- Design
- Post hoc risk-prediction model from a randomised trial cohort
- Population
- 649 PEXIVAS participants with ANCA vasculitis in remission
- Primary outcome
- First relapse within 2 years of remission
- Effect
- Optimism-corrected AUC 0.71 (95% CI 0.707 to 0.710); calibration slope 0.99
Using data from 649 participants in the PEXIVAS trial who reached remission, all with eGFR below 50 or diffuse alveolar haemorrhage at presentation, investigators built a model to estimate the risk of first relapse within two years. One hundred relapsed in that time.
The final model used sex, induction with oral cyclophosphamide or rituximab, ANCA subtype, non-haemorrhagic respiratory or mucous membrane and eye involvement at presentation, and eGFR at remission. Discrimination was moderate (optimism-corrected AUC 0.71) and calibration good (slope 0.99).
Relapse risk varies widely, and decisions about maintenance duration and monitoring intensity currently rest on judgement. A structured estimate could help explain that variation to patients and target closer follow-up.
The model comes from a severe-disease trial population, has not been externally validated, and an AUC of 0.71 leaves substantial misclassification. It should support, not replace, clinical judgement.
- Factors linked to relapse include ANCA subtype, induction agent, sex, and eGFR at remission.
- Upper respiratory, mucosal or eye involvement at presentation contributed to relapse risk.
- Use any risk estimate to guide monitoring frequency, not to stop maintenance on its own.
- This model derives from severe disease with kidney or lung involvement.
Why it matters
It turns a vague sense of relapse risk into a number that can be discussed.
Don't overread it
Not externally validated, and built from patients with severe renal or pulmonary disease.
The statistics, in plain English
An AUC of 0.71 means that, for a random pair of one patient who relapsed and one who did not, the model ranks the relapsing patient higher about 71% of the time. That is moderate, not precise, discrimination.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for rheumatology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free