Article Text

FRI0573 Integration of Time-Averaged DAS28 Fits Better Joint Destruction In Rheumatoid Arthritis Than One-Time DAS28 and Identifies A Significant Joint-Destructive Association of HLA-DRB1*04:05 Which is Independent of ACPA and DAS28
  1. H. Tsuji1,
  2. C. Terao2,
  3. K. Yano3,
  4. K. Ikari3,
  5. M. Hashimoto4,
  6. M. Furu4,
  7. H. Ito5,
  8. T. Fujii4,
  9. W. Yamamoto6,
  10. N. Yamakawa7,
  11. K. Ohmura1,
  12. A. Taniguchi3,
  13. S. Momohara3,
  14. H. Yamanaka3,
  15. F. Matsuda2,
  16. T. Mimori1
  1. 1Department of Rheumatology and Clinical Immunology
  2. 2Department of Center for Genomic Medicine, Kyoto University Graduate School of Medicine, Kyoto
  3. 3Institute of Rheumatology, Tokyo Women's Medical University, Tokyo
  4. 4Department of the Control for Rheumatic Diseases
  5. 5Department of Orthopedic Surgery, Kyoto University Graduate School of Medicine, Kyoto
  6. 6Department of Health Information Management, Kurashiki Sweet Hospital, Kurashiki
  7. 7Department of Immunology and Rheumatology, Katsura Hospital, Kyoto, Japan


Background Disease Activity Score 28 (DAS28) is one of correlates of joint destruction in rheumatoid arthritis (RA) as well as genetic components, RA-related autoantibodies, and disease duration. Joint destruction is considered to reflect the cumulative history of the disease activities, but one-time DAS28 is used in many studies. Relationship among the correlates is still unclear, while we previously showed that HLA-DRB1*04:05 is associated with joint destruction independently of anti-citrullinated protein antibody (ACPA).

Objectives We hypothesized the superiority of time-averaged DAS28 to one-time DAS28 to fit the model to assess joint destruction. We also hypothesized that DRB1*04:05 might be associated with joint destruction independently of ACPA and DAS28.

Methods We recruited RA patients in the KURAMA (n=204) and IORRA cohort (n=557) whose data of modified Sharp/van der Heijde score (SHS), consecutive DAS28, and other correlates of SHS were available. We calculated time-averaged DAS28 and modeled SHS using time-averaged DAS28 and the other covariates. We randomly picked up DAS28 in each patient and constructed 1000 sets of DAS28 to model SHS. We empirically evaluated fitting of the model using time-averaged DAS28 among the 1000 models using one-time DAS28. Next, we analyzed an association between DRB1*04:05 and SHS of ACPA (+) patients in KURAMA (n=100) and IORRA (n=472).

Results In conditioned for the autoantibodies and disease duration, the time-averaged DAS28 showed significant improvement of model fitting in comparison with one-time DAS28 (R2 in KURAMA: p=0.037, R2 in IORRA: p=0.001). Generalized linear regression model in ACPA (+) patients showed a significant association between DRB1*04:05 and SHS (overall p=0.0002) in conditioned for time-averaged DAS28 and the other covariates. We observed this significant association even in patients whose time-averaged DAS28 was in low disease activity or remission.

Conclusions Time-averaged DAS28 fits SHS better than one-time DAS28 and usage of time-averaged DAS28 as a covariate would increase the power of studies to identify novel correlates of SHS. DRB1*04:05 demonstrates an association with joint destruction which is independent of ACPA and even DAS28 by using time-averaged DAS28, suggesting necessity of strong treatment for RA patients carrying DRB1*04:05 even when they are in low disease activity or remission.

  1. Navarro-Compán V, et al. Rheumatology. 2015 54(6):994–1007.

  2. Terao C, et al. Arthritis Rheumatol. 2015 67(7):1744–50.

Disclosure of Interest None declared

Statistics from

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.