The ravulizumab ARTEMIS trial has been terminated at 555 patients. Read alone it is one failed study. Read against the registry it is the sixth attempt to take a rare-disease drug into a large indication, and the sixth to stop.
Alexion built a Phase 3 across 130 sites to test whether a single dose of ravulizumab given before cardiac surgery could prevent kidney damage afterwards. The study registered in February 2023, dosed its first patient in April, and had an estimated completion date of February 2027.
It is now marked terminated — 555 patients enrolled against a published target of 736.
No reason is recorded. Alexion has not announced one.
01Six attempts, six terminations
Ravulizumab is Ultomiris, and it works. It is approved in paroxysmal nocturnal haemoglobinuria, atypical haemolytic uraemic syndrome, generalised myasthenia gravis and neuromyelitis optica spectrum disorder. Those four labels are intact and none of what follows touches them.
What follows is everything Alexion tried outside them.
ARTEMIS alone accounts for 42.2% of that. The largest bet came last.
02What the ravulizumab ARTEMIS numbers show
Enrolled 555 (registry at termination)
Shortfall 736 − 555 = 181
The design was published in full. A global executive steering committee of independent experts was set up in 2023 and met every six months. The primary endpoint was MAKE90: a sustained 25% fall in eGFR, initiation of kidney replacement therapy, or death from any cause within 90 days of bypass.
That is not a soft endpoint, and it is not a small study. Three-quarters of it was completed.
The sequence is worth holding still for a moment. Interim results were presented at a surgical meeting. The registry then recorded the study as terminated.
03The reasons that were given
Four of the six terminations carry a stated cause in the registry, and they are not the same kind of reason.
Two were stopped because the drug was not working. Another two were stopped for reasons that say nothing about whether it worked. The last two say nothing at all.
The TMA study is worth a second look on its own. It activated 97 sites and enrolled 16 patients. That is one patient for every six sites, which points at a design problem rather than a drug problem. The trigger-associated population was far harder to identify and randomise than the protocol assumed.
04Where the mechanism works, and where it did not travel
Look at the two lists side by side and the dividing line is not therapeutic area. It is what complement is doing in the disease.
In PNH, complement destroys red cells. It drives the microangiopathy in aHUS. In generalised myasthenia gravis and NMOSD it does the damage at the neuromuscular junction and in the optic nerve. Block C5 and the disease process stops, because complement is the disease process.
Now the other column. In cardiac surgery, complement activation happens on bypass. So does ischaemia-reperfusion injury, haemodilution, hypotension, embolism, nephrotoxic drug exposure and inflammation from the circuit itself. Complement is one contributor. In ALS, in COVID-19 ARDS, in dermatomyositis and in lupus nephritis, the same is true in different proportions.
Complement is a participant → six terminations, 1,316 patients
This is the same shape we set out in the OX40 class collapse, where a cleaner mechanism cleared the safety bar and still could not beat an entrenched standard of care. Mechanistic elegance is not demonstrated clinical benefit. Here it is the same lesson from the supply side: a target that works beautifully in a disease it defines does not automatically work in a disease it merely joins.
05Why the largest study came last
CSA-AKI was the biggest prize on the list by a wide margin.
Ultomiris is a rare-disease franchise. Its four labels cover populations measured in tens of thousands. Cardiac surgery on cardiopulmonary bypass is performed on hundreds of thousands of patients a year, and a substantial share of them have pre-existing chronic kidney disease. There is no approved targeted pharmacological prevention for cardiac surgery-associated acute kidney injury, which is why the protocol paper opens by saying so.
A single pre-operative dose. One infusion, one procedure, an enormous eligible population — and no incumbent to displace. If terminal complement inhibition was ever going to move from rare disease into volume medicine, this was the route.
06What this changes for the franchise
Nothing about the approved base, and that matters. Ultomiris keeps its four labels and its revenue.
What changes is the growth story. A drug confined to rare disease grows by finding more patients inside small populations, by dosing convenience, and by holding share. It does not grow by entering a market of hundreds of thousands. That door has now been tried six times.
The rare-disease base is also no longer uncontested. Complement is now a crowded target: C5 competitors, proximal C3 and factor B inhibitors, and oral options changing what a PNH patient will accept. A franchise that cannot expand has to defend, and defending is harder when the competitive set is growing and the expansion budget has just produced its sixth termination.
07Witfire Risk Score
Evidence Integrity at 6.5 against Capital Position at 1.5. The company can afford this easily and has explained it least where it cost most.
08What to watch, with numbers attached
- Any stated reason appearing on the ARTEMIS record. Futility and portfolio pruning are different findings, and only the sponsor can settle which one this was.555
- Publication of the ARTEMIS interim analysis in full. The abstract exists. The dataset would show whether MAKE90 moved at all.MAKE90
- Any new ravulizumab indication entering Phase 2 or 3. Silence across four quarters would confirm the expansion strategy has been retired rather than paused.0
- Results posting on the six terminated records. Registries require it. None of these has posted results, and 1,316 patients participated.1,316
- Whether any other C5 inhibitor enters an acute surgical indication. If nobody follows, the field has read the same six results the registry shows.6 for 6
09Verdict on the ravulizumab ARTEMIS termination
Three figures carry this.
The useful reading is not that ravulizumab failed. It did not. It works in four diseases and continues to.
The reading is that a mechanism can be genuinely correct and still not travel. Complement inhibition earned its four approvals in diseases where complement is the whole story. Six times it was asked to work in diseases where complement is one strand of a knot. Six times the answer came back the same, whether by futility rule, by monitoring committee, by recruitment collapse or by a sponsor quietly closing the file.
ARTEMIS was the most expensive way to ask that question. It is also the one where nobody has said what the answer was.
10FAQ
What was the ravulizumab ARTEMIS trial?
A Phase 3 study of one pre-operative dose of ravulizumab in adults with chronic kidney disease undergoing non-emergent cardiac surgery on cardiopulmonary bypass, with major adverse kidney events at 90 days as the primary endpoint. It ran at 130 sites and enrolled 555 of a planned 736 participants.
Why was it terminated?
The ravulizumab ARTEMIS record carries no reason, and no company statement has been made. Interim results were presented in a 2026 surgical journal supplement before the termination was posted.
How many ravulizumab trials have been terminated?
Six Alexion-sponsored trials outside the approved indications: ALS, COVID-19 ARDS, thrombotic microangiopathy with a trigger, dermatomyositis, lupus nephritis and IgA nephropathy, and CSA-AKI. Together, 1,316 patients across 536 site activations.
Is Ultomiris still approved?
Yes, in PNH, aHUS, generalised myasthenia gravis and NMOSD. None of the terminations affects those labels. What stopped is expansion beyond them.
What does this say about complement inhibitors generally?
Terminal complement inhibition has a consistent record where complement drives the disease and no completed success where it is one contributor among several. That distinction, rather than therapeutic area, separates the four approvals from the six terminations.
Computed in this brief: the 1,316-patient and 536-site totals; the 42.2% ARTEMIS share; the 75.4% enrolment against target and the 181-patient shortfall; the 30-month run to primary completion and the approximately 16-month early stop against the previously estimated February 2027 completion; the ratio of one enrolled patient per six activated sites in the TMA study.
Disclosure: Editorial analysis, not investment or medical advice. The Witfire Risk Score weights Regulatory Exposure 30%, Competitive Displacement 25%, Capital Position 20%, Evidence Integrity 15% and Execution & Credibility 10%.
Put this framework on your own file.
The same method used in these briefs, applied privately to a decision you are actually facing — a target, a partner, a facility, a filing.
- CMC & manufacturing risk assessment
- Regulatory / CRL readiness review
- Due diligence on a target or partner
- Investor brief or market analysis
- Commissioned editorial
- Something else
The signal before the market catches up.
A low-noise briefing on the pharma events that change strategy, risk and value. No more than one a week.
