Autoinflammatory & rare systemic diseases
Developing efficient predictive models for the diagnosis of VEXAS syndrome
Authors' conclusion
Multi-feature models that efficiently separate VEXAS cases from controls were successfully developed.
Study context
- Population
- Patients who underwent UBA1 mutation testing
- Participants
- 144 patients
- Comparison
- Clinical and laboratory features in UBA1 positive vs negative patients; multi-feature least absolute shrinkage and selection operator (LASSO) models to predict pathogenic UBA1 mutations
- Outcome
- Discrimination of models predicting pathogenic UBA1 mutations (AUC)
- Follow-up
- Symptom onset and time of UBA1 mutation testing
Methodological detailsLASSO prediction models, two timepoints: symptom onset and UBA1 mutation testing
- LASSO prediction models
- Two timepoints: symptom onset and UBA1 mutation testing
Performance
As reported in the abstract
Model discrimination (AUC)
| Test or model | AUC | AUC (both timepoints) |
|---|---|---|
| 12-feature model (symptom onset) | 0.86 | — not stated in the abstract |
| 5-feature model (symptom onset) | 0.81 | — not stated in the abstract |
| 16-variable model (testing) | 0.92 | — not stated in the abstract |
| 5-feature model (testing) | 0.84 | — not stated in the abstract |
| Single feature model | — not stated in the abstract | 0.78 |
AUC
12-feature model (symptom onset)0.86
5-feature model (symptom onset)0.81
16-variable model (testing)0.92
5-feature model (testing)0.84
Single feature model— not stated in the abstract
AUC (both timepoints)
12-feature model (symptom onset)— not stated in the abstract
5-feature model (symptom onset)— not stated in the abstract
16-variable model (testing)— not stated in the abstract
5-feature model (testing)— not stated in the abstract
Single feature model0.78
— not stated in the abstract
Further findings
Features associated with UBA1 mutation
- Skin rash, chondritis and monocytopenia significantly associated with UBA1 mutation at both timepoints
- Macrocytosis, uveitis and pulmonary disease significantly associated at time of testing
Original publication
Developing efficient predictive models for the diagnosis of VEXAS syndrome. Semin Arthritis Rheum. Published 16 July 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →