Merck and Moderna Say Their Melanoma Vaccine Worked, but Two Months On the Public Still Hasn't Seen by How Much

America’s first therapeutic cancer vaccine reached patients in 2010, and four years later the company that made it was in bankruptcy court. Provenge, Dendreon’s prostate-cancer treatment built from each patient’s own immune cells, won FDA approval on a median survival gain of just over four months, and it cost $93,000 for a course of treatment. When Dendreon filed for Chapter 11 in November 2014, PharmaTimes listed what had gone wrong: a modest benefit at a whopping price, $620 million in convertible debt coming due, and newer prostate drugs, Zytiga and Xtandi, that took the market Provenge was supposed to own. A cancer vaccine can clear every regulatory bar and still deliver far less than its launch promised. Keep that in mind this autumn, because the industry has a new one.

On August 19, Merck and Moderna announced that their phase 3 melanoma trial, INTerpath-001, met its primary endpoint of recurrence-free survival and its key secondary endpoint of distant metastasis-free survival. The product is intismeran autogene, a personalized mRNA therapy given alongside Merck’s Keytruda, and Moderna chief executive Stéphane Bancel called the results “a pivotal moment for the field of cancer research.” The release said the improvement was statistically significant and clinically meaningful. For the phase 3 it gave no hazard ratio, no confidence interval, no event count and no follow-up time, though it found room to repeat the smaller phase 2b trial’s 49 percent reduction in the hazard of recurrence or death. Two months on, the public has seen nothing more. Nobody reading the companies’ materials can tell whether “pivotal” was earned, and the question TrialSite News put this week, whether this is a genuine breakthrough or an expensive promise, can’t be answered from anything Merck or Moderna has published.

The idea itself is clever. Tumors pile up mutations, and some of those mutations produce protein fragments the immune system has never seen, called neoantigens. After surgery, the patient’s tumor is sequenced and Moderna writes a synthetic mRNA coding for up to 34 of that patient’s own neoantigens, given as up to nine doses, one every three weeks. The goal is to train immune cells to find any stray cancer the surgeon missed, while Keytruda takes the brakes off those same cells. INTerpath-001 was built to test this properly: 1,137 patients with completely resected stage IIB to IV melanoma, randomized 2:1, double-blind, with a placebo. That is a big step up from the trial that started the excitement.

The 0.053 problem

KEYNOTE-942, the phase 2b, was open-label, with 157 patients randomized 2:1, which left 50 people in the control arm. Patients and their doctors knew who was getting the experimental product, which is exactly what blinding is meant to prevent. Only the highest-risk patients, stage IIIB to IV, were eligible, and the study was funded by Moderna in collaboration with Merck. When the first result came out in December 2022, the trade press reported a cancer vaccine that “shows promise in trials”. The peer-reviewed version in The Lancet gave the primary result as a hazard ratio of 0.561, roughly a 44 percent lower hazard of recurrence or death, with a 95 percent confidence interval of 0.309 to 1.017 and a two-sided p-value of 0.053. The interval crosses 1, the value that means no difference between the arms, and the p-value misses the two-sided 0.05 threshold most readers have in mind when they hear a trial succeeded. Phase 2b studies often use a looser bar, and the program’s updates have leaned on the friendlier arithmetic: the companies’ three-year update in December 2023 reported its result with a one-sided nominal p-value of 0.0095.

Longer follow-up made the curves look better. At a median of 60.3 months, 68.8 percent of patients on the combination were alive without recurrence, against 49.1 percent on Keytruda alone, and the recurrence-free hazard ratio settled at 0.51, the same number the August release reused. Overall survival, the endpoint patients care about most, came in at a hazard ratio of 0.47. It was also an exploratory endpoint in a trial with 50 controls, which makes it a signal worth testing and not yet a finding.

The phase 3 also enrolled a broader group, opening eligibility to stage IIB and IIC. Melanoma stages don’t line up neatly with risk, since some stage II tumors recur as often as early stage III ones, but if the phase 3 population recurs less often on average, the same relative effect prevents fewer recurrences in absolute terms. A smaller relative effect shrinks the gain further, and that’s what analysts expect. According to MedPath, the general consensus is a recurrence-free hazard ratio of about 0.65, a 35 percent lower hazard instead of the 49 percent the companies keep quoting from the smaller trial. Citi’s Geoff Meacham wrote in August that without the hazard ratio it was “difficult to assess the magnitude, durability and commercial relevance of benefit in adjuvant melanoma.” When the old 49 percent sits next to the new announcement, readers will assume the bigger trial produced a similar number, and nothing public shows that it did.

Nine doses for people with no visible cancer

Side effects weigh more here than in most cancer trials because of who gets the drug. In The Lancet report, grade 3 or worse treatment-related adverse events hit 25 percent of the combination arm against 18 percent on Keytruda alone. At three years, the most common events attributed to the mRNA product were fatigue in 60.6 percent of patients, injection-site pain in 56.7 percent, and chills in 49 percent. These patients have no visible cancer after surgery. Some carry hidden disease, which is why they’re treated at all, yet in the phase 2b about half of those on Keytruda alone were still recurrence-free five years later. For someone who may never relapse, nine extra injections and a higher chance of a serious reaction is a very different bargain than it is for someone whose tumor is growing.

For a phase 3 with 1,137 patients, the companies have so far published one sentence on safety: the profile was consistent with earlier studies, “with no new safety signals observed.” This is the same company whose COVID shot, mRNA-1273, led pathologists to publish cases of lymphohistiocytic myocarditis after vaccination. Intismeran is a different product with a different payload, and nobody has reported a heart problem with it. Repeated mRNA dosing in people with no visible disease should still come with full cardiac and immune safety tables. Until October 24, all the public has is an adjective.

Who gets paid if it works

Moderna’s need for a win is right there in its filings. In the second quarter of 2026 the company reported revenue of $145 million and a net loss of $782 million, and it paid $950 million in July to settle litigation. Lower COVID vaccine sales in the U.S. and South America were offset by deliveries to the United Kingdom, and research spending fell after what the company called “the wind-down of several late-stage programs.” A company that once sold pandemic shots by the billion now has much of its future riding on a melanoma readout.

Merck’s stake is quieter and probably bigger. Intravenous Keytruda faces loss of exclusivity in 2028, and Merck’s hedge is Keytruda Qlex, the under-the-skin version approved in September 2025, which has patents of its own. Both arms of INTerpath-001 got Keytruda, so intismeran was tested as an add-on to Merck’s drug and never as a competitor. Moderna’s filing also describes a newer phase 3 in high-risk stage I lung cancer that tests intismeran alone and with Keytruda Qlex. As strategy, pairing a personalized add-on with the patent-protected version of Merck’s flagship is a tidy way to extend a franchise, whatever the product ends up doing for patients.

No list price exists. According to the Chinese tech outlet QbitAI, the U.S. investment bank William Blair modeled a wholesale price as high as $475,000 per patient, matching the 2017 launch price of the CAR-T therapy Kymriah, though the outlet notes that sources disagree on whether that figure covers one dose or a full course. By QbitAI’s count, Keytruda adds up to about $221,000 over nine cycles, which would put the combination near $696,000 in that scenario. Provenge’s $93,000 already looked steep for four extra months of median survival. If the phase 3 effect comes in closer to 0.65 than 0.51 and is spread across lower-risk patients, the cost of each recurrence prevented climbs fast, and insurers and taxpayers would pay most of it.

INTerpath-001 is a better trial than the one that started the hype. It was blinded and placebo-controlled, and it met its endpoints. But meeting an endpoint is a threshold, not a size, and Merck and Moderna have spent two months collecting headlines on the threshold alone. Until the size is public, “expensive promise” is the more accurate description.

The size arrives on Saturday, October 24, at 4:30 p.m. Madrid time, when Georgina Long presents the trial as late-breaking abstract LBA1 at the ESMO Presidential Symposium. Look for these on her slides: the hazard ratio with its confidence interval, the absolute gap in recurrence-free survival both overall and among the stage IIB and IIC patients, the median follow-up, grade 3 or worse adverse events by arm, and whether overall survival appears at all. Beating the analysts’ 0.65 will move Moderna’s share price, but the absolute gap is what matters to patients. After Long leaves the stage, the next move belongs to Merck and Moderna, who still have to set a price for intismeran and file it with the FDA.

Sources

1. Merck: INTerpath-001 phase 3 topline announcement (August 19, 2026)
2. The Lancet: Weber et al., KEYNOTE-942 randomised phase 2b (2024)
3. HMP Global Learning Network: five-year KEYNOTE-942 results
4. Moderna and Merck: three-year KEYNOTE-942 update (December 2023)
5. MedPath: intismeran phase 3 data at the ESMO Presidential Symposium
6. Moderna: second-quarter 2026 results, SEC filing exhibit
7. QbitAI: William Blair’s price estimate for intismeran (August 2026)
8. PharmaLive: As exclusivity loss looms, Merck wins subcutaneous approval for Keytruda
9. PharmaTimes: Dendreon files for Chapter 11 (November 2014)
10. MDedge: Vaccine approved for prostate cancer
11. Chemical & Engineering News: Moderna and Merck’s cancer vaccine shows promise in trials (2022)
12. American Journal of Clinical Pathology: Lymphohistiocytic myocarditis and Moderna mRNA-1273 vaccine (2022)
13. TrialSite News: Moderna’s mRNA cancer vaccine, a genuine breakthrough or an expensive promise?