Lupus & antiphospholipid syndrome
Machine learning identifies clusters of longitudinal autoantibody profiles predictive of systemic lupus erythematosus disease outcomes
Authors' conclusion
Four discrete SLE patient longitudinal autoantibody clusters were predictive of long-term disease activity, organ involvement, treatment requirements and mortality risk.
Study context
- Population
- Patients with systemic lupus erythematosus (SLE) from a multinational inception cohort, data within 15 months of diagnosis and at 3-year and 5-year follow-upinception cohort
- Participants
- 805 patients
- Exposure
- Longitudinal autoantibody clusters 1, 3 and 4 vs cluster 2
- Cluster 1n = 137
- Cluster 2n = 376reference
- Cluster 3n = 80
- Cluster 4n = 212
- Outcome
- Clinical outcomes and survival by longitudinal autoantibody cluster
- Follow-up
- 10-year follow-up
Methodological detailsmultinational, machine learning, K-means clustering on PCA-transformed longitudinal autoantibody profiles, Cox proportional hazards model adjusted for age of disease onset
- Multinational
- Machine learning
- K-means clustering on PCA-transformed longitudinal autoantibody profiles
- Cox proportional hazards model adjusted for age of disease onset
Main finding
As reported in the abstract
Survival (lower survival vs cluster 2)
Each group vs cluster 2AHR, event not clear (lower survival), adjusted for age of disease onset (95% CI)P value
Cluster 1
2.60(1.12–6.05)
P = 0.03Cluster 3
2.87(1.22–6.74)
P = 0.02Observational data: associations, not effects.
Further findings
Clinical outcomes by cluster · year 10
- Cluster 1: highest cumulative disease activity and immunosuppressants/biologics use
- Cluster 2: lowest disease activity, frequency of LN and immunosuppressants/biologics use
- Cluster 3: highest frequency of seizures and hypocomplementaemia
- Cluster 4: also high disease activity
Original publication
Machine learning identifies clusters of longitudinal autoantibody profiles predictive of systemic lupus erythematosus disease outcomes. Ann Rheum Dis. Published 21 April 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →