Background/Purpose: Albeit lupus occurring on each TNFi have been reported, no epidemiological study has been conducted to assess the link between lupus onset and each TNFi drug. Indeed, the risk may differ because of different structural properties. Moreover, there are reports of infliximab-induced lupus without positive rechallenge with etanercept, and etanercept seems beneficial in systemic lupus erythematosus patients with arthritis and seritis. The aim of this study was to describe the cases of TNFi-related lupus reported in the French PharmacoVigilance Database (FPVD), and to assess the putative association with each TNFi.

Methods:

All spontaneous reports of TNFi-related lupus recorded in the FPVD between January 2000 and December 2012 were described. We conducted disproportionality analyses (case/non-case method) to assess the link between lupus and exposure to TNFi. Cases were all reports of lupus recorded during the study period. Non-cases were all other reports recorded during the same period. Exposure to TNFi was searched in cases and non-cases. Reporting odds ratios (ROR) were calculated to assess the association. We used exposure to isoniazid (well-known as lupus inducer) as positive control and acetaminophen as negative one. Sensitivity analyses were performed to test for event-related competition bias (removing reports of infections from the model) and for drug-related competition bias restricting to the marketing period of each TNFi and withdrawing well-known lupus-inducers. These were identified from the Chang and Gershwin list (J Autoimmunity 2010), updated through MEDLINE search until 2012 to detect new signals (comparative studies or ≥3 reports). 

Results:

During the study period, 309 671 spontaneous reports were colligated in the FPVD, of which 5213 (1.68%) involved TNFi. Among these TNFi reports, 39 were lupus in 37 patients: 25 involved infliximab, 9 adalimumab, and 5 etanercept. Male:female sex-ratio was 0.1 and mean age was 44.9 ± 14.4 years. Seventeen patients were treated for rheumatoid arthritis, 15 for inflammatory bowel disease, 4 for ankylosing spondylitis. Median delay from TNFi introduction to lupus onset was 11 months (range: 1 – 84 months). Cutaneous and rheumatologic involvements were the more frequent. Antinuclear autoantibody were positive in all the patients with this data reported (n=35). Anti-DNA antibodies were positive in 21/27 patients (77.8%). Improvement was observed after TNFi withdrawal (data available for half of the reports). Association between TNFi exposure and lupus was significant for all TNF-alpha antagonists pooled together (ROR=7.72, 95%CI[5.50-10.83]) and for isoniazid (3.5 [1.44–8.49]) but not with acetaminophen (0.28 [0.12–0.63]). It was similar for infliximab (10.97 [7.27-16.56]) and adalimumab (9.03 [4.64-17.58]) but was only 4.02 [1.66-9.75] for etanercept. Sensitivity analyses lead to similar results.

Conclusion:

Albeit confidence intervals slightly overlap probably because of lack of power, the association of etanercept and lupus occurrence is estimated the half of the association with monoclonal antibodies in all analyses. Etanercept should be preferred patients who experienced lupus while exposed to monoclonal antibody TNFi.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/is-the-risk-of-tumor-necrosis-factor-inhibitor-induced-lupus-the-same-with-monoclonal-antibodies-and-soluble-receptor-a-casenon-case-study-in-a-nationwide-pharmacovigilance-database/