49 Cluster analysis of patients with SLE in the adelphi lupus disease specific programme. (April 2019)

• Record Type: Journal Article
• Authors: Gairy, Kerry; Hoskin, Ben; Bell, David; Massey, Olivia; Amelio, Justyna; Liakos, Alex
• Is Part Of: Lupus science & medicine. Volume 6(2019)supplement 1
• Page Start: A36
• Page End: A37
• Publication Date: 2019-04
• Subjects: Systemic lupus erythematosus -- Periodicals; 616.772005
• DOI: 10.1136/lupus-2019-lsm.49
• Languages: English
• ISSNs: 2398-8851

Abstract

Abstract : Background: Organ system involvement in systemic lupus erythematosus (SLE) is highly variable. Better understanding of the symptom and organ involvement patterns in patients with SLE may facilitate making more accurate prognoses and individualized management decisions. This analysis aimed to categorize patients with SLE into clusters according to their organ system involvement. Methods: This was a secondary, descriptive analysis (208683) that utilized survey data collected from adult patients with SLE and their physicians, enrolled in the 2015 Adelphi Real World Disease Specific Programme (US and Europe) (205086). Latent class modeling based on current organ involvement was used to generate clusters of patients with similar manifestations; characteristics of each cluster were compared using the chi-square test for categorical variables and Kruskal-Wallis test for ordered/numeric outcomes. Results: Overall, 1376 patients (n=1196 [87.0%] female; mean [standard deviation (SD)]) age, 42.1 [13.6] years) were included in the analysis. Four patient clusters were generated: Cluster 1 (n=250, 18.2%), lowest overall organ burden (predominantly mucocutaneous); Cluster 2 (n=670, 48.7%), joint and skin SLE (predominantly mucocutaneous and musculoskeletal) with limited renal/hematologic involvement; Cluster 3 (n=150, 10.9%), highest frequency of renal/hematologic involvement; Cluster 4 (n=306, 22.2%), highest frequency of mucocutaneous, musculoskeletal, constitutional, cardiorespiratory and neuropsychiatric involvement, but without renal involvement. Key results are summarized in the table 1 . Significant between-cluster differences were observed for disease severity (p<0.0001; highest: Cluster 3); number of affected organ systems (p<0.0001; highest: Clusters 3 and 4); number of flares in prior 12 months (p<0.0001; highest: Clusters 3 and 4); disease progression (p<0.0001; most compromised: Cluster 3); time since diagnosis (p<0.001; longest: Cluster 3); and ethnicity (p<0.01; black race most prevalent: Cluster 3). Overall, the most commonly experienced symptoms were pain/inflammation and skin symptoms (p<0.0001; highest: Cluster 4). Frequency of organ involvement increased over time in Clusters 3 and 4 but decreased in Cluster 1. Statistically significant differences were observed between clusters in the number of previous treatments and treatment classes (both p<0.01). Activity impairment generally increased, while fatigue severity worsened, across the clusters (both <0.0001). Conclusions: This analysis provides important insights on potential clinically meaningful subsets of SLE (per organ system involvement) using real-world evidence. The highest disease burden was observed in Clusters 3 and 4, confirming the extensive impact of SLE irrespective of renal involvement. Limitations included the absence of serological findings or disease activity indices for cluster formation or comparison. Funding Source(s): Study funded by GlaxoSmithKline; ownership of data retained by Adelphi Real World.