A 42% Cut in Melanoma Relapse Means 12 More Per 100 at Two Years

Picture 100 people whose melanoma has been cut out and who then take Keytruda (pembrolizumab), an immunotherapy drug approved in the United States for use after surgery for stage IIB to III melanoma. In a large new trial reported by Merck and Moderna, an estimated 30 of every 100 such people had seen the cancer return, or had died, by the two-year mark. That is the comparison group, and the number to hold onto.

Among people who also received an experimental, individually made mRNA treatment, the figure was 17.8 in 100. The gap is about 12 people. Vaccinologist Helen Petousis-Harris, who declares industry research funding and advisory-board roles, told the NZ Science Media Centre:

"After two years, around 7 out of 10 patients receiving pembrolizumab alone were alive without their cancer returning. Adding the personalised vaccine increased this to just over 8 out of 10. That means approximately 12 additional patients in every 100 were alive without recurrence at two years. This is a meaningful improvement, particularly for people at high risk of their cancer returning, but it is not a cure."

Assoc Prof Helen Petousis-Harris, vaccinologist, University of Auckland

The treatment is intismeran autogene (also V940 or mRNA-4157). Headlines call it a cancer vaccine, which can mislead: it is given after surgery to people who have already had melanoma, to reduce the chance of recurrence. It does not prevent the disease. Merck and Moderna reported the detailed results on Thursday, Oct. 22, 2026, with a headline figure of 42% lower risk of recurrence or death. What that means for real people is less simple.

Two years in, about 12 more people per 100 are recurrence-free

The trial, called INTerpath-001, enrolled 1,137 people with stage IIB to IV cutaneous melanoma that surgery had completely removed. Twice as many were randomly assigned to the combination as to Keytruda plus placebo injections, double-blind, meaning neither patients nor doctors knew who received which. Both groups took Keytruda for about a year.

According to the companies, the combination lowered the risk of recurrence or death by 42%. That is a hazard ratio of 0.58, a measure of how quickly events pile up in one group compared with the other, over a median follow-up of 24.6 months. The 95% confidence interval, the range in which the true value plausibly sits once chance is allowed for, was 0.45 to 0.75, entirely below 1.0.

The two-year figures give the same message in people rather than ratios. The companies reported 82.2% alive and recurrence-free with the combination, against 70.0% with Keytruda alone. Their intervals (79.3% to 84.8%, and 65.0% to 74.5%) do not overlap. As event rates, 17.8% against 30.0% is a 40.7% lower rate at two years, which agrees with the 42% hazard ratio.

Put another way, roughly one extra person stays recurrence-free for every eight or nine treated, at two years; Dr Lennard Lee of the University of Oxford gave approximately nine.

Head counts beat percentages because a relative cut is applied to a starting risk, and this control group started high, at about 30 in 100 by two years. Someone at lower risk would, if the same proportional gain held, get a smaller absolute one (the same lens applies to hormone-therapy risk figures). Stage-by-stage numbers are unpublished; the companies' release said only that improvements were "generally consistent across pre-specified patient subgroups", and Merck's Jane Healy told BioSpace that benefit did not differ by stage, without giving numbers.

All of these figures come from the companies' own release. The paper is accepted by the New England Journal of Medicine but unpublished; the full presentation is Saturday, Oct. 24, at the European Society for Medical Oncology (ESMO) congress in Madrid.

No visible benefit in the first six months

The most surprising line in the release is not the 42%. It is that in the first six months the combination was comparable with Keytruda alone: a hazard ratio of 0.99 (95% CI 0.66 to 1.49) for recurrence or death, which shows no detectable difference.

After six months, a pre-specified sensitivity analysis showed a 58% lower risk (HR 0.42; 95% CI 0.30 to 0.58). A sensitivity analysis is a supporting check, not the main result. So the reported 42% blends a flat start with a stronger later effect.

The 58% is a lower event rate in a later window, not a share of patients who respond.

Moderna's chief development officer, Dr David Berman, told TIME the immune system takes time to be activated, which is a company executive's explanation, not a trial finding. The practical lesson is to treat "42%" as an average over two years, not a promise about any given month.

Survival, stage breakdowns and a vaccine-alone arm are still missing

Overall survival, meaning how long people live, is "not yet mature", the companies said, though "a favorable trend" has been observed. Karen Knudsen, chief executive of the Parker Institute for Cancer Immunotherapy, asked Scientific American the plain question: did the vaccinated live longer, and how was their quality of life?

Lee drew the distinction with a metaphor.

"It's a bit like buying a better car that breaks down less often, although it costs more."

Dr Lennard Lee, associate professor, University of Oxford (declares he co-founded the NHS cancer vaccine launchpad)

No price has been announced in the sources reviewed. Lee added that the trial shows the risk of recurrence falls, while whether more people live longer and better is still awaited.

Next, the design cannot isolate the vaccine. Both groups received Keytruda, so, as Petousis-Harris pointed out, how it works alone is unknown. The population was narrow too: fully resected cutaneous melanoma. Professor Christopher Sweeney of Adelaide University (who declares consulting payments from Merck) said people with visible disease need separate study, and Professor Alan Melcher of the Institute of Cancer Research cautioned against assuming other cancers will follow. On stage, a William Blair note, as summarised by BioSpace, read efficacy as similar across lower and higher disease groups: one secondhand reading, not a table.

RBC also pointed to a rival: immunotherapy before surgery (neoadjuvant), which it said gives similar efficacy at potentially much less cost.

Then the trade-off. Treatment-related side effects occurred in 95.8% of the combination arm (723 of 755 patients) and 90.0% of the Keytruda-alone arm (342 of 380), about six more per 100. The most common were fever (52.2%), injection-site pain (51.0%) and fatigue (38.5%). Serious ones, grade 3 to 5, were 14.2% against 13.4%, under one person per 100 apart. BioPharma Dive reported that about 7% of vaccine recipients stopped treatment over an adverse event, against 5% in the comparator group.

Against that, BioPharma Dive noted that even with Keytruda, one-third or more of people with late-stage melanoma will die or see disease return within five years.

Moderna's shares fell against a number from a different trial

The numbers reached the public early. On Oct. 22, results appeared around 12:30 p.m. ET on the news site Oncology Central and came down shortly afterward, per Investing.com and TIKR. BioPharma Dive described screenshots posted on social media and Scientific American said the data leaked there, but none of the coverage reviewed identifies who leaked it.

The companies' release followed at 2:59 p.m. EDT, about two and a half hours later. Moderna's shares closed at $202.63, down 17.57% from $245.81, after a trading halt. The Motley Fool reported 49.3 million shares traded against a three-month average of 19.9 million; intraday, the drop ran from about a fifth to more than a quarter. Merck, which owns Keytruda, closed down 1.76%.

A positive result moved the stock down because investors held a different yardstick. Wall Street expected a 40% to 50% relative reduction, BioPharma Dive reported. RBC said investors had anchored on the earlier trial's 0.51 and hoped for 0.50 to 0.59, putting 0.58 at the low end. The shares had already risen more than sixfold in 2026, including a 177% jump on Aug. 19, per 24/7 Wall St, when the first positive result came with no numbers.

The 49% comes from KEYNOTE-942, a Phase 2b trial presented at ASCO on June 1, 2026; when it first reported on Dec. 13, 2022, the companies had said 44%.

The 49% and the 42% come from different trials.

Read it as arithmetic. The Phase 2b hazard ratio of 0.51 lies inside the Phase 3 interval, and 0.58 lies inside the Phase 2b interval, which is about twice as wide. That does not make the results the same or different; it makes them hard to line up. Berman himself expected a broader population to make the effect harder to see.

Analysts read the data three ways. RBC said approval seems likely but commercial success is far less certain; BMO said the fall likely reflected inflated expectations rather than a true indictment of efficacy; TD Cowen said the findings substantially clear the bar for a practice-changing dataset. Dr Hussein Tawbi of MD Anderson, an investigator in the trial, told BioPharma Dive why the result matters regardless:

"We have tried really hard. I have led clinical trials. None of them has improved over PD-1."

Dr Hussein Tawbi, professor of medicine, MD Anderson Cancer Center

A share price measures expectations, not patient benefit.

For a patient, access is the plainer question. The treatment is not approved anywhere, and the companies say only that they are "engaging with regulatory authorities". On Aug. 19, Bloomberg reported chief executive Stéphane Bancel saying approval could come as soon as 2027; ABC reported that Melanoma Institute Australia said a regulatory review could take several years. INTerpath-001 is fully enrolled, though related studies, such as a Phase 2 in first-line metastatic melanoma, were listed as recruiting.

Each dose is made for one person. A tumor sample is sequenced against normal DNA, and an algorithm picks up to 34 mutations most likely to provoke an immune response, Merck's Jane Healy told Bloomberg. It takes about six weeks, during which patients stay on Keytruda. Professor Georgina Long, the principal investigator, offered an image:

"It's like a wanted poster that gets printed into the body so … the immune cells can see."

Prof Georgina Long, principal investigator, Melanoma Institute Australia

ABC News calls it an Australian-led study, run at 157 sites worldwide; the NZ Science Media Centre notes New Zealand patients took part. Long told ABC: "We did not expect it to be that high. We knew it would be important and significant, but this is phenomenal." In the UK, Lee estimated that 3,000 to 4,000 patients a year could potentially be eligible.

Questions for an oncologist might start here. What is my stage, and was it completely removed? Is Keytruda part of my plan, and is it approved for my stage where I live? Is there a trial I could join? Is tissue from my surgery stored and suitable for sequencing? What does follow-up look like in the first two years, when the release says recurrence most often happens?

And sun protection stays. Dr Jessamy Tiffen of the Centenary Institute told ABC: "And don't throw away the sunblock just yet. This vaccine is in no way preventative and is intended for advanced stage II-IV melanoma patients … the best way to prevent melanoma is still to protect yourself from the sun." A true preventive, like the HPV vaccine against cervical cancer, works before disease starts.

The group to watch is still the one that got only Keytruda. Its curve will be redrawn when overall survival is reported, and that will show whether the 12 extra people at two years stay well, or only relapse later.