Is it possible to predict which patients treated with biologic agents for rheumatic diseases will develop anti-drug antibodies?

Résumé 0

Objectives Biologic agents are one of the main treatments for auto-immune diseases such as rheumatoid arthritis (RA), ankylosing spondylitis (AS), and psoriasis arthritis (aPso). These drugs are often associated with good control of the disease, but in some cases, patients develop anti-drugs antibodies (ADABs) which can lead to a failure in the control of the disease. Why only certain patients develop these antibodies is not yet really understood. The aim of this study is to look for clinical and biological predictors for the development of ADABs when tested in a real life cohort. Method In this retrospective study, 297 patients, followed at the unit of rheumatology of the Centre Hospitalier Universitaire Vaudois (CHUV) for a RA, SA, aPso and some with connective tissue disease, were included. The patients had to have at least one measurement of ADABs and drug trough level since 2013 to January 2016. They also had to be exposed to a treatment of anti-TNF agent or non-anti-TNF agent such as rituximab or tocilizumab. The method used for detecting the ADABs was a sandwich ELISA. The reproducibility of the ELISA method and the cut-off for ADABs have been tested among patients exposed and non-exposed to the medication. Results 63 patients out of 297 developed ADABs, which represents 21% of the total cohort. In univariable analyses, many clinical and biological predictors were significantly associated with ADABs. In multivariable analyses, only four predictors remained significantly associated with ADABs. Clinical predictors independently associated with the development of ADABs were treatment with a monoclonal anti-TNF agent (OR:26, 95% CI:2.6-264), and previous exposure to other biologic agents (OR:5.9, 95% CI: 1.1-30). Two laboratory predictors were also independently associated with the development of ADABs an undetectable through level of the medication (OR:34, 95% CI: 7.2-160) and an high TNF through levels (OR:4.2, 95%CI:1.1 -15). In patients exposed and tested for ADABs against more than one bDMARD, the percentage of ADABS was: 33 % (not significantly than against one agent: (p-value = 0.08)). Conclusion Our study confirms that ADABs can be found in a significant number of patients treated with biological DMARDs. The clinical predictors for developing such ADABs are limited and in line with those found in previous publications.