Background Tapering of biologics has been shown feasible, effective and safe in rheumatoid arthritis (RA) patients with stable low disease activity. Tapering of biologics is done using a trial-and-error disease activity guided strategy, because it is not known in advance which patient can successfully taper. Prediction of successful dose reduction or discontinuation of a biologic could prevent flares and would save time and drug exposition. In the last years, several studies have investigated biomarkers for predicting successful tapering. However, these results have not yet been systematically summarised.
Objectives To systematically review studies that address prediction of successful dose reduction or discontinuation of biologics in RA patients.
Methods A literature search was performed in PubMed, EMBASE and Cochrane Library (November 2015). Clinical (un)controlled trials were included if they investigated the predictive value of clinical patient characteristics, biochemical tests or imaging measures assessed prior to protocolised dose reduction or discontinuation of a biologic (abatacept, adalimumab, certolizumab, etanercept, golimumab, infliximab, rituximab and tocilizumab) in RA patients with a low disease activity state. Two review authors independently selected studies, extracted data and assessed risk of bias. Qualitative best-evidence synthesis was separately performed for the prediction of successful dose reduction and successful discontinuation. A biomarker was classified as potential predictor if the univariate association was either strong (odds ratio (OR) or hazard ratio (HR) >2.0 or <0.5) and/or significant (p<0.05). If no univariate OR or HR was provided, other univariate association measures, multivariate results or textual conclusions were used. Potential predictors that were studied multiple times and were consistently (≥75% of findings in same direction) associated with successful dose reduction or discontinuation of a biologic were regarded as predictors.
Results Out of 3029 non-duplicate articles, 16 articles on 15 cohorts were included. A total of 17/55 and 33/65 biomarkers were studied in more than one cohort for the prediction of successful dose reduction and successful discontinuation of a biologic. Three predictors were identified: higher adalimumab trough level for successful dose reduction; and lower Sharp/van der Heijde erosion score and shorter symptom duration at the start of a biologic for successful discontinuation. Noteworthy, all three biomarkers were only studied twice, meaning that more frequently investigated biomarkers yielded no consistent predictors. In addition, 13/38 and 8/32 biomarkers studied in only one cohort were classified as potential predictor for successful dose reduction and discontinuation, respectively.
Conclusions The predictive value of a wide variety of biomarkers for successful dose reduction or discontinuation of a biologic has been investigated, but we have found only three biomarkers that were predictive in just two studies. The strength of the evidence is limited by the low quality of included studies and the likelihood of publication bias and multiple testing. Therefore, validation of the found potential predictors in well-designed predictive studies is needed.
Disclosure of Interest None declared