Pathogenesis of pulmonary vasculitis

Vasculitis is inflammation of blood vessels and can affect any type of vessel in any organ. Pulmonary vasculitis usually is a component of a systemic small vessel vasculitis. Three major forms of small vessel vasculitis that often affect the lungs are Wegener's granulomatosis, microscopic polyangiitis, and Churg-Strauss syndrome. These forms of vasculitis are strongly associated with antineutrophil cytoplasmic autoantibodies (ANCA) directed against enzymes contained in the primary granules of neutrophils and peroxidase-positive lysosomes of monocytes. This review discusses the evidence for a pathogenic role of ANCA. In vitro, ANCAs can activate cytokine-primed neutrophils and monocytes resulting in oxygen radical formation and release of lysosomal enzymes. In vivo, antimyeloperoxidase ANCA has been shown to induce crescentic glomerulonephritis and systemic vasculitis. Overall, the available data suggest that ANCA are indeed a pathogenic factor in the development of small-vessel vasculitis. Antiglomerular basement membrane (anti-GBM) disease also causes pulmonary vasculitis through immune attack on alveolar capillaries and glomerulonephritis through antibody mediated injury to glomerular capillaries. Thus, there is evidence that antibodies are important pathogenic factors in both ANCA disease and anti-GBM disease, however, there are also indications that T cells may play important pathogenic roles in both categories of disease as well.     

Antineutrophil cytoplasmic autoantibodies and pathophysiology: new insights from animal models

PURPOSE OF REVIEW: Since the discovery of antineutrophil cytoplasmic autoantibodies (ANCA) and their association with the occurrence of several types of small-vessel vasculitis, a causal relation between the two has been suggested. Various in vitro and in vivo experimental data provide indirect evidence in support of this view. This article comprises a review of the animal models that have been used to investigate the pathogenesis of ANCA-associated vasculitis, and focuses on recent developments in this field. RECENT FINDINGS: Xiao et al. provide definite proof of the pathogenic potential of ANCA in a novel mouse model of myeloperoxidase (MPO)-ANCA-associated vasculitis, in which transfer of splenocytes or IgG from MPO-/- mice immunized with murine MPO, to naive wild-type or Rag2-/- (lacking mature B and T lymphocytes) mice causes a disease remarkably similar to its human counterpart. In addition, preliminary studies by Smyth et al. show that immunization of Wistar Kyoto rats with human MPO induces antihuman MPO antibodies that cross-react with rat MPO, as well as a disease closely resembling human small-vessel vasculitis. Another murine ANCA model is the SCG/Kj mouse. A recent publication by Neumann et al., however, puts an important limitation on the use of this mouse model for the study of ANCA-associated vasculitis, demonstrating multiple immune complex deposits in the spontaneously occurring vascular lesions. SUMMARY Recently developed animal models of MPO-ANCA-associated vasculitis convincingly demonstrate that MPO-ANCA are pathogenic. Whether similar strategies can be used to develop an appropriate model for proteinase 3-ANCA-associated vasculitis remains to be investigated.     

Pathophysiology of ANCA-associated vasculitides: are ANCA really pathogenic?

The strong relation between antineutrophil cytoplasmic autoantibodies (ANCA) and primary vasculitic syndromes suggests a pathophysiological role for ANCA. Experimental evidence for the pathogenic potential of ANCA has been derived from in vitro studies that demonstrate that ANCA can activate tumour necrosis factor alpha primed neutrophils, monocytes and/or endothelial cells. The binding of ANCA to primed neutrophils results in activation of these cells by a process that is largely dependent on engagement of beta-2 integrins and on the interaction of the Fc portion of ANCA. An Fc-independent mechanism is, however, also operative. In experimental animal models, it has been demonstrated that immunisation with myeloperoxidase (MPO) induces MPO-ANCA. The induction of ANCA, however, is not sufficient to induce vasculitis in rats since immune complexes first have to be deposited along the vessel wall before lesions develop. When MPO-deficient mice are, however, immunised with murine MPO, anti-MPO immunoglobulins are purified and subsequently injected into mice that are not deficient for MPO, systemic vasculitis and glomerulonephritis is induced. These experiments suggest that ANCA indeed induces vasculitis. Risk factors for breaking self-tolerance to ANCA antigens are genetic factors, drugs, chemical substances and/or infectious agents.     

Opportunistic infections are preceded by a rapid fall in antineutrophil cytoplasmic antibody (ANCA) titer in patients with ANCA associated vasculitis

OBJECTIVE: To study the clinical course and changes in antineutrophil cytoplasmic antibody (ANCA) titers in patients with ANCA associated vasculitis (AAV) who developed opportunistic infections. METHODS: Among the patients with AAV tested in the Immunopathology Laboratory at the Massachusetts General Hospital between 1989 and 1998, all patients who experienced opportunistic infections (n = 16) were included. We retrospectively studied their clinical features and examined the relationship between changes in ANCA titer and the onset of the opportunistic infections. ANCA titers were measured by antigen-specific ELISA. RESULTS: Of the 16 AAV patients with opportunistic infection, 15 had no evidence for active vasculitis at the time of the infection. Among these 15 patients, opportunistic infections were associated with a steep fall in ANCA titers. There was no consistent pattern of change in C-reactive protein levels. In 7 patients, the immunosuppressive regimen was increased for new clinical findings shortly before the diagnosis of an opportunistic infection, despite the absence of histologic documentation of active vasculitis. Three of these 7 patients died. One patient, who did not experience a significant fall in ANCA titer. i.e., less than 4-fold from his prior peak, was simultaneously found to have Pneumocystis carinii pneumonia and biopsy proven active vasculitis. CONCLUSION: Our data strongly suggest that opportunistic infections in patients with AAV are associated with negative or rapidly falling ANCA titers. Therefore, changes in ANCA titer can help distinguish opportunistic infections from vasculitis flares when patients with AAV present with indeterminate clinical findings.     

Alternating antineutrophil cytoplasmic antibody specificity: drug-induced vasculitis in a patient with Wegener's granulomatosis

We describe a patient who presented with Wegener's granulomatosis associated with antineutrophil cytoplasmic antibodies (ANCA) directed against proteinase 3 (PR3) with a cytoplasmic immunofluorescence pattern (cANCA), whose ANCA type changed to antimyeloperoxidase antibodies with a perinuclear immunofluorescence pattern (pANCA) when treated with propylthiouracil, and changed back to anti-PR3 antibodies with cANCA after the medication was discontinued. The patient developed flares of vasculitis symptoms associated with rises in either type of ANCA. Tests for antimyeloperoxidase ANCA were repeatedly negative before the drug was started, strongly implicating the drug as the cause of the episode. This case demonstrates that patients with idiopathic ANCA-positive vasculitis may quickly develop a superimposed drug-associated ANCA-positive vasculitis. Iatrogenic vasculitis should be suspected when a patient with idiopathic vasculitis with one type of ANCA develops the other type of ANCA.     

Is Wegener's granulomatosis an autoimmune disease?

Wegener's granulomatosis is a multisystem disease characterized by granulomata of the respiratory tract and systemic necrotising vasculitis. There is a strong and specific association with autoantibodies directed against proteinase 3, a constituent of neutrophril azurophilic granules. Antibody titers correlate with clinical disease activity and predict relapses. The disease responds favorably to immunosuppressive therapy. The pathogenicity of antineutrophil cytoplasmic antibodies (ANCA), however, remains unproven. In vitro, the expression of proteinase-3 and other ANCA antigens on the surface of neutrophils and monocytes can be induced by priming with proinflammatory cytokines. Antineutrophil cytoplasmic antibodies are then able to activate these leukocytes, stimulating degranulation, the production of reactive oxygen species, and the secretion of further cytokines. Neutrophils activated by ANCA, and possibly ANCA alone, directly damage endothelial cells in vitro. An animal model of proteinase 3-ANCA-induced vasculitis has not been found. Antineutrophil cytoplasmic antibodies directed against another antigen, myeloperoxidase, are not sufficient to cause vasculitis but they promote damage in certain animal models. Thus, a considerable amount of evidence supports the notion that Wegener's granulomatosis is an autoimmune disease.     

Resolution of anti-neutrophil cytoplasmic antibody-associated vasculitis after resection of gastric cancer

We report a case of anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis in a 62-year-old patient with gastric cancer. The myeloperoxidase–anti-neutrophil cytoplasmic antibody (MPO–ANCA) level was threefold above normal preoperatively. Vasculitis was seen on renal biopsy. Gastric resection revealed well-differentiated adenocarcinoma and vasculitis. The MPO–ANCA level returned to normal post-operatively. Although ANCA-associated vasculitis occasionally accompanies malignant tumors, this is the first documented case of concurrent gastric cancer-associated and ANCA-associated vasculitis, with post-operative resolution of the vasculitis.     

Antineutrophil cytoplasmic antibody vasculitis associated with mycobacterium avium intracellulare infection

A variety of possible associations between infection and antineutrophil cytoplasmic antibody (ANCA) associated vasculitis have been reported. We describe a 75-year-old woman who presented with chronic nonproductive cough, migratory polyarthralgias, and microscopic hematuria. She had an elevated perinuclear ANCA and antimyeloperoxidase antibody. She had a positive PPD test and a cavitary lesion in the right upper lung lobe; biopsy of the lung lesion showed granulomatous vasculitis, but the culture grew Mycobacterium avium intracellulare (MAI). There are clinical and histiologic similarities between ANCA vasculitis and pulmonary MAI infection. Treatment of vasculitis with immunosuppressive agents could be detrimental in patients with MAI infection. Thus, when ANCA associated vasculitis is considered, mycobacterium infection should be excluded before starting immunosuppressive therapy.     

Resolution of anti-neutrophil cytoplasmic antibody-associated vasculitis after resection of gastric cancer

Abstract

We report a case of anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis in a 62-year-old patient with gastric cancer. The myeloperoxidase–anti-neutrophil cytoplasmic antibody (MPO–ANCA) level was threefold above normal preoperatively. Vasculitis was seen on renal biopsy. Gastric resection revealed well-differentiated adenocarcinoma and vasculitis. The MPO–ANCA level returned to normal post-operatively. Although ANCA-associated vasculitis occasionally accompanies malignant tumors, this is the first documented case of concurrent gastric cancer-associated and ANCA-associated vasculitis, with post-operative resolution of the vasculitis.     

Characterization of biologically active antineutrophil cytoplasmic antibodies induced in mice:pathogenetic role in experimental vasculitis

Objective. To investigate the pathogenetic role of antineutrophil cytoplasmic antibodies (ANCA) in Wegener's granulomatosis (WG). Methods. BALB/c mice were immunized with human IgG ANCA from a patient with WG. Control mice were immunized with normal human IgG. Levels of mouse ANCA and other autoantibodies were determined. Mouse ANCA were tested for their ability to induce adhesion and respiratory burst of neutrophils. The mouse lungs and kidneys were examined for the development of vasculitis. Results. Mice immunized with human ANCA developed anti-human ANCA and anti–anti-human ANCA (mouse ANCA), while the controls did not develop these antibodies. Mouse ANCA were capable of inducing adhesion of neutrophils to fibronectin and activating the respiratory burst in neutrophils. Moreover, the mice that were immunized with human ANCA developed perivascular mononuclear cell infiltrates in the lungs, suggesting vasculitis. Conclusion. The results suggest a pathogenic role of ANCA in WG, and may imply that activation of neutrophils is the initiating event in the development of vasculitis in WG.     

Antineutrophil cytoplasmic antibodies induce human monocytes to produce oxygen radicals in vitro

OBJECTIVE: Antineutrophil cytoplasmic antibodies (ANCA) are believed to play a pathogenetic role in necrotizing small-vessel vasculitis. While the involvement of neutrophils in this disease has been extensively studied in vitro, we undertook to analyze thoroughly the contribution of monocytes to tissue destruction in systemic vasculitis. METHODS: Monocytes obtained from normal human individuals were stimulated by ANCA isolated from patients with active vasculitis. The formation of oxygen radicals was measured by a fluorometric assay using 2',7'-dichlorofluorescin diacetate. RESULTS: ANCA induced monocytes to produce oxygen radicals, resulting in a mean 43% increase (range 21-84%) in oxygen radical formation compared with normal IgG. The formation of reactive oxygen species was time and concentration dependent and was also induced by ANCA F(ab')2 fragments. Normal nonspecific IgG or their corresponding F(ab')2 fragments induced no release or very little release of oxygen radicals. Preincubation of monocytes with the Fcy receptor type II-blocking monoclonal antibody IV.3 before addition of ANCA greatly reduced formation of oxygen radicals. Using ligand affinity chromatography with proteinase 3 (PR3) and myeloperoxidase (MPO), ANCA were further purified by depletion of patient IgG. The stimulation of monocytes with these pure PR3- and MPO-ANCA confirmed that cellular activation was specifically induced by ANCA. CONCLUSION: These results show that ANCA induce the formation of reactive oxygen species in human monocytes. These findings support the notion that ANCA specifically activate monocytes by several mechanisms to participate in the inflammatory process of ANCA-associated vasculitis.     

Anti-neutrophil cytoplasmic antibodies and their clinical significance

Anti-neutrophil cytoplasmic antibodies (ANCA) are a group of autoantibodies that cause systemic vascular inflammation by binding to target antigens of neutrophils. These autoantibodies can be found in serum from patients with systemic small-vessel vasculitis and they are considered as a biomarker for ANCA-associated vasculitis (AAV). A conventional screening test to detect ANCA in the serum is indirect immunofluorescence study, and subsequently confirmed by enzyme-linked immunosorbent assay. A positive staining of ANCA can be classified into three main categories based on the staining patterns: cytoplasmic, perinuclear, and atypical. Patients with granulomatosis with polyangiitis (GPA) mostly have a positive cytoplasmic staining pattern (c-ANCA) whilst a perinuclear pattern (p-ANCA) is more common in microscopic polyangiitis (MPA) and eosinophilic granulomatosis with polyangiitis (EGPA) patients. Atypical pattern (a-ANCA) is rarely seen in patients with systemic small-vessel vasculitis but it can be found in other conditions. Here, techniques for ANCA detection, ANCA staining patterns and their clinical significances are reviewed.     

Antineutrophil cytoplasmic antibody (ANCA)-associated systemic vasculitis after immunisation with bacterial proteins

OBJECTIVE: There is circumstantial evidence for a role for infections in the development of the small vessel vasculitides associated with antineutrophil cytoplasmic antibodies (ANCA). The aim of this study was to determine whether the immunisation of rats with bacterial proteins could result in circulating ANCA, T cells with specificity for ANCA antigens, and a systemic vasculitis. METHODS: Adult male Wistar rats were immunised with pasteurised sonicated S. aureus (n = 7), E. coli (n = 8), purified protein derivative (PPD, n = 5), myeloperoxidase (MPO, n = 5) or phosphate-buffered saline (PBS, n = 5), in complete and in incomplete Freund's adjuvant. ANCA were assayed by indirect immunofluorescent (IIF) examination of normal rat neutrophils, and in ELISAs using human proteinase 3 (PR3), MPO and bactericidal/permeability-inreasing protein (BPI). The T cell response to PR3, MPO and BPI was assessed by a whole blood T cell proliferative assay in vitro, and by a delayed type hypersensitivity (DTH) response in vivo. Kidney and bowel were examined histologically for evidence of vasculitis and colitis. RESULTS: One rat from each group immunised with S. aureus or E. coli developed pauciimmune segmental glomerular sclerosis. The rat immunised with E. coli had additionally an arteritis affecting renal interlobular and gut vessels. This rat had circulating C-ANCA, that produced granular cytoplasmic neutrophil fluorescence with central accentuation, but the target antigen could not be determined in ELISAs using human PR3, MPO or BPI. In animals immunised with S. aureus or E. coli, there was no significant T cell proliferative or DTH response specific for human PR3, MPO or BPI. CONCLUSION: The development of ANCA and vasculitis in a rat immunised with bacterial proteins indicates that the relationship between infections and ANCA should be investigated further.     

The prevalence and target antigens of antithyroid drugs induced antineutrophil cytoplasmic antibodies (ANCA) in Chinese patients with hyperthyroidism

OBJECTIVE: Antithyroid drugs such as propylthiouracil (PTU) and methimazole (MMI) are common medications in Chinese patients with hyperthyroidism and PTU-induced antineutrophil cytoplasmic antibody (ANCA) positive vasculitis has been reported. The current cross-sectional study aimed to investigate the prevalence and the target antigens of ANCA in Chinese patients with hyperthyroidism pre- and post-antithyroid medication therapy. METHODS: Sera from 216 patients with hyperthyroidism in our hospital were collected from January to July in 2002. Patients were divided into four groups: untreated (n = 34); treated with PTU (n = 62); treated with MMI (n = 77); and treated with both PTU and MMI (n = 43). Indirect immunofluorescence (IIF) assay was used to detect ANCA and ANA. Antigen-specific ELISAs were used to detect antigen specificities. The known antigens included myeloperoxidase (MPO), proteinase 3 (PR3), human leukocyte elastase (HLE), lactoferrin, bactericidal/permeability-increasing protein (BPI), cathepsin G and azurocidin. RESULTS: 33/216 sera were IIF positive, 20 of the 33 samples were ANCA positive, 11 samples were ANA positive, and two samples were both P-ANCA and ANA positive. The prevalence of positive ANCA in patients receiving PTU (14/62, 22.6%) was significantly higher than that of untreated patients (1/34, 2.9%) and patients treated with MMI (0/77, 0), P < 0.017. Of the 22 IIF-ANCA positive samples, 12 (54.5%) sera recognized lactoferrin, seven (31.8%) sera recognized HLE, four sera recognized MPO and azurocidin respectively, three sera recognized PR3 and cathepsin G respectively, and one serum recognized BPI. Six of the 22 (27.3%) patients with ANCA positive had clinical evidence of vasculitis. All patients with MPO-ANCA and two of the three patients with PR3-ANCA had clinical vasculitis. CONCLUSION: PTU is associated with the production of ANCA in patients with hyperthyroidism. PTU-induced ANCA are due to polyclonal activation of B cells. Anti-MPO and anti-PR3 antibodies may associate the occurrence of clinical vasculitis.     

Clinical value of antineutrophil cytoplasmic antibodies

This article presents a brief review of the clinical value of antineutrophil cytoplasmic antibodies (ANCA) in the diagnosis and management of patients with various forms of vasculitis. ANCA assay methodology and testing recommendations are reviewed. The patterns of reactivity of ANCA observed by indirect immunofluorescence, the antigens recognized by ANCA, and the sensitivity, specificity, and positive predictive value of ANCA for diagnosis of different vasculitides are described. In addition, the spectrum of drugs and nonvasculitic diseases that are associated with ANCA and need to be differentiated from true ANCA-positive vasculitides are reviewed. When properly utilized and cautiously interpreted, ANCA are a useful, noninvasive diagnostic tool in the differential diagnosis of vasculitis.     

Antineutrophil cytoplasmic antibodies (ANCA) and systemic vasculitis: update of assays, immunopathogenesis, controversies, and report of a novel de novo ANCA-associated vasculitis after kidney transplantation

OBJECTIVES: To characterize antineutrophil cytoplasmic antibodies (ANCA), their major autoantigens, disease associations, and pathophysiology in systemic vasculitides. To describe a patient with a novel de novo ANCA-associated vasculitis after kidney transplantation. METHODS: We reviewed and compiled the literature on ANCA-related topics and systemic vasculitis. Laboratory and clinical data from a cadaveric kidney transplant patient who developed necrotizing vasculitis involving glomerular capillaries, with crescent formation associated with P-ANCA and myeloperoxidase, were analyzed. RESULTS: Large-scale multi-center testing of patient and normal sera by the European ANCA Assay Standardization Project using immunofluorescence assays and enzyme immunoassays indicate the assays have good sensitivity and specificity, and diagnostic utility for ANCA-associated vasculitis. A few investigations covering basic and clinical research with ANCA remain controversial: whether endothelial cells do or do not express a 29-kd neutral serine protease termed proteinase-3 (PR-3), the target of ANCA in most individuals with Wegener's granulomatosis, and whether anti-myeloperoxidase (MPO) ANCAs recognize a restricted number of epitopes on MPO. This issue has relevance for using monoclonal antibodies to treat patients with vasculitis who have adverse effects from immunosuppressive drugs. The two allelic forms of FcgammaRIIa (H131/R131) and the two of FcgammaRIIlb (NA1/NA2) are discussed as possible inheritable genetic elements for vasculitic disorders and for signaling responses. Stimulatory and costimulatory molecules, and cytokine profiles of T lymphocytes are characterized to show that these cells are actively involved in the ANCA-associated vasculitides. The patient described had a de novo ANCA associated small vessel vasculitis which developed after renal transplantation. CONCLUSIONS: There have been significant advances in the development of sensitive and specific ANCA assays. The immunopathogenetic mechanism of ANCA involves the constitutive FcgammaRs, ligands, and signaling responses to activate cytokine-primed neutrophils. This may lead to the generation of reactive oxygen intermediates, degranulation, and secretion of intracellular granule contents, and ultimately inflammation and vasculitis.     

A Ruptured Jejunal Arterial Aneurysm in a Young Woman Undergoing Chronic Hemodialysis Due to Myeloperoxidase-antineutrophil Cytoplasmic Antibody-associated Vasculitis

A 21-year-old woman was admitted to our hospital because of massive intestinal bleeding. She started hemodialysis due to myeloperoxidase antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) at 18 years of age. Her ANCA titers remained stable; however, her C-reactive protein increased on 5 mg/day prednisolone before admission. Computed tomography angiography revealed a ruptured jejunal arterial aneurysm. Transcatheter arterial embolization, blood transfusion and the reinforcement of steroid therapy resolved her symptoms of AAV. Our case of a young patient with AAV and medium-sized arterial vasculitis is rare and emphasizes that the ANCA titer does not always rise, especially in patients with nonrenal vasculitis flare-ups.     

Smoking habits in patients diagnosed with ANCA associated small vessel vasculitis

OBJECTIVES: To study the prevalence of smoking at onset of symptoms in patients with small vessel vasculitis associated with antineutrophil cytoplasmic antibodies (ANCA). METHODS: A retrospective study, in 197 patients with ANCA associated vasculitis, of the history of cigarette smoking at onset of symptoms. Prevalence of smoking in patients with ANCA associated vasculitis was compared with age-specific values for the general population in Germany. RESULTS: 27 (14%) patients were smokers at the time of first disease manifestation (p < 0.001, compared with the entire population); 54 (27%) had smoked previously with 1-110 pack-years (median 18) but had stopped > or = 2 years before onset of vasculitis; 116 (59%) patients were lifelong non-smokers. At onset of symptoms, active smokers were younger (median age 42 years) than patients with vasculitis (median 54 years, p < 0.01, Mann-Whitney U test) with a lower percentage of women (15%, p < 0.005, Fisher's exact test) than in the entire group (47%). Smokers, non-smokers, or ex-smokers did not differ in organ manifestation, mortality, and development of end stage renal disease and relapse rate. CONCLUSIONS: The proportion of active smokers in the group of patients with ANCA associated vasculitis is significantly lower than in the entire population. Cigarette smoking may be associated with a reduced risk of ANCA associated vasculitis.     

Anti-neutrophil cytoplasmic autoantibodies induce neutrophils to degranulate and produce oxygen radicals in vitro.

Anti-neutrophil cytoplasmic autoantibodies (ANCA) are in the circulation of most patients with pauci-immune necrotizing vasculitis and pauci-immune crescentic glomerulonephritis. The current study demonstrates an effect of these autoantibodies on neutrophil function in vitro. ANCA cause normal human neutrophils to undergo an oxidative burst and degranulate. Both ANCA phenotypes (i.e., cytoplasmic-pattern ANCA and myeloperoxidase-specific ANCA) induce neutrophil activation. ANCA sera and purified immunoglobulins significantly increase the release of reactive oxygen species when compared with controls. ANCA, in a dose-dependent manner, induce the release of primary granule contents. These effects are markedly enhanced by priming neutrophils with tumor necrosis factor. Flow cytometry studies demonstrate the presence of myeloperoxidase on the surface of neutrophils after cytokine priming, indicating that primed neutrophils have ANCA antigens at their surfaces to interact with ANCA. These observations suggest an in vivo pathogenetic role for ANCA. We propose that, in patients with necrotizing vasculitis, ANCA-induced release of toxic oxygen radicals and noxious granule enzymes from cytokine-primed neutrophils could be mediating vascular inflammation.     

Relationship between anti-neutrophil cytoplasmic antibody determined with conventional binding and the capture assay, and long-term clinical course in vasculitis

OBJECTIVES: To evaluate the relationship between anti-neutrophil cytoplasmic antibody (ANCA) measured with two different methods and long-term clinical course in vasculitis. DESIGN: Retrospective determination of ANCA with two different assays for detection of PR3-ANCA, conventional direct binding ELISA and capture ELISA using monoclonal antibodies against PR3. The 245 ANCA determinations were performed from frozen blood samples collected three to four times a year in each patient. SETTING: Department of Nephrology at a Swedish University Hospital. SUBJECTS: A total of 10 ANCA-positive patients with vasculitis caused by Wegener's granulomatosis (WG) or microscopic polyarteritis (MPA) and a very long follow-up time (mean 9 years, range 5-15.5 years). RESULTS: The total number of episodes with active vasculitis was 29 and all of them (100%) were detected by the capture technique whilst the conventional technique detected 23 (79%). The mean number of episodes with active disease requiring treatment with steroids and cytotoxic drugs was three per patient (range 1-6). At the time of clinical relapse of the vasculitis disease, the ANCA titre using the capture technique was either increasing or showed a very high value in all cases. The pattern of capture ANCA response could be subdivided into three categories: a close (four patients), an intermediate (three patients), and no (three patients) relationship between capture ANCA level and long-term clinical course. CONCLUSION: Detection of PR3-ANCA by the capture ELISA showed a higher sensitivity than that obtained by the direct ELISA in diagnosing relapse during follow-up of patients with vasculitis. The specificity of the capture ANCA was, however, low, as high levels occurred in patients without clinical disease activity.