Background: It has been well known that anti-TNF-α treatment for patients with rheumatoid arthritis (RA) is associated with anti-nuclear antibody (ANA) development. Using data of 110 RA patients treated with infliximab (IFX), we previously reported that ANA development along with ANA levels at base line were associated with poor outcomes of IFX.¹ However, no replication studies have been reported. In addition, whether the findings are true to general biological disease-modifying anti-rheumatic drugs (bDMARDs) is uncertain.

Objectives: To replicate an association between poor response and development of ANA during IFX treatment in patients with RA. To analyse whether the association is found in other bDMARDs.

Methods: We analysed a data of Japanese RA patients treated with (n=721) or without (bDMARDs-naïve; n=1050) bDMARDs (IFX, etanercept ETN, adalimumab ADA, golimumab GLM, certolizumab pegol CZP, tocilizumab TCZ, abatacept ABT) as a first line bDMARD from a single comprehensive RA cohort. All of the participants were not included in our previous study. We conducted multiple logistic linear regression analysis to assess effects of ANA development on treatment outcomes. We further analysed correlates of ANA development. ANAs were determined by indirect immunofluorescence with HEp-2 cells, and ANA development was defined as more than two times increase from baseline values.

Results: We found that ANA development at 3 months after starting bDMARDs was significantly associated with cessation of bDMARDs due to insufficient response within a year (OR 3.70, p=0.037). We further found that RA patients who did not develop ANA at 3 months but developed ANA within a year showed a significant association with treatment failure between 12 and 24 months (OR 7.11, p=0.044). ANA levels at baseline showed significant association with or tendency of insufficient response in both situations (OR 1.21 and 1.69, respectively), independently on ANA development. We still found associations of ANA development after conditioning on IFX usage, indicating that the associations are not limited to IFX. Female was associated with ANA development (OR 1.85) and high levels of ANA at baseline (OR 2.41, p=0.006), which is consistent with previous data in healthy population.² bDMARDs use (OR 2.09, p=0.056) was also a risk for ANA development. Among bDMARDs, anti-TNF agents, especially IFX, were risk factors of ANA development (OR 6.34, p<0.0001).

Conclusions: ANA development during treatment is associated with poor response to bDMARDs, which is not limited to IFX. Female and IFX usage are risk factors for ANA development.

References

1. Yukawa N, et al. Arthritis Research & Therapy 2011;13:R213.
2. Kuramoto N, et al. Sci Rep 2017;7:6911.

Disclosure of Interest: Y. Ishikawa: None declared, M. Hashimoto: None declared, H. Ito Grant/research support from: Pfizer Japan Inc., Mitsubishi Tanabe Pharma Corporation, Astellas Pharma Inc., Chugai Co., Ltd., Bristol-Myers Squibb, M. Tanaka: None declared, N. Yukawa: None declared, T. Fujii Grant/research support from: Pfizer Japan Inc., Ono Pharmaceutical Co., Ltd., Mitsubishi Tanabe Pharma Corporation, Eisai Co.,Ltd., Astellas Pharma Inc., T. Mimori Grant/research support from: Pfizer Japan Inc., Mitsubishi Tanabe Pharma Corporation, Eisai Co.,Ltd., Astellas Pharma Inc., AYUMI Pharmaceutical Co., Chugai Co., Ltd., Bristol-Myers Squibb, Daiichi Sankyo Company, Ltd., Nippon Shinyaku Co., Ltd., Takeda Pharmaceutical Company, Ltd., C. Terao: None declared

DOI: 10.1136/annrheumdis-2018-eular.1232