The press release Merck and Moderna issued on Wednesday contains no hazard ratio, no confidence interval, no event count, and no median follow up. Moderna's stock rose 177 percent and closed at $174.38. That is the best single day in the company's history. The previous record was 27.81 percent, set on February 25, 2020, which was the week the world figured out what was coming. The company has now had a better day selling a document with no numbers in it than it had at the dawn of the pandemic that made it famous.

The genuinely irritating part is that the underlying result is probably real.

Here is what was actually said. INTerpath-001 is a randomized, double blind, placebo and active comparator controlled Phase 3 trial. It enrolled 1,137 patients with completely resected stage IIB, IIC, III, or IV cutaneous melanoma, randomized two to one to intismeran autogene plus pembrolizumab, or pembrolizumab alone. Intismeran at 1 mg every three weeks for up to nine doses. Pembrolizumab at 400 mg every six weeks for about a year. At a prespecified interim analysis the trial met its primary endpoint of recurrence free survival and its key secondary endpoint of distant metastasis free survival. Overall survival is still accruing. Safety was consistent with prior reports. Data will be presented at a medical meeting that has not been named.

That is the whole disclosure. Everything else written this week is inference, including some of mine.

Intismeran autogene, formerly mRNA-4157 and before that V940, is not a vaccine in the sense your flu shot is a vaccine. The tumor comes out in the operating room. Someone sequences it, compares it against the patient's own germline, and picks up to 34 mutations that look like a T cell might plausibly notice them. A factory in Marlborough then prints an mRNA strand encoding those 34 things and packages it in lipid. The vial that results fits exactly one human being on earth. If the patient recurs before it ships, it is expensive waste that matches nobody.

The useful thing about a data free announcement is that it gives you a window to write down your thresholds before you know the answer. Two firms published theirs this week. Keep this table and check it against whatever gets presented.

This is the part of the week's coverage that got the least attention and deserves the most.

In February 2023 the FDA granted Breakthrough Therapy Designation to this combination on the strength of a hazard ratio of 0.56 whose confidence interval crossed 1.0. Moderna then hoped to file for accelerated approval on that same readout. The FDA closed that avenue in 2024 and made them finish the Phase 3.

Consider what the counterfactual looks like. Accelerated approval in 2024 on 157 patients. A confirmatory trial running in the background while the drug sells. If the effect had regressed toward the null, which was well inside the confidence interval, we would be several years into treating people with a bespoke and expensive therapy on a signal that did not hold, and we would be having the usual argument about how hard it is to pull a drug once oncologists have built practice around it. We have had that argument many times. It never goes well.

Instead the agency held the line, the sponsors ran the trial, and the trial worked. Everybody got a better answer than they would have gotten by the fast route, including the sponsors, who now have a Phase 3 win instead of a contested accelerated approval. It is worth saying out loud during a period when regulatory caution is mostly discussed as an obstacle. Sometimes the obstacle is the thing that makes the result mean something.

Assume the data hold. The interesting question stops being immunological and becomes logistical.

Moderna's chief development officer David Berman told the Globe the window from tumor sampling to a finished personalized dose is about six weeks. He compared it to the 42 days from COVID sequence to first drug, which is a fair comparison and also a slightly worrying one, because that 42 days was a heroic effort by a company with the full attention of the planet, and this needs to be Tuesday.

Every dose requires a tumor specimen of adequate quality, a sequencing run, a neoantigen prediction pipeline, and a manufacturing slot. Then it requires all of that to finish inside the window when the patient is disease free and waiting. The trial protocol allowed no more than 13 weeks between the resection that rendered the patient disease free and the first dose of pembrolizumab. That is a clock, and in the trial it was running in an academic center with a dedicated coordinator.

Now run it at a community oncology practice in a county with one pathologist. The technology is real. The supply chain is the part nobody has built.

This is the same problem CAR T has, and CAR T has never solved it. Years after approval, access still tracks proximity to an academic center more tightly than it tracks who needs the drug. An individualized mRNA therapy is easier to make than an engineered cell and harder to make than anything else in oncology. Where it lands on that spectrum will determine whether this is a treatment or a headline.

In August 2025 the Department of Health and Human Services cancelled 22 BARDA funded mRNA vaccine projects worth close to 500 million dollars, on the stated view that the platform was a poor investment for respiratory pathogens. Twelve months later the same platform produced the first positive Phase 3 in individualized cancer immunotherapy.

These are not the same product and I want to be careful about that. A prophylactic vaccine against a circulating respiratory virus and a therapeutic vaccine encoding 34 mutations from one person's melanoma differ in target, in mechanism, and in how you would evaluate either. Someone could hold both positions without contradiction, and some serious people do.

What they share is the manufacturing base, the regulatory literacy, the lipid nanoparticle chemistry, and the people who know how to run the machines. You cannot defund a platform in one indication and expect it to stay fully staffed in another. That is not an argument about vaccine policy. It is an argument about industrial capacity, which is duller and harder to get out of.