Klinische Wochenschrift

Klin Wochenschr (1991) 69: 5 5 ~ 5 5 7 002321739100138L

9 Springer-Verlag 1991

Antigen Specificities and Clinical Distribution of ANCA in Kidney Diseases P. Lesavre, L.H. Noel, D. Chauveau, C. Geffriaud, and J.P. Grfinfeld Department of Nephrology and I N S E R M U90, H6pital Necker, Paris

Summary. The antigenic specificity and clinical distribution of the antineutrophil cytoplasmic antibodies (ANCA) in kidney diseases have recently been extensively studied. In patients with systemic vasculitis, the great predominance of two major ANCA antigens, proteinase 3 (PR3) and myeloperoxidase (MPO), is now established. PR3 and MPO are colocalized in the azurophilic granules of neutrophils and translocated to the cell surface during activation, and thus are able to interact with autoantibodies after neutrophil preactivation. Furthermore, by comparison of amino acid and DNA sequences, it has been shown that PR3 is identical to myeloblastin, which has been described independently and is involved in the control of growth and differentiation of leukemic cells. Aside from the two major ANCA antigens, a number of neutrophil cytoplasmic antigens recognized by ANCA have been identified, including human leukocyte elastase, lactoferrin, CAP57, and cathepsin G. These rare ANCA specificities occur in a limited number of patients. The variety of ANCA antigen specificities contrasts, however, with the fact that the vast majority of ANCA-positive sera are monospecific for one single ANCA antigen. With regard to clinical distribution, ANCA have major diagnostic significance in the four conditions in which they are frequently detected: We* Preprint of a lecture to be read at the 22nd Congress of the "Gesellschaft fiir Nephrologie ", Heidelberg, September 1518, 1991 (Editor: Prof. Dr. E. Ritz, Heidelberg)

Abbreviations: A N C A = antineutrophil cytoplasm antibodies; PR3 = proteinase 3 ; MPO = myeloperoxidase; CAP57 = cationic antimicrobial protein 57 kDa; W G = W e g e n e r ' s granulomatosis; CSS = Churg and Strauss syndrome; M P A = microscopic periarteritis; N C G N = necrotic and crescentic glomerulonephritis; IIF = indirect immunofiuorescence; HLE = h u m a n leukocyte elastase; G B M = glomerular basement membrane; IgAN = IgA nephropathy; HSP = Henoch-Sch6nlein purpura

gener's granulomatosis (WG), Churg and Strauss Syndrome (CSS), microscopic periarteritis (MPA), and necrotic and crescentic glomerulonephritis (NCGN). However, the initial dichotomy between MPO-associated vasculitis (NCGN, MPA) and that associated with anti-PR3 antibodies (WG) appears far from absolute. Key words: Antineutrophil cytoplasm antibodies (ANCA) - Proteinase 3 - Myeloperoxidase - Wegener's granulomatosis -Necrotizing and crescentic glomerulonephritis - Systemic vasculitis Periarteritis nodosa - Microscopic periarteritis Goodpasture's syndrome - IgA nephropathy

The identification of a new autoantibody family directed against components of the polymorphonuclear neutrophil cytoplasm (ANCA) and associated with vasculitis and rapidly progressive glomerulonephritis has led to considering these diseases as possible autoimmune disorders. In addition, ANCA constitute a diagnostic tool and a guideline for therapy during follow-up. ANCA were originally identified in 1982 by Davies [6] under indirect immunofluorescence (IIF). They were subsequently observed by Hall [13], and a clear association with Wegener's granulomatosis was further described by Van der Woude in 1985 [32], who also suggested a correlation between the level of ANCA and disease activity. Several methods of detection have been developed. Routine determination is obtained by IIF. Two ANCA patterns are observed [33, 34], namely cytoplasmic ANCA (cANCA) and perinuclear ANCA (pANCA), thus suggesting that at least two different antigens are involved.

P. Lesavre et al. : A N C A Antigen Specificity and Clinical Associations

Antigen Specificity of ANCA The Two Major ANCA Antigens: Proteinase 3 and Myeloperoxidase Identification of the antigens recognized by ANCA is in progress. An increasing number of studies using ELISA, western blot analysis, subcellular fractionation, protein purification, and amino acid and DNA sequencing have allowed better definition of the ANCA autoantigens. The large predominance of the two main antigens, proteinase 3 (PR-3) and myeloperoxidase (MPO), is now well established. Proteinase 3 The cANCA antigen is a DFP binding serine protease of 29 kDa [11, 23], contained in azurophilic granules and released after degranulation by f-metleu-phe (fMLP). Lfidemann et al. [21] proposed that the cANCA antigen is proteinase 3 (PR-3), a constituent of neutrophil primary granules characterized by Kao et al. [19] and distinct from elastase and cathepsin G. Finally, the identity between PR-3 and the target of Wegener's granulomatosis was established by Jenne et al. [14] and confirmed by Jennette et al. [15] and by us. We used PR-3 purified by dye ligand chromatography on Orange A [19] and immunoaffinity on immobilized mouse monoclonal anti-PR-3 (CLB 12.8 kindly provided by R. Goldschmeding), eluted at pH 11.6 in accordance with our observation that PR-3 precipitates at acid pH. The revised PR-3 sequence suggested that this serine protease is identical to AGP7 [14, 35] and to p29 [12, 23]. Recently Gupta et al. [19] have proposed that PR-3, p29, and AGP7 are identical to myeloblastin (mbn) present in HL-60 human leukemic cells and involved in the control of growth and differentiation of leukemic cells [1]. The missing 5' end nucleotide sequence of PR-3 was very recently achieved, and HL60 cells expressed a single m R N A species, indicating that no differential splicing takes place and that PR-3, p29, AGP7, and mbn correspond to a unique protein (Y. Cayre, personal communication). Since most autoantibodies against PR-3 interfere with its enzymatic function, Jenne et al. [14] have proposed that fulminating WG may reflect enhanced granulocyte differentiation. The major plasma inhibitor of PR-3 is el antitrypsin, the molecular weight of the PR-3-eI-AT complex being 83 kDa. This may very well corre-

553

spond to the 91 kDa antigen recognized in several cANCA-positive sera described by Daha et al. [2] in sputum or degranulation supernatant of neutrophils likely to be in contact with plasma proteins including ~I-AT. Myeloperoxidase The original report by Falk and Jennette [8] showed that most pANCA are directed against myeloperoxidase (MPO). This was further confirmed by Cohen Tervaert et al. [5]. Falk et al. showed that the MPO epitope recognized by patient antibodies is expressed by the native molecule (130 kDa) but not by the isolated chains after denaturation and reduction (65, 43, 15 kDa). The inhibitory effect of anti-MPO IgG from patients on MPO enzymatic activity seems marginal, suggesting that the epitope recognized by patient antibodies and the enzymatic site are distinct. By IIF microscopy using ethanol-fixed neutrophils as substrate, two major ANCA types are recognized, strongly indicating that cANCA and pANCA correspond to different antigens. Indeed, the two major ANCA antigens, PR-3 and MPO, respectively, correspond to cANCA and pANCA. However, there are some exceptions. In particular, both patterns may be associated with other rare specificities, and some anti-MPO antibodies exhibit a cANCA pattern. Whether this latter fact is due to irreproducibility in the artifactual redistribution of MPO after ethanol fixation or to a true variability in anti-MPO specificities is not yet clear. Functional Analogy Between the Two Major ANCA Antigens The two major ANCA antigens, PR-3 and MPO, are colocalized in the azurophilic granules of PMN. Both are translocated to the cell surface during PMN activation. This was first observed in the rabbit [2]. Thus the two major ANCA antigens are able to interact directly with ANCA after P M N preactivation. Because of their cationic charge both PR-3 and MPO interact with cell membranes and glomerular basement membrane (GMB) [16, 17]. The enzymatic activity of PR-3 and MPO may be involved in the vascular lesions after ANCAamplified release. Indeed, PR-3 has a capacity exceeding that of elastase or cathepsin G to degrade elastin and induce emphysema in the hamster [19]. The myeloperoxidase - H z O 2 - halide system is also directly involved in tissue injury [16, 18]

554

P. Lesavre et al. : A N C A Antigen Specificity and Clinical Associations

through the generation of hypochlorous acid and chloramine radicals. MPO may also play an indirect role, since hypochlorous acid induces proenzyme activation of collagenase and gelatinase [30].

Rare ANCA Antigens Besides the two major A N C A antigens, a number of P M N cytoplasmic antigens recognized by A N C A have been identified. However, the limited number of patients in whom autoantibodies with rare A N C A specificity are observed (probably less than 5%) prevents any definitive conclusion on their diagnostic and pathogenetic values.

culitic patients and is associated with a p A N C A staining pattern. Pozzi et al. [28] observed anti-LF antibodies in less than 5% of ANCA-positive patients with necrotizing glomerulonephritis and in some cases of lupus nephritis. In their experience anti-LF antibodies were not associated with either MPO or PR-3 antibodies. Conversely, Dolman et al. observed a patient with a combination of A N C A directed against LF, MPO, and elastase. Although rare, the anti-LF A N C A are of theoretical importance with regard to the mechanisms of vasculitis, since LF is the only known A N C A antigen contained in specific granules.

Cathepsin G CAP57 CAP57 is an antimicrobial catinic protein contained in a subpopulation of azurophilic granules in normal P M N [27]. A N C A specific for CAP57 have been demonstrated by Falk et al. CAP57 A N C A produce cANCA staining by IIF and were blocked by monoclonal anti-CAP57. In such cases patient antibodies bound to a 57-kDa band in western blot analysis and to HPLC-purified CAP57 in ELISA. The exact frequency of this particular A N C A and its cooccurence with anti-PR-3 antibodies is not yet fully established. Elastase Autoantibodies to human leukocyte elastase (HLE), a constituent of the azurophilic granule, have been described independently by Goldschmeding etal. [11] in vasculitis and by N/issberger et al. [22] in systemic lupus erythematosus. In general these anti-HLE antibodies produce a perinuclear IIF staining pattern (pANCA). They usually occur in the absence of anti-MPO or anti-PR-3 antibodies in less than 0.5% of the pANCA-positive sera; two exceptional patients, treated with propylthiouracil for hyperthyroidism, were found to have antibodies to HLE and PR-3 [7]. Lactoferrin Antibodies to lactoferrin (LF) were first reported by Thomson [31] in vasculitis. It is also the experience of other groups, including ours, that anti-LF A N C A may occur as single specificity in rare vas-

A N C A directed against cathepsin G, a 25-kDa neutral protease of the azurophilic granules, seems very unusual but has been observed by Flesh et al. [10] in WG patient sera selected on the basis of atypical IIF A N C A staining (homogeneous noncANCA pattern). The presence of anti-cathepsin G antibodies was demonstrated by ELISA and usually occurs in combination with anti-elastase and anti-MPO antibodies. Eosinophil Peroxidase Dolman et al. [7] found antibodies directed against eosinophil peroxidase (EPO) reacting only with eosinophils by IIF and precipitating the two bands at 58 kDa and 14 kDa identical to those precipitated by polyclonal rabbit anti-EPO antibodies.

Single Versus Multiple A N C A Specificities The vast majority of ANCA-positive sera are monospecific for one single A N C A antigen. It is striking that anti-PR-3 and anti-MPO antibodies almost never coocur in one given serum. Among more than 400 ANCA-positive sera tested, we have found such an association in only one serum. With regard to the rare specificities, anti-HLE and antiLF antibodies usually occur as a single specificity but are occasionally associated with anti-PR-3 and anti-MPO, respectively. Finally, anti-CAP57 antibodies may be isolated or associated with anti-PR3 ANCA. The limited number of patients with multispecific A N C A does not allow firm conclusions but may favor the hypothesis that A N C A are mainly directed against PR-3 or MPO and that multiple A N C A specificities in one single serum

P. Lesavre et al. : ANCA Antigen Specificity and Clinical Associations

"'~'"1 ;::::::." Anti c~ granules (PR3) Anti MP0

:i:!:!:i Z':':" .:.:.:.: ??:.:.:.:.:.: :-Z.:.

(52) [27) (24) 17) {13) (51) (42)(100){3f])

Fig. 1. Antigen specificity and frequency of ANCA in Wegener's granulomatosis (WG), microscopic periarteritis (MPA), crescentic and necrotizing glomerulonephritis (RPGN), Churg and Strauss syndrome (CSS), periarteritis nodosa (PAN), Goodpasture syndrome (aGNM), systemic lupus erythematosus (SLE), Berger's disease (IgAN), and Henoch Sch6nlein purpura (HSP). The proportions of patients' sera studied are indicated (positivity for anti-PR3 antibodies [], antiMPO [], or negativity by both ELISAs [])

probably reflect a secondary response to antigens that are released from P M N as the vasculitis progresses.

Diagnostic Value of ANCA in Kidney Diseases We have determined A N C A by immunofluorescence and specific ELISA for PR3 and MPO in a large series of sera referred to our laboratory. Results are presented in Fig. I and are coherent with published results from other groups.

Wegener's Granulomatosis A N C A were first observed in the sera of patients with viral infections [6] or presenting with glomerulonephritis and vasculitis [13]. Their association with WG was first recognized in 1985 [32] and then confirmed by various groups [3, 11, 21, 24]. Usually the A N C A associated with W G are directed against PR3, with a cANCA pattern by IF. However, a certain number of authentic W G are associated with anti-MPO antibody and a p A N C A pattern. This percentage varies from 20% [3, 24] to 40% in our series of consecutive patients studied from 1989 to 1990 (Fig. 1). Overall, the presence of PR3 or MPO antibody is observed in more than

555

90% of the cases. We also found about 5 to 10% of Wegener patients with A N C A by IF and negativity of the two specific ELISA, particularly in children and in WG patients with hepatic or digestive localization. We have studied 17 patients followed up in our department and diagnosed with W G by the presence of granulomas and necrotizing and destructive upper respiratory tract (URT) lesions. A N C A were detected in all by IF and PR3 ELISA was positive in ten (59%), eight males and two females. U R T lesions were present in all, necrotizing crescentic glomerulonephritis (NCGN) in eight, lung involvement in four. Histologically demonstrated granuloma was present in seven patients (renal 3, nasal 3, lung 1). MPO ELISA was positive in three females (18%) and was associated with N C G N in all and lung involvement in two. All had granulomas (renal and nasal 2, occular 1). Finally, four patients, three males and one female, were A N C A positive by IF but negative with PR3 and MPO ELISA. Cholestasis was present in two, cytolytic hepatitis in one, and ileo-cecal ulceration in one. All of them had granuloma (lung 1, renal 1, nasal 2).

Systemic Vasculitis The presence of A N C A in patients with other forms of systemic vasculitis led to the hypothesis of a common mechanism for those diseases. The classification of systemic vasculitis is not yet completely clear [9, 20]. It is, however, possible to individualize, in addition to WG, periarteritis nodosa (PAN) and Churg and Strauss syndrome (CSS). Other forms of systemic vasculitis involving the small and medium-sized arteries do not correspond to any of these three anatomical entities. It has been proposed to term these forms "microscopic periarteritis" (MPA) when they associate systemic vasculitis and N C G N [29]. We have observed, as shown in Fig. 1, that MPA and CSS are very frequently associated with A N C A but less frequently in typical macroscopic PAN. These results are similar to those obtained by others [4]. The A N C A specificity is indicated in the figure. They are exclusively anti-MPO in CSS and are anti-MPO twice as frequently as anti-PR3 in MPA. In a specific study of patients followed in this department, we assessed the clinical profile of the pulmonary renal syndrome associated with antiMPO ANCA. Thirteen patients (5 males, 8 females) were selected on the basis of acute renal failure and pulmonary hemorrhage, diagnosed on bronchial alveolar lavage and positive MPO by

556

P. Lesavre et al. : ANCA Antigen Specificity and Clinical Associations

ELISA. On renal biopsy, twelve had glomerular capillaritis without granuloma; only one patient had extraglomerular vasculitis. In four patients, pulmonary involvement was delayed, occurring two to five years after the remission of the glomerular capillaritis. Patients were classified as MPA, 10; WG, 1 ; classical PAN 1 ; and SLE, 1.

minority of patients with IgAN extends the spectrum of disease associated with this ANCA. However, the absence of specific IgA binding to neutrophil granule components in IgAN or HSP and the failure to correlate with clinical features implies that ANCA are not involved in the immunopathogenesis of glomerular injury in IgA-associated glomerulopathies.

Conditions Rarely Associated with ANCA Conclusion

Antiglomerular Basement Membrane (GBM) Nephritis." We have observed [26] IgG class ANCA with anti-MPO specificity in about 15% of Goodpasture's syndrome. The coexistence of ANCA and antiglomerular basement membrane antibodies (~GBM) has recently been reported. However the frequency, antigen specificity and clinicopathological significance of this association remains to be defined. We have prospectively studies sera from 52 patients with ~GBM disease characterized by IgG along the glomerular basement membrane and associated anti-GBM in the serum. ANCA were detected by IIF in 7 cases. All 7 ~GBM diseaserelated ANCA were characterized as MPO in ELISA and by inhibition experiments. In 5 cases, ANCA were detected at presentation and in 2, ANCA developed during follow-up. The renal histology in the ANCA-positive patients could not be distinguished from classical ~GBM disease. However, the late occurrence of ANCA, after the disappearance of c~GBM antibodies, was associated with clinical and histological evidence of leukocytoclastic vasculitis. Thus, eGBM disease-associated ANCA occur in a substantial minority of patients (13.5%); MPO is the major neutrophil granule antigen in these cases. The detection of anti-MPO antibodies in this setting does not appear to be related to significantly different glomerular histology but may identify a distinct subset with coexisting systemic vasculitis.

IgA Nephropathy: In a cross-sectional study, sera from 2 of 100 patients with IgAN but none of 30 with HSP had raised IgG binding to MPO, and specific IgA ANCA could not be demonstrated. In 1 of the positive cases the anti-MPO antibody titers did not fluctuate in association with clinical features during extended follow up (8 years). A further 20 IgAN and 10 HSP patients were studied longitudinally in different clinical phases; none developed ANCA during macroscopic hematuria, declining renal function, or rash [25]. The finding of IgG anti-MPO antibodies in a

The role of the two major ANCA antigens, PR3 and MPO, in the vasculitic process including Wegener's granulomatosis is now well established. Ongoing studies will elucidate the role and specificity of ANCA in other pathological conditions such as ulcerative colitis and rheumatoid arthritis. Acknowledgements. The authors are grateful to Drs. D. O'Donoghue, P. Ronco, P. Vanhille, E. Alline, S. Houhou, F. Kirhaoui, P. Nusbaum and F. Lallemand for their contributions. We wish to thank M.A. Monod for preparing the manuscript.

References 1. Bories D, Raynal MC, Solomon DH et al. (1989) Downregulation of a serine protease, myelobastin, causes growth arrest and differentiation of promyelocytic leukemia cells. Cell 59: 959-968 2. Brown WJ, Shannon WA Jr, Snell WJ (1983) Specific and azurophilic granules from rabbit polymorphonuclear leukocytes. III: Cell surface localization of granule membrane and content proteins before and after degranulation. J Cell Biol 96:104~1046 3. Cohen Tervaert JW, Van der Woude F J, Fauci AS, Ambrus JL, Velosa J, Keane WF, Meijer S, Van der Giessen M, The TH, Van der Hem GK, Kallenberg CGM (1989) Association between active Wegener's granulomatosis and anticytoplasmic antibodies. Arch Intern Med 149 : 2461 2465 4. Cohen Tervaert JW, Goldschmeding R, Elema JD, Limburg PC, Van der Giessen M, Huitema MG, Koolen MI, Hen6 RJ, The TH, Van der Hem GK, Von dem Borne AEGK, Kallenberg CGM (1990a) Association of autoantibodies to myeloperoxidase with different forms of vasculitis. Arthritis Rheum 33:1264-1272 5. Cohen Tervaert JW, Goldschmeding R, Elema JD et al. (1990 b) Autoantibodies against myeloid lysosomal enzymes in crescentic glomerulonephritis. Kidney Int 37 : 799-806 6. Davies D J, Moran JE, Nial JF, Ryan GB (/982) Segmental necrotizing glomerulonephritis with antineutrophil antibody: possible arbovirus aetiology. BMJ 265:606 7. Dolman KM, Goldschmeding R, Sonnenberg A et al. (1990) ANCA related antigens. Acta Pathol Microbiol Immunol Scand 98 :28 8. Falk R J, Jennette JC (1988) Anti-neutrophil cytoplasmic autoantibodies with specificity for myeloperoxidase in patients with systemic vasculitis and idiopathic necrotizing and crescentic glomerulonephritis. N Engl J Med 318:1651 1657

P. Lesavre et al. : ANCA Antigen Specificity and Clinical Associations 9. Fauci AS, Haynes BF, Katz P (1978) The spectrum of vasculitis: clinical, pathological, immunologic, and therapeutic considerations. Ann Intern Med 89:660 676 10. Flesch BK, Lampe M, Rautman A e t al. (1991) Anti-elastase cathepsin G and lactoferrin antibodies in sera with cANCA or with atypical fluorescence staining pattern. In: The Third International Workshop on ANCA. Washington, November 1990 (abstr). Am J Kidney Dis (in press) II. Goldschmeding R, van der Schoot CE, ten Bokkel Huinink D et al. (1989) Wegener's granulomatosis autoantibodies identify a novel diisopropylfluorophosphate-binding protein in the lysomes of normal human neutrophils. J Clin Invest 84:1577-1587 12. Gupta SK, Niles JL, McCluskey RT et al. (1990) Identity of Wegener's autoantigen (p29) with proteinase 3 and myeloblastin. Blood 76 : 2162 13. Hall JB, Wadham BMCN, Wood JC et al. (1984) Vasculitis and glomerulonephritis: a subgroup with an antineutrophil cytoplasmic antibody. Aust NZ J Med 14:277 278 14. Jenne DE, Tschopp J, Lfidemann J e t al. (1990) Wegener's autoantigen decoded. Nature 346 : 520 15. Jeunette JC, Hoidal JH, Falk RJ (1990) Specificity of antineutrophil cytoplasmic autoantibodies for proteinase 3. Blood 75:2263-2264 16. Johnson RJ, Couser WG, Chi EY et al. (1987) New mechanisms for glomerular injury. Myeloperoxidase hydrogen peroxide- halide system. J Clin Invest 79:1379-1387 17. Johnson RJ, Couser WG, Alpers CE et al. (1988a) The human neutrophil serine proteinases, elastase and cathepsin G, can mediate glomerular injury in vivo. J Exp Med 168:1169-1174 18. Johnson RJ, Guggenheim SJ, Klebanoff SJ et al. (1988) Morphologic correlates of glomerular oxidant injury induced by the myeloperoxidase hydrogen peroxide - halide system of the neutrophil. Lab Invest 58:294-301 19. Kao RC, Wehner NG, Skubitz KM et al. (1988) Proteinase 3. A distinct human polymorphonuclear leukocyte proteinase that produces emphysema in hamster. J Clin Invest 82:1963-1973 20. Lie JT (1987) The classification and diagnosis of vasculitis in large and medium-sized blood vessels. Pathol Ann 22:125-162 21. Lfidemann J, Utecht B, Gross WL (1990) Anti-neutrophil cytoplasm antibodies in Wegener's granulomatosis recognize an elastinolytic enzyme. J Exp Med 171:357 362 22. N~issberger L, Jonsson H, Sj6holm AG et al. (1989) Circulating anti-elastase in systemic lupus erythematosus. Lancet I (letter) : 509 23. Niles JL, McCluskey RT, Ahmad MF et al. (1989) Wegener's granulomatosis autoantigen is a novel neutrophil serine proteinase. Blood 74:1888 1893

557

24. N611e B, Specks V, Lfidermann J, Rohrbach MS, De Remee RA, Gross WL (1989) Anticytoplasmic autoantibodies: their immunodiagnostic value in Wegener's granulomatosis. Ann Intern Med 111:28-40 25. O'Donoghue D, Nusbaum P, Halbwachs-Mecarelli L, Lesavre P (1991a) Antimyeloperoxidase antibodies of IgG isotype in igA nephropathy (abstract). Am J Kidney Dis (in press) 26. O'Donoghue P, Vanhille P, Lesavre P, Noel LH (1991b) Antimyeloperoxidase antibodies in antiglomerular basement membrane disease (abstract). Am J Kidney Dis (in press) 27. Pereira A, Spitznagel J, Winton EF et al. (1990) The ontogeny of a 57-kD cationic antimicrobind protein of hmnan polymorphonuclear leukocytes: localization to a novel granule population. Blood 76 : 825 834 28. Pozzi C, Radice A, Rota Set al. (1991) Clinical significance of anti-lactoferrin antibodies in renal disease. In : The Third International Workshop on ANCA. Washington, November 1990 (abstract). Am J Kidney Dis (in press) 29. Savage COS, Winearls CG, Evans D J, Rees A J, Lockwood CM (1985) Microscopic polyarteritis: presentation, pathology and prognosis. Q J Med 56:467-483 30. Shah SV, Baricos WH, Basci A (1987) Degradation of human glomerular basement membrane by stimulated neutrophils. Activation of a metallo-proteinase(s) by reactive oxygen metabolites. J Clin Invest 79:25 31 31. Thomson RA, Lee SS (1989) Antineutrophil cytoplasmic antibodies. Lancet I (letter):670 671 32. Van der Woude FJ, Daha MR, van Es LA (1989) The current status of neutrophil cytoplasmic antibodies. Clin Exp Immunol 78:143 148 33. Wiik A (1989) Delineation of a standard procedure for indirect immunofluorescence detection of ANCA. APMIS 97 (Suppl 6): 12 34. Wiik A, Van der Woude FJ (1990) The new ACPA/ANCA nomenclature. Neth J Med 36:107 108 35. Wilde CG, Snable JL, Griffith JE et al. (1990) Characterization of two azurophil granule proteases with active site homology to neutrophil elastase. J Biol Chem 265:2038-2041 Received: April 1, 1991 Accepted: May 11, 1991 Prof. P. Lesavre M.D., Ph.D. Department of Nephrology, H6pital Necker, 161 rue de S6vres, F-75743 Paris C6dex 15, France