C1q-bearing immune complexes detected by a monoclonal antibody to human C1q in rheumatoid arthritis sera and synovial fluids

Summary Using a monoclonal antibody directed against the C-chain of human C1q, we detected C1q-bearing immune complexes (IC) in sera and synovial fluids of rheumatoid arthritis (RA) patients. In a sandwich-ELISA, C1q-bearing IC were captured by the solid-phase monoclonal antibody and then detected with peroxidase-labeled F(ab)₂-antibodies to either human IgG or IgM. The results of this assay were compared to an ELISA-modification of the C1q-solid-phase binding assay (C1q-SPBA). C1q-bearing IC were detected in 81.1% of RA-sera and the 65.2% of RA-synovial fluids. IgG as well as IgM was present in 72.6% of the sera and 70% of the synovial fluids which were positive in both assays. Most RA sera that were only positive for C1q-bearing IC, contained IgG alone (81.5%). The corresponding synovial fluids showed IgG alone (53%) or both IgG and IgM (41.1%). IgM alone (25%) could be detected in sera, e.g. in juvenile forms of RA. The levels of IC were higher in synovial fluid than in paired serum. In comparison to normal human serum (NHS) and patients with osteoarthritis, complement activity (CH50 titers) and C1q-values in patients with RA were frequently elevated. Since the formation of C1q-bearing IC is an indicator for the classical complement pathway activation, an assay with monoclonal anti-C1q antibody may be a useful tool in the diagnosis of rheumatoid diseases.     

Isotypes of rheumatoid factors in rheumatoid arthritis and chronic liver diseases

We studied isotype-specific rheumatoid factors (RFs) to clarify their significance in rheumatoid arthritis (RA) and to verify the difference in RF isotypes between RA and chronic liver diseases (CLD). Isotype-specific RFs in RA and in CLD were measured by enzyme-linked immunosorbent assay (ELISA). Most sera (n = 51, 94.1%) from RA patients contained some kind of RF isotypes (92.1% for IgM RF, 76.4% for IgG RF, and 43.1% for IgA RF), and seronegative RA by ELISA was seen in only 11.8% (n = 6). The most characteristic combination of RF isotypes in active RA was IgG, IgA, and IgM. This combination of RF isotypes changed to IgG plus IgM, according to the diminution of RA activity; then, we found only IgM RF in inactive RA. The titers of each RF isotype also decreased in parallel with the activity of RA. IgA RF seemed to be the most sensitive factor for evaluating the activity of RA. In CLD, almost the same high frequency (n = 49, 89.8% for IgM RF, 59.2% for IgG RF), with the same titer levels seen in RA, was observed. On the other hand, IgA RF was significantly lower in frequency (n = 9, 18.4%) and in titer, compared with the finding in RA. Surprisingly, even in CLD, true seronegativity by ELISA was also found in very few patients (n = 4, 8.1%). In CLD, positive RFs detected by agglutination assay were seen more often in chronic hepatitis than in liver cirrhosis. In RA patients, significant associations of IgA RF and the serum concentration of IgA, and IgG RF and the serum concentration of IgG, were observed. On the other hand, in CLD patients, significant associations of IgG RF and the serum IgG concentration, and of IgM RF and the serum IgM concentration, were observed. These results indicated that IgA RF in active RA is the most characteristic RF isotype distinguishing it from other nonrheumatic diseases, as well as from inactive RA. RF isotypes reflected the background polyclonal B-cell activation in different manners in both diseases. In CLD, RF isotypes seemed to be disease-related immunological disorders reflecting disease progression. Received: February 17, 2000 / Accepted: July 5, 2001     

Serum immune complexes containing IgA appear to predict erosive arthritis in a longitudinal study in rheumatoid arthritis.

Fifty seven patients with rheumatoid arthritis (RA) were studied longitudinally, and the presence of rheumatoid factor (RF) and various types of immune complexes (IC) was correlated with joint activity and the presence of extra-articular features (EAF). In a cross sectional study it was found that the levels of circulating IC and RF correlated significantly with joint disease activity and the presence of EAF. Longitudinally, levels of IC measured by the C1q binding activity and IC containing IgG and IgM correlated significantly with fluctuations in joint disease activity, whereas IC containing IgG and IgA correlated with the occurrence of EAF. RF and IC levels, however, did not predict the clinical course of the disease. IC containing C3 and C4 were found infrequently and were only present in patients with active rheumatoid vasculitis (RV). The continuous presence of these IC appeared to be linked to the recurrence of vasculitis, irrespective of treatment. Significantly more erosions of hands and feet were found after one year follow up in those RA patients who presented early (disease duration less than one year) who initially had a raised serum IgA IC level (r = 0.72; p less than 0.005).     

IgA containing immune complexes in rheumatoid vasculitis and in active rheumatoid disease

Serum and synovial fluid (SF) from 68 patients with rheumatoid arthritis (RA) were studied for the presence of immune complexes (IC) and the results correlated with extraarticular features and/or disease activity. IC were measured by the 125I Clq binding assay (ClqBA) and with one detecting IgG, IgA, C3 or C4 in IC. Disease activity correlated significantly with IgG or IgA containing and Clq binding IC. The IgA containing IC were found only in 25% of the patients, including all but one case of rheumatoid vasculitis, but otherwise only in seropositive active RA. C3 and C4 IC did not correlated with disease activity, seropositivity or vasculitis. IC in serum did not correlate with SF levels, but C4 containing IC were more frequent in SF (60%) than in serum (30%). Thus serum IC did not reflect SF levels. Patients with vasculitis showed more IC in the sera than in SF.     

Inhibition of C1q binding to antigen-antibody complexes by a factor in rheumatoid arthritis serum

Summary The sera and synovial fluids of patients with rheumatoid arthritis (RA) contain a factor which decreases the binding of C1q to antigen-antibody complex (IC). Several lines of evidence suggest that this factor is distinct from the documented C1q inhibitor which is a chondroitin sulphate. (1) It binds to IC rather than to C1q. (2) It is resistant to digestion with chondroitinase ABC. (3) The addition of chondroitin sulphate to serum does not inhibit the binding of IC to C1q. The observation that three purified IgM and IgG rheumatoid factors (RF) did not reduce C1q binding to IC indicates that the factor is not RF. The ability of RA sera to reduce IC binding to C1q was inversely correlated with their ability to prevent immune precipitation (PIP), and directly with levels of an inhibitor of PIP. These data suggest that the factor which binds to IC and reduces C1q binding may be responsible for the excessive immune precipitation which occurs in RA sera.     

QUANTITATIVE DETECTION OF CLASS-SPECIFIC RHEUMATOID FACTORS USING MOUSE MONOCLONAL ANTIBODIES AND THE BIOTIN/STREPTAVIDIN ENHANCEMENT SYSTEM

An enzyme-linked immunosorbent assay (ELISA) for the quantitativedetection of rheumatoid factors (RF) was developed using mousemonoclonal antibodies against human IgG, IgA and IgM togetherwith the biotin/streptavidin enhancement system. One hundredand eight patients with rheumatoid arthritis (RA) of whom 47had a positive serum latex agglutination assay, 100 healthycontrols and 95 diseased controls (25 systemic lupus erythematosus(SLE), 25 ankylosing spondylitis, 20 osteoarthritis and 25 bronchialasthma) were evaluated for the presence of IgG-, IgA- and IgM-RFby ELISA. Elevated levels of IgG-, IgA-, and IgM-RF could bedemonstrated in respectively 94%, 91% and 98% of serum samplesfrom RA patients with a positive latex test and in 72%, 69%and 8% of serum samples from RA patients with a negative latextest. In the latter patient group, increased levels of one ormore RF isotypes could be detected m 82% of the patients. Exceptfor the presence of IgG-RF in serum of 20% of the SLE patients,increased levels of RF isotypes were present in less than 5%of the diseased controls. Highly significant correlations werefound between serum IgM-RF levels as detected by ELISA and thoseof the latex and the Rose-Waaler agglutination assays. Positivecorrelations were also found between the serum levels of thethree RF isotypes investigated and between the levels of RFisotypes as measured in serum and synovial fluid of the samepatient. Compared to agglutination assays, the ELISA for thequantitative detection of RF isotypes is a more reproducibleand sensitive test which avoids some of the problems encounteredin earlier ELISA methods. KEY WORDS: Rheumatoid factor, ELISA, Rheumatoid arthritis     

Relation between bone erosions and rheumatoid factor isotypes.

Total rheumatoid factor (RF) activity and individual RF isotypes were measured in 62 patients with rheumatic diseases. Retrospective analysis of available x rays showed an association between IgA RF and the occurrence of periarticular bone erosions in hands. In contrast, IgG RF and IgM RF did not show any significant association with erosions. Furthermore, a close correlation was observed between the RF isotype levels in simultaneously drawn serum and synovial fluid samples. The possible significance of IgA RF in the pathogenesis of bone erosions is discussed.     

IgG and IgM rheumatoid factor synthesis in rheumatoid synovial membrane cell cultures

The detection of rheumatoid factors (RFs) in synovial membranes and fluids of patients with rheumatoid arthritis (RA) has suggested that local production of these antiimmunoglobulin autoantibodies may have a role in the pathogenesis of synovitis. To quantitate RF synthesis in the rheumatoid synovial membrane, 12 synovial specimens were obtained from patients with seropositive RA, 5 from patients with seronegative RA, and 6 from patients with other arthritides. Single cell suspensions were cultured, and supernatants were analyzed for IgG, IgM, IgG-RF, and IgM-RF by solid-phase radioimmunoassays. IgM-RF was detected in all of the 12 seropositive culture supernatants, and IgG-RF was detected in 8 of the 12. Addition of cycloheximide to the cultures resulted in a greater than or equal to 40% decreased in the amount of IgM-RF. A similar decrease in IgG-RF occurred in the 4 cultures in which the largest amounts of IgG-RF were detected. IgM-RF synthesis represented 7.3 +/- 0.7% (mean +/- SEM) of the total IgM produced, and IgG-RF represented 2.6 +/- 1.1% (mean +/- SEM) of the IgG synthesized in those cultures with detectable IgG-RF. Cultures of synovial membrane cells (SMC) from seronegative RA patients or patients with other arthritides did not contain detectable amounts of IgM-RF or IgG-RF. Selective synthesis of RF by seropositive synovium was suggested by the observation that the fractions of synthesized IgM with RF activity were greater in the SMC supernatants than in paired sera in all cases, and the fractions of IgG with RF activity were greater in the SMC supernatants of 3 of the 4 cases in which substantial amounts of IgG-RF were produced. Comparison of the percentages of newly synthesized IgM with RF activity in paired cultures of SMC and peripheral blood mononuclear cells similarly indicated selective synthesis of IgM-RF by the synovium. These results demonstrate active and selective synthesis of both IgG-RF and IgM-RF by seropositive SMC. However, RFs account for only a minor fraction of the total Ig produced.     

Intra-articular and circulating immune complexes and antiglobulins (IgG and IgM) in rheumatoid arthritis; correlation with clinical features.

Solid phase radioimmunometric methods have been used to assay immune complexes and IgG and IgM antiglobulins in paired samples of synovial fluid and serum from patients with rheumatoid arthritis (RA) or osteoarthrosis. Over 60% of RA patients had some increase in complexes in their sera, while nearly 90% had synovial fluid complexes. Moreover, the levels of complexes within the joint were much higher than in the serum. Both IgG and IgM antiglobulins were raised in most RA patients.The levels of IgG antiglobulins--and to a less extent IgM antiglobulins--were nearly always higher in synovial fluid than in the corresponding serum sample.A strong correlation was found between the levels of immune complex and IgG antiglobulin. A marked association was seen between the presence of subcutaneous nodules and increased IgG antiglobulins.     

Effects of gold therapy on the synthesis and quantity of serum and synovial fluid IgM, IgG, and IgA rheumatoid factors in rheumatoid arthritis patients

Eleven patients with active rheumatoid arthritis were monitored prospectively while receiving up to 1 gm of gold sodium thiomalate. There was a significant decrease in serum and synovial fluid IgG, IgA, and IgM rheumatoid factor (RF) levels over the period of study. Comparison of changes in serum RF and total immunoglobulin levels indicated a selective effect on RF production. These observations were supported by changes in the spontaneous in vitro production of IgM-RF and total IgM by peripheral blood mononuclear cells. Studies of synovial membrane synthesis showed a downward trend in immunoglobulin and RF production, but this did not reach statistical significance. A differential effect on the various RF classes was also noted. The most profound effect was on IgM-RF production; whereas, changes in IgG-RF production were least affected. These results suggest a selective and differential effect of gold salts on RF production.     

Anti-cyclic citrullinated peptide antibody isotypes in rheumatoid arthritis: association with disease duration, rheumatoid factor production and the presence of shared epitope

OBJECTIVE: Anti-cyclic citrullinated peptide (anti-CCP) antibodies of IgG isotype are specific diagnostic markers of rheumatoid arthritis (RA). Recent evidence also points to their direct involvement in the pathophysiology. Little information is available, however, regarding the isotype distribution of anti-CCP antibodies and the characteristics of IgA and IgM anti-CCP. METHODS: IgG, IgA and IgM anti-CCP2 and rheumatoid factor (RF) levels were measured in the sera of 119 RA patients and 118 controls, including patients with other rheumatic diseases and healthy subjects. We analyzed the diagnostic performance of IgA and IgM anti-CCP2 antibodies and their relationship with IgG anti-CCP2, RFs, disease duration and the presence of HLA-DRB1 shared epitope (SE) alleles. RESULTS: Patients with RA had significantly higher serum IgA and IgM anti-CCP2 antibody levels than healthy subjects and patients with other rheumatic diseases (p<0.0001). IgG, IgA and IgM anti-CCP2 antibodies were present in 74.8%, 52.9% and 44.5% of RA patients, and their diagnostic specificity was 95.8%, 95.8% and 91.6%, respectively. The presence of anti-CCP2 antibodies was significantly associated with SE alleles (p=0.03). The frequency of IgM anti-CCP2 positivity was lower in longstanding disease compared to early RA (p=0.03). CONCLUSION: IgA and IgM anti-CCP2 antibodies are present in RA patients, and they are similarly specific for RA as IgG anti-CCP2. The higher frequency of IgM anti-CCP2 antibodies in early RA suggests that they are mostly generated during the first phase of immune response; nonetheless, their production seems to be sustained in some patients. Further analysis of IgM and IgA anti-CCP2 antibodies may provide insights into the pathogenesis of RA.     

Serum and synovial fluid IgG,IgA and IgM antigammaglobulins in rheumatoid arthritis

Antigammaglobulins of IgG, IgA and IgM classes were measured in normal individuals and in patients with osteoarthritis or rheumatoid arthritis. Serum IgG and IgA and synovial fluid IgG antigammaglobulin levels were significantly higher in patients with rheumatoid arthritis than in other individuals, with highest levels occurring in patients with positive latex fixation tests. IgM antigammaglobulins were elevated only in patients with latex positive rheumatoid arthritis. Increased serum levels of IgG, IgA and IgM antigammaglobulins were each associated with clinical findings of severe rheumatoid arthritis. Increased levels of serum and synovial fluid IgG and IgM antigammaglobulins were each associated with diminished serum and synovial fluid complement levels.     

Anti-IL-8 autoantibodies and complexes in rheumatoid arthritis: polyclonal activation in chronic synovial tissue inflammation

The chemokine interleukin-8 (IL-8) is frequently associated with inflammatory diseases, and autoantibodies against IL-8 are present in the periphery at elevated levels in such conditions as rheumatoid arthritis (RA). Circulating free anti-IL-8 IgG autoantibodies correlate with inflammatory parameters and disease severity in RA. In this study, correlations were sought between these disease parameters and other antibody subclasses. We assayed IgM, IgA and IgG anti-IL-8 antibodies and IL-8 immunoglobulin immune complexes in the serum of 29 healthy controls and 56 patients with defined RA, and compared the results with clinical and humoral disease parameters. IgG and IgM antibodies directed against IL-8 were present in all samples. In the disease groups, all isotypes of free anti-IL-8 antibodies correlated with increasing humoral disease parameters like CRP and CIC and their related anti-IL-8 immune complexes. Samples which contained high titers of anti-IL-8 antibody subclasses and complexes were RF subclass-positive, while IgM RF-negative sera showed low levels of anti-IL-8 and complexes. Detectable levels of IgG and IgA RF were found in all sera. Patients with extra-articular organ manifestation showed significantly increased free IgA and IgA/IL-8 complexes, with no correlation to the IgA RF titer or IgA hypergamma-globulinemia. The highest titers were seen in two RA cases with vasculitis and in one patient with colitis. Polyclonal activation of the humoral antibody system, which normally precedes the resolution of an inflammatory response, can itself lead to secondary stimulation of inflammatory processes via immune complex formation. In the immune pathology of RA, it degenerates into a persistent chronic inflammation accompanied by progressive joint destruction. The presence of elevated IgA subclass anti-IL-8 autoantibodies in RA patients with extra-articular manifestations suggests these autoantibodies as a clinically useful marker of disease severity and extra-articular manifestations. Received: 10 June 1998 / Accepted: 5 October 1998     

Correlation between rheumatoid factor and IL-6 activity in synovial fluids from patients with rheumatoid arthritis

Synovial fluids (SF) and sera (S) from patients with rheumatoid arthritis (RA) were examined for IgM, IgM-rheumatoid factor (IgM-RF), albumin and interleukin-6 (IL-6) activity. The quotient of SF/S IgM-RF was elevated compared with that of SF/S albumin in 7 patients with seropositive RA, although the quotient of SF/S IgM was lower than that of SF/S albumin. SF IL-6 activity was much higher than serum IL-6 activity in all the 7 RA patients. In synovial fluids from 22 seropositive RA patients, SF IL-6 activity was significantly correlated with the SF IgM-RF, IgG-RF and IgA- less than RF, but not with SF IgM, IgG or IgA. Moreover, SF IgM-RF as well as SF IL-6 activity was significantly correlated with the Westergren erythrocyte sedimentation rate (ESR) or the Lansbury articular index. These results indicate that IL-6 and RF might be produced within the rheumatoid joints as a result of abnormal immune system activation, which is associated with the disease activity of RA. Three of the 4 seronegative RA patients, however, showed high SF IL-6 without detectable levels of SF IgM-RF, indicating that IL-6 alone is not sufficient for IgM-RF production.     

Absence of cyclic citrullinated peptide antibody in nonarthritic patients with chronic hepatitis C infection

OBJECTIVE: The increased prevalence of rheumatoid factor (RF) in patients with chronic hepatitis C virus (HCV) infection markedly diminishes the diagnostic specificity of serum rheumatoid factor (RF) for rheumatoid arthritis (RA) in patients with HCV. Cyclic citrullinated peptide (CCP) antibody, a highly specific biomarker for RA in the general population, may have better diagnostic utility for RA in the HCV population. To investigate if CCP antibody retains its specificity for RA in HCV infection, we determined the prevalence of CCP antibodies and examined the relationship between RF production and CCP antibody levels in a population of nonarthritic patients with chronic HCV infection. METHODS: CCP antibody and IgM, IgG, and IgA RF isotypes were determined by ELISA in serum from nonarthritic patients with chronic HCV infection. RESULTS: In a series of 50 HCV patients, IgG-RF, IgM-RF, and IgA-RF were detectable in 52%, 26%, and 14%, respectively, with a total seropositivity rate of 54%. Marginally elevated CCP antibody was detected in a single patient (2%). By regression analysis, serum levels of CCP antibodies did not correlate with RF levels. CONCLUSION: In contrast to RF, CCP antibody is not increased in HCV infection. CCP antibody may have improved utility for the diagnosis of RA in this patient population.     

Elevated Levels of Immunoglobulins and Immune Complexes in the Bile of Patients with Primary Sclerosing Cholangitis

The concentration of immunoglobulins and immune complexes was determined in the serum and bile of four patients with primary sclerosing cholangitis, only one of whom had chronic ulcerative colitis, and in four postcholecystectomy control patients. The result of the immunoglobulin studies demonstrated increased serum levels of IgM in primary sclerosing cholangitis as compared to control patients. The IgG, IgA, and IgM in bile were all significantly elevated as compared to controls (p = 0.02). Immune complexes in bile and serum were also determined in both groups by the C1q and conglutinin binding assays. Immune complexes in bile were elevated in three of four of the patients with primary sclerosing cholangitis. These findings further characterize primary sclerosing cholangitis in terms of biliary immunoglobulins and immune complexes but do not resolve whether these changes are an epiphenomenon or reflect a primary role of immunoglobulins and immune complexes in the pathogenesis of this disorder.     

Prospective study of early rheumatoid arthritis. I. Prognostic value of IgA rheumatoid factor.

Thirty-three patients with early arthritis, 28 of whom developed classical/definite rheumatoid arthritis (RA), were followed up for two to four years. Rheumatoid factor (RF) levels of the IgM, IgA, and IgG isotypes were measured in serum and synovial fluid by an ELISA technique developed in our laboratory. All seven patients who presented with raised IgA RF developed erosions of their hands and wrists. This was significantly different from the remaining 26. By contrast none of the five patients who presented with isolated elevation of IgM RF developed erosive disease. The patients with raised IgA RF needed significantly more treatment with 'specific' drugs than the remaining 26. It is suggested that the detection of IgA RF in early RA indicates poor prognosis, justifying a more aggressive treatment at an early stage.     

Isolation and analysis of complement activating aggregates from synovial fluid of patients with rheumatoid arthritis using monoclonal anti-C3d antibodies.

The complement activating aggregates in synovial fluids of patients with rheumatoid arthritis (RA) have been isolated using monoclonal IgM anti-C3d antibodies attached to solid phases, and the content of the material bound has been analysed. High levels of aggregated IgG bearing C3d were found in RA synovial fluids, and IgG was the major immunoglobulin bound from such synovial fluids by anti-C3d Sepharose. A strong correlation was shown between levels of aggregated IgG bearing C3d and complement activation, as judged by C3d levels. Significant (but less strong) relationships were also observed between C3d levels and both complement consuming and C1q binding activity. C3d levels and levels of aggregated IgG bearing C3d were both significantly associated with the numbers of polymorphonuclear leucocytes (PMNs) found in RA synovial fluids. From these results it is concluded that the aggregated immunoglobulins bearing C3d (particularly IgG) isolated from RA synovial fluids are responsible for activating complement and attracting PMNs into the joint space. Radioimmunoassay showed no correlation, however, between levels of aggregated IgG (or IgM) bearing C3d and rheumatoid factor (RF) activity bound by anti-C3d. In addition, the material bound by anti-C3d Sepharose from most synovial fluid polyethylene glycol precipitates did not contain either IgM or IgG RF. Thus both techniques show that the majority of complexes bearing C3d do not contain RF. As the complement fixing aggregates apparently contain only immunoglobulin and complement components the results raise the problem of how the aggregates are formed. It is suggested that RA IgG may remain aggregated after either antigen or antibody (RF) has dissociated from the complex.     

Effect of circulating immune complexes on the binding of rheumatoid factor to histones

OBJECTIVE: To determine whether the reaction of rheumatoid factor (RF) with solid phase histone is due to the simultaneous presence of circulating immune complexes (CICs) or aggregated IgG. METHODS: Serum samples from 56 patients with seropositive rheumatoid arthritis (RA) and 50 random blood bank donors were used. Binding of immunoglobulins to histone was determined by enzyme linked immunosorbent assay (ELISA) and by western blots. Aggregated IgG was obtained by heating at 61(o)C for 30 minutes. RESULTS: Among the RA sera tested by ELISA, 54% were positive for histone binding by IgM, IgG, or IgA and 20% by IgM only. Heating of normal sera caused a significant enhancement in the binding of IgG to histone (p<0.001). This binding had a non-cognate behaviour-that is, it was destroyed by pepsin treatment of serum and was not significantly inhibited by competition with free histone. The same behaviour was seen for IgM, IgG, and IgA binding from RA sera. However, cognate IgG antibody binding to histone was inhibited by free histone and was resistant to pepsin digestion. Addition of heat aggregated IgG to RA sera or pretreatment of histone with aggregated IgG caused a significant increase in IgM binding to histone. CONCLUSION: IgM, IgG, and IgA RF bind to solid phase histone as a result of attachment to histone of immune complexes or aggregated IgG and not as a result of a cognate reaction with histone.     

Population study of the importance of rheumatoid factor isotypes in adults.

Blood samples collected from 13,858 randomly selected subjects participating in a health survey in Iceland from 1974 to 1983 were tested for rheumatoid factor. Samples that were positive in a sensitive RF screening test were analysed further by the Rose-Waaler technique and an isotype specific enzyme linked immunosorbent assay (ELISA). In 1987 the 173 available participants who were RF positive and 156 matched RF negative controls were evaluated clinically for rheumatoid diseases. RF levels and isotype patterns were more persistent in the patients with rheumatoid arthritis (RA) than in RF positive subjects who did not have overt RA. The prevalence of RA was only 19% in the participants who were RF positive in 1987. Forty per cent of the participants who had a persistent (four to 13 years) increase of IgA RF combined with either IgM or IgG RF were diagnosed as having RA. A positive correlation was found between RF levels and various manifestations of RA. This association was stronger for the IgA and IgG RF isotypes than for IgM RF. Excluding RF positivity as a diagnostic parameter, RA was diagnosed in 33 of the participants and 20 (61%) of these patients had increased levels of IgM and IgA RF. Patients with RA with bone erosions in their hands had higher levels of IgA RF than patients without erosions, but an association was not found between bone erosions and other RF isotypes. None of the RF negative participants who were symptom free when the original blood sample was taken developed RA during the four to 13 year follow up period. In contrast, five symptom free RF positive participants developed RA during this period. These five patients had all had increased levels of at least two RF isotypes before the onset of their symptoms. It is concluded that the IgA and IgG RF isotypes have a closer association with the clinical parameters of RA than IgM RF. Furthermore, increases in RF can precede clinical manifestations of RA and this applies in particular to the IgA and IgG RF isotypes.