-year recurrence-free survival rates were 68.8% with the combination and 49.1% with pembrolizumab alone.
These figures need careful interpretation. A hazard ratio compares the instantaneous rates at which events occur in the two groups among patients who have not yet experienced the event. It incorporates information across the follow-up period rather than simply comparing the proportions affected at a particular time. A 49% reduction in this measure does not mean that 49 recurrences are prevented for every 100 patients treated. The difference between the estimated five-year recurrence-free survival rates was 19.7 percentage points.
The five-year analyses were descriptive and were not designed as confirmatory hypothesis tests. They strengthen the evidence of sustained benefit but do not constitute an independent confirmation or establish that the treatment improves overall survival.
Safety deserves an equally careful account. In the initial publication, treatment-related adverse events of grade 3 or higher were reported in 25% of patients receiving the combination and 18% receiving pembrolizumab alone. The effects attributed to intismeran most frequently included fatigue, injection-site pain and chills, generally mild or moderate.
The trial was small and open-label, meaning that both clinicians and patients knew which treatment had been assigned. A larger, double-blind phase 3 trial was needed to assess benefit and safety more precisely.
A Positive Phase 3 Trial: INTerpath-001 enrolled 1,137 patients with completely resected, high-risk stage IIB–IV cutaneous melanoma. Participants were randomly assigned in a two-to-one ratio to receive intismeran plus pembrolizumab or placebo plus pembrolizumab. The trial was double-blind, meaning that neither patients nor investigators knew the treatment assignment. Because both groups received pembrolizumab, the study specifically assessed the benefit of adding the personalised therapy.
The companies announced the positive outcome on 19 August and released the first detailed figures on 22 October. The presentation in Madrid on 24 October added information on overall survival, subgroup findings, safety and manufacturing, but did not represent longer follow-up. The efficacy results came from the same data cut-off of 30 June 2026, with a median follow-up of 24.6 months.
The combination reduced the hazard of recurrence or death by 42%, with a hazard ratio of 0.58 and a 95% confidence interval of 0.45–0.75. For distant metastasis or death, the relative reduction was 38%, with a hazard ratio of 0.62 and a 95% confidence interval of 0.45–0.85. Both results were statistically significant, with p values below 0.001 and equal to 0.002, respectively.
Both endpoints count death from any cause as an event. Recurrence-free survival also counts any recurrence of the disease, whereas distant metastasis-free survival counts the development of metastases at sites distant from the original tumour.
The primary analysis followed the intention-to-treat principle. Patients were analysed in the groups to which they had originally been assigned, regardless of whether they received or completed the planned treatment.
At two years, estimated recurrence-free survival was 82.2% with the combination and 70.0% with pembrolizumab plus placebo—a difference of 12.2 percentage points. Estimated distant metastasis-free survival was 88.5% and 80.2%, respectively, a difference of 8.3 percentage points. These estimates help convey the absolute magnitude of the benefit at that point in follow-up, beyond the relative reductions.
The Benefit Was More Apparent After Six Months: The ESMO presentation added an important qualification. In a prespecified analysis examining separate time periods, there was no clear difference between the groups in recurrence or death during the first six months. The hazard ratio was 0.99, with a 95% confidence interval of 0.66–1.49. Beyond six months, it was 0.42, with a confidence interval of 0.30–0.58.
A similar pattern was observed for distant metastasis or death. During the first six months, the hazard ratio was 1.01, with a 95% confidence interval of 0.58–1.79. Beyond six months, it was 0.50, with a confidence interval of 0.34–0.73.
The observed benefit was concentrated after the first six months. This analysis does not, however, establish that the timing was caused solely by the manufacturing process or by the time needed to develop an immune response. Nor does it justify replacing the main 42% relative reduction with the 58% reduction observed after six months. These analyses address different questions, and the primary result remains the estimate across the entire follow-up period.
Overall Survival Remains Unresolved: The additional findings clarify another essential question. For death from any cause, the hazard ratio was 0.77, with a 95% confidence interval of 0.45–1.31 and a p value of 0.33. The prespecified threshold for statistical significance was not met.
The point estimate favours the combination, but uncertainty remains substantial. The trial has not yet established that patients live longer, although such a benefit cannot be ruled out. The data are still immature, and follow-up continues.
According to information released after the presentation, the results have been accepted for publication in The New England Journal of Medicine. At the time of the presentation, the full paper had not yet been published. Its eventual publication will allow closer examination of the methods, statistical analyses and findings.
Understanding the Safety Findings: In phase 3, grade 3–5 treatment-related adverse events were reported in 14.2% of patients receiving the combination and 13.4% receiving the comparison regimen. Under the standard toxicity grading system, grade 3 denotes severe events, grade 4 life-threatening events and grade 5 death. This grading system is distinct from the separate regulatory classification of a serious adverse event, which is based on criteria such as death, a life-threatening outcome or hospitalisation.
When all grade 3–5 adverse events were counted, regardless of whether they were attributed to treatment, the proportions were 23.6% and 20.3%. Treatment-related adverse events of any grade occurred in 95.8% of patients in the combination group and 90.0% in the comparison group. Treatment-related adverse events led to discontinuation of at least one study drug in 14.0% and 8.7%, respectively.
Events with a fatal outcome were reported in eight patients (1.1%) in the combination group and none in the comparison group. Two deaths were considered treatment-related and both were attributed to pembrolizumab. Both occurred before the first dose of intismeran. No deaths were attributed to intismeran.
The most frequently reported adverse events in the combination group included fever, injection-site pain and fatigue. Fever and injection-site reactions were generally low-grade and transient. Immune-mediated adverse events of particular interest for pembrolizumab occurred in similar proportions in the two groups, at 36.0% and 34.7%.
Both sides of the balance matter. The combination was associated with more treatment-related adverse events overall and more treatment-related discontinuations, although the rates of grade 3–5 treatment-related events were similar. The absence of new safety signals does not mean that important risks are absent, and the safety profile should not be equated with that of a routine preventive vaccine.
The Workshop and Its Timelines: Manufacturing a different medicine for each person requires a different production model. The Madrid presentation provided further evidence of its feasibility, while also revealing an important obstacle before manufacturing even begins.
Of the 1,749 people screened for participation, 302 (17.3%) were excluded because their tumour sample was inadequate. Among the 757 patients assigned to the combination group, intismeran was successfully manufactured for more than 99%. The maximum of 34 neoantigens was identified for treatment design in 91% of the 1,137 randomised patients. The median number of doses administered was nine.
These figures describe distinct steps. Manufacturing was highly reliable among patients who passed screening, but access to suitable tumour tissue limited participation for a proportion of candidates. It would be misleading to present the manufacturing success rate as though it applied to everyone screened.
A median interval of 64 days to the first intismeran dose was also reported. This should not simply be labelled “manufacturing time”, because the available summary does not adequately specify the starting point or all the stages included in that interval.
Introducing the treatment into routine care will require an assessment of tissue collection and preservation, genomic analysis, logistics, cost and the capacity to coordinate the entire pathway. For publicly funded healthcare systems, including Spain’s, the question will not simply be whether the treatment works, but how to deliver it promptly, sustainably and equitably.
Beyond Melanoma: Intismeran’s clinical development programme includes trials in non-small-cell lung cancer, bladder cancer and renal cell carcinoma. Phase 3 lung cancer studies are under way, including one in patients with high-risk stage I disease. Each trial must establish separately whether the approach offers a benefit in that setting.
Not every study continues. INTerpath-007, which investigated cutaneous squamous cell carcinoma, was terminated after enrolling 46 patients. The stated reason was a business decision, which does not, by itself, establish either a lack of efficacy or a safety problem.
Another question is what this strategy can achieve in tumours with limited immune activity. The terms “hot” and “cold” broadly describe differences in that activity. They are not rigid categories and cannot, on their own, predict how a therapeutic vaccine will perform.
Experience with autogene cevumeran, another personalised mRNA therapy developed by BioNTech and Genentech, illustrates why each setting must be evaluated separately. Its phase 2 trial in surgically treated colorectal cancer was designed to assess the vaccine as adjuvant monotherapy in patients with high-risk stage II or stage III disease and detectable circulating tumour DNA. The trial was terminated in August 2026 following a recommendation from the independent data monitoring committee. The committee identified a numerical imbalance in overall survival and considered further continuation unlikely to change the efficacy outcome.
That finding does not, by itself, prove that the vaccine caused harm. Nor can the outcome simply be attributed to the absence of pembrolizumab or to the tumour being “cold”. The study evaluated a specific strategy in a particular patient population, without an accompanying immune checkpoint inhibitor. BioNTech’s statement on the colorectal cancer trial.
In pancreatic cancer, a phase 1 trial studied autogene cevumeran after surgery, alongside atezolizumab and chemotherapy. Eight of the 16 vaccinated patients developed a detectable vaccine-induced T-cell response. In an update presented in 2026, after a median follow-up of 4.2 years, seven of those eight patients were still alive, compared with two of the eight who had not developed a detectable response.
The observation supports further research but does not establish that the vaccine prolongs life. The study was small, and comparing patients who developed an immune response with those who did not is not equivalent to a randomised comparison of efficacy. Differences between the groups may also reflect other biological characteristics. A phase 2 trial is evaluating the strategy in a larger population with a comparison group. American Association for Cancer Research, overview of pancreatic cancer vaccine research.
The Regulatory Path: In 2023, the US Food and Drug Administration granted Breakthrough Therapy designation to V940 in combination with pembrolizumab for the adjuvant treatment of high-risk melanoma following complete resection. The combination was also granted access to the European Medicines Agency’s PRIME programme, which supports promising medicines addressing unmet medical needs.
These designations facilitate development and dialogue with regulators, but they do not constitute marketing authorisation.
The companies have reported ongoing discussions with regulatory authorities. The phase 3 findings provide stronger clinical evidence, but intismeran remains investigational.
A Measured Conclusion: A tailor knows that a suit is judged when it is worn, not when its pattern is drawn. The results presented in Madrid provide evidence that adding a personalised neoantigen therapy to pembrolizumab improves recurrence-free and distant metastasis-free survival in people with surgically treated, high-risk melanoma. The durability of that benefit, its effect on overall survival and the requirements for routine clinical use still need to be established more fully.
Other tumours will need their own evidence. The aim is not to dress everyone in the same suit, but to discover how far a shared approach can go while adapting treatment to the characteristics of each person’s cancer. If the evidence continues to support it, the challenge will be to make that individualised approach available to many.
Today’s research is tomorrow’s therapy