Most therapeutic antibodies are large, around 150 kilodaltons, Y-shaped, and built to circulate. That size is an asset in the bloodstream and a liability in dense, inflamed tissue. On 10 August, MoonLake reported positive Phase 3 results for a drug built on the opposite premise.

What IZAR-1 showed

Sonelokimab was tested in biologic-naïve adults with active psoriatic arthritis. The primary endpoint was superiority to placebo on ACR50, a 50% improvement on the American College of Rheumatology composite measure, at Week 16, and it was met. Reported responses included ACR20 of 66.5%, minimal disease activity in 41.2% of patients, and PASI90 in 61%, meaning a 90% improvement on the Psoriasis Area and Severity Index. The blinded safety analysis was consistent with earlier studies with no new signals, and dropout to Week 16 was low. Patients continue to Week 52 to test durability.

What a Nanobody actually is

Camels, llamas and alpacas make an unusual class of antibody: one that has dropped the light chain altogether and binds its target using a single heavy-chain domain. Isolate that domain and you have a VHH, or Nanobody, a fully functioning antigen-binding unit roughly a tenth the size of a conventional antibody.

Sonelokimab strings three of them together with flexible glycine-serine spacers into a single molecule of about 40 kilodaltons. Two of the domains bind IL-17A and IL-17F, the signalling proteins that drive inflammation in psoriatic disease, covering all three of the dimers those proteins form: A/A, A/F and F/F. The third domain does something quite different. It binds human albumin, the most abundant protein in blood, and that serves two purposes. Albumin has a long half-life, so hitching a ride keeps the drug in circulation long enough for convenient subcutaneous dosing. And albumin leaks into inflamed tissue wherever fluid accumulates, so the drug concentrates where the disease actually is.

Why size might matter

The argument for a smaller molecule is tissue penetration. Psoriatic arthritis involves inflamed joints and entheses, the points where tendon meets bone, which are poorly vascularised and hard for a large protein to reach. A 40-kilodalton molecule that also rides albumin into swollen tissue should, in theory, reach higher local concentrations than a 150-kilodalton antibody circulating past.

That is the theory. Whether it beats the established IL-17 inhibitors is a question only a head-to-head trial can answer, and IZAR-1 was placebo-controlled.

Why it matters

Molecular format is becoming a design variable in its own right. For thirty years the interesting question in biologics was which target to hit. Increasingly it is what shape of molecule to hit it with, and where in the body that shape can go. Nanobodies, antibody-drug conjugates and bispecifics are all answers to the same question.

Cosmael ThinkLab commentary

The albumin-binding domain is the part that deserves attention. It does nothing to the disease. It is a delivery decision, using the body’s own transport protein as a shuttle to a place the drug needs to reach. That kind of thinking, where pharmacokinetics is engineered into the molecule rather than managed around it, is where a good share of the near-term gains in biologics are likely to come from.

There is a nice piece of evolutionary opportunism here as well. Nobody designed the camelid single-domain antibody; it is an accident of one lineage’s immune system that turned out to be a superb engineering starting point. Much of biotechnology is this: finding something nature already built and repurposing it.

Keep the framing honest, though. These are topline numbers in a press release, against placebo, at Week 16. The comparison that matters is against the IL-17 inhibitors already on the market, and that comparison has not yet been run.

Sources: MoonLake Immunotherapeutics topline release via GlobeNewswire (10 August 2026); MoonLake Immunotherapeutics investor relations; MoonLake science pages on the sonelokimab Nanobody format.