Moderna and Merck's Personalized mRNA Vaccine Cuts Melanoma Recurrence 42% in Phase 3
This Thursday, Moderna and Merck released the full results of INTerpath-001, the pivotal Phase 3 trial of intismeran, their personalized mRNA vaccine for melanoma. The headline number: in 1,137 patients whose high-risk melanoma had been surgically removed, adding the custom-brewed vaccine to Merck's Keytruda cut the risk of recurrence or death by 42% compared with Keytruda alone. It is the first Phase 3 success for any personalized cancer vaccine — and, I'd argue, the most important clinical result mRNA technology has produced since the COVID shots.
The vaccine is not a vaccine in the way we usually mean it. For each patient, scientists sequence the tumor, hunt for mutations that create proteins no healthy cell should carry, and synthesize an mRNA recipe that teaches the immune system to recognize those so-called neoantigens. Up to nine doses, three weeks apart, given alongside Keytruda. The whole process — from analyzing the tumor's genetics to the first dose — takes roughly six weeks. Moderna brews it at a dedicated facility in Massachusetts that is currently supplying thousands of patients across multiple international studies.
Let me be precise about the 42%, because relative numbers are how excitement travels and absolute numbers are how patients live. At two years, 82.2% of patients who got vaccine plus Keytruda were alive without recurrence, versus 70.0% on Keytruda alone — a 12.2 percentage-point gain. In a setting where Keytruda alone already works well, that is a meaningful improvement, not a miracle cure. Distant metastasis-free survival improved by 38% (88.5% versus 80.2%). The trial is continuing to track participants, and there is no overall-survival data yet — that is the number everyone is still waiting for.
The reactions from researchers who had nothing to sell were the part I paid attention to. Dr. Hussein Tawbi of MD Anderson, one of the trial investigators, called it "a big win" that "the field has been looking for." Karen Knudsen of the Parker Institute called it "a new frontier." Dr. Michael Atkins of Georgetown said it was "a big step forward for science, for oncology, because this vaccine will apply to many different cancers." The cautions were equally instructive: Professor Alan Melcher in London called the results "very impressive" but warned against assuming the same approach works across all cancers, and Dr. Jessamy Tiffen noted that this is not a preventive shot — "don't throw away the sunblock just yet."
Regulators are next. The companies say they are engaging with regulatory authorities about next steps, could seek approval as early as next year, and the program reportedly holds FDA Breakthrough Therapy designation. Prediction markets are already pricing it: Polymarket traders currently put the chance of FDA approval for intismeran by December 31, 2027 at around 64%. My own number, below, is 60%. The case for it is a clean Phase 3 win on both recurrence-free and distant-metastasis endpoints at a prespecified interim analysis, with no new safety signals. The case against is that this is genuinely new territory for regulators — every dose is unique, a batch of one — and no filing date has been announced. That operational novelty is the real uncertainty, not the biology.
What I keep coming back to is the information structure of the thing. The COVID vaccines were one recipe shipped to billions; intismeran is 1,137 recipes shipped to 1,137 people. It is medicine compiled like software — tumor DNA in, neoantigen predictions out, mRNA synthesized to order. If the COVID era proved mRNA could manufacture at civilization scale, this trial suggests it can also manufacture at individual scale. The questions now are the unglamorous ones: cost (undisclosed; personalized medicine is expensive), manufacturing throughput, and whether the approach generalizes beyond melanoma — lung, bladder, and kidney trials are already underway. Moderna's shares fell 18% when the full results landed, though they remain more than 200% above their late-August level — the market had priced a miracle and got something better: a real, complicated, working first version.
My prediction: the FDA approves intismeran for adjuvant melanoma by the end of 2027, 60% confidence. Not because the science is finished — overall survival is still pending — but because a 42% relative risk cut with no new safety signals is exactly the kind of result regulators move on.