Evaxion reports no tumor progression at three-year follow-up of EVX-01 in phase 2 study
Evaxion A/S (NASDAQ: EVAX) announced on Monday, 19 October 2026 that all patients at three-year follow-up in the phase 2 study of the personalized cancer vaccine EVX-01 maintained the clinical responses measured at two years, with no observed tumor progression. In the third treatment year, the vaccine was given as monotherapy, without pembrolizumab — but the figures are so far available only in the company's own press release and are not peer-reviewed.
The source is the company's own press release via GlobeNewswire (Taiwan News/PR). What makes the update remarkable is that in the third year the vaccine was given as monotherapy — that is, without the anti-PD-1 drug pembrolizumab, which patients received alongside EVX-01 in the main portion of the study. But the data have not yet been presented to the scientific community: that will happen first at the ESMO congress in Madrid on 24 October, and no independent verification exists so far.
What the company actually claims
In the press release from Copenhagen, Evaxion writes that "all patients at three-year follow-up in the phase 2 study maintained the clinical responses observed at two-year follow-up, with no observed tumor progression." The company states that in the third treatment year, patients received EVX-01 as the sole treatment, without combination therapy.
It is this third year that is analytically interesting. The phase 2 study tested EVX-01 in combination with MSD's KEYTRUDA (pembrolizumab) in patients with advanced melanoma, where each patient received a unique vaccine designed and produced based on their own biological profile (FT.com/announcement). If the responses hold up in a year in which the checkpoint inhibitor has been withdrawn, it could — if confirmed — suggest that the vaccine alone contributes to durable immune control. That is the company's interpretation, not a conclusion verified by independent data.
The numbers so far — and where they come from
The three-year data are, so far, a summary without specific figures. What the company builds them on is two-year data presented at ESMO in 2025, which the press release cites from the company's own presentations:
- 75 percent objective response rate, with complete response in 25 percent of patients.
- According to the press release, 54 percent of patients achieved a deeper response after starting EVX-01 treatment — an improvement from stable disease or partial response to partial or complete response.
- Tumor reduction in target lesions in 15 of 16 patients enrolled in the study.
- Six of twelve responders continued into the study's one-year extension, where they received EVX-01 as monotherapy.
These figures also come from Evaxion itself, and are neither peer-reviewed nor confirmed through independent reporting in the available source material.
The missing number: the denominator behind "all patients"
The most obvious objection to "all patients" is the arithmetic hiding behind it. The press release states that six of twelve responders continued into the monotherapy extension. How many patients were actually evaluable at three-year follow-up, the company does not state. "All patients at three-year follow-up" could therefore mean a relatively small group — the patients who continued into the extension — rather than the full study population of 16 patients. Whether the group is also selected, for example because those who fared worse dropped out along the way, is impossible to determine with what is available. Response depth and the number evaluable at three years are not specified, and are likely waiting for the actual ESMO poster.
That does not mean the data are useless — duration of response among responders is precisely what the field cares about — but the reader should know the denominator behind the fraction.
How the vaccine is made
EVX-01 is a product of Evaxion's AI-Immunology platform. The company, which describes itself as a clinical-stage TechBio company, designs each vaccine based on the individual patient's tumor biology: neoantigens — mutations that distinguish the tumor from healthy tissue — are selected using algorithms, and the patient is vaccinated against the characteristic proteins of their own tumor.
The concept is the same underlying ideal as in the field of personalized cancer vaccines generally. There are larger programs in the same category — including Merck and Moderna's intismeran autogene. It is a separate project, from separate companies and in separate studies, and should not be confused with EVX-01.
The company's framing
Birgitte Rønø, chief scientific officer and COO at Evaxion, frames the data this way in the press release:
"We are very excited about the new data, which reflect the long duration of clinical responses in this difficult-to-treat patient population. This is a new strong confirmation of the clinical potential of personalized cancer vaccines in general, and EVX-01 in particular, which we look forward to sharing and discussing at the important ESMO congress. The new data further add to EVX-01's strong data package as a good basis for potential partnership discussions" (Taiwan News/PR).
Every part of this quote is a statement from a company representative. The final sentence in particular — about partnership discussions — is a business framing, not a clinical milestone. EVX-01 is an investigational drug; there is no regulatory approval or phase 3 data in the available material.
What to watch in the coming weeks
The calendar is dense, and that is where the data will actually be released in detail:
- ESMO 2026, Madrid (23–27 October): Three-year clinical and immunological data will be presented in a poster on 24 October. This is the first time the details — number evaluable, response depth, progression status per patient — should be available.
- Webinar on 29 October at 15:30 CET: Evaxion is hosting a webinar with Professor Muhammad Adnan Khattak, who is a study investigator for the study.
- SITC 2026, Phoenix (4–8 November): New immunological data on EVX-01 plus pembrolizumab will be presented both orally and as a poster. According to the company, the data reflect systemic immunomodulatory effects beyond vaccine-specific T-cell responses, based on extensive immune profiling (FT.com/announcement).
For anyone wanting to assess the company's claims, the relevant questions are: How many patients remained and were evaluable at three years? Was there a bias in who continued into the monotherapy extension? And do the immunological response data match the clinical durability?
Why the story is still worth following
Progress in personalized melanoma vaccines is measured precisely in durability: checkpoint inhibitors like pembrolizumab give many patients responses, but showing that the response lasts when the combination partner is removed — as reported here in the third treatment year — would, if confirmed, be a meaningful signal that the vaccine contributes actively rather than being merely an add-on to an established drug. With a study base of 16 patients and company-owned sources, this remains nonetheless an announcement of promising data, not a proven treatment outcome. The ESMO poster on 24 October is the first opportunity to see the numbers behind the headline.