Background Systemic lupus erythematosus (SLE) is heterogeneous autoimmune disease. Identification of patient clusters may be useful for the management of the disease. The aim of this study is to describe different SLE clusters according to sociodemographic, clinical and serological characteristics.

Methods GLADEL 2.0 is an ongoing Latin-American observational cohort initiated in 2019. Variables chosen at cohort entry to stratify patients and construct clusters were selected from sociodemographic and cumulative clinical and serological variables. Hierarchical cluster analyses were performed by the Ward method on a distance matrix using the Gower’s method.

Results A total of 560 SLE patients were included in this analysis. Three clusters were identified. Cluster 1 (n=269) was characterized by more cutaneous, articular, renal and serosal involvement; serological manifestation was positive anti-dsDNA. Cluster 2 (n=194) was represented by patients who rarely had renal involvement and the most frequent clinical manifestations were cutaneous and hematological; the most frequent serological manifestations were the presence of antiphospholipid antibodies (aPLs). Cluster 3 (n=97) was characterized by a lower frequency of clinical and serological involvements, with the exception of neurological domain. Clusters 1 and 2 share hematologic manifestations and hypocomplementemia (table 1).

Conclusions In this cohort, three clusters were identified. Cluster 1 patients were characterized by renal, articular, cutaneous and serositis involvement, anti-dsDNA antibodies and hypocomplementemia, Cluster 2 patients were characterized by hematologic, cutaneous involvement, aPLs and hypocomplementemia. Cluster 3 patients presented fewer serological findings but a higher frequency of neurological involvement. Follow up of these patients will allow for elucidation of relationship of these clusters with SLE outcomes.