Novo Nordisk has stopped two Phase 3 IL-6 heart failure trials on the advice of an independent monitoring committee. The company notified trial investigators on Friday 4 September 2026. Trade press reported the decision on Monday 7 September. Both studies, HERMES and ATHENA, tested ziltivekimab, a monoclonal antibody directed against interleukin-6. The decision leaves one late-stage cardiovascular readout standing in a programme that once carried four.
What Novo Nordisk stopped, and when
The stop followed a data monitoring committee review of both IL-6 heart failure trials rather than a safety signal. Clinical Trials Arena reported the committee’s reasoning on 7 September. The committee reviewed the totality of the data. It concluded that neither trial was likely to reach a different outcome from ZEUS. A Novo Nordisk spokesperson told Reuters the committee had found a “low likelihood” of a different result.
Novo Nordisk told the outlet how participants would be handled. “All patients will be followed through to their three-months follow-up visit as already planned,” the company said. That wording matters to sites still holding open visits. It commits the sponsor to the existing follow-up schedule rather than an immediate close-out.
LSDN found no Novo Nordisk company announcement or stock exchange release covering the two discontinuations as at 7 September 2026. The news reached the market through reporting by STAT News, Endpoints News, Clinical Trials Arena and Reuters. Endpoints reported it first, according to Clinical Trials Arena. STAT reported that investigators were told on the Friday. That is a lower disclosure threshold than the company applied to ZEUS in July.
Inside the halted IL-6 heart failure trials
The design of both IL-6 heart failure trials was published four months ago. Mendieta and colleagues set it out in the European Journal of Heart Failure. The paper appeared online on 10 May 2026. Its authors include Paul M Ridker, who led the CANTOS inflammation trial. Mark C Petrie of the University of Glasgow is the senior author.
HERMES
HERMES, registered as NCT05636176, was the outcomes study of the pair. It enrolled patients with a left ventricular ejection fraction above 40 per cent. Entry required high-sensitivity C-reactive protein of at least 2 mg/L. Participants were in New York Heart Association classes II to IV. The primary endpoint combined cardiovascular death, heart failure hospitalisation and urgent heart failure visits.
The design paper records 4,769 participants enrolled against a 4,900 target as at 26 March 2026. That is 97 per cent of the planned population, across 848 sites in 46 countries. The trial was event driven and sized for roughly 845 adjudicated primary events. The UK Health Research Authority record for HERMES carries a larger planned figure. It lists 5,600 participants across about 720 sites. That was the original target. The design paper records a reduction to 4,900 in September 2025. The stated reasons were a minimum exposure requirement and a higher than anticipated event rate.
ATHENA
ATHENA, registered as NCT06200207, tested symptoms rather than events. It enrolled 673 participants across 173 sites in 15 countries. Its primary endpoint was the change in Kansas City Cardiomyopathy Questionnaire clinical summary score at 12 months. Entry required a score below 80 points and a six-minute walk distance of at least 100 metres.
The timing of the ATHENA stop is worth noting. The design paper records that the trial completed enrolment on 25 September 2025. ATHENA ran to a fixed 12-month duration rather than to an event count. On those published figures, its last scheduled primary endpoint visits fell close to the futility decision. LSDN has seen no sponsor comment on that overlap.
One further point of record is worth setting down. Novo Nordisk’s ZEUS release of 31 July 2026 named HERMES and ARTEMIS as the trials planned to continue. It did not mention ATHENA. LSDN draws no inference from that omission and found no company explanation for it.
Why the ZEUS result made this foreseeable
ZEUS reported on 31 July 2026 and framed everything that followed. Novo Nordisk said the trial randomised more than 6,300 patients. All had atherosclerotic cardiovascular disease, chronic kidney disease and inflammation. Ziltivekimab 15 mg once monthly was compared with placebo. The hazard ratio for major adverse cardiovascular events was 0.99. The 95 per cent confidence interval ran from 0.88 to 1.11.
The safety picture was not neutral. Novo Nordisk reported comparable overall adverse event rates between the arms. It also reported a higher proportion of serious infections on ziltivekimab, consistent with IL-6 inhibition. All-cause mortality did not differ between the groups. LSDN examined the full ZEUS readout and its implications at the time.
Martin Holst Lange is executive vice president, chief scientific officer and head of research and development at Novo Nordisk. Ziltivekimab “produced the expected biological effect”, he said in the company’s announcement. That “did not result in MACE benefits in this population”, he added.
Analysts read the result as a problem for the target rather than for the molecule. Andy Hsieh of William Blair set out that view to PharmaVoice. He said the bearish read-through was “more limited to hsCRP’s surrogacy for cardio protection” under IL-6 inhibition. That remains one analyst’s reading rather than a settled position.
What the halt leaves standing
ARTEMIS
ARTEMIS, registered as NCT06118281, continues. It tests ziltivekimab in patients admitted with acute myocardial infarction. The UK Health Research Authority record lists approximately 10,000 participants across 24 countries. The study is event driven and each participant is followed for about 25 months. Novo Nordisk has guided to a readout in the first half of 2027.
ARTEMIS asks a different question from the halted IL-6 heart failure trials. Its population is post-infarction rather than chronic heart failure. Inflammatory burden after an acute event is not the same as in stable disease. That distinction is real. The burden of proof on the one remaining readout is nonetheless heavier than it was in July.
HERMES stopped at 97 per cent of its enrolment target. A substantial dataset therefore exists in both halted studies. LSDN found no sponsor statement on whether either will be presented or published. Investigators and sites will want that answer early.
Exposure to the target beyond Novo Nordisk
Novo Nordisk is not the only sponsor with capital committed to IL-6 in cardiovascular disease. CSL is running the Phase 3 POSIBIL6ESKD trial of clazakizumab, another anti-IL-6 antibody. That study is in patients with end-stage kidney disease on dialysis and runs to 2029. Its population is adjacent to, rather than the same as, the renal patients enrolled in ZEUS.
Eli Lilly bought into the same antibody on 18 February 2026. It paid CSL 100 million US dollars upfront for rights in additional indications. Milestone and royalty terms were not disclosed. CSL retained the dialysis cardiovascular indication. Neither company had commented publicly on the HERMES and ATHENA stops as at 7 September 2026.
UK sites and the treatment gap the IL-6 heart failure trials targeted
Both the halted and the continuing studies recruited in the UK. HERMES holds a favourable opinion from South Central Berkshire B Research Ethics Committee, reference 23/SC/0004, dated 31 July 2023. ARTEMIS was reviewed by London Surrey Borders Research Ethics Committee, reference 24/LO/0060, dated 27 March 2024. UK sites therefore face close-out work on HERMES while ARTEMIS carries on.
The clinical gap the drug aimed at is well defined in England. NICE guideline NG106 was last updated on 3 September 2025. It tells clinicians to consider an MRA and an SGLT2 inhibitor for heart failure with preserved ejection fraction. Those two classes are the current disease-modifying options.
Nothing in the IL-6 heart failure trials changes that position. Preserved and mildly reduced ejection fraction remain areas with a short list of proven agents. A stop on those grounds creates no evidence for or against any other approach. It removes a candidate that had reached 97 per cent enrolment in its outcomes study.
Disclosure and market context
The disclosure contrast on 7 September is hard to miss. Novo Nordisk issued a company announcement that day on its STEP Young Phase 3 paediatric data. After 68 weeks, 40.4 per cent of children on semaglutide were no longer classified as having obesity. That figure assumes full adherence to treatment. That release makes no reference to ziltivekimab or to the halted IL-6 heart failure trials.
Market reporting described Novo Nordisk shares as slipping on the day. Percentage figures differed between outlets, so LSDN reports the direction only. The reading offered across that coverage was consistent. Investors weighed the loss of two late-stage cardiovascular assets more heavily than a positive paediatric readout.
The pattern is familiar from earlier this month. Novartis reported on 4 September that its Phase 3 Lp(a)HORIZON trial of pelacarsen missed its primary endpoint. Both programmes tested whether moving a blood marker moves hard clinical outcomes. Both answered no. That is the question the surviving IL-6 readout now has to settle on its own.
