Background/Purpose: Systemic Lupus Erythematosus (SLE) is a chronic, heterogeneous, multisystem autoimmune disease characterized, in part, by the presence of autoantibodies. The spectrum of disease presents a challenge when discussing disease course, outcomes and treatment. While autoantibody clusters have been described in adult lupus, there is little data examining relationships among specific autoantibody clusters and disease manifestations in pediatric SLE (pSLE) patients. The objective of this study is to identify distinct autoantibody clusters and their role in the development of unique disease profiles within an exclusively pSLE cohort.

Methods: We conducted a cross-sectional analysis of pSLE patients in the CARRA legacy database (of N=989), ages 1-19 years. Patients without a complete autoantibody profile (including anti-dsDNA, anti-Sm, anti-RNP, anti-Ro, and anti-La antibodies) were excluded. A cohort of 198 patients were analyzed. Two-step cluster analysis was performed to identify subsets of patients with similar autoantibody patterns. Additional frequency analysis and logistic regression were completed to assess for cluster/disease manifestation associations.

Results: Demographic variables were similar between the entire pSLE cohort and the subset of 198 patients included in our analysis. Anti-dsDNA was the most prevalent autoantibody (66.2%), followed by anti-Smith (50%), Anti-RNP (47%), anti-RO (45%), and anti-LA (23.2%). The most prevalent manifestations were immunologic disorder (86.9%), hematological disorder (65.7%), arthritis (64.1%), renal disorder (54%), and malar rash (45.5%). Four distinct antibody clusters were identified: Cluster 1 (all 5 antibodies present, n=53), Cluster 2 (anti-RNP, anti-Smith, anti-dsDNA, n=50), Cluster 3 (anti-dsDNA only, n=51), and Cluster 4 (anti-Ro, anti-La, anti-RNP and anti-dsDNA, n=44). The distribution of African American patients was statistically different across clusters, with the lowest percentage in cluster 3 (p=0.0067). There were roughly twice as many white patients in cluster three as in the other three clusters with a p-value approaching significance (p=0.0634). Patients in Cluster 1 were on average younger than patients in Cluster 2 at age of onset and at first visit to a rheumatologist. Clusters 2 and 3 had the highest frequency of neuropsychiatric manifestations (40% and 50%, respectively). The remaining clinical manifestations were equally distributed among clusters. For individual antibodies, patients with an anti-La antibody were 61% less likely to develop arthritis and 79% less likely to develop serositis. Patients were three times more likely to have hematological manifestations when anti-RNP antibodies were present, and 73% less likely if anti-Smith antibodies were present.

Conclusion: Our data suggests that autoantibodies in pSLE do tend to cluster, which could potentially represent subsets of disease. However, in our sample, there was not a robust correlation amongst clinical manifestations and clusters. Future research is needed to better define these clusters in larger populations.

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/cluster-analysis-of-autoantibodies-and-their-relationship-with-demographic-and-clinical-features-in-pediatric-systemic-lupus-erythematosus-in-the-childhood-arthritis-and-rheumatology-research-all/