
OR WAIT null SECS
Atopic dermatitis is a heterogeneous inflammatory skin disease driven by immune dysregulation across multiple T cell axes — Th2, Th17, Th22, and others — making it notoriously difficult to treat consistently across patient populations. Current approved biologics and small molecules largely intervene downstream, targeting individual cytokines or intracellular signaling cascades. Effective as these agents are, disease recurrence upon treatment discontinuation remains a significant limitation for many patients.
The OX40 pathway represented a potential workaround. By intervening upstream — at the interface between antigen-presenting cells and T cells — OX40-targeted therapies theoretically offered not just symptom control but the prospect of durable remission, including suppression of effector memory T cells long thought responsible for disease persistence after treatment withdrawal. Two agents advanced into late-phase development targeting this pathway: rocatinlimab, an anti-OX40 receptor monoclonal antibody developed by Kyowa Kirin with Amgen, and amlitelimab, an anti-OX40 ligand monoclonal antibody developed by Sanofi.
Rocatinlimab demonstrated clinically meaningful efficacy in the pivotal ROCKET-IGNITE and ROCKET-HORIZON Phase 3 trials, results of which were published in The Lancet in November 2025.¹ IGNITE enrolled 769 adults with moderate-to-severe atopic dermatitis, including patients previously treated with a biologic or systemic JAK inhibitor. At week 24, EASI-75 was achieved by 42% of patients receiving the 300 mg dose and 36% receiving 150 mg, compared with 13% on placebo (P < .001 for both). vIGA-AD 0/1 was achieved by 24% and 19% of the respective dose groups versus 9% on placebo (P < .001 and P = .002, respectively). In the HORIZON monotherapy trial, 33% of patients on rocatinlimab 300 mg achieved EASI-75 versus 14% on placebo (difference 19.1%; 95% CI, 12.4–25.2), with vIGA-AD 0/1 reached by 19% versus 7%.¹
However, rocatinlimab was associated with adverse events including pyrexia, chills, and aphthous ulcers, and its efficacy was not demonstrated to be superior to dupilumab in head-to-head data. In March 2026, Kyowa Kirin and Amgen announced the discontinuation of all rocatinlimab clinical trials following a planned safety review that identified 1 newly confirmed and 1 suspected case of Kaposi sarcoma across the global program, in addition to a previously confirmed case — findings the companies concluded suggested a potential mechanistic link to OX40 pathway modulation.²
The key mechanistic distinction separating rocatinlimab from amlitelimab is not simply target specificity but the downstream consequence of that distinction. Rocatinlimab binds the OX40 receptor on activated T cells and, critically, depletes those cells via antibody-dependent cellular cytotoxicity. Amlitelimab targets the OX40 ligand on dendritic cells and is explicitly non-depleting, blocking signaling through steric hindrance rather than cellular elimination. Sanofi has continued to advance its program despite disclosing 1 case of Kaposi's sarcoma in the ATLANTIS long-term safety study — occurring in a patient with known risk factors — and presented 3-trial Phase 3 data from COAST 1, COAST 2, and SHORE as late-breaking research at the 2026 American Academy of Dermatology Annual Meeting in Denver.³⁻⁵
Across all 3 trials, amlitelimab was evaluated at dosing intervals of every 4 weeks (Q4W) and every 12 weeks (Q12W). In COAST 1 (n = 601), vIGA-AD 0/1 was achieved by 21.1% of patients in the Q4W arm and 22.5% in the Q12W arm, versus 9.2% on placebo (P ≤ .01 for both). EASI-75 rates were 35.9% (Q4W) and 39.1% (Q12W) versus 19.1% on placebo (P < .001 for both). In COAST 2, vIGA-AD 0/1 rates were 25.3% and 25.7% for Q4W and Q12W, respectively, compared with 14.8% on placebo (P ≤ .025 for both); EASI-75 and peak pruritus numerical rating scale response reached nominal significance, though the more stringent vIGA-AD 0/1 with barely perceptible erythema endpoint did not. The combination therapy SHORE trial produced the highest response rates across the program, with vIGA-AD 0/1 achieved by 28.7% (Q4W) and 32.3% (Q12W) versus 16.8% on placebo, and EASI-75 rates approaching 47–48% in treated arms versus 32.3% for placebo.⁵ A key feature across all 3 datasets was progressive efficacy with no evidence of plateau at week 24, suggesting continued improvement beyond the primary endpoint window. In the ATLANTIS long-term safety analysis, vIGA-AD 0/1 rates increased from 35.4% at week 24 to 50.3% at week 52, with EASI-75 rising from 62.9% to 76.5% over the same period.³ COAST 2 did not meet its European co-primary estimand endpoints, a distinction with potential regulatory implications outside the United States.
Whether amlitelimab's mechanistic and safety profile is sufficiently differentiated from rocatinlimab's to overcome the shadow cast by the class-level Kaposi's sarcoma signal is an active point of debate among clinicians. HCPLive spoke with 3 leading dermatologists to explore both sides of this question, what the Phase 3 data show, and what comes next for OX40 targeting in atopic dermatitis.
Featured experts:
Armstrong’s disclosures include AbbVie, Almirall, Arcutis, ASLAN, Beiersdorf, Boehringer Ingelheim, Bristol Myers Squibb, Dermavant, Dermira, EPI, Incyte, Janssen, LEO Pharma, Eli Lilly, Nimbus, Novartis, Ortho Dermatologics, Parexel, Pfizer, Regeneron, Sanofi, Sun, and UCB. Soung’s disclosures include AbbVie, Amgen, Arcutis, Aslan, Bristol Myers Squibb, Coval Biopharma, Dermavant, Eli Lilly, Johnson & Johnson, KoBio Labs, National Psoriasis Foundation, Novartis, Ortho Dermatologic, Oruka, Pfizer, Regeneron/Sanofi, and UCB. Song’s disclosures include AbbVie, Alphyn Biologics, Amgen, Apogee, Arcutis, Bristol Myers Squibb, Boehringer Ingelheim, Dermavant, DermBiont, Galderma, Incyte, Janssen, LEO Pharma, MoonLake, Novartis, Ortho, Pfizer, Regeneron, Sanofi, Sun Pharmaceutical Industries, Timber Pharmaceuticals, and UCB.
Related Content: