Preferential decrease in IgG4 anti-citrullinated protein antibodies during treatment with tumour necrosis factor blocking agents in patients with rheumatoid arthritis W H Bos, 1,2 G M Bartelds, 2 M Vis, 3 A R van der Horst, 4 G J Wolbink, 1,2 R J van de Stadt, 2 D van Schaardenburg, 2,3 B A C Dijkmans, 2,3 W F Lems, 3 M T Nurmohamed, 2,3 L Aarden, 1 D Hamann 4 c Additional figures are published online only at http:// ard.bmj.com/content/vol68/ issue4 1 Sanquin Research, Amsterdam, The Netherlands; 2 Jan van Breemen Institute, Amsterdam, The Netherlands; 3 VU University Medical Center, Department of Rheumatology, Amsterdam, The Netherlands; 4 Sanquin Diagnostic Services, Amsterdam, The Netherlands Correspondence to: Dr W H Bos, Department of Immunopathology, Sanquin Research, PO Box 9190, 1006 AD Amsterdam, The Netherlands; w.bos@sanquin.nl Accepted 24 April 2008 Published Online First 29 April 2008 ABSTRACT Objective: To investigate the dynamics of IgG1 and IgG4 anti-citrullinated protein antibody (ACPA) subclasses during anti-tumour necrosis factor (TNF) treatment in patients with rheumatoid arthritis (RA). Methods: IgG, IgG1 and IgG4 ACPA levels were determined by ELISA on anti-citrullinated fibrinogen (ACF) and IgG1 : IgG4 ACPA ratios were calculated. A pilot study was performed in 28 ACF-positive patients treated with infliximab for one year. Confirmation of the results was obtained using a cohort of 180 consecutive patients treated with adalimumab for 28 weeks. Results: The median reduction in ACF levels was 31% for total IgG, 29% for IgG1, 40% for IgG4 and 22% for the IgG4 : IgG1 ACF ratio in the infliximab cohort. In adalimumab-treated patients, ACF levels declined 14% for total IgG and IgG1, and 36% for IgG4 ACF; the IgG4 : IgG1 ratio was reduced by 24% (all percentage values p,0.05). The decrease in antibody levels was correlated with the clinical response; European League Against Rheumatism good responders had the greatest decline in antibody levels and this effect was most pronounced for IgG4 (48% reduction). The IgG4 : IgG1 ACF ratio preferentially decreased in patients with adequate therapeutic adalimumab levels. Conclusion: ACPA subclass distribution is modulated by effective anti-inflammatory treatment. The preferential decline of IgG4 ACPA, reflected by the decreased IgG4 : IgG1 ratio, suggests a beneficial effect of anti-TNF treatment on chronic antigenic stimulation by citrullinated proteins. This effect may be directly anti-TNF mediated or the result of effective dampening of the inflammation in the rheumatoid joint. Rheumatoid arthritis (RA) is a chronic inflamma- tory disease, which may lead to joint destruction. 1 One of the characteristics of the disease is the presence of autoantibodies. Anti-citrullinated pro- tein antibodies (ACPA) comprise a group of antibodies highly specific for RA; among those described are antibodies against cyclic citrullinated peptide, 2 citrullinated fibrinogen, 3 citrullinated alpha-enolase 4 and mutated citrullinated vimen- tin. 5 They share a similar high sensitivity and specificity for RA and are present in early and even preclinical disease. 67 A pathophysiological role for ACPA in RA has been suggested, 8 and indeed, in a serum transfer model of collagen-induced arthritis, ACPA have been shown to enhance arthritis. 9 Two papers have recently reported on ACPA IgG subclass distribution, with similar results. 10 11 Both show IgG1 as the main IgG subclass, as expected in a T-helper cell type 1-driven disease. Unexpectedly, IgG4 anti-citrullinated fibrinogen (ACF) and anti- cyclic citrullinated peptide antibodies were the second most frequent IgG subclass. In the latter paper, differences in isotype usage have been implied in the transition of undifferentiated arthritis to RA. 11 One explanation for the high frequency of IgG4 ACPA in RA might be that during prolonged antigenic stimulation a shift in the IgG4 : IgG1 antibody ratio occurs that finally results in an IgG4-dominated response. 12–16 The introduction of anti-TNF agents has revo- lutionised RA treatment, effectively dampening inflammation in the rheumatic joint. 17 Despite its proposed pathophysiological role in RA, data on the effect of anti-TNF treatment on ACPA levels in RA are controversial, because most studies reported modest or no effect of anti-TNF treatment on IgG ACPA levels, 18–26 and data on the dynamics of IgG4 and IgG1 ACPA subclasses are lacking. If the presence of IgG4, in the context of IgG1, is a measure of chronic antigenic stimulation, as proposed by Aalberse et al, 13 a reduction of chronic antigenic stimulation may lead to a preferential decrease of IgG4 ACPA. As both IgG1 and IgG4 ACPA levels might be affected by anti-TNF treatment, the relative contribution of both sub- classes in an individual patient can best be studied by calculating the IgG4 : IgG1 ratio. The aim of the present study was to investigate the dynamics of predominant ACPA subclasses during anti-TNF treatment and to relate the changes in ACPA levels to treatment response. METHODS Patients Two prospective observational study cohorts were used in the present study. The first cohort consisted of consecutive RA patients treated with infliximab for at least one year at Slotervaart Hospital, Amsterdam, The Netherlands. Twenty- eight of 51 ACF-positive patients described pre- viously had serum available for further analyses (nine non-responders, 15 moderate responders and four good responders after 46 weeks of treat- ment). 26 Serum was collected on the morning Extended report 558 Ann Rheum Dis 2009;68:558–563. doi:10.1136/ard.2008.088401