Starting mepolizumab early in EGPA was linked to more remission and less damage

Design

Retrospective multicentre cohort

Population

64 patients with EGPA (16 early mepolizumab, 48 conventional), Japan

Primary outcome

Clinical remission at 1 year (BVAS 0, prednisolone ≤4 mg/day)

Effect

69% vs 27% (P = 0.006); lower VDI over time (β −0.36, 95% CI −0.70 to −0.03)

A retrospective multicentre cohort study in Rheumatology, published on 1 October, used data from a Japanese vasculitis cohort to compare 16 patients with eosinophilic granulomatosis with polyangiitis who started mepolizumab within three months of remission induction with 48 managed conventionally.

Clinical remission at one year — no vasculitis activity with prednisolone at 4 mg/day or less — was reached by 69% with early mepolizumab and 27% without (P = 0.006). At two years, all 10 evaluable early-mepolizumab patients were in remission against 29% of the comparison group, and 40% vs 9% had stopped glucocorticoids. Vasculitis damage scores were lower over time with early mepolizumab.

The size of these differences is striking, but the groups were not randomised, the early group was small, and patients chosen for early biologic therapy may have differed in disease features, access and era of treatment. The finding supports considering mepolizumab during induction rather than waiting for relapse or steroid dependence, but it is not proof.

• Consider mepolizumab during remission induction in EGPA rather than only after relapse.
• Track cumulative glucocorticoid dose; reducing it was one of the main gains here.
• Record a Vasculitis Damage Index at baseline and over time.
• Check local availability and cost; access to mepolizumab in India is limited.

Why it matters

Glucocorticoid damage is a major source of long-term harm in EGPA, and earlier steroid-sparing therapy may reduce it.

Don't overread it

Retrospective and non-randomised, with only 16 patients in the early group; it cannot show cause.

The statistics, in plain English

With 16 patients in one arm and 10 evaluable at two years, a single patient moves the percentages substantially. The damage difference has a confidence interval reaching almost zero (−0.70 to −0.03).

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.