Analysis of anti-neutrophil cytoplasmic antibodies (ANCA): frequency and specificity in a sample of 191 homozygous (PiZZ) alpha1-antitrypsin-deficient subjects

BACKGROUND

ANCA are autoantibodies directed against polymorphonuclear cell antigens, mainly proteinase 3 (PR3) and myeloperoxidase (MPO), which are implicated in the pathogenesis of small-vessel necrotizing vasculitis. Alpha1-antitrypsin is the main inhibitor of neutral serine proteinase [i.e. human leukocyte elastase (HLE) and PR3] present in PMN alpha-granules (alphaGr). An association first reported by us between PR3 ANCA and the deficient PiZZ phenotype in ANCA-positive systemic vasculitis, now widely confirmed by others, led us to study the incidence and specificity of ANCA among PiZZ subjects.

METHODS

We tested a population of 191 PiZZ (273 sera) for ANCA activity versus 272 PiMM matched control subjects using alphaGr or antigen-specific ELISA [PR3, HLE, MPO, lactoferin (LF) and bactericidal/ permeability increasing protein (BPI)].

RESULTS

The incidence of antibodies directed against alphaGr and HLE but not PR3, MPO, LF or BPI was increased in the PiZZ as compared to the PiMM group (Fisher probability respectively P < 0.0001 and P < 0.05).

CONCLUSIONS

ANCA not directed against classical antigens (MPO and PR3) may be found in PiZZ patients. However, these patients do not develop systemic vasculitis features. Therefore, alpha1-antitrypsin deficiency is not sufficient to induce ANCA positive vasculitides, and may only act as a second hit amplifying factor.

Prevalence of rheumatic manifestations and antineutrophil cytoplasmic antibodies in haematological malignancies. A prospective study

OBJECTIVE

To evaluate the prevalence of antineutrophil cytoplasmic antibodies (ANCA) and rheumatic manifestations associated with chronic haematological malignancies.

METHODS

Two groups of patients were prospectively studied (group I: 60 patients with myelodysplastic syndromes and group II: 140 patients with lymphoid malignancies) for clinical 'immune' manifestations and ANCA.

RESULTS

In the myelodysplastic group, six patients had ANCA-negative systemic medium-size vasculitis, one had systemic vasculitis with cytoplasmic ANCA, one relapsing polychondritis, one giant cell arteritis, one polymyalgia rheumatica, one polyarthritis and two fasciitis. In group II, two patients had ANCA-negative systemic vasculitis, two had leucocytoclastic vasculitis associated with tuberculosis, two had polyarthritis, one polymyalgia rheumatica and one giant cell arteritis. Six sera were ANCA-positive with perinuclear pattern in four cases, atypical pattern in one and cytoplasmic pattern in one. Two sera had anti-myeloperoxidase (MPO) specificity, and others had no known specificity; none had anti-proteinase 3 (PR3) specificity. Global prevalence of ANCA in our cohort was 3%, similar to the French general population.

CONCLUSION

Polyarteritis nodosa-type systemic vasculitis and polymyalgia rheumatica were the most frequent findings (18%) in myelodysplastic syndromes and particularly in chronic myelomonocytic leukaemia. ANCA were not helpful for the diagnosis of vasculitis. Vasculitis associated with infection, in particular tuberculosis, must be ruled out.

Alpha1-antitrypsin phenotypes and anti-neutrophil cytoplasmic auto-antibodies in inflammatory bowel disease

OBJECTIVES

Alpha1-antitrypsin (alpha1-AT) is encoded by a highly polymorphic gene with over 75 codominantly expressed alleles at the protease inhibitor (Pi) locus classified as normal, deficient, dysfunctional or null. The aim of this study was to determine in patients with inflammatory bowel disease: (i) the prevalence of anti-neutrophil cytoplasmic auto-antibodies (ANCA) and their antigen specificities; (ii) alpha1-AT Pi phenotypes; and (iii) possible associations between ANCA, disease activity and deficient alpha1-AT alleles.

DESIGN

Study of 95 consecutive patients with ulcerative colitis (UC) and 63 patients with Crohn's disease (CD).

METHODS

Diagnosis and disease activity were determined by clinical, endoscopic and histological criteria. ANCA by indirect immunofluorescence (IIF) and Pi phenotyping by isoelectric focusing were performed for all patients. Positive IIF sera were tested in antigen-specific enzyme-linked immunosorbent assay: proteinase 3 (PR3), myeloperoxidase (MPO), lactoferrin (LF), lysozyme, human leucocyte elastase (HLE), cathepsin G and bactericidal/permeability increasing protein (BPI).

RESULTS

Sixty-one patients with UC (64.2%) and only 11 with CD (17.5%) had ANCA (P < 0.001). Antigen specificities were PR3 (7/61), MPO (3/61), LF (6/61), HLE (1/63) and BPI (10/61) in UC, and PR3 (2/11) and BPI (2/11) in CD. Three PiZ alleles were found, matching the prevalence in the normal French control population. No relationship was found between the presence of ANCA, antibody specificity, disease activity and deficient alpha1-AT alleles.

CONCLUSION

ANCA are more frequent in UC than CD and do not correlate with disease activity. ANCA and protease/antiprotease imbalance do not appear to modulate the clinical course of inflammatory bowel disease.

Effect of protein A immunoadsorption in nephrotic syndrome of various etiologies

Protein A immunoadsorption (IA) has proved effective in reducing proteinuria in patients with nephrotic syndrome after recurrence of focal and segmental glomerulosclerosis (FSGS) in kidney transplants. The effect of IA in nephrotic syndrome of other etiologies remains unknown. Nine patients with nephrotic syndrome secondary to membranous nephropathy (four cases), diabetes mellitus (one case), IgA nephropathy (two cases), and amyloidosis (two cases) had three to five IA of 2.5 plasma volumes over 4 to 8 d. Patients received no concomitant immunosuppressive treatment, and antihypertensive drugs were left unchanged. Proteinuria decreased from 12.64 +/- 5.49 to 3.35 +/- 2.2 g/24 h (mean +/- SD) in all patients after three to five IA. Hematocrit decreased from 37.32 to 32.64% (12.5% hemodilution) and serum albumin from 25.43 to 18.6 g/L (26.4% decrease). Proteinuria returned to baseline levels within 1 mo, as described in recurrent FSGS following transplantation. When serum albumin balance was controlled by albumin infusion after IA in two patients, comparable decreases in proteinuria were observed. Therefore, IA is effective in producing short-term reduction of proteinuria in nephrotic syndromes related not only to FSGS but also to membranous and IgA nephropathies, diabetes mellitus, and amyloidosis, which suggests that IA removes a nonspecific circulating hemodynamic-altering or permeability-increasing factor.

Alpha-1-antitrypsin phenotyping in ANCA-associated diseases: one of several arguments for protease/antiprotease imbalance in systemic vasculitis

The vasculitic lesions observed in Wegener's granulomatosis may be partly the consequence of proteases released following activation of neutrophils by ANCA. The activity of these proteases, including proteinase 3 (PR3) and elastase, is normally closely restricted to the inflammation site by a large excess of circulating alpha-1-antitrypsin (alpha1AT). Patients with ANCA-positive systemic vasculitis may exhibit a protease/antiprotease imbalance either genetically determined in the rare patients with deficient alpha1AT phenotypes, or more often acquired through both alpha1AT inactivation in various pathological conditions and possible inhibition of PR3/alpha1AT complexation by anti-PR3 ANCA. This imbalance may at least contribute to disease spreading or aggravation.

Antineutrophil cytoplasmic antibodies (ANCA) in chronic graft-versus-host disease after allogeneic bone marrow transplantation

We studied the usefulness of monitoring antineutrophil cytoplasmic antibodies (ANCA) in chronic graft-versus-host disease (cGVHD), a major complication of allogeneic bone marrow transplantation. Antigen-specific ELISA and indirect immunofluorescence (IIF) were used to search for ANCA in 47 allogeneic bone marrow graft recipients who developed cGVHD and in 43 who did not (controls). Eight patients exhibited ANCA IIF positivity in the cGVHD group, but none in the controls. Specificity was confirmed in antigen-specific assays in only two cGVHD patients, both showing antilactoferrin (anti-LF) activity. One of these patients was followed-up, and the antilactoferrin antibodies were found only at the time of active but limited cGVHD. Among three ANCA IIF-positive patients, two had antinuclear autoantibodies and three antineutrophil alloantibodies secondary to blood transfusion, which may have been responsible for false ANCA IIF positivity. It is concluded that ANCA determination is not useful in patients with cGVHD. Polyclonal activation of B lymphocytes could result in ANCA activity during cGVHD. False-positive ANCA could be due to allo-immunization following blood transfusion. Rare patients may present antilactoferrin antibodies of unknown clinical significance.

Recurrence of immunoglobulin A nephropathy with immunoglobulin A antineutrophil cytoplasmic antibodies following renal transplantation

We previously reported the presence of immunoglobulin A (IgA) antineutrophil cytoplasmic antibodies (ANCAs) in patients presenting IgA nephropathy (IgAN), particularly when associated with Henoch-Schonlein purpura. Most of the patients exhibited IgA ANCAs directed at an unknown 50-kd neutrophil protein but no IgG ANCAs. A subgroup of patients presented IgG as well as IgA ANCAs, suggesting an overlap syndrome between Henoch-Schonlein purpura and microscopic polyangiitis. We aimed at confirming the correlation of IgA ANCA titer with disease activity in a patient presenting IgAN relapse following kidney transplantation. The ANCAs were searched for by isotype- and antigen-specific enzyme-linked immunosorbent assay. Specificity was confirmed by antigen-specific enzyme-linked immunosorbent assay and Western blot analysis. At the onset of the disease in 1989, the patient presented with ANCAs of IgA and IgG class with specificity for myeloperoxidase and no rheumatoid factor. End-stage renal failure developed 1 year afterward. In 1991, he received a cadaveric renal allograft, and 9 months later developed acute nephrotic syndrome with rapidly progressive renal failure and recurrence of IgAN on the kidney transplant. An increase in IgA but not IgG ANCAs was found on clinical relapse after kidney transplantation. We conclude that rare patients may present an overlap syndrome between IgG ANCA-positive systemic vasculitis and IgAN, characterized by the presence of IgG and IgA anti-myeloperoxidase antibodies.

Anti-native and recombinant myeloperoxidase monoclonals and human autoantibodies

Myeloperoxidase (MPO) is one of the main antigen targets of anti-neutrophil cytoplasmic antibodies (ANCA) in systemic vasculitides. It has been suggested that anti-MPO antibodies may recognize a single epitope on recombinant MPO. If confirmed on native MPO, this might allow specific therapeutic intervention with anti-idiotypic MoAbs to prevent antibody antigen interaction which is thought to cause activation of neutrophils and vasculitis. We searched for restriction in the epitope recognition profile in 50 patients with anti-MPO autoantibodies, using both native and recombinant MPO. Mouse monoclonals were purified and tested in competition assays. At least four epitopes were identified on native MPO using these monoclonals and only two were conserved on recombinant MPO. We found that human MPO autoantibody response was not restricted to a single epitope on native MPO, as all sera tested did not show the same profile in competitive studies with monoclonals. Furthermore, 30% of human anti-native MPO sera failed to recognize rMPO.

Anti-lactoferrin autoantibodies: relation between epitopes and iron-binding domain

Anti-neutrophil cytoplasm antibodies (ANCA) have been found in the sera of patients presenting systemic necrotizing microscopic vasculitis, i.e. Wegener's granulomatosis and microscopic polyangiitis. Lactoferrin (LF) is one of the antigens rarely recognized by ANCA, and anti-LF autoantibodies are found in several autoimmune conditions, including rheumatoid vasculitis, rheumatoid arthritis, systemic lupus erythematosus, ulcerative colitis, primary sclerosing cholangitis and Crohn's disease. We analysed the epitopes recognized by human anti-LF antibodies to test whether the heterogeneity of clinical presentation might be due to a different epitope recognition profile. Several monoclonal antibodies were raised and used in competition studies with six human sera. Four distinct epitopes were identified on LF, and LF binding of only one of six sera was inhibited by one of the monoclonals. Thus, anti-LF autoreactivity appears to be polyclonal and not restricted to an immunodominant epitope. Specific epitope profiles cannot be determined in these autoimmune conditions. We hypothesized that the interaction of anti-LF antibodies with the LF iron binding domain might contribute to pathogenesis by inhibiting iron chelation after neutrophil activation, thereby providing increased iron availability for endothelial cell damage. The relation of anti-LF mouse monoclonals or polyclonal human or rabbit antibodies to the LF iron-binding domain was studied in competition assays between 59Fe and these antibodies. Preincubation of LF with monoclonals or anti-LF human sera did not affect the binding of 59Fe on LF. 59Fe-binding kinetic studies showed that rabbit anti-LF polyclonal, but not mouse monoclonals or human anti-LF positive sera, was capable of inhibiting iron binding on LF. Therefore, anti-LF autoantibodies did not appear to modulate LF iron-binding activity. We conclude that LF is a rare antigen specificity for ANCA and that the clinical and pathophysiological relevance of anti-LF autoreactivity remains uncertain.

Anti-neutrophil cytoplasm antibodies in patients with ACR criteria for polyarteritis nodosa: help for systemic vasculitis classification?

The american college of rheumatology (ACR) proposed in 1990 revised clinical criteria for systemic vasculitis classification to define homogeneous group of patients for clinical trials. However, microscopic polyarteritis (MPA) was not clearly identified from polyarteritis nodosa (PAN). Since anti-neutrophil cytoplasm antibodies (ANCA) are markers of disease activity of small vessel vasculitides including MPA, we tested the clinical significance of ANCA in 24 patients with PAN according to the ACR 1990 criteria. Two of 24 patients had ANCA, as defined by indirect immunofluorescence on normal human neutrophils, antigen-specific ELISA and Western blot analysis. However, they exhibited histologically proven small vessel but not medium vessel vasculitis. Furthermore, they had neither artery microaneurysms nor large organ injury consequent upon large vessel occlusion. Although they satisfied ACR criteria for PAN, they probably were misclassified and should be considered as MPA. We conclude that: (i) ANCA are not found in patients with classical PAN in the absence of MPA features; (ii) caution should be exercised when defining PAN according to the ACR 1990 criteria; (iii) ANCA may help systemic vasculitis classification.

High immunoreactivity of lactoferrin contaminating commercially purified myeloperoxidase

Three sets of experiments were performed to investigate the quality of myeloperoxidase (MPO) preparations and anti-MPO reagents. In the first experiment, two groups of three and four mice were immunized with commercially purified MPO (Calbiochem). Immunization was performed in PBS in the first group and in acetate buffer in the second. From the first group, five monoclonals were raised, and their specificities examined by ELISA and immunoblotting. Surprisingly, these antibodies reacted with lactoferrin (LF) and not MPO. In the second group, 13 monoclonals were raised; six of these reacted with MPO and seven reacted with LF. In a second set of experiments, MPO and LF reactivity were tested in different buffer conditions in the ELISA procedure. Slight variations in the detection of contaminating LF were found. In a third experiment, polyclonal reagents directed against MPO and LF were tested in MPO immunoblotting studies. A polyclonal anti-MPO reagent reacted not only with MPO but also with contaminating material including LF. The anti-MPO polyclonal reagent also reacted with LF on immunoblotting. We conclude that: (i) caution should be exercised when defining anti-neutrophil cytoplasm specificities of human sera and monoclonals by ELISA, (ii) the low concentration of contaminating LF in the commercially purified reference MPO preparation should be taken into consideration since it appears to have high immunoreactivity, (iii) changes in MPO immunoreactivity may occur under different buffer and pH conditions.

Autoantibodies to lactoferrin and histone in systemic vasculitis identified by anti-myeloperoxidase solid phase assays

We aimed at confirming the antigen specificity recognized by anti-neutrophil cytoplasm antibodies (ANCA) in patients presenting systemic vasculitis with anti-myeloperoxidase (MPO) activity on ELISA. Thirty-five consecutive patients with reactivity in anti-MPO ELISA and systemic microscopic vasculitides were tested in slot and Western blot analyses. Eleven of 35 sera exhibited binding in Western blot studies with the MPO preparation used in the ELISA: five sera bound at the size of MPO, but five sera reacted with a 78 kD species (p78) co-purifying with MPO, and one serum blotted both MPO and p78. Sequence analysis and antigen-specific assays including Western blot studies showed that p78 is lactoferrin. All anti-lactoferrin positive sera, but no anti-MPO positive sera, also exhibited anti-nuclear binding on HEp2 cells with specificity for histone. We concluded that: (a) a subgroup of patients presenting systemic vasculitis with false anti-MPO reactivity on ELISA had anti-lactoferrin antibodies; (b) anti-lactoferrin was associated with anti-nuclear activity with specificity for histone; (c) these patients had systemic vasculitis without histological evidence of immune complex deposition.

Absence of antineutrophil cytoplasmic antibodies in giant cell arteritis

OBJECTIVE

To determine the prevalence of antineutrophil cytoplasmic antibodies (ANCA) in patients with active giant cell arteritis (GCA).

METHODS

23 patients with GCA were selected according to ACR 1990 criteria. Sera were harvested in all patients at an active stage of the disease and during followup (1 to 3 sera/patient for a total of 50 sera). ANCA positivity was searched for by indirect immunofluorescence (IIF) and enzyme linked immunosorbent assay (ELISA) using a neutrophil extract, and antigen specificity was determined by proteinase 3 (PR3), lactoferrin (LF) and myeloperoxidase (MPO) ELISA:

RESULTS

Only 1/23 patients exhibited reactivity in IgG ANCA ELISA and IIF, with borderline anti-MPO reactivity in ELISA which was not inhibited by preincubation with MPO in the liquid phase, and no reactivity in Western blot analysis. Specificity could not be demonstrated in another patient who had positive IgG ANCA ELISA but negative ANCA IIF and negative antigen specific ELISA: All other patients were ANCA negative.

CONCLUSION

As our patients with GCA did not exhibit typical ANCA when validated antigen specific assays were used, careful laboratory controls and clinical evaluation would seem essential in cases of apparent ANCA positivity.