Sanofi Abandons Eczema Approval for Amlitelimab, Dealing Another Blow to the OX40 Class

Sanofi is dropping efforts to win approval for its eczema drug amlitelimab in atopic dermatitis — an unusual retreat given the drug actually showed efficacy in late-stage testing.

The company concluded that amlitelimab, an OX40-ligand blocker, would not offer “meaningful improvements” over the current standard of care. While it demonstrated efficacy across its Phase 3 programme, it missed a co-primary endpoint in the COAST 2 study, and safety flagged too: two patients developed Kaposi sarcoma, a rare cancer.

What OX40 does

OX40 is a receptor appearing on T cells after they are activated, and its ligand appears on the cells presenting antigen to them.

The interaction functions as a secondary confirmation. A T cell receiving it survives longer, proliferates more and develops into a memory cell; without it the response is weaker and shorter.

Blocking that interaction therefore dampens ongoing immune responses without preventing them entirely — attractive in principle for chronic inflammatory disease, where the aim is reducing a persistent response rather than abolishing immunity.

Why Kaposi sarcoma is the specific signal

The cancer that appeared is not a random finding, and its identity is what makes the safety signal serious.

Kaposi sarcoma is caused by a herpesvirus that most infected people carry without consequence. It causes disease almost exclusively when T-cell immunity is impaired — which is why it became the defining cancer of advanced HIV, and why it appears in transplant recipients on immunosuppression.

Its appearance is therefore a direct readout of T-cell surveillance failing. In a drug designed to modulate T-cell responses, two cases is not an unexplained coincidence — it is the predicted consequence of the mechanism working more than intended.

Another hit for the class

The decision compounds the troubles of the OX40 class. Kyowa Kirin previously terminated its own OX40 inhibitor, rocatinlimab, after patients also developed Kaposi sarcoma.

Two independent programmes producing the same rare malignancy is what converts an observation into a class effect. A single occurrence in one trial could be chance; the same specific cancer in two separate drugs targeting the same pathway points to the mechanism.

“Industry momentum for OX40s will likely wind down” following amlitelimab’s discontinuation, alongside rocatinlimab’s, said GlobalData’s Tanuj Sircar.

Why efficacy was not enough

Eczema is a crowded, competitive market anchored by Sanofi and Regeneron’s own blockbuster Dupixent — a high bar a me-too mechanism has to clear.

The competitive position makes the risk-benefit calculation unforgiving. Approving a drug with a rare cancer signal might be defensible if patients had nothing else; when an established therapy with a well-characterised safety record already works well, a new option must be clearly better to justify any additional risk.

Sanofi owning Dupixent sharpens that further. The company would have been introducing a riskier alternative to its own successful product, competing with itself while adding liability — which is a straightforwardly poor proposition.

What happens to the drug

Sanofi has also halted amlitelimab development in asthma, alopecia and hidradenitis suppurativa. One programme survives: a Phase 2 in celiac disease, with results expected in the second half of 2026.

Keeping only that one is consistent with the risk calculus. Celiac disease has no pharmacological treatment at all, so the comparator is nothing rather than an effective biologic — and a safety signal that disqualifies a drug in a well-served indication may be acceptable where patients have no alternative.

What the eczema market now looks like

The competitive bar amlitelimab failed to clear is worth describing, because it explains why a working drug was abandoned.

Dupixent transformed the field by blocking signalling from two cytokines central to the inflammatory pathway driving atopic dermatitis. It works in a substantial majority of patients, is generally well tolerated, and established that targeting that pathway is the productive approach.

Competitors have since arrived targeting adjacent points in the same pathway, along with oral JAK inhibitors offering the convenience of a pill and carrying class safety warnings that limit their use.

What that leaves for a new entrant is a narrow space. Patients inadequately controlled on existing biologics are a real population and a smaller one than the market as a whole, and demonstrating benefit specifically in them requires a trial designed around that group rather than around treatment-naive patients.

Amlitelimab was tested broadly and showed efficacy without clear superiority, which is the outcome that leaves a drug with nowhere commercially useful to go. A mechanism genuinely distinct from the dominant pathway might have found the gap — and OX40 blockade, whatever its safety profile, was not different enough in effect to make the case.

Amlitelimab’s exit is a reminder that in a mature market, efficacy alone is not enough. Business news, not medical advice.