Merck and Moderna’s melanoma vaccine posts a phase 3 win, even after an early leak
Merck and Moderna’s intismeran autogene entered ESMO with a headline problem and left with a clinical one: the vaccine’s phase 3 melanoma data were disclosed two days early after an unexpected leak, forcing the companies to present results ahead of schedule. The market reaction was immediate, but the underlying question remained more durable than the noise around it — whether adding a personalized cancer vaccine to pembrolizumab can shift the natural history of high-risk melanoma.
The answer from the 24-month readout is yes, at least in recurrence control. In patients with certain melanoma, the combination produced a recurrence-free survival rate of 82.2% versus 70% with Keytruda alone. The hazard ratio was 0.58, reflecting a 42% reduction in the risk of recurrence or death. For a disease defined by early relapse and the danger of microscopic residual disease escaping immune pressure, that kind of separation is not a cosmetic difference. It suggests the vaccine is doing what the biology promised: sharpening immune recognition of tumor-specific neoantigens while pembrolizumab removes the brakes on T-cell activity.
Distant metastasis control improved, but not as dramatically as phase 2 had implied
The more complicated part of the dataset sits in distant metastasis-free survival. At 24 months, the vaccine arm reached 88.5% compared with 80.2% in the pembrolizumab-only arm. The hazard ratio was 0.62, translating to a 38% reduction in the risk of distant metastasis or death. That is still clinically meaningful, because distant spread is the event that most often converts a surgically managed melanoma into a systemic, life-limiting cancer. Once melanoma seeds organs beyond the original site, the treatment stakes rise sharply and the margin for recovery narrows.
The softer-than-expected DMFS result mattered because phase 2 had raised the bar. In that earlier setting, the signal had amplified expectations for how deeply the regimen might suppress dissemination beyond the primary tumor bed. Phase 3 did not erase the benefit, but it did make the conversation less about triumphalism and more about proportion. The combination appears to be helping patients avoid metastatic relapse, yet the magnitude of that benefit sits below the higher hopes that had accumulated around the program.
Shweta Jain, vice president of global clinical development and oncology clinical research at Merck, framed the totality of the data as persuasive and clinically consequential. She pointed to the reduction in distant metastasis or death as especially important, arguing that lowering both local recurrence and distant spread matters in a regimen intended to move beyond disease control toward curative potential. The phrasing fits the therapeutic logic. A vaccine that trains the immune system against tumor-specific neoantigens should, in theory, be most valuable when residual disease is still small enough to be contained by host immunity and checkpoint blockade.
The leak changed the venue, not the dataset
The early disclosure forced Merck and Moderna to reshuffle their ESMO plans, but the presentation itself did not appear to change. Jain said there was no rewriting and no substantive alteration to the Saturday presentation. The companies had intended to unveil the data in Madrid under the spotlight of the Presidential Symposium, and the leak stripped away that timing advantage. What remained was the clinical package itself, which Jain described as convincing and “reasonable” on the phase 3 numbers.
The tone from Merck was notably measured rather than defensive. Jain repeatedly returned to the idea that the team could not control the leak, only the science. That posture makes sense in oncology, where long development timelines and compressed public attention often collide. Trial integrity depends on the data, not the sequence of headlines. Once the readout existed, the debate centered on what the numbers showed for patients at meaningful endpoints — not on the choreography of how the results reached the public.
For Merck, the result also validates a partnership that dates to 2016, when the companies began working together on mRNA vaccines. Jain described the collaboration as a meeting of institutional strengths: pembrolizumab’s extensive clinical experience on one side, Moderna’s platform expertise in mRNA and neoantigen design on the other. The strategy is technically coherent. Pembrolizumab can sustain antitumor immune activity, while a personalized vaccine can expand the repertoire of T cells recognizing tumor-specific targets. In melanoma, where immune escape is a constant threat, that dual pressure may be exactly what is needed to deepen and prolong remission.
Merck is treating the result as one piece of a wider oncology plan
The company is now preparing for a hoped-for FDA filing next year, with Jain saying Merck is working as fast as possible to get the package ready for regulators. That timeline places intismeran into a broader corporate strategy rather than a one-off win. Merck has already launched phase 2 studies of the vaccine in renal cell carcinoma, bladder cancer, and non-small cell lung cancer, all settings in which immune modulation and tumor antigenicity could make the platform relevant if the melanoma data hold up.
The immediate clinical question is whether the phase 3 melanoma benefit can be translated cleanly into regulatory approval and then into everyday practice. Melanoma specialists will want to know which patients derive the most durable benefit, how the vaccine performs across disease burden categories, and whether the recurrence advantage persists beyond the current timepoint. Those details were not part of this report. The present data still leave room for scrutiny, especially around the gap between the stronger recurrence-free survival signal and the less forceful distant metastasis-free survival result.
For now, the core finding is straightforward: adding intismeran autogene to pembrolizumab improved melanoma outcomes versus pembrolizumab alone, with a particularly notable reduction in recurrence risk and a meaningful, if smaller, reduction in distant metastatic events. The leak changed the timing of the disclosure, not the clinical signal. What remains unresolved is how far this platform can be pushed beyond melanoma, and whether later readouts in other tumors can reproduce the same degree of immune control without the benefit of a highly immunogenic disease setting.