You are here

Article Text

Article menu

Download PDFPDF

S4d – Reproduction
S4D:6 Sle comprises four immune-phenotypes, which differ regarding hla-drb1 and clinical associations
  1. LM Diaz Gallo1,
  2. E Lundström1,
  3. V Oke1,
  4. K Elvin2,
  5. YL Wu3,
  6. J Gustafsson1,
  7. A Jönsen4,
  8. D Leonard5,
  9. A Zickert1,
  10. G Nordmark5,
  11. AA Bengtsson4,
  12. J Sandling5,
  13. L Rönnblom5,
  14. I Gunnarsson1,
  15. CY Yu3,
  16. L Padyukov1 and
  17. E Svenungsson1
  1. 1Department of Medicine, Rheumatology Unit, Karolinska Institutet/Karolinska University Hospital, Stockholm, Sweden
  2. 2Department of Clinical Immunology and Transfusion Medicine, Unit of Clinical Immunology, Karolinska Institutet/Karolinsk, Stockholm, Sweden
  3. 3The Research Institute at Nationwide Childrens Hospital, Columbus, Ohio, USA
  4. 4Department of Clinical Sciences, Section of Rheumatology, Lund University, Skåne University Hospital, Lund, Sweden
  5. 5Department of Medical Sciences, Section of Rheumatology, Uppsala University, Uppsala, Sweden

Abstract

SLE is a heterogeneous disease including diverging clinical symptoms, autoantibodies and genetic susceptibility. Hitherto unrecognised patterns may define sub-phenotypes with different pathogenesis and specific treatment needs. Based on autoantibody profile we therefore investigated phenotypic clusters and explored cluster associations with clinical manifestations and one of the most important genetic risk factors for SLE, HLA-DRB1 alleles.

908 SLE Caucasian patients and 3654 age- gender- and ethnicity-matched healthy controls (HC) were included. We determined the occurrence of 13 autoantibodies: dsDNA, nucleosomes, ribosomal P, RNP68, RNPA, Sm, Sm/RNP, SSA52, SSA60, SSB, aCL-IgG/IgM and aB2GP1. HLA-DRB1 typing was performed by sequence-specific primer polymerase chain reaction assay. Cluster analysis was done using Gower distance matrix, followed by partition around medoids cluster calculation and Silhouette metric for number of clusters validation. Chi-square test, odds ratios (OR), 95% confidence intervals and false discovery rate p value (p) were calculated for the association tests.

Four clusters were defined based on autoantibody occurrence.

  1. 29%, dominated by anti–SSA52/60/SSB positivity is strongly associated with HLA–DRB1*03 when compared to HC (4.1[3.4–4.9] p=6.4E–56) and other clusters (OCs) (2.9[93.3–3.6] p=1.1E–19). Discoid lesions were more common vs OCs (1.8[1.3–2.6] p=0.02).

  2. 29%, dominated by anti–SmRNP/Sm/DNA/RNPA/RNP68/nucleosome, was specifically associated with HLA–DRB1*15 when compared to HC (1.7[1.6–2.1] p=5.7E–6) and other clusters (1.5[1.1–1.9] p=0.01). Nephritis was common vs OCs (1.9[1.4–2.7) p=2.E–03)

  3. 24%, dominated by anti–B2GP1/aCL–IgG/IgM, was associated with HLA–DRB1*04 when compared with other clusters (1.8[1.4–2.4] p=2E–4). More thrombotic events vs OCs were observed in this group (1.84 [1.3–2.6] p=0.01)

  4. 18% was negative for the 13 tested autoantibodies and was not associated with any specific HLA–DRB1 alleles and it was not associated as risk factor for any of the evaluated clinical manifestations.

We demonstrate that immune-phenotypes/clusters in SLE can fit into a frame of HLA-DRB1 alleles and that the overall association between SLE and HLA-DRB1*03 and HLA-DRB1*15 seems to be driven mainly by clusters 1 and 2, respectively. We also confirm previous observations that autoantibody clusters associate with clinical symptoms. We believe that these results could be used to redefine SLE, determine predictive biomarkers and inclusion criteria for clinical trials.

  • Auto antibodies
  • Immune-phenotypes
  • HLA-DRB1

http://dx.doi.org/10.1136/lupus-2018-abstract.25

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.