Pemphigus: the antibodies fell, the blisters did not

Design

Phase 3, multicentre, randomised, double-blind, placebo-controlled trial (2:1)

Population

222 adults with newly diagnosed or relapsing moderate-to-severe pemphigus (190 vulgaris, 32 foliaceus)

Primary outcome

Complete remission on minimal prednisone by week 30, sustained 8 weeks or more

Effect

35.5% (44/124) vs 30.3% (20/66); OR 1.19 (95% CI 0.60 to 2.41), P=0.60

Two hundred and twenty-two adults with moderate-to-severe pemphigus, mostly pemphigus vulgaris, were randomised 2:1 to weekly subcutaneous efgartigimod PH20 or placebo, every patient also starting prednisone at 0.5 mg/kg daily. The primary endpoint — complete remission on 10 mg of prednisone or less, sustained at least eight weeks, by week 30 — was reached by 35.5 per cent on the drug and 30.3 per cent on placebo. Cumulative prednisone exposure was no different either.

What makes this trial worth reading rather than filing is the dissociation. Efgartigimod did exactly what it is designed to do: blocking the neonatal Fc receptor cut total IgG and both anti-desmoglein-1 and anti-desmoglein-3 titres quickly. Pemphigus Disease Area Index scores did not move with them.

So for pemphigus, at least at this dose and over 30 weeks, autoantibody level is not a sufficient target. That is a caution worth carrying into how titres are used in ordinary follow-up: a falling desmoglein titre on any treatment is a pharmacodynamic observation, and the skin remains the endpoint.

• Do not treat a falling desmoglein titre as evidence that a pemphigus patient is improving; score the skin and mucosa.
• Corticosteroid plus rituximab remains the standard first-line combination for moderate-to-severe disease.
• Adverse events were commoner on efgartigimod (89.1% vs 76.0%) though serious events were not, and there were no deaths.
• Pemphigus foliaceus contributed only 32 of the 222 patients; the primary analysis was in pemphigus vulgaris.
• Counsel patients asking about FcRn blockers that the phase 3 result in pemphigus was negative, whatever the mechanism suggests.

Why it matters

It breaks the assumed chain from autoantibody level to disease activity in the disease where that chain looked strongest.

Don't overread it

A negative trial at this dose and duration does not rule out benefit in a different regimen or a different population.

The statistics, in plain English

An odds ratio of 1.19 with a confidence interval from 0.60 to 2.41 includes 1.0 comfortably: no benefit was shown, and the interval is wide enough that a modest benefit is not excluded either. With 190 patients contributing to the primary endpoint this trial could only have detected a large effect. The more informative result is the mechanistic one — antibodies down, disease unchanged — which does not depend on the sample size.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.