Background Several studies have shown that down-titration of infliximab is feasible in a part of rheumatoid arthritis (RA) patients with stable RA treatment and stable disease activity.1.2 However since down-titration can also cause flares predictors for successful down-titration are warranted. There is an association between infliximab serum trough levels and clinical effect, and therefore they might aid in predicting successful down-titration.3,4
Objectives To determine whether baseline infliximab serum trough levels can predict successful down-titration or discontinuation in RA patients long term treated with infliximab with stable low disease activity.
Methods RA patients treated with infliximab who had stable low disease activity defined as a DAS28 of <3.2 and stable RA treatment for more than 6 months were included in a prospective cohort study. Inclusion was from the 1st of January 2010 until 1st of April and patients were followed for a year. In all patients, infliximab was down-titrated with 25% of the original dose (3mg/kg) every 8-12 weeks without interval change until discontinuation or flare. At baseline infliximab serum trough level and anti-infliximab antibodies (AIA) were measured. Infliximab levels below 1.0 mg/l, between 1.0 and 5.0 and above 5.0 were categorized as low, normal and high levels. Sensitivity and specificity of low, normal and high levels of serum trough levels at baseline were calculated for the prediction of successful infliximab cessation, impossibility to stop treatment and successful partial down-titration respectively. ROC curves were generated for discontinuation and down-titration.
Results In 51 RA patients, with a mean baseline DAS28 of 2.5 (sd 0.7) and median disease duration of 12 years [9-18], 8 patients could stop infliximab, 22 had partial down titrated and the rest returned to the baseline dose after 1 year. Sensitivity and specificity of low trough levels for successful cessation was 50% (95% confidence interval(CI)=16-84)) and 56% (95%CI=40-71), for high levels and partial down titration 26% (95%CI=10-48) and 82% (95%CI=63-94) and for normal levels and remaining on the same dose 20% (95%CI=6-44) and 58% (95%CI=39-75) respectively. Sensitivity and specificity for AIA and discontinuation was 13 (95%CI=3-53) and 84 (95%CI=69-93). Area under the curves for infliximab level as predictor of successful discontinuation and down-titration are 0.44 and 0.65 respectively.
Conclusions In our cohort baseline infliximab serum trough levels were not useful as an aid in the clinical decision making for down-titration for prediction of successful down-titration or cessation of infliximab.
Van der Maas A, et al. Ann Rheum Dis 2012 Nov;71(11):1849-54
Tanaka Y, et al. Ann Rheum Dis 2010;69:1286-1291
Wolbink GJ, et al. Ann Rheum Dis 2005;64:704-707
Bartelds GM, et al. JAMA 2011, 305:1460-1468
Disclosure of Interest A. Van Der Maas: None Declared, B. van den Bemt: None Declared, F. van den Hoogen: None Declared, P. van Riel Grant/research support from: Merck, Pfizer, Abbott, BMS, Roche, A. den Broeder: None Declared