Immune complexes contain immunoglobulin A rheumatoid factor in serum and synovial fluid of patients with polyarticular juvenile rheumatoid arthritis.

SIR, Immune complexes (IC) are believed to play a role in the pathogenesis of juvenile rheumatoid arthritis (JRA) w1x. Their prevalence varies from 39 to 79% depending on the method used for their detection w2x. Immunoglobulin M (IgM)and IgG-containing IC activate the classical complement pathway, whereas IgA-containing IC activate an alternative complement pathway. Several groups, including ours, have shown complement activation through either or both pathways in patients with JRA w3, 4x. IgA rheumatoid factor (RF) has rarely been studied in IC isolated from patients with JRA. It may thus be interesting to determine if the IC in patients with JRA contain IgA RF. Sera were obtained from 61 patients fulfilling the American College of Rheumatology criteria for JRA w5x. Synovial fluid specimens were available from seven of them. Sera from 25 healthy children were used as controls. Complement component C3 was measured by turbidimetery (Behring, Germany). Disease activity was defined based on previously described criteria w4x. Serum and synovial fluid specimens were tested for IC using a C3d-based enzyme-linked immunosorbent assay (ELISA) w6x, using the mean+ 2 S.D. (2.0+ 2 3 1.1 mguml) of the controls as the cut-off. From the specimens testing positive, IC were precipitated using polyethylene glycol (PEG) w7x and tested for IgA and IgM RF using ELISA w8, 9x. For the IgA RF assay, a known positive specimen from a patient with rheumatoid arthritis was used as the standard; absorbance readings for its doubling dilutions yielded a sigmoid curve and the concentration at the beginning of the upper plateau was assigned a value of 100 arbitrary units (AUuml) of IgA RF. For IgM RF, a WHO standard (WHO, Geneva) was used and concentrations were expressed as IUuml. Wilcoxon’s rank sum test, Fisher’s exact test and the x test were used for analysis. Of the 61 patients (42 male; median age 11 yr, median disease duration 3 yr) studied, 25 (15 male), 17 (14 male) and 19 (13 male) had polyarticular, pauciarticular and systemic onset types of JRA, respectively. Eight patients had RF detectable by latex agglutination. Serum C3 levels (mean" S.D.) were: polyarticular type (180.7" 79.3 mgudl), pauciarticular type (198.8" 88.0 mgudl) and systemic onset type (232" 110.4 mgudl). In 33 of the 61 (54%), the sera had detectable IC (control 3u25; P< 0.0001). The prevalence of IC in polyarticular, pauciarticular and systemic onset types was 64% (16u25), 59% (10u17) and 37% (7u19), respectively. The prevalence of IC was higher in patients with active disease (30u49 vs 3u12 in inactive disease; P< 0.05). Of the 33 patients with IC, IgA RF and IgM RF were detected in PEG-precipitated IC in nine and 12 patients, respectively. Four patients had both IgA RF and IgM RF. IgA RF was detected more often among patients with polyarticular onset type (8u16) than in those with pauciarticular onset (0u10; P< 0.001) and systemic onset (1u7; P= ns) types. The prevalence of IgM RF was similar in the three onset groups (7u16, 2u10 and 3u7, respectively). Of the seven synovial fluid specimens (all from patients with IC in the serum; with IgM RF and IgA RF in two and four patients, respectively), six contained IC; of these, none contained IgM RF, whereas four had IgA RF. Synovial fluid C3 levels were below 50% of the simultaneous serum C3 concentration (Table 1).