Amlitelimab

Amlitelimab is Sanofi's anti-OX40L antibody for moderate-to-severe atopic dermatitis (AD), with three Phase 3 readouts now in hand: COAST 1 (Sept 2025, monotherapy) and SHORE plus COAST 2 (Jan 2026, with topical corticosteroids) all met their US co-primary endpoints, but the absolute response rates fell below dupilumab benchmarks and below analyst expectations, and COAST 2 missed its EU co-primary [8][9]. Sanofi has guided global regulatory submissions for H2 2026, with two further Phase 3 studies (AQUA and the durability-focused ESTUARY) also expected to read out in H2 2026 [9]. The investment thesis now pivots almost entirely from 'can it beat dupilumab on response' (it cannot, at the approved comparator dose) to 'can ESTUARY prove durable off-treatment response,' the only credible differentiator versus Dupixent and the second wave of IL-13 blockers.

Amlitelimab (SAR445229) is a fully human anti-OX40L monoclonal antibody developed by Sanofi (acquired via Kymab in 2021), not yet approved anywhere. The Phase 3 program in AD is branded OCEANA and comprises five studies: COAST 1 (NCT06130566, n=601), COAST 2, SHORE, AQUA, and the long-term ESTUARY withdrawal study (NCT06407934, n=1541). Dosing in all Phase 3 trials is 250 mg subcutaneous every 4 weeks (Q4W) or every 12 weeks (Q12W), after a 500 mg loading dose; pediatric patients <40 kg receive 125 mg [8]. COAST 1 (monotherapy, Sept 2025): at Week 24, vIGA-AD 0/1 (validated Investigator Global Assessment of clear or near-clear skin, 5-point scale) was 21.1% (Q4W) and 22.5% (Q12W) versus 9.2% placebo; EASI-75 (75% reduction in Eczema Area and Severity Index) was 35.9% (Q4W) and 39.1% (Q12W) versus 19.1% placebo [8]. SHORE (with background topical corticosteroids, Jan 2026): vIGA-AD 0/1 was 28.7% (Q4W) and 32.3% (Q12W) versus 16.8% placebo; EASI-75 was 48.1% and 46.8% versus 32.3% placebo [9]. COAST 2: met US co-primary endpoints with similar magnitude but did not achieve statistical significance on EU co-primary endpoints [9]. NCT06241118 (n=636) enrolls biologic- or oral JAK-experienced patients (primary endpoint vIGA-AD 0/1 with ≥2-point drop from baseline at Week 36 for EU/Japan filings) [2]. NCT05769777 (n=999) is the long-term open-label safety extension [3]. AQUA and ESTUARY (the durability readout) are due H2 2026, and Sanofi has explicitly guided BLA + MAA submissions for H2 2026 [9]. Asthma development continues in parallel (NCT06033833, n=335 safety extension) [4]. A Phase 1 DDI (drug-drug interaction) study (NCT06686628) completed in 23 healthy volunteers assessed CYP450 substrate effects. No publicly disclosed FDA breakthrough therapy or fast-track designation.

OX40L sits on the surface of dendritic cells and other antigen-presenting cells. When OX40L docks with the OX40 receptor on a T-cell, it sends a costimulatory 'keep going' signal that expands that T-cell population and pushes it toward the Th2 program, the type of immune response that drives allergy, eczema, and asthma. Blocking OX40L cuts the costimulation, slowing the entire downstream cascade rather than picking off one cytokine at a time. Dupilumab, the standard of care, blocks the IL-4 receptor alpha subunit and shuts down IL-4 and IL-13 signaling, but it acts after T-cells are already activated. Amlitelimab works one step earlier, at the costimulation handshake. Genetic support for the target is reasonable but not airtight: Open Targets links TNFSF4 (the gene for OX40L) to lupus (~0.59), asthma (~0.58), and atopic eczema (~0.55), and GWAS identified the TNFSF4 locus as a lupus risk gene more than a decade ago. The unresolved biological question, and the one ESTUARY is built to answer, is whether broad immune dampening at this upstream node produces durable remission after drug withdrawal, the property hinted at in the STREAM-AD Phase 2b maintenance data [5]. The Phase 3 monotherapy and combination response rates show the on-treatment efficacy ceiling is below cytokine blockade; the mechanism's commercial value now depends entirely on off-treatment persistence.

The OCEANA Phase 3 program is unusually large for atopic dermatitis and tests both Q4W and Q12W maintenance dosing after a 500 mg loading dose, structurally enabling a 'least frequent injection in class' label claim if approved. The key endpoints rely on standard regulatory measures: vIGA-AD 0/1 (validated Investigator Global Assessment for AD, a 5-point clinician-rated scale where 0 = clear and 4 = severe; 0/1 means essentially clear or near-clear skin, typically combined with a ≥2-point drop from baseline) is the FDA-favored primary; EASI-75 (75% reduction in the Eczema Area and Severity Index, a validated composite of rash extent and severity) is the EU/Japan co-primary and the analyst-watched efficacy headline. ESTUARY (NCT06407934, n=1541) compares two amlitelimab maintenance regimens to withdrawal after induction; primary endpoint is the proportion of induction responders who maintain vIGA-AD without relapse [1]. This is the trial Sanofi's entire differentiation story rests on. NCT06241118 (n=636, still recruiting in mid-2026) targets biologic- and oral JAK-experienced patients with primary endpoint vIGA-AD 0/1 plus ≥2-point drop at Week 36 (EU/Japan submission anchor) [2]. NCT05769777 (n=999) is the long-term open-label safety extension and will carry most of the chronic-use safety database [3]. Design concerns: the Phase 3 Q4W/Q12W dosing was chosen for convenience but produced lower absolute response than Phase 2b's 250 mg Q4W with loading dose (EASI-75 ~40% at Week 16, ~54% at Week 24 [5]) would have predicted, suggesting some sensitivity to induction strategy. COAST 2's EU co-primary miss [9] is a structural concern for the MAA pathway and means EMA review will scrutinize pooled and subgroup data heavily.

Our model puts this drug's approval odds at 42%. That starts from a 61% historical baseline for Phase 3 drugs in this area, then shifts based on ten facts about the trial and sponsor. The estimate is helped by a non-randomized design and light or open-label blinding, and held back by smaller-than-typical enrollment and weak or limited earlier-phase results. The remaining facts land near average, so they leave the number close to where the base rate set it.

Efficacy risk (now partly resolved, partly worse than expected): COAST 1 monotherapy EASI-75 was 35.9% (Q4W) / 39.1% (Q12W) at Week 24 [8]. The closest dupilumab monotherapy benchmark is SOLO-1 and SOLO-2 at Week 16: EASI-75 51% (SOLO-1) and 44% (SOLO-2) on 300 mg Q2W [11]. The dupilumab + topical corticosteroid benchmark (CHRONOS) was EASI-75 ~69% at Week 16 [12], versus amlitelimab + TCS in SHORE at 48.1% (Q4W) at Week 24 [9]. So even on the more generous Week 24 comparison, amlitelimab trails dupilumab by roughly 10 percentage points monotherapy and roughly 20 points in combination. That gap is why analysts called the Phase 3 data disappointing despite endpoints being met. Durability is now the only viable differentiation thesis; ESTUARY (H2 2026 readout) is the binary [1]. Regulatory risk: COAST 2's EU co-primary miss [9] means the MAA pathway is harder than the BLA; expect EMA to require strong pooled or subgroup analyses. Safety risk: OX40 costimulation matters for immune memory and pathogen response. Phase 2b and Phase 3 safety to date has been clean, but long-term immune dampening still raises questions about infection (notably herpes zoster, a class issue across IL-4/13 blockade as well) and malignancy surveillance; NCT05769777 is the database that will surface those signals [3]. Competition risk: rocatinlimab (Amgen/Kyowa Kirin, anti-OX40) is now further along clinically than the writeup previously implied. HORIZON (Sept 2024) met co-primary endpoints with EASI-75 32.8% at Week 24, and IGNITE (March 2025) hit EASI-75 42.3% at the high dose [10]. Rocatinlimab and amlitelimab will likely launch into the same AD market in 2027-2028, fighting for the same dupilumab-experienced patient pool. Other competitors include lebrikizumab (Eli Lilly, anti-IL-13, approved), tralokinumab (LEO, anti-IL-13, approved), nemolizumab (Galderma, anti-IL-31, approved), and the oral JAKs. Commercial risk: Dupixent posted €13.07B (≈$14B at 2024 average rates) in 2024 [13], dominating the AD market. Payers will not pay a premium for a non-superior biologic. Sanofi needs ESTUARY to deliver real durability for premium positioning; without it, amlitelimab becomes a second-tier biologic in a crowded class.

Watch ESTUARY. With three US Phase 3 readouts in hand and submissions planned H2 2026, approval risk has compressed materially, but the commercial case is now entirely a durability story. Sanofi has publicly guided peak sales potential of over $5.4 billion for amlitelimab as part of its second-generation immunology portfolio [14]; sell-side analyst consensus has been more cautious post-COAST 1, with several houses cutting peak forecasts below management's number after the Phase 3 efficacy fell short of expectations [15]. The two near-term catalysts that matter: (1) ESTUARY topline in H2 2026, the only data that can support drug-holiday positioning and a real premium versus dupilumab biosimilars, and (2) AQUA topline (H2 2026), which adds to the registration package. The asthma indication is the underweighted second leg of the story: NCT06033833 is the long-term safety extension [4], and Sanofi has separate Phase 3 asthma readouts due later this decade. A successful asthma label would roughly double addressable patient population and is the gap between $5B and $8B peak scenarios, but is dependent on durability and Th2-axis breadth holding up across indications. EMA pathway: COAST 2's EU co-primary miss [9] complicates the MAA but does not block it; Sanofi has not disclosed whether MAA filing precedes, accompanies, or follows the H2 2026 BLA. Strategic logic for Sanofi: Dupixent's biosimilar exposure begins around 2031; Dupixent contributed €13.07B of €41.1B in 2024 net sales (≈32% of total revenue) [13], so franchise renewal is existential, not optional. Check back on ESTUARY topline.