Relapse and infection scores may help decide who should extend rituximab maintenance in ANCA vasculitis

Design

External validation of prediction models in trial cohorts

Population

217 patients with ANCA-associated vasculitis from MAINRITSAN1 and MAINRITSAN2 after 18 months of rituximab

Primary outcome

Relapse-free and serious infection-free survival

Effect

Relapse score HR 1.91 (95% CI 0.97 to 3.76); infection score HR 2.93 (1.28 to 6.72)

Long-term follow-up of the MAINRITSAN trials showed rituximab prevents relapse in ANCA-associated vasculitis, and extending beyond 18 months may help high-risk patients. This study externally validated published relapse and infection prediction models in 217 patients from MAINRITSAN1 and MAINRITSAN2, who had received rituximab for 18 months.

Relapse occurred in 76 patients, with 72-month relapse-free survival of 55% (95% CI 48 to 64), and serious infection in 33 patients (20% at 72 months). The relapse score (male sex, age above 60, ANCA positivity, relapsing disease, ENT involvement, prednisone dose) predicted major relapse (p = 0.04) but gave a borderline hazard ratio for major relapse-free survival of 1.91 (95% CI 0.97 to 3.76; p = 0.059), with the association mainly in PR3-ANCA. The infection score (male sex, structural lung disease, diabetes, infections during maintenance, gammaglobulin level) identified higher risk (HR 2.93, 95% CI 1.28 to 6.72).

The relapse model's validation was modest and the infection model performed better. Neither tells you whether extending rituximab improves outcomes; they could support individualised discussions with the patient about the balance of relapse and infection risk.

• 72-month relapse-free survival was 55%; serious infection was 20% at 72 months.
• The infection score identified higher risk (HR 2.93, 95% CI 1.28 to 6.72).
• The relapse score was borderline (HR 1.91, 95% CI 0.97 to 3.76), mainly in PR3-ANCA.
• Use the factors to frame a discussion about extending rituximab, not to decide.
• Check immunoglobulin levels and infection history when considering extension.

Why it matters

It puts numbers to the trade-off at the point where many clinicians decide whether to continue rituximab.

Don't overread it

These models predict risk but have not been shown to improve outcomes when used to guide extension.

The statistics, in plain English

A hazard ratio of 1.91 with an interval of 0.97 to 3.76 crosses 1, so the relapse score did not clearly separate groups in this validation. The infection score's interval of 1.28 to 6.72 excludes 1 but is wide.

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.