Merck and Moderna Bet the Melanoma Vaccine Can Travel

Merck and Moderna’s custom mRNA cancer vaccine delayed melanoma’s return and spread after surgery in a 1,137-patient Phase 3 trial, the companies said on August 19. The shot, intismeran autogene, was given with Keytruda, Merck’s PD-1 drug, and beat Keytruda alone in patients whose tumors had already been cut out.

On October 15, Moderna president Stephen Hoge told an industry summit in Boston that the partners still have “a lot of exploring to do and need to prioritize where we explore.” They have not published hazard ratios. Those figures are due in Madrid on October 24.

A Landmark Win That Still Hides Its Numbers

INTerpath-001 enrolled people with completely resected stage IIB to IV cutaneous melanoma and no prior systemic therapy, then randomized them 2:1 to intismeran plus Keytruda or Keytruda alone for about a year. At a planned interim look, the combination improved recurrence-free survival, the primary endpoint, and distant metastasis-free survival, a key secondary endpoint. Overall survival was not mature. The companies called it the first positive Phase 3 readout for an individualized neoantigen therapy and for an mRNA cancer therapy, and the first Phase 3 study to beat Keytruda alone in adjuvant melanoma.

Gary Keblish, a public school teacher diagnosed with melanoma in 2019, went on an early version of the same regimen after a surgeon removed his tumor. Speaking by video at the Boston meeting, he likened the ride to Ms. Frizzle’s Magic School Bus.

Seatbelts, everyone!

Gary Keblish, public school teacher, industry summit in Boston, October 15

He is still teaching. The companies plan to talk with regulators about filing intismeran with Keytruda once the full dataset is public. In the United States, melanoma is on track for about 112,000 new cases and more than 8,500 deaths in 2026, according to figures Merck circulated with its five-year study.

Along with our partners at @Merck, we announced positive results from the Phase 3 INTerpath-001 study of intismeran autogene, our investigational individualized neoantigen therapy, in combination with pembrolizumab as adjuvant treatment in patients with completely resected Stage… pic.twitter.com/oaVDmxeHmx

— Moderna (@moderna_tx) August 19, 2026

WHAT WE KNOW

• The trial: INTerpath-001 (NCT05933577) randomized 1,137 patients 2:1 after complete resection of stage IIB to IV cutaneous melanoma.
• The regimen: Intismeran 1 mg in the muscle every 3 weeks for up to 9 doses, plus pembrolizumab 400 mg in the vein every 6 weeks for up to 9 cycles.
• The result: Recurrence-free survival and distant metastasis-free survival both improved versus Keytruda alone at a pre-specified interim analysis, with no new safety signals named.

WHAT IS UNCONFIRMED

• Effect size: Hazard ratios, p-values, and absolute recurrence rates remain embargoed until the Madrid session.
• Survival: Overall survival is still following and was not mature at the interim look.
• Filing clock: The companies will engage regulators on submissions; no application has been accepted and no decision date exists.

A five-year look at the smaller study that justified Phase 3 is the only public measure of how large the effect might be. In KEYNOTE-942, 157 patients with resected stage III or IV melanoma were assigned 2:1 to the same pairing, using a 200 mg every-3-weeks Keytruda schedule. At a median 60.3 months of follow-up, the combination produced a 49 percent lower risk of recurrence or death (HR 0.51, 95% CI 0.294 to 0.887) and a 59 percent lower risk of distant metastasis or death (HR 0.411, 95% CI 0.200 to 0.843) versus Keytruda alone.

An exploratory overall-survival analysis favored the combination (HR 0.471), but the confidence interval ran from 0.165 to 1.345 and rested on 14 deaths, so it does not settle the survival question. Immune-related side effects showed up in 45.2 percent of combination patients and 44 percent on Keytruda alone. David Berman, Moderna’s chief development officer, said the five-year follow-up gave “a more complete picture of the durability” of the pairing.

Why Merck Is Staying on Keytruda’s Map

Intismeran is not being developed as a replacement for checkpoint blockade. It is being built as an add-on in the settings where Merck already knows PD-1 drugs work, which Dean Li, Merck’s head of research, laid out on the Boston stage as a ranking, not a slogan.

We know exactly where I-O works, where a checkpoint inhibitor works, where it doesn’t, where it almost worked and didn’t make it.

Dean Li, head of research, Merck, industry summit in Boston, October 15

Hot tumors, packed with mutations, are where Keytruda has done its best work. Melanoma after surgery is the cleanest version of that idea: the visible mass is gone, leftover cells are few, and the immune system has a chance to learn a wanted list before the cancer comes back. Adjuvant lung, kidney, and bladder trials are already running on the same logic.

Hoge pushed the ranking one step earlier. “There’s an incredible opportunity to think about how we bring immunotherapy into earlier interventions and really intercept cancer so that we don’t present with stage 3, stage 4 disease later,” he said. That is why a new Phase 3 study in high-risk stage I non-small cell lung cancer includes an intismeran-alone arm, a rare break from the Keytruda pairing.

Li contrasted that map with groups that have gone after cold tumors or bulky metastatic disease, settings where PD-1 drugs often stall. “We have chosen not to do this,” he said. “But I can tell you: Anyone has positive data and it’s out there, you will never see Moderna and Merck pivot so fast.”

Keytruda’s U.S. compound protection ends in 2028, which is not a secret inside Merck. A custom shot that is only sold with Keytruda, in patients Merck already treats, is a way to keep the franchise useful after surgery even as biosimilars line up. That is a business reason to stay on the map Li described. It is also why a miss in a cold tumor would be more than a scientific disappointment.

BioNTech’s Monotherapy Bet Collapsed First

The other mRNA house placed a different wager, and it landed nine days after Merck and Moderna’s melanoma topline. On August 28, BioNTech said it would stop BNT122-01, a Phase 2 study of autogene cevumeran as adjuvant monotherapy in people with ctDNA-positive, resected high-risk stage II or stage III colorectal cancer. An independent data board had crossed a futility line in October 2025 and, on a later look, found a numerical imbalance in overall survival between arms. Continuing, the board said, was unlikely to change the outcome. No new safety signals were flagged. A Genentech spokesperson, the Roche unit that co-develops the shot, confirmed there were more deaths in the vaccine arm in that specific group.

Colon cancer after surgery, treated with a custom mRNA shot and no checkpoint drug, is a colder immune problem than resected melanoma treated with Keytruda. The split does not prove intismeran will work in lung or bladder. It does show why Li is refusing to treat every solid tumor as the next melanoma. BioNTech and Genentech had already dropped a separate Phase 2 study of the same vaccine plus nivolumab in adjuvant muscle-invasive urothelial cancer earlier in 2026, citing a shifting treatment field. Their pancreatic combination study, IMcode003, is still open.

Merck and Moderna are running small adjuvant tests in pancreatic and gastric cancer, which is the closest they will come to Li’s “almost worked” pile until someone else posts a win. Those are probes, not the Phase 3 spine.

How a Custom Shot Gets Made in Six to Eight Weeks

Each dose is a different drug. Surgeons take tumor tissue, a blood draw supplies a healthy-cell control and HLA type, and an algorithm ranks mutations that look most likely to provoke T cells. Up to 34 of those neoantigens are written into one mRNA strand, packed in lipid nanoparticles of the same family as Spikevax, and shipped frozen. One batch covers nine intramuscular doses, given every three weeks.

A Society for Melanoma Research poster from Moderna and Merck staff puts typical time from an adequate sample to delivery at a clinic at 6 to 8 weeks. About 1 mm³ of tumor is the working minimum, often as 20 unstained slides, plus about 3 mL of blood. Doses can sit refrigerated for up to 24 hours after thaw, or at room temperature for up to 6 hours. Merck chief executive Rob Davis said in a mid-October interview that he expects that six-week clock to shorten.

Drug plants are built to stamp identical pills. This process designs, synthesizes, and quality-tests a new product for every patient, then has to be sure the right vial reaches the right person. That factory problem, not the melanoma p-value, is what people who manufacture medicines keep circling. Moderna has automated scheduling against live plant constraints; the remaining bottleneck is a diagnostic network that can move tissue on a clock.

FROM BIOPSY TO THE FIRST SHOT

1. Sample: Tumor tissue and blood are collected after a patient is signed up, then shipped to a certified sequencing lab.
2. Read: Technicians extract DNA and RNA, call tumor-only mutations, and type the patient’s HLA.
3. Pick: A machine-learning ranker scores candidates by variant frequency, RNA expression, sequence conservation, and HLA fit, then keeps up to 34.
4. Make: Those sequences go onto one mRNA strand, wrapped in lipid nanoparticles, and pass potency, particle-size, and endotoxin checks.
5. Ship: A single patient-specific batch of nine frozen doses leaves for the clinic, typically 6 to 8 weeks after an adequate sample arrives.

On October 7, Tempus, the Chicago precision-medicine firm, said it had expanded a multi-year pact to handle collection and transfer of tumor samples and, subject to approvals, the sequencing that feeds design. Said Francis, Moderna’s chief business officer and head of individualized neoantigen therapies, called a reliable diagnostic network the requirement for taking a custom therapy out of trials. Jannie Oosthuizen, who leads global oncology at MSD, said every treatment is designed around one tumor’s biology, “from collecting a patient’s tumor sample and generating sequencing data to designing and manufacturing an individualized therapy.” Eric Lefkofsky, Tempus’s founder and chief executive, put it as an end-to-end clinical workflow that has to work in the real world. Terms were not disclosed.

Nine Open Trials Stay Inside Known Wins

Merck and Moderna said in June they had nine Phase 2 and Phase 3 studies of intismeran under way with Keytruda, with other drugs, and as monotherapy. The list is still clustered where PD-1 drugs already have a foothold. Metastatic melanoma and metastatic squamous lung are on it, which is not the same as walking into a cold, immunotherapy-deaf tumor. Non-muscle-invasive bladder cancer pairs the vaccine with BCG, the local immune standard there, instead of Keytruda.

THE INTERPATH PROGRAM

A Phase 1 adjuvant study of intismeran plus FOLFIRINOX in resectable pancreatic cancer is due as poster 3151P on October 25, presented by David J. Pinato of London. That is the cold-tumor probe, not the registration path. Cutaneous squamous cell carcinoma appeared on an older operations poster and is not in the nine-trial list Merck issued in June.

The Algorithm Learns From Thousands of Immune Systems

The picker is imperfect, which Hoge said out loud. Intismeran is a wanted poster, not a complete census of a tumor. In KEYNOTE-942, the combination roughly doubled the share of new expanded T-cell clonotypes versus Keytruda alone (median summed frequency 0.030 versus 0.016). Patients who stayed recurrence-free on the combination had a median of 42 unique new clones at long-term follow-up, against 20 in those whose cancer came back. A subset of those new clones traced back to antigens encoded in the shot, which is the mechanism doing what it says on the label.

It’s a real privilege to have all that information across the now several thousand people in our Phase 3 program. I’m looking forward to digging into it.

Stephen Hoge, president, Moderna, industry summit in Boston, October 15

That dataset is the second half of the wager. If the melanoma win is only a melanoma win, intismeran is a specialty add-on for a hot, mutation-rich skin cancer. If the same immune traces teach the software to pick better antigens in lung or kidney, the 6-to-8-week factory starts to look like a platform. Hoge’s “several thousand” people in Phase 3 are, in that sense, a training set Merck and Moderna paid to enroll.

Keblish’s case sits at the other end of that stack. A teacher with resected melanoma, treated while the software was still crude, is still working. The next patients will be enrolled under a picker that has seen far more immune systems than the one that built his vials. Whether that matters in a bladder or a lung is the question the nine-trial map is actually asking.

Madrid Gets the Hazard Ratios on October 24

Georgina Long, the INTerpath-001 principal investigator and medical director of Melanoma Institute Australia, will present the study as LBA1 in Presidential Symposium on October 24 in the Alicante Auditorium, Hall 6, starting at 4:30 p.m. CEST (10:30 a.m. EDT). Inge Marie Svane of Herlev, Denmark, is the invited discussant. Moderna will host an investor webcast at 1:00 p.m. EDT the same day.

FROM A PHASE 2 SIGNAL TO MADRID

1. 2019: Gary Keblish is diagnosed with melanoma and enters an early intismeran trial after surgery; KEYNOTE-942 begins enrolling resected stage III and IV patients.
2. June 1, 2026: Five-year KEYNOTE-942 follow-up at ASCO shows a 49 percent cut in recurrence or death and a 59 percent cut in distant metastasis or death versus Keytruda alone.
3. August 19, 2026: INTerpath-001 meets RFS and DMFS at interim in 1,137 patients; hazard ratios held for a medical meeting.
4. August 28, 2026: BioNTech and Genentech stop the colorectal monotherapy trial after a survival imbalance.
5. October 7, 2026: Tempus expands its role in sample logistics and, pending approvals, sequencing for a possible commercial launch.
6. October 15, 2026: Hoge and Li describe a Keytruda-map strategy and decline to chase cold tumors unless someone else posts a win.
7. October 24, 2026: Long presents LBA1 in Madrid; pancreatic Phase 1 data follow a day later.

Long’s session will finally put a number on the melanoma win. Filing talks come after that. The lung, kidney, and bladder studies will take years, and they will test a picker trained mostly on skin cancer. The factory still has to turn a new drug for every patient in 6 to 8 weeks without mixing up the vials.

Disclaimer: This article is news reporting and analysis of company trial announcements, scientific-meeting plans, and manufacturing disclosures. It is for information only and is not medical advice, a treatment recommendation, or an endorsement of any unapproved vaccine or immunotherapy. Readers making decisions about melanoma, surgery, Keytruda, clinical trials, or other cancer care should consult a licensed oncologist or their own cancer-care team, who can weigh personal pathology, staging, and eligibility. Trial results, hazard ratios, enrollment status, and any future regulatory filings reflect the companies’ and meeting organizers’ statements on the dates named above and can change as data mature or reviews proceed.