Abstract
Purpose Systemic lupus erythematosus (SLE) is a heterogeneous disorder with diverse manifestations. This study tried to classify patients with SLE by combining laboratory values when they were classified as SLE.
Methods We performed hierarchical cluster analysis with laboratory results at the time of classification of SLE. Linear discriminant analysis was performed to construct a model for predicting the clusters.
Results The cluster analysis using data of 389 patients with SLE yielded 3 clusters with different laboratory characteristics. Cluster 1 had the youngest age at diagnosis and showed significantly lower lymphocyte, hemoglobin, platelet count and complement levels, and the highest erythrocyte sedimentation rate (ESR) and anti-dsDNA antibody. Cluster 2 showed higher white blood cell (WBC), lymphocyte and platelet, and lower ESR and anti-dsDNA antibody. Cluster 3 revealed the highest titer of antinuclear antibody and lower WBC and lymphocyte count. For 171 months follow-up, Cluster 1 showed higher number of cumulative manifestations compared to Cluster 2 and 3 with higher prevalence of malar rash, alopecia, arthritis and renal disease. In addition, the dose of glucocorticoids and the proportion taking immunosuppressive agents were higher in Cluster 1 than Cluster 2 or Cluster 3. However, damage index and mortality didn’t differ significantly among 3 Clusters.
Conclusions Cluster analysis using initial laboratory results could identify 3 Clusters which had a distinct clinical characteristic in patients with SLE, with 84.5% accuracy. Although organ involvements and management patterns differ among the Clusters, damage and mortality didn’t differ.
