Shares of Alumis fell about 55% after its experimental drug for systemic lupus erythematosus failed a mid-stage trial — another entry in a long record of lupus disappointments.

Envudeucitinib, a TYK2 inhibitor, missed all primary and secondary objectives across three doses versus placebo in a 408-patient, 48-week study. The result nearly erased gains Alumis had banked from positive psoriasis data in January.

Why lupus defeats drugs

Systemic lupus erythematosus is among the hardest diseases in which to demonstrate a treatment effect, and the reasons are structural rather than bad luck.

Lupus is extraordinarily heterogeneous. It can attack skin, joints, kidneys, blood, heart, lungs or brain in almost any combination, and two patients carrying the same diagnosis may have little in common clinically. A drug effective against one manifestation may do nothing for another, and trials enrolling across that spectrum dilute any real signal.

The disease also relapses and remits spontaneously, producing large placebo responses that swamp modest treatment effects. Compounding this, most patients remain on background steroids and immunosuppressants, so any new drug must demonstrate benefit on top of therapy already suppressing the disease.

The result is a field where only a small number of drugs have been approved in decades, despite sustained effort.

What TYK2 inhibitors do

TYK2 belongs to the JAK family of enzymes that transmit signals from inflammatory cytokine receptors into the cell. Blocking it interrupts several inflammatory pathways at once.

The class attracted interest because selective TYK2 inhibition promised the benefits of JAK inhibition with a better safety profile — broader JAK inhibitors carry warnings around clots, infections and cardiovascular events. A TYK2 inhibitor reaching the market in psoriasis appeared to validate that thesis.

The class has struggled outside psoriasis

That validation has not travelled well. Multiple TYK2 inhibitors have struggled in lupus and in inflammatory bowel disease, and even the approved TYK2 drug has underperformed sales expectations in the indications where it works.

A pattern is emerging: the mechanism appears sufficient in psoriasis, a relatively homogeneous disease driven by a well-defined cytokine axis, and insufficient in conditions with more complex or redundant inflammatory drivers. Lupus, driven substantially by interferon signalling and autoantibody production, may simply require more than partial interruption of one signalling node.

The subgroup Alumis is pointing to

Alumis highlighted “robust” responses in patients with a high interferon gene signature — those overexpressing interferon-controlled genes — a group the company says constitutes a majority of moderate-to-severe SLE patients. It plans to discuss a Phase 3 with regulators.

The biological rationale is legitimate. Type I interferon is a well-established driver in lupus; an approved lupus therapy targets that pathway directly, and the interferon signature is a recognised biomarker rather than an invented one. TYK2 sits in interferon signalling, so a stronger effect in interferon-high patients is mechanistically coherent.

Why coherent is not the same as convincing

The problem is statistical rather than biological. When a trial misses every primary and secondary endpoint at every dose, subgroup analyses lose most of their evidentiary weight. Trials contain many subgroups, and some will show apparent benefit by chance alone — which is precisely why regulators treat post hoc subgroup findings as hypothesis-generating rather than as evidence.

The honest test is whether the interferon-high analysis was pre-specified with a defined threshold, and whether the effect size is large enough to survive a properly powered trial in that population alone. Companies routinely present subgroups after a failure; occasionally they are right, and the way to find out is to run the trial.

What it means for Alumis

The company says it does not foresee “significant changes” to its overall timeline and still plans a US psoriasis filing by year-end.

That is the more consequential fact for the business. Psoriasis is where the drug has positive data and where the class has demonstrably worked, and an approval there would give Alumis a commercial product. But psoriasis is also a crowded market with multiple effective biologics and an established oral competitor, so the commercial ceiling is considerably lower than lupus would have offered — which is what the 55% share decline reflects.

The wider read

Two things follow from this failure. The narrower one is about the TYK2 class: partial interruption of one signalling node appears adequate for some inflammatory diseases and not others, and the boundary is becoming clearer through expensive trials.

Why the psoriasis-to-lupus leap keeps failing

Alumis is far from the first company to reason from a psoriasis success to a lupus programme, and the pattern of failure is instructive.

Psoriasis is unusually tractable for an immune-mediated disease. It is driven by a well-characterised cytokine axis, the affected tissue is visible and directly measurable, response can be scored objectively by looking at the skin, and the disease does not spontaneously remit and relapse on the timescale of a trial. Drugs that interrupt the relevant pathway work, and the trial reliably detects it.

Lupus violates every one of those conditions. Multiple pathways contribute, disease activity is assessed through composite indices combining clinician judgement with laboratory values, organ involvement varies between patients, and spontaneous fluctuation is the norm. A mechanism can be genuinely active in lupus and still fail to produce a detectable signal in a trial designed around such an endpoint — which is why the field has struggled to distinguish drugs that do not work from trials that cannot measure whether they do.

The broader one concerns lupus. Trial after trial has faltered, and the field is still some distance from reliably helping patients with a disease that predominantly affects young women and causes substantial organ damage over time. Business and clinical news, not investment or medical advice.