OX40 atopic dermatitis development is over. Two companies, two mechanisms, 175 days. The press releases said “discontinued”; the registry says what that word cost.
OX40 atopic dermatitis development ended this month. Sanofi said on 24 July it will not file amlitelimab, four months after Kyowa Kirin halted every rocatinlimab trial on safety grounds.
Both decisions were reported. Neither company published the arithmetic, so here it is: 8,185 patients across 1,893 sites sit in studies that have been terminated or announced for wind-down.
Every figure below comes from the trial registry or is computed from registry figures with the working shown.
01OX40 atopic dermatitis: 175 days from exit to abandonment
Amgen went first. On 30 January it terminated the collaboration and returned global rights to Kyowa Kirin. Kyowa Kirin said at the time it planned a regulatory submission in the first half of 2026.
Thirty-two days later that plan was gone. Kyowa Kirin halted every trial. A safety review had found emerging malignancies with possible viral or immune-related links, including Kaposi sarcoma. The company was blunt about it: risks may outweigh benefits in the populations studied.
The efficacy was never the problem. Phase 3 ROCKET data in roughly 1,500 adults had shown durable, clinically meaningful responses.
03 Mar 2026 Kyowa Kirin discloses the malignancy signal
Kyowa Kirin described the safety review as conducted over the preceding several weeks. That places it around or after Amgen’s exit. Fierce Biotech reported that Amgen’s decision may have rested on the data alone. Questions about where the drug would sit in a crowded market were already live. Nobody outside the two companies can say which reading is right, and neither has claimed the other.
02Half of rocatinlimab’s Phase 3 exposure is in terminated studies
ROCKET-ASCEND is the one that matters. Patients rolled into it from every parent study: IGNITE, HORIZON, SHUTTLE, ASTRO, ORBIT and VOYAGER. That made it the long-term safety database for rocatinlimab as a whole.
Its scheduled completion date was 1 July 2026. The registry now records completion and termination on the same date.
The prurigo nodularis study tells you something narrower but sharper. It ran to its 24-week primary endpoint on 17 April. It recorded a completion date of 25 June. It is marked terminated. The data exists. It will not be filed.
03Sanofi announced a wind-down the registry has not recorded
Sanofi said on 24 July that amlitelimab would not be submitted for review in atopic dermatitis. Checked again on 10 August, seventeen days later, four studies covered by that decision still carry active statuses.
04The safer mechanism died too, and not of safety
This is the part with consequences beyond one class.
Rocatinlimab targets the OX40 receptor and depletes pathogenic T cells. Amlitelimab targets the ligand, OX40L, and blocks the pathway without depleting anything. The second design existed precisely to avoid the risks the first one carried. The field spent two years arguing about whether it would.
It did. Amlitelimab was not stopped for safety. Sanofi cited efficacy and safety data insufficient to improve on current standard of care. Regulators were not cited as having raised concerns.
Amlitelimab OX40 ligand, non-depleting → could not beat dupilumab
So the mechanistic question was answered and it did not matter. A cleaner mechanism cleared the safety bar and then failed the commercial one. In a market where IL-4 and IL-13 biologics already work, “different upstream target” is not by itself a reason for a payer to move.
We made the same argument from the other direction in the Tanabe and Zenas case. Obexelimab’s whole pitch was inhibition without depletion. Mechanistic elegance is not demonstrated clinical superiority. Here is the class-level proof.
05What survives, and where
Nine hundred and sixty-four patients against 8,185. The mechanism is not extinct. It now lives in indications where the standard of care is weak enough that a new upstream target still has something to beat.
The celiac readout in the second half of this year is the one to watch. It is the closest thing left to a clean test of whether OX40L blockade does anything useful outside dermatology. It arrives with the class narrative already written.
06Witfire Risk Score
Evidence Integrity at 4.0 against Competitive Displacement at 9.5. Both companies told the truth quickly about a class that was already finished.
07What to watch, with numbers attached
- Registry status of the four amlitelimab AD studies. When these flip to terminated, the wind-down is real rather than announced.4,839
- Amlitelimab celiac readout. The only near-term test of whether OX40L blockade works where standard of care is weak.2H 2026
- Publication of the full rocatinlimab safety dataset. Both companies committed to a joint analysis. Whether the malignancy signal is receptor-specific or axis-wide decides what anyone can build here next.3,090
- Any new OX40-axis IND. Silence for the next four quarters would confirm the axis is closed rather than paused.0
- Kyowa Kirin’s replacement pipeline disclosure. It has the most concentrated exposure and the least room to absorb it.32 days
08Verdict
Three numbers carry this.
The useful lesson is not that OX40 failed. Targets fail constantly, and a malignancy signal is a good reason to stop.
It is that the two molecules failed for unrelated reasons and arrived at the same place. One could not clear safety. The other cleared it and could not clear dupilumab. A class can be killed twice. The second death says more about the market than the mechanism, because even a clean version of this idea had nowhere profitable to land.
Any sponsor building on a novel upstream target in a crowded indication now has to answer that before the first patient is dosed.
09FAQ
Why was rocatinlimab discontinued?
Kyowa Kirin halted all trials on 3 March 2026. A safety review had found emerging malignancies with possible viral or immune-related links, including Kaposi sarcoma. Efficacy in the ROCKET programme had been positive.
Why did Sanofi discontinue amlitelimab?
Sanofi said on 24 July 2026 that it would not file in atopic dermatitis. It cited efficacy and safety data insufficient to improve on current standard of care. It was a positioning decision, not a safety signal, and regulators were not cited as having raised concerns.
How many patients were in the terminated OX40 trials?
8,185 across 1,893 sites, computed from registry records. That is 3,090 in two terminated rocatinlimab Phase 3 studies, 256 in two terminated amlitelimab Phase 2 studies, and 4,839 in the four amlitelimab studies covered by the July announcement.
Is any OX40 drug still in development?
Not in atopic dermatitis. Amlitelimab continues in celiac disease with a readout due in the second half of 2026. An asthma extension and an investigator-led systemic sclerosis study also remain open.
What replaces OX40 drugs for atopic dermatitis?
IL-4 and IL-13 biologics and oral JAK inhibitors remain the primary systemic options. The end of OX40 atopic dermatitis development removes the most advanced novel mechanism from the late-stage pipeline.
Computed in this brief: the 175-day and 32-day intervals; terminated and completed patient and site totals for each molecule; the 48.1% share of rocatinlimab Phase 3 exposure in terminated studies; the 4,839-patient and 1,038-site wind-down total; the 8,185-patient and 1,893-site class total; and the 964-patient surviving total outside atopic dermatitis.
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.
