ANCA and anti-GBM antibodies in RPGN

ANCA and anti-GBM antibodies can occur together in RPGN. These patients are rare and have two distinct populations of antibodies and the ANCA specificity tends to be for myeloperoxidase more often than expected.

Further characterization of an animal model of systemic vasculitis

The demonstration of vasculitis and anti-myeloperoxidase antibodies in the mercuric chloride treated Brown Norway rat provides a useful, though limited, animal model of systemic vasculitis. We describe some preliminary experiments on the effect of transfer of serum from mercuric chloride treated rats and of two forms of immunotherapy: intravenous immunoglobulin and an anti-CD4 antibody. Transfer of serum did not lead to tissue injury and neither of the two forms of therapy proved beneficial in this model.

Pooled intravenous immunoglobulin in the management of systemic vasculitis

Investigation into the therapeutic use of intravenous immunoglobulin (IVIg) in systemic vasculitis was prompted by the detection of anti-idiotype antibodies reactive with ANCA in IVIg and the proven ability of IVIg to reduce the incidence of coronary artery aneurysms in Kawasaki disease. The efficacy and safety of IVIg (Sandoglobulin) was assessed in an open study of 26 patients with active systemic vasculitis. Eight weeks after IVIg 13 patients were in full and 13 in partial remission, clinical benefit was maintained in 18 twelve months later and was reflected by changes in C-reactive protein and ANCA.

Lactoferrin co-purifies with myeloperoxidase and is recognised by anti-neutrophil cytoplasm antibodies

Thirty-five sera with binding greater than 20% in a myeloperoxidase (MPO, Calbiochem) ELISA were tested in Western blot analyses. 5/35 blotted MPO, but 5/35 blotted lactoferrin (LF) contaminating the commercial MPO preparation. All the anti-LF positive sera, but none of the anti-MPO positive sera, also exhibited anti-nuclear binding on Hep2 cells. Three of the patients with anti-LF antibodies had vasculitis affecting areas additional to the pulmonary-renal involvement which characterised the patients with anti-MPO antibodies.

Activation of normal neutrophils by anti-neutrophil cytoplasm antibodies

Anti-neutrophil cytoplasm antibodies (ANCA) are markers of systemic vasculitis for which a pathogenetic role has been postulated. We have examined the effect of these autoantibodies on the function of normal human neutrophils in vitro. In the presence of ANCA positive sera luminol-amplified chemiluminescence was significantly increased compared to the values seen in the presence of normal or anti-double stranded DNA positive sera (P < 0.01). Five of six ANCA positive F(ab)2 preparations also produced significant neutrophil activation as demonstrated by the chemiluminescence response. This response was totally abrogated by the addition of neutrophil cytoplasm extract, containing the ANCA antigen. Addition of inhibitors to the chemiluminescence system demonstrated that the chemiluminescence response was inhibited by azide and salicylhydroxamic acid and reduced by histidine, suggesting that the chemiluminescence response was due to activation of myeloperoxidase, with generation of singlet oxygen. The chemotactic response to f-Met-Leu-Phe, a bacterial chemotactic peptide, was significantly augmented in the presence of ANCA. Chemotaxis to zymosan-activated serum and chemokinesis was not affected. Phagocytosis was also unaffected. We propose that neutrophil activation and modulation of neutrophil migration by ANCA may be of pathogenetic significance in systemic vasculitis.

Spectrum of autoimmune responses in systemic vasculitis

Growing evidence supports the inclusion of primary forms of systemic vasculitis within the category of autoimmune diseases. Thus their association with a family of autoantibodies, related to their specificity or neutrophil cytoplasm antigens, has been identified and the role these autoantibodies play in pathogenesis is now being explored. Different members of the autoantibody family have now been characterised in terms of their molecular specificity and the isotype of immunoglobulin of which they are composed. These properties may be a useful adjunct in the classification of systemic vasculitis (see table I). It is the purpose of this review to put a perspective on the specificity and pathogenicity of autoantibodies found in patients with systemic vasculitis and to outline new strategies for the treatment of these conditions, based on the involvement of autoimmune mechanisms in the development of these disorders.

Association of IgM with IgG ANCA in patients presenting with pulmonary hemorrhage

ANCA are markers for systemic vasculitis such as Wegener's granulomatosis (WG) and microscopic polyarteritis (MPA) and are usually of the IgG isotype. However, IgM ANCA may rarely occur in isolation, and in these circumstances, we have found that they are associated clinically with a syndrome of pulmonary hemorrhage and systemic vasculitis. How frequently IgM ANCA may occur in conjunction with IgG has not previously been investigated. We report here a study of 24 consecutive patients with IgG ANCA-positive systemic vasculitis (14 WG, 10 MPA) in whom we determined whether IgM ANCA occurred in association with IgG ANCA, and if so, whether this had clinical importance. Eight patients were found to have IgM ANCA as well as IgG ANCA, and of these, seven presented with severe pulmonary hemorrhage. None of the IgM ANCA-negative patients presented with pulmonary hemorrhage. Although the occurrence of pulmonary hemorrhage in ANCA positive vasculitis was closely correlated with the presence of IgM ANCA, the antigen specificity of these IgM autoantibodies was variable, since both myeloperoxidase (4 patients), PR3 (3 patients), and an unknown ANCA antigen (1 patient) were found to be targets. We conclude that knowledge of ANCA isotype may have important clinical and therapeutic implications for the management of patients with systemic vasculitis.

T and B cell responses to neutrophil cytoplasmic antigens in systemic vasculitis

The systemic vasculitides (SV) are characterized by the presence of autoantibodies to neutrophil cytoplasmic antigens (ANCA). The role of T cells in SV is uncertain. We studied human and murine T cell responses to human neutrophil cytoplasmic antigens in vitro. T cells from mice immunized with the neutrophil extract showed dose-dependent antigen-specific proliferation, restricted by the MHC class II E molecule. Myeloperoxidase (MPO) was not an important target antigen for murine T cells. Peripheral blood lymphocytes (PBLs) were obtained from 36 patients with SV, 31 before the start of immunosuppressive therapy, and from 11 healthy controls. T cell responses to the neutrophil extract in vitro did not differ between patients and controls: there were only low levels of antigen-specific proliferation, and this could not be amplified by in vitro selection. In 3 patients and 2 normals, PBLs were also tested after the depletion of CD8+ cells; this did not unmask T cell reactivity to neutrophil extract. The lack of demonstrable T cell reactivity to this antigen preparation may indicate that T cells do not play an important effector role in these diseases. A solid-phase spot ELISA was adapted to demonstrate autoantibody-producing B cells in vitro. Low numbers of ANCA-producing B cells could be demonstrated in the majority of patients. B cells producing antibody to MPO could be demonstrated in most patients and in three laboratory staff, but not in normals from outside the laboratory. In 2 patients, sequential B cell spot ELISAs were performed during the introduction of therapy, and autoantibody-producing B cells rapidly decreased in number. This assay may therefore be useful in monitoring the effects of treatment at the cellular level.

IgG subclass distribution and relative functional affinity of anti-myeloperoxidase antibodies in systemic vasculitis at presentation and during follow-up

Circulating IgG autoantibodies to myeloperoxidase (MPO) are associated with renal vasculitis and have been implicated in its pathogenesis. However, raised levels of these autoantibodies may persist during clinical remission. We tested whether this paradox could be explained by immunoglobulin subclass switching during disease evolution, since different subclasses have different immunological and biochemical properties. Sera with anti-myeloperoxidase (anti-MPO) activity from 33 patients with active disease and 20 anti-MPO positive follow-up sera were studied by an ELISA using a panel of anti-human IgG subclass monoclonal reagents previously calibrated on human myeloma proteins. Anti-MPO subclass distribution in initial sera was: IgG1, 31 (94%); IgG2, 10 (30%); IgG3, 24 (73%); and IgG4, 22 (67%). IgG3 anti-MPO decreased during follow-up (P less than 0.02), with no change in IgG1 and IgG4. Relative functional affinity of anti-MPO antibodies in purified IgG subclasses was studied by the diethylamine method. IgG3 fractions consistently had a greater affinity for MPO than the other subclasses. Sequential studies in four patients demonstrated an affinity maturation for IgG1 and IgG4 anti-MPO as IgG3 anti-MPO disappeared. We conclude that dynamic changes of subclass distribution and affinity may explain discrepancies between anti-MPO antibody titre and disease expression.

Plasma exchange in focal necrotizing glomerulonephritis without anti-GBM antibodies

To determine whether plasma exchange was of additional benefit in patients treated with oral immunosuppressive drugs for focal necrotizing glomerulonephritis (without anti-GBM antibodies), we performed a randomized controlled trial with stratification for renal function on entry. Forty-eight cases were analyzed, 25 in the treatment group (plasma exchange, prednisolone, cyclophosphamide and azathioprine) and 23 in the control group (drug therapy only). There was no difference in outcome in patients presenting with serum creatinine less than 500 mumol/liter (N = 17), or greater than 500 mumol/liter but not on dialysis (N = 12), all but one of whom had improved by four weeks. However, patients who were initially dialysis-dependent (N = 19) were more likely to have recovered renal function (P = 0.041) if treated with plasma exchange as well as drugs (10 of 11) rather than with drugs alone (3 of 8). Long-term follow-up showed that improvement in renal function was generally maintained. The results of this trial confirm that focal necrotizing glomerulonephritis related to systemic vasculitis responds well to immunosuppressive drugs when treatment is started early, and suggest that plasma exchange is of additional benefit in dialysis-dependent cases.

The effect of anti-neutrophil cytoplasm autoantibodies on the signal transduction in human neutrophils

The effect of anti-neutrophil cytoplasm autoantibodies (ANCA) on neutrophil activation was studied by pre-incubating neutrophils with IgG and F(ab')2 prepared from ANCA patients with Wegener's granulomatosis or microscopic polyarteritis. We measured the generation of inositol triphosphate (IP3) (a product of hydrolysis of membrane phospholipid which acts as a second intracellular messenger), and the translocation of protein kinase C (PKC) upon stimulation by a chemotactic peptide, fMet-Leu-Phe (fMLP). ANCA+ F(ab')2 did not induce a significant increase in IP3 generation. Nonetheless, ANCA+ F(ab')2 and ANCA+ IgG pretreatment of human neutrophils reduced the production of inositol phosphates upon subsequent fMLP stimulation compared with experiments performed when cells were pretreated with F(ab')2 and IgG prepared from ANCA- healthy subjects. A significantly reduced generation of IP3 and inositol biphosphate (IP2) was observed. ANCA+ F(ab')2 pretreatment of neutrophils inhibited fMLP-stimulated IP3 generation in a dose-dependent manner. The membrane-bound PKC activity upon stimulation by FMLP and PMA was reduced in neutrophils pretreated with ANCA+ F(ab')2 and IgG. These results indicate that ANCA affect in vitro signal transduction (IP3 generation, and translocation of PKC) in human neutrophils. Apparently, further activation of signal transduction by chemotactic peptide is significantly blunted in cells pre-incubated with ANCA+ F(ab')2 but not with F(ab')2 from healthy controls.

Immunoregulation of mercuric chloride-induced autoimmunity in Brown Norway rats: a role for CD8+ T cells revealed by in vivo depletion studies

Mercuric chloride (HgCl2) induces the production of autoantibodies to glomerular basement membrane (GBM) in the Brown Norway (BN) rat. The autoimmune response is self-limiting and thereafter the animals are resistant to rechallenge with HgCl2. Resistance can be transferred to naive animals by spleen cells from HgCl2-treated rats. A similar state of resistance can be induced with a low dose of HgCl2, insufficient in itself to induce autoimmunity. We have examined the role of CD8+ T cells in the immunoregulation of this experimental model by depleting this subset in vivo. We have also used inhibition studies in a solid-phase radioimmunoassay in an attempt to demonstrate any effect of anti-idiotypic antibodies in the spontaneous resolution of the anti-GBM antibody response. The initial induction and spontaneous resolution of anti-GBM antibodies were unaffected by depletion of CD8+ T cells. However, CD8-depleted animals were no longer resistant to rechallenge with HgCl2. Cell transfer studies showed that spleen cells from CD8-depleted animals conferred less resistance to HgCl2 than those from animals which had received control antibody. CD8 depletion also reduced the resistance induced by pretreatment with low-dose HgCl2. Studies in which peak sera were pre-incubated with post-recovery sera before testing in a solid-phase anti-GBM radioimmunoassay did not support an important role for anti-idiotypic antibodies. We conclude that CD8+ T cells play an important role in the resistance to rechallenge with HgCl2 in the BN rat, although they are not required for the induction or spontaneous resolution of the initial autoimmune response. Demonstration of the reversal of a suppressive phenomenon in vivo using an anti-CD8 monoclonal antibody is unusual.

Antineutrophil cytoplasmic autoantibodies: the nephrologist's perspective

Serologic tests for antineutrophil cytoplasmic autoantibodies (ANCA) provide diagnostic and prognostic information in patients with nephritis and systemic vasculitis. Four case histories are discussed that illustrate ANCA positivity in patients with renal impairment caused by interstitial, rather than glomerular, nephritis. Renal biopsy should continue to be used to diagnose the cause of renal impairment in ANCA-positive patients.

Vascular damage in Wegener's granulomatosis and microscopic polyarteritis: presence of anti-endothelial cell antibodies and their relation to anti-neutrophil cytoplasm antibodies

To define mechanisms of vascular injury in Wegener's granulomatosis and microscopic polyarteritis, anti-endothelial cell antibodies (AECA) were sought in serum from 168 patients, all of whom had anti-neutrophil cytoplasm antibodies (ANCA) detectable by indirect immunofluorescence. Using an ELISA with human umbilical vein endothelial cells (HUVEC), IgG AECA were demonstrated in 59% and IgM AECA in 68% of patients. Pretreatment of HUVEC with tumour necrosis factor (TNF), IL-1 or interferon-gamma (IFN-gamma) led to increased binding. Adsorption of AECA/ANCA-containing serum with HUVEC or neutrophils demonstrated that AECA and ANCA recognized different targets. von Willebrand factor (vWf) antigen levels in the patient samples were markedly elevated, with a mean of 3.10 +/- 1.89 U/ml (control population mean 1.04 +/- 0.36 U/ml), suggesting widespread endothelial cell damage. Studies using an 111In-labelled HUVEC release assay with 29 AECA-containing sera did not demonstrate complement-mediated cytotoxicity, even following activation of HUVEC with TNF. Four out of 16 AECA-containing sera tested showed antibody-dependent cellular cytotoxicity with unfractionated peripheral blood mononuclear cells. These data suggest that patients with Wegener's granulomatosis or microscopic polyarteritis can develop AECA to constitutively expressed but cytokine modulated determinants on HUVEC. These antibodies do not appear to support complement-mediated cytotoxicity, but a proportion can support antibody-dependent cellular cytotoxicity, suggesting that they may contribute to vascular injury.

IgG subclass distribution of autoantibodies to neutrophil cytoplasmic antigens in systemic vasculitis

The IgG subclass distribution for both ANCA-specific, and total IgG subclasses, of sera from 48 patients with active systemic vasculitis was determined by solid-phase assay using monoclonal anti-human subclass reagents. To control for varying performance of the subclass reagents, the assays were calibrated with sera affinity-depleted into subclass specific fractions, whose ANCA IgG subclass distribution was determined by a common reagent. In 10 patients whose ANCA persisted during clinical remission, the ANCA IgG subclass distributions between active and remission phases of disease were compared. The total IgG subclass distribution of sera from patients with active vasculitis was similar to that found in a normal population. The ANCA IgG subclass distribution of the same sera showed relative enrichment for IgG3 and depletion of IgG2 (P less than 0.05). When the active and clinical remission ANCA IgG subclass distributions were compared, ANCA IgG3 levels had fallen and ANCA IgG2 levels were increased during remission (P less than 0.01).

Treatment of systemic vasculitis with pooled intravenous immunoglobulin

The therapeutic effect of a course of high-dose, pooled, intravenous immunoglobulin (IVIg) on disease activity and circulating antineutrophil cytoplasm antibodies (ANCA) was investigated in 7 patients with systemic vasculitis. 5 had active disease despite conventional immunosuppression, and 2 had not received any treatment. All 7 had clinical improvement, which was sustained in 6 and transient in 1. The fall in ANCA concentrations to a mean of 51% of the pre-treatment values was maintained during follow-up. C-reactive protein concentration also dropped considerably. IVIg seemed to confer a useful therapeutic effect without adverse reaction.

Anti-idiotypes against anti-neutrophil cytoplasmic antigen autoantibodies in normal human polyspecific IgG for therapeutic use and in the remission sera of patients with systemic vasculitis

Anti-neutrophil cytoplasmic antigen (ANCA) activity was inhibited in 15 out of 21 sera from patients with acute systemic vasculitis following incubation with normal polyspecific IgG for therapeutic use (IVIg). ANCA antibodies reacted with IVIg through idiotypic-anti-idiotypic interactions, as shown in competitive binding assays using F(ab')2 fragments from IVIg and affinity chromatography of ANCA IgG on Sepharose-bound F(ab')2 fragments from IVIg. Co-incubation of sera from patients with acute systemic vasculitis with paired autologous remission stage sera also resulted in inhibition of ANCA activity in acute sera. Remission sera contain IgM and IgG capable of interacting with beta and or alpha idiotypes of ANCA IgG from acute sera. Anti-idiotypic IgM may account for the lack of expression of ANCA activity in whole serum from patients in remission from systemic vasculitis, which were found to contain high titres of ANCA IgG. These observations suggest that remission of systemic vasculitis is associated with the generation of anti-idiotypes against autoantibodies rather than the suppression of production of ANCA autoantibodies. IVIg may modulate the activity of systemic vasculitis in vivo.

Inability to detect circulating anti-idiotypic antibodies in human anti-GBM antibody-mediated disease using a splenic PFC assay

Anti-idiotypic antibodies are antibodies that are directed against the variable region of the corresponding antibody molecule and have been postulated to have a regulatory role in the immune response. Circulating anti-idiotypic antibodies have been reported in several antibody-mediated autoimmune diseases. We have established the following detection system for anti-idiotypic antibodies in human anti-glomerular basement membrane (GBM) disease. Spleen cells harvested from BALB/c mice immunised with human GBM were incubated with GBM-coated sheep red blood cells in the presence of guinea pig complement. Each spleen cell that produced anti-GBM antibodies resulted in a surrounding plaque where antibodies lysed the GBM-coated RBC. The number of plaques could be inhibited by the addition of heterologous anti-idiotypic antibodies to the plaquing mixture. Anti-idiotypic antibodies were then sought in both acute and convalescent phase sera from patients with anti-GBM disease and in laboratory staff who repeatedly handled GBM and sera containing anti-GBM antibodies. Anti-GBM antibodies were removed from all material before testing and affinity-purified preparations contained concentrations of IgG corresponding to the range that resulted in inhibition when affinity-purified rabbit anti-idiotypic antibody was examined. No anti-idiotypic antibodies could be demonstrated in any of the sera or IgG preparations examined. We conclude that if circulating anti-idiotypic antibodies are present in patients with anti-GBM disease, they must be infrequent, or at concentrations below the limits of detection of this assay (less than 1 mcg/mL). The observations of anti-idiotypic antibodies in other antibody-mediated diseases need to be confirmed.

Experimental autoimmune glomerulonephritis induced by homologous and isologous glomerular basement membrane in Brown-Norway rats

In order to study disease mechanisms and potential forms of therapy in glomerulonephritis, a model of experimental autoimmune glomerulonephritis (EAG) has been developed in the rat. We have examined the response of Brown-Norway (BN) rats to a single i.m. injection of collagenase-solubilised homologous (Sprague-Dawley, SD) or isologous (BN) glomerular basement membrane (GBM), with and without complete Freund's adjuvant (CFA). There was a dose-dependent circulating anti-GBM antibody response to all preparations of rat GBM. Animals given either antigen alone at a dose of 2 mg/kg developed circulating anti-GBM antibodies, which reached peak values by 6 weeks (63 +/- 5% following SD GBM; 53 +/- 8% following BN GBM), but did not develop glomerular deposits of IgG or nephritis. Animals given 2 mg/kg SD GBM in CFA developed greater concentrations of anti-GBM antibody by 6 weeks (122 +/- 20%) together with linear deposits of IgG on glomerular and tubular basement membranes (TBM), albuminuria (mean 7 mg/24 h), and variable focal segmental necrotising glomerulonephritis with mild interstitial nephritis. The same dose of BN GBM in CFA produced similar concentrations of circulating antibody (144 +/- 26%), with linear deposits of IgG on GBM but rarely TBM, little albuminuria, and variable mild focal glomerulonephritis. Other strains injected with SD GBM in CFA showed a variable circulating anti-GBM antibody response, which was similar to that of BN rats in PVG and DA rats but lower in LEW and WAG rats. Linear deposits of IgG on the GBM were detected in a proportion of PVG and DA rats, but not in LEW or WAG rats.(ABSTRACT TRUNCATED AT 250 WORDS)

Anti-neutrophil cytoplasmic antibodies and disease activity during long-term follow-up of 70 patients with systemic vasculitis

Anti-neutrophil cytoplasmic antibodies (ANCA) have been identified as a diagnostic marker in primary systemic vasculitis, and immunofluorescence assays identify two patterns of binding: cytoplasmic (C-ANCA and perinuclear (P-ANCA). We have examined retrospectively the use of such assays in long-term monitoring of disease activity, in order to determine the relationship between presence of ANCA and relapse, and to assess their suitability as a guide to therapy. Seventy patients were studied over a period of 50 months, using clinical and laboratory criteria for the diagnosis of relapse and the internationally standardised immunofluorescence assay for detection of ANCA. In 19 patients C- or P-ANCA were detectable throughout the study period; six of these (with C-ANCA) relapsed. In 18 patients ANCA were undetectable during long-term follow-up; none of these patients relapsed. In 33 patients C- or P-ANCA were intermittently present; nine of these relapsed and all had C-ANCA detectable at the time of relapse. In six of the nine cases, relapse was accompanied or closely preceded by reappearance of C-ANCA. We conclude that continuing presence and reappearance of ANCA may identify patients who are at risk of relapse and who are most likely to benefit from long-term immunosuppressive therapy.

Plasmapheresis and subsequent pulse cyclophosphamide versus pulse cyclophosphamide alone in severe lupus: design of the LPSG trial. Lupus Plasmapheresis Study Group (LPSG)

A group of clinics are collaborating in the Lupus Plasmapheresis Study Group (LPSG) to investigate whether repeated plasmapheresis prior to pulse cyclophosphamide improves the therapeutical results in severe systemic lupus erythematosus (SLE). The underlying rationale is the hypothesis that plasmapheresis 1) eliminates pathogenic autoantibodies and immune complexes and 2) induces compensatory lymphocyte activation via feedback mechanisms between circulating antibodies and their respective clones ("antibody rebound"). It should be possible to utilize this enhanced activity for increased clonal deletion if pulse cyclophosphamide is applied shortly after plasmapheresis. Accordingly, in a randomized study, the LPSG will be comparing the repeated application of pulse cyclophosphamide alone with a treatment involving repeated plasmapheresis prior to the cyclophosphamide pulses in severe SLE. A third arm of the study will be gathering experience with a more intensified procedure. This overview summarizes the most important details of the planned study.

The specificity of anti-neutrophil cytoplasm autoantibodies in systemic vasculitides

Autoantibodies to neutrophil cytoplasmic antigens are found in patients with small or medium-size vessel systemic vasculitis such as Wegener's granulomatosis and microscopic polyarteritis. We studied the specificity of anti-neutrophil cytoplasm autoantibodies (ANCA) by examining the binding of sera from patients with different forms of systemic vasculitis, to neutrophil extract size-fractionated by gel filtration for high performance liquid chromatography. Sequential fractions were collected for solid-phase radioimmunoassay. Patients with Wegener's granulomatosis showed two peaks of activity over fractions of 12 and 2 kD. Patients with microscopic polyarteritis had two different peaks of activity over fractions of 100 and 25 kD. Patients with Takayasu's arteritis and dermal lymphocytic vasculitis were studied as controls (all were ANCA negative by indirect immunofluorescence studies and solid-phase radioimmunoassay). However, the patient with lymphocytic vasculitis showed a distinct peak of activity over 22 kD and patients with Takayasu's arteritis had a single peak of activity over 200 kD. These results suggest the binding pattern of sera to fractionated neutrophil extract could be useful in differentiating between different diseases within the spectrum of systemic vasculitis. Cryptic antigen recognition in certain vasculitides may only be revealed by this procedure. We conclude that the association of ANCA and systemic vasculitis may not be limited to diseases of small or medium-size vessels, but could also include large vessel disease such as Takayasu's arteritis.

Autoantibodies to endothelial cells and neutrophil cytoplasmic antigens in systemic vasculitis

The interaction of circulating autoantibodies with the endothelium may be an important mechanism in the pathogenesis of systemic vasculitis. In a prospective study, we looked for circulating antiendothelial cell autoantibodies (AECA) and anti-neutrophil cytoplasm autoantibodies (ANCA) in 80 patients with suspected systemic vasculitis. AECA were measured using an isotype-specific cellular ELISA incorporating human umbilical vein endothelial cells. ANCA activity was determined by indirect immunofluorescence and radioimmunoassay. Sequential studies were performed on sera from four cases with dual positivity, where autoantibody binding was compared with von Willebrand factor (vWF) concentration and disease activity. IgG AECA were significantly higher in the 27 ANCA-positive sera as compared with normal controls (P = 0.027) with IgG (P = 0.009) and IgA (P = 0.046) AECA isotypes correlating with ANCA positivity; in contrast, no differences were found between AECA levels in the ANCA-negative sera and the normal controls. Cross-inhibition studies pointed to the co-existence of two autoantibody populations. An association between autoantibody binding, disease activity and vWF concentration was found for both ANCA and AECA. Some patients with systemic vasculitis have detectable AECA that recognize different epitopes to ANCA and like ANCA, their titre correlates with disease activity and thus they may have a pathogenetic role in these conditions.