Coexistence of Anti-Glomerular Basement Membrane Antibodies and Myeloperoxidase-ANCAs in Crescentic Glomerulonephritis

Coexistence of Anti–Glomerular Basement Membrane Antibodies and Myeloperoxidase-ANCAs in Crescentic Glomerulonephritis Abraham Rutgers, MD, PhD, Marjan Slot, MD, Pieter van Paassen, MD, PhD, Peter van Breda Vriesman, MD, PhD, Peter Heeringa, PhD, and Jan Willem Cohen Tervaert, MD, PhD ● Background: In a substantial proportion of patients with crescentic glomerulonephritis (CGN), both anti– glomerular basement membrane (GBM) antibodies and antineutrophil cytoplasmic antibodies (ANCAs) with specificity for myeloperoxidase (MPO-ANCA) are detected. In the present study, we questioned whether histological and clinical features of patients with both ANCA and anti-GBM antibodies differ from those of patients with either ANCA or anti-GBM alone. Methods: We reviewed the Limburg renal biopsy registry (1978 to 2003; n ⴝ 1,373) for cases of CGN. The presence of linear fluorescence on renal biopsy and the presence of ANCA and/or anti-GBM antibodies were measured. Subsequently, we assessed patient characteristics and follow-up and compared histological findings among the different groups. Results: We identified 46 MPO-ANCA–positive, 10 doublepositive, and 13 anti-GBM–positive patients. Mean ages were 63, 64, and 52 years (P ⴝ 0.04), and serum creatinine levels were 5.0, 10.3, and 9.6 mg/dL (445, 910, and 850 ␮mol/L), respectively (P ⴝ 0.01). Granulomatous periglomerular inflammation was found in either MPO-ANCA– or double-positive patients, but not in anti-GBM–positive patients with CGN without MPO-ANCAs. Patient survival among the 3 groups was different, although not statistically significant (log rank P ⴝ 0.17, with 75%, 79%, and 100% alive at 1 year, respectively). Renal survival analysis showed significant differences among the 3 groups (P ⴝ 0.04, with 65%, 10%, and 15% off dialysis therapy at 1 year, respectively). Conclusion: In patients with both anti-GBM antibodies and MPO-ANCAs, histological findings differ from those of patients with anti-GBM antibodies only. However, renal survival in these patients is not better than that in anti-GBM–positive patients and is worse compared with patients with MPO-ANCAs only. Am J Kidney Dis 46: 253-262. © 2005 by the National Kidney Foundation, Inc. INDEX WORDS: Glomerulonephritis; antineutrophil cytoplasmic antibodies (ANCAs); anti– glomerular basement membrane (GBM) antibodies.

C

RESCENTIC glomerulonephritis (CGN) is a renal biopsy finding clinically characterized by the presence of autoantibodies or antibody complexes, such as anti–glomerular basement membrane (GBM) antibodies (type I CGN), immune complexes (type II CGN), or antineutrophil cytoplasmic antibodies (ANCAs; type III CGN).1 Histologically, these antibody-mediated forms of CGN are characterized by vasculitis of the glomerular capillaries, resulting in the formation of cellular crescents in most glomeruli.2 In patients with anti-GBM–mediated CGN, immunofluorescence typically shows linear immunoglobulin deposition along the GBM with or without complement deposition, and in patients with immune-complex–associated CGN, immunofluorescence shows widespread immune complex deposition with complement activation, whereas immunofluorescence in those with ANCA-associated CGN shows absent or scanty (ie, pauciimmune) immune deposits.1,3 Anti-GBM antibodies are associated with Goodpasture disease (antiGBM antibody disease with pulmonary hemorrhage) and renal-limited anti-GBM nephritis, whereas ANCAs are associated with Wege-

ner granulomatosis, Churg-Strauss syndrome, microscopic polyangiitis, or idiopathic necrotizing crescentic glomerulonephritis.4 The pathogenic effect of anti-GBM antibodies has been shown by Lerner et al5 in their 1967 report, in which they showed that human antiGBM immunoglobulins could induce CGN in squirrel monkeys. No such data exist for ANCAs, and pathogenicity has been questioned. However, for 1 of 2 types of ANCA associated with CGN (ie, myeloperoxidase [MPO]-ANCA), From the Department of Clinical and Experimental Immunology, Cardiovascular Research Institute Maastricht, Maastricht University, The Netherlands. Received December 22, 2004; accepted in revised form May 4, 2005. Originally published online as doi:10.1053/j.ajkd.2005.05.003 on June 21, 2005. Supported in part by a grant from ZonMW (M.S). Address reprint requests to Peter Heeringa, PhD, Department of Clinical and Experimental Immunology, Maastricht University, PO Box 616, 6200 MD Maastricht, The Netherlands. E-mail: [email protected] © 2005 by the National Kidney Foundation, Inc. 0272-6386/05/4602-0008$30.00/0 doi:10.1053/j.ajkd.2005.05.003

American Journal of Kidney Diseases, Vol 46, No 2 (August), 2005: pp 253-262

253

254

RUTGERS ET AL

Xiao et al6 showed in mice that MPO-ANCA can induce CGN. Interestingly, a substantial proportion of patients with anti-GBM–mediated CGN also are positive for ANCAs (predominantly MPOANCAs).7 It is unclear whether ANCA-associated CGN predisposes to the development of anti-GBM disease or ANCA positivity occurs in the course of anti-GBM CGN.8 In addition, the respective pathophysiological roles of ANCAs and anti-GBM antibodies in double-positive patients with CGN have not been clarified. However, it was suggested that double-positive patients have a better prognosis than patients with anti-GBM nephritis.9,10 Conversely, Heeringa et al11 showed that MPO-ANCAs increase the severity of anti-GBM–mediated glomerulonephritis in a rat model of anti-GBM nephritis. In the present study, we questioned whether histological and clinical features of CGN in patients with both ANCA and anti-GBM antibodies differ from those in patients with either ANCA or anti-GBM alone. METHODS

Study Population In this study, we reviewed all entries (1978 to 2003; n ⫽ 1,373) in our regional renal biopsy registry (Limburg, The Netherlands)12 for the presence of either pauci-immune CGN or CGN with linear fluorescence. CGN is defined as the presence of any crescentic glomerulus in which 25% or more of Bowman space was occupied by extracapillary cellular proliferation in the renal biopsy specimen.13 Biopsies that showed CGN with immune complex deposits were excluded from further study. When no adequate material for immunofluorescence could be obtained or immunofluorescence failed for any other reason, cases were still included and serological tests were performed. We then analyzed retrospectively the presence of anti-GBM antibodies and ANCAs in stored serum samples obtained at the time of renal biopsy. Based on the Dutch central bureau for statistics (www.statline.nl), the population served by the registry was estimated to be 720,000 on average during the study period (750,000 in 2000 and 690,000 in 1980). For the purpose of this report, we restricted subsequent analysis to MPOANCA– and/or anti-GBM–positive patients because the number of proteinase 3 (PR3)-ANCA– and anti-GBM double-positive patients is too little to draw conclusions.

Histological Scoring Renal biopsy specimens were processed for light microscopy and immunofluorescence as described elsewhere.12 Biopsy specimens had to contain 4 or more glomeruli per section to qualify for evaluation. To assess the activity and chronicity of the lesions, serial 1-␮m sections were stained

with silver methenamine and counterstained with hematoxylin eosin. Numbers of normal glomeruli and numbers of glomeruli with global sclerosis were counted. Each glomerulus was assessed separately for the presence of fibrinoid necrosis and cellular and fibrous crescents (the latter 2 had to cover at least 25% of the glomerular surface area). Percentages of affected glomeruli were calculated by dividing the number of glomeruli with lesions by the total number of glomeruli. Interstitial lesions, such as interstitial inflammation, interstitial fibrosis, and tubular atrophy, were graded semiquantitatively on a scale of 0 to 3 (absent, mild, moderate, and severe, respectively). The presence of periglomerular granulomatous inflammation was noted. Periglomerular granulomas were defined as described by Wegener14 in 1939 and Former15 in 1950 and included a periglomerular round cell infiltrate (lymphocellular/plasmacellular) with locally epitheloid-like cells, partial (⬎50%) to total destruction of Bowman capsule, and (remnants of) glomerular capillaries in the center (Fig 1). Giant cells sometimes were detected, as well. For calculation of activity scores (modified from16), the presence of fibrinoid necrosis and cellular crescents was scored in a range of 0 to 3, in which 0 stands for absence; 1, less than 25% of glomeruli involved; 2, 25% to 50% of glomeruli involved; and 3, greater than 50% of glomeruli involved. Additionally, the grade of interstitial inflammation was included in the activity score, resulting in a maximum score of 9. The chronicity score (modified from16) included the percentage of glomeruli with fibrous crescents and grade of interstitial fibrosis and tubular atrophy, resulting in a maximum chronicity score of 9. The presence of granulomas was not included in the activity or chronicity score. Immunofluorescence findings were categorized into 4 groups: (1) no immune deposits; (2) scanty nonlinear immune deposits; (3) linear immunofluorescence, as found in anti-GBM nephritis; and (4) a final group in which insufficient material was present for immunofluorescent evaluation. All biopsies were scored by one nephropathologist blinded to clinical diagnosis (P.v.B.V.).

ANCA and Anti-GBM Testing ANCAs were evaluated by using indirect immunofluorescence on ethanol-fixed neutrophil granulocytes (INOVA Diagnostics, San Diego, CA) and MPO-ANCA– and PR3ANCA–specific enzyme-linked immunosorbent assays (ELISAs; Euro-Diagnostica BV, Arnhem, The Netherlands).17 The presence of ANCAs was confirmed by means of a capture ELISA specific for PR3-ANCA17 or MPOANCA.18 Anti-GBM antibodies were tested by means of commercial ELISA (Euro-Diagnostica).

Diagnostic Criteria and Clinical Characteristics At the time of renal biopsy, patient sex, age, and blood pressure were obtained; 24-hour urine was collected and analyzed for volume and total protein content; and serum samples were obtained for determination of creatinine and antibody titers. Anuria and oliguria are defined as urinary output less than 50 or 400 mL/24 h, respectively. Anti-GBM–mediated glomerulonephritis was diagnosed based on evidence of anti-GBM antibody formation. This

ANTIBODIES IN CRESCENTIC GLOMERULONEPHRITIS

255

Fig 1. (A) Periglomerular granulomas surrounding a sclerotic and crescentic glomerulus, with complete destruction of Bowman capsule. (B) A similar process, with incomplete destruction of Bowman capsule. (Methenamine silver staining; original magnification: [A] ⴛ400, [B] ⴛ200.)

could be shown by the presence of linear fluorescence of immunoglobulin G along the GBM in the renal biopsy specimen and/or the presence of specific circulating antiGBM antibodies.1 MPO-ANCA–associated glomerulonephritis was diagnosed based on the presence of pauci-immune CGN in combination with the presence of MPO-ANCA, detected by means of ELISA. Patients with anti-GBM– mediated CGN who tested positive for MPO-ANCAs by means of ELISA are referred to as double positive.

(survival without end-stage renal disease) are calculated with subjects who died with independent renal function censored. Variables included in survival analyses were disease groups and the presence of periglomerular granulomatous inflammation. Correlations were tested by using Pearson analysis, and all analyses were performed using the statistical programs SPSS, version 10.1 (SSPS Inc, Chicago, IL), and GraphPad Prism, version 4.00 (GraphPad Software Inc, San Diego, CA).

Statistical Analysis

RESULTS

Data are shown as mean ⫾ SD unless stated otherwise. Analysis of variance was used to test significant differences among groups, and Bonferroni-corrected post testing was performed as indicated. Cross-tabulated data were tested by using chi-square analysis. Survival curves for patient and renal survival were calculated by using Kaplan-Meier estimates for survival distribution. The end point for renal survival analysis was the start of renal replacement therapy or death caused by renal failure. Differences between groups in survival were analyzed by means of log-rank test with survival time as a dependent variable. Renal survival rates

From 1978 to 2003, a total of 127 of 1,373 (9.2%) biopsies showed pauci-immune or antiGBM–mediated CGN. Forty-seven specimens were PR3-ANCA positive, 46 specimens were MPO-ANCA positive, 5 specimens were both MPO-ANCA and PR3-ANCA positive, 10 specimens were double positive (ie, anti-GBM and MPO-ANCA positive), and 13 specimens were anti-GBM positive (Table 1). One case identified

Table 1. Patient Characteristics at Time of Biopsy

Age (y) Male sex (%) Systolic/diastolic blood pressure (mm Hg) Proteinuria (g/24 h) Anuria/oliguria (%) Serum creatinine (mg/dL)

MPO-ANCA (n ⫽ 46)

Double Positive (n ⫽ 10)

Anti-GBM (n ⫽ 13)

P

63 ⫾ 12.7 67 150/83 2.0 (0.0-1.1) 12 5.0 ⫾ 2.9

64 ⫾ 8.7 80 158/84 1.3 (0.0-2.6) 63 10.3 ⫾ 5.6

52 ⫾ 20.6 39 145/85 2.3 (0.6-5.4) 50 9.6 ⫾ 8.1

0.043 0.082 0.45/0.9 0.315 ⬍0.01 0.01

NOTE. Values expressed as mean ⫾ SD, median (range), or percent. P indicates significance level calculated by means of analysis of variance statistics over the 3 groups. To convert serum creatinine in mg/dL to ␮mol/L, multiply by 88.4.

256

RUTGERS ET AL Table 2. Histopathologic Characteristics of the 3 Groups

Active

Chronic

No. of glomeruli Normal glomeruli* Necrotic glomeruli* Cellular crescents* Fibrous crescents* Activity score Interstitial inflammation* Interstitial fibrosis* Tubular atrophy* Chronicity score Periglomerular granuloma (%)

MPO-ANCA

Double Positive

Anti-GBM

P

16 (5-65) 30 (0-89) 4 (0-20) 18 (0-75) 27 (0-100) 2.8 (0-5) 1.4 (0-3) 1.3 (0-3) 1.5 (0-3) 4.3 (0-8) 11

12 (4-21) 16 (0-44) 9 (0-33) 27 (0-75) 31 (0-100) 3.7 (2-7) 1.7 (0-3) 1.3 (0-2) 1.7 (0-3) 4.3 (0-8) 40

10 (5-17) 20 (0-75) 9 (0-60) 29 (0-90) 43 (0-100) 3.5 (1-6) 1.5 (0-3) 1.2 (0-2) 1.3 (0-2) 4.4 (0-7) 0

0.12 0.14 0.16 0.24 0.31 0.18 0.66 0.92 0.53 0.99 0.02

NOTE. Granulomatous inflammation is absent in the GBM group. Numbers in parentheses indicate maximum and minimum values. *Expressed as percentage of total glomeruli.

positive for both PR3-ANCA and anti-GBM antibodies was excluded because we restricted the analysis to patients with MPO-ANCA and antiGBM. Overall annual incidences of MPO-ANCA, double positive, and anti-GBM single-positive CGN were calculated to be 2.6, 0.6, and 0.7 per million per year (based on an average population of 720,000), respectively. When analyzing cases from 1978 to 1991 and 1991 to 2003, a doubling in number of CGN-positive biopsy specimens was observed (24 to 45 specimens). Patient characteristics at the time of biopsy differed among groups (Table 1). Mean age was higher in the MPO-ANCA–positive and doublepositive groups compared with the anti-GBM– positive group (63 and 64 versus 52 years, respectively; P ⫽ 0.04), as was the percentage of men (67%, 80%, and 39%, respectively; P ⫽ 0.08). However, serum creatinine levels and percentages of patients presenting with oliguria or anuria were lower in the MPO-ANCA singlepositive group (5 mg/dL [445 ␮mol/L] and 12%, respectively) compared with the double-positive (10.3 mg/dL [910 ␮mol/L] and 63%, respectively) and anti-GBM–positive groups (9.6 mg/dL [850 ␮mol/L ] and 50%; P ⬍ 0.01 and P ⫽ 0.01, respectively; Table 1). Histopathologic findings also differed among the 3 groups (Table 2). Activity scores tended to be higher for the anti-GBM–positive and doublepositive groups, whereas chronicity scores tended to be lower in these 2 groups compared with the

MPO-ANCA–positive group. Percentage of normal glomeruli correlated with serum creatinine level at diagnosis (P ⬍ 0.01; r ⫽ 0.5). Periglomerular granulomatous inflammation was observed in only MPO-ANCA–positive and double-positive patients (11% and 40% of patients), but not those with anti-GBM–mediated CGN. In 3 patients, giant cells were found. Patients with periglomerular granulomas (n ⫽ 9) had similar creatinine levels (7.6 ⫾ 6.8 mg/dL [673 ⫾ 602 ␮mol/L] versus 6.4 ⫾ 4.9 mg/dL [564 ⫾ 435 ␮mol/L]) and the same chronicity index scores (4.0 ⫾ 1.9 versus 4.3 ⫾ 2.3) compared with patients without granulomas (n ⫽ 58), respectively. However, these patients had less proteinuria (protein, 0.781 ⫾ 0.573 versus 2.821 ⫾ 2.478 g/24 h; P ⫽ 0.02) and a higher activity index score (4.4 ⫾ 1.2 versus 2.8 ⫾ 1.5; P ⫽ 0.03). Immunofluorescence studies of renal biopsy specimens showed that MPO-ANCA–positive patients did not have linear immunofluorescence (by definition), but in 33% of biopsy specimens, scanty immune deposits were noted. In doublepositive patients, immunofluorescence in 50% of biopsy specimens showed scanty immune deposits, and in 40%, showed linear fluorescence. Within this double-positive group, patients with linear fluorescence presented with a greater percentage of cellular crescentic glomeruli compared with patients with scanty immune deposits (55% versus 10%; P ⬍ 0.05, respectively). Serum creatinine levels also were higher, although

ANTIBODIES IN CRESCENTIC GLOMERULONEPHRITIS

257

Table 3. Immunofluorescence Findings Immunofluorescence Pattern

No deposits (% of group) Scanty, not linear deposits Linear deposits No material

MPO-ANCA

Double Positive

Anti-GBM

27 (59)

0 (0)

0 (0)

15 (33) 0 (0) 4 (9)

5 (50) 4 (40) 1 (10)

2 (15) 10 (77) 1 (8)

NOTE. Values expressed as number (percent).

this did not reach significance (13.8 ⫾ 7.7 mg/dL [1,219 ⫾ 678 ␮mol/L] versus 7.5 ⫾ 2.4 mg/dL [665 ⫾ 209 ␮mol/L]; P ⫽ 0.2). In 77% of the anti-GBM–positive group, linear fluorescence could be observed (Table 3). Table 4 lists therapeutic regimens and patient follow-up data. Data for 7 of 69 patients were not accessible for analysis (all MPO-ANCA singlepositive patients), and these patients subsequently were excluded from survival analysis. In 38% of cases, treatment was clinically considered futile and thus not started (Table 4). Average patient follow-up was 5.1 years (range, 2 days to 22 years). Figure 2 shows patient survival for the 3 different patient groups within 1 year. Patient survival among the 3 groups was different; the anti-GBM group had the best patient cumulative survival (100% at 1 year), followed by the double-positive group (79% at 1 year) and the MPO-ANCA–positive group (75% at 1 year), although differences did not reach statistical significance (P ⫽ 0.17). Figure 3 shows renal survival for the 3 groups within 1 year. Combined log-rank analysis showed significant differences among the 3 groups (P ⫽ 0.04). Pairwise comparison per Table 4. Therapeutic Regimens for the Different Patients

Therapy

MPO-ANCA

Double Positive

Other Cytostatic therapy Cytostatic therapy and plasma exchange Unknown

9 (20) 29 (63)

2 (20) 3 (30)

5 (38) 0 (0)

4 (9) 4 (9)

4 (40) 1 (10)

7 (54) 1 (8)

Anti-GBM

NOTE. Values expressed as number (percent). Other includes patients treated with steroids alone. Cytostatic therapy includes steroids and cyclophosphamide or, in 1 case, mycophenolate mofetil.

Fig 2. Kaplan-Meier curve of overall patient survival within 1 year. Combined log-rank analysis showed no significant differences among the 3 groups (P ⴝ 0.17). Cumulative patient survival at 1 year was 75.3% for the MPO-ANCA group, 78.8% for the doublepositive group, and 100% for the anti-GBM group. Censored subjects, thick black squares. There were 29, 5, and 11 patients still at risk at 1 year, respectively.

group showed a significant difference between the MPO-ANCA and double-positive groups, but not the anti-GBM group (P ⫽ 0.01 and P ⫽ 0.17, respectively). At diagnosis, 28% of MPO-ANCA– positive patients were dialysis dependent compared with 60% and 69% of double-positive and anti-GBM–positive patients, respectively (P ⫽ 0.02). MPO-ANCA–positive patients had better cumulative renal survival (64% at 1 year) than double-positive (10% at 1 year) and anti-GBM– positive patients (15% at 1 year).

258

RUTGERS ET AL

Fig 3. Kaplan-Meier curve of overall renal survival; patients who died with independent renal function within 1 year were censored. Combined log-rank analysis showed significant differences among the 3 groups (P ⴝ 0.04). Pairwise comparison per group showed a significant difference between the MPO-ANCA group and the doublepositive group, but not the anti-GBM group; (P ⴝ 0.01 and P ⴝ 0.17, respectively). Cumulative renal survival at 1 year was 63.7% for the MPO-ANCA group, 10.0% for the double-positive group, and 15.4% for the anti-GBM group. Censored subjects, thick black squares. There were 19, 0, and 2 patients still at risk at 1 year, respectively.

DISCUSSION

In our study, we found that double-positive patients with CGN had severe renal involvement at diagnosis similar to patients with anti-GBM CGN and more severe than patients with MPOANCAs only. However, histological findings showed that periglomerular granulomatous inflammation was found in only MPO-ANCA– and double-positive patients. Finally, patient and renal survival analysis showed that doublepositive patients did not have a better prognosis than anti-GBM antibody–positive patients, contrary to reports in the literature, but in line with our results obtained in our experimental animal model. We observed that 43% of patients with antiGBM–mediated CGN were positive for MPOANCAs. These findings are somewhat higher compared with other groups (Table 5). Combining available data from the literature, we calculated an overall percentage of MPO-ANCA positivity of 25% (excluding studies without clear ANCA testing).7,9,19-27

In our series, periglomerular granulomas were found exclusively in MPO-ANCA–positive patients and not in anti-GBM single-positive patients. Periglomerular granulomas were described first by Wegener14 in 1939 in 3 patients with systemic vasculitis, granulomatous upper respiratory tract disease, and glomerulonephritis, later described as Wegener granulomatosis. Godman and Churg28 also found this type of granulomatous inflammation in 2 of 7 patients described, which they considered not to be specific for Wegener disease. In 1950, Former15 already had described 5 patients with periglomerular granulomas, including 2 patients with bacterial endocarditis, showing that this type of lesion is not found exclusively in patients with Wegener granulomatosis. At that time, ANCA testing was not available and thus exclusion of (coexisting) ANCAs in their patients with endocarditis and CGN was not possible.29,30 More recently, Bajema et al31 found periglomerular granulomas in approximately 10% of 157 renal biopsy specimens from patients with

ANTIBODIES IN CRESCENTIC GLOMERULONEPHRITIS

259

Table 5. Percentage of MPO-ANCA– and PR3-ANCA–Positive Patients in the Group of Anti-GBM–Positive Patients Reported in the Literature Total No. of Anti-GBM–Positive patients

MPO-ANCA

PR3-ANCA

Year

Reference

136 100 13 9 160 12 23 37 67 79 121

31 (42) 25 (25) 54 (7)* 0 (0) 16 (25) 17 (2) 30 (7) 30 (11)† 30 (20)‡ 32 (25) 32 (39)§

10 (14) 12 (12)

2002 1997 1996 1995 1995 1992 1992 1991 1990 2003 2004

20 7 23 24 25 19 26 9 21 27 22

0 (0) 2 (3) 17 (2) 0 (0)

5 (4) 4 (1)

NOTE. Values expressed as percent (number of patients). *Includes only patients with pulmonary hemorrhage and nephritis, no division between PR3- or MPO-ANCA was made. †MPO-ANCA specificity was tested by testing indirect immunofluorescence on MPO-deficient neutrophils. ‡Only indirect immunofluorescence-ANCA was performed in these samples. §Includes 4 patients who had both MPO- and PR3-ANCAs.

ANCA-associated systemic vasculitis. Periglomerular granulomas are found in MPO-positive and PR3-ANCA–positive patients.32 In our study, we found that patients with anti-GBM nephritis and granulomas were always MPO-ANCA positive, as well. Bajema et al31 found that renal granulomas were not indicative of severe renal involvement because renal functioning was not different from patients without granulomatous inflammation. We also found that creatinine levels were similar, as were percentages of normal glomeruli, in our patients with and without periglomerular granulomatous inflammation. However, we found that patients with periglomerular granulomatous inflammation presented with lower proteinuria (protein, 7.81 ⫾ 5.73 versus 2.821 ⫾ 2.478 g; P ⬍ 0.001) and a higher activity index score (4.4 ⫾ 1.2 versus 2.8 ⫾ 1.5; P ⫽ 0.003), perhaps indicating a more aggressive and active disease process. Larger series are needed to provide a more definite answer to whether periglomerular granulomatous inflammation in patients with anti-GBM CGN is restricted to those who also test positive for ANCAs. In 20% of our MPO-ANCA–positive patients, 20% of double-positive patients, and 38% of anti-GBM–positive patients, therapy was withheld or consisted of merely a short course of steroids. Nowadays, standard treatment for patients with CGN includes steroids and cytostatic therapy. In patients with MPO-ANCA CGN with

severe renal failure and all patients with antiGBM CGN, plasma exchange is added to this regimen. In most of our patients for whom therapy was withheld, renal function was considered lost and side effects of therapy outweighed the expected treatment benefit.33,34 Patient survival in our groups was 75%, 79%, and 100% alive at 1 year for the MPO-ANCA– positive, double-positive, and anti-GBM–positive groups, respectively. These numbers are similar to those in the literature and are dependent on renal function at diagnosis.34 At diagnosis, 28% of MPO-ANCA–positive patients were dialysis dependent compared with 60% and 69% of double-positive and anti-GBM–positive patients, respectively. Cumulative renal survival was 64%, 10%, and 15% at 1 year for the MPO-ANCA–positive, double-positive, and antiGBM–positive groups, respectively. None of our patients on dialysis therapy recovered renal function. This is contrary to data in the report by Bosch et al,9 in which double positivity was linked to better renal prognosis. They showed that none of their dialysis-dependent patients with anti-GBM CGN recovered independent renal function, whereas 45% of double-positive patients on dialysis therapy recovered renal function. Major differences between our population and the one studied by Bosch et al9 include the percentage of anti-GBM patients presenting with oliguria and/or anuria (50% versus 88%, respec-

260

tively) and use of plasma exchange next to immunosuppression in the double-positive group (40% versus 100%, respectively). The latter might suggest that patients in our study group were not treated optimally, and therapy should always include prompt plasma exchange. However, in contrast to the findings of Bosch et al,9 Levy et al34 also suggested that double-positive patients respond to therapy poorly and thus are associated with poor renal outcome; these patients received plasma exchange in addition to immunosuppressive therapy.34 Bonsib et al19 also showed poor prognosis and treatment response in a small study. Here, we show that renal survival in double-positive patients is not better and may be worse than in patients with anti-GBM CGN. It is not clear whether ANCA-associated CGN predisposes to the development of anti-GBM disease or that ANCA positivity occurs in the course of anti-GBM CGN.8 Our findings that characteristics of double-positive patients are similar to those of the MPO-ANCA group (eg, patient age and sex) suggest that MPO-ANCA disease occurs first. The finding that disease and histological characteristics in double-positive patients contain elements of MPO-ANCA disease (ie, granulomatous periglomerular inflammation), as well as of anti-GBM disease (ie, serum creatinine level, oliguria/anuria, and renal survival at 1 year), is suggestive of coexistent disease; thus, MPO-ANCA disease first, followed by concomitant anti-GBM disease. We observed 1 patient in our clinics who presented with doublepositive disease in which, in retrospect, MPOANCA positivity could be shown before the development of anti-GBM disease. A similar patient was reported by Serratrice et al.35 The pathophysiological explanation for the occurrence of anti-GBM antibody formation as a result of MPO-ANCA–mediated glomerulonephritis and/or vasculitis is speculative, but could include the following factors. MPO-ANCAs have been shown to activate (primed) neutrophils, resulting in the production of reactive oxygen species and degranulation. MPO present outside inflammatory sites normally is cleared and inactivated by ceruloplasmin.36 However, the presence of MPO-ANCAs interferes with these mechanisms, leaving a circulating highly reactive enzyme.37 This reactive enzyme produces MPO-derived oxidants, which increases proteol-

RUTGERS ET AL

ysis of the GBM.38 Moreover, Kalluri et al40 showed that oxidants can alter the hexameric structure of the GBM, therewith exposing immunologic epitopes (ie, the Goodpasture epitope39) embedded in the basement membrane.40 More specifically, it recently was shown that the Goodpasture epitope can be cleaved from the ␣3 chain of type IV collagen by matrix metalloproteinase9,41 an enzyme that can be activated by MPO.42 In patients with MPO-ANCA–mediated CGN, MPO-derived oxidants have been shown in kidney biopsy specimens.43 Overall, our results regarding patient and renal survival in patients with either anti-GBM or both anti-GBM and anti-MPO antibodies are in agreement with data recently reported by Levy et al.22 However, in our study, patients were selected based on the presence of CGN, rather than positive serological test results. In conclusion, in patients with both anti-GBM antibodies and MPO-ANCAs (ie, double-positive patients), histological findings differ from those in patients with anti-GBM antibodies only. However, renal survival in these patients is not better than that in anti-GBM antibody–positive patients and is worse compared with patients with MPO-ANCA only. ACKNOWLEDGMENT The authors thank H. van Rie and P. Heerings, Clinical and Experimental Immunology, University Hospital Maastricht, for excellent collection and technical preparation of biopsy material and all participating physicians and patients for their generous contributions to this report (including Dr F. de Heer and Dr G. H. Verseput, Maaslandziekenhuis, Sittard; Dr W. Grave, Dr J. Wirtz, and Dr S. Boorsma, St Laurentiusziekenhuis, Roermond; Dr J. Wolters and Dr L. A. M. Frenken, Atrium Medisch Centrum, Heerlen; Dr E. Zeppenfelt, Landgraaf; Dr K.M.L. Leunissen, University Hospital Maastricht, Maastricht).

REFERENCES 1. Glassock RASG, Cohen AH: Primary glomerular diseases, in Rector F (ed): The Kidney. Philadelphia, PA, Saunders, 1991, pp 1182-1279 2. Hricik DE, Chung-Park M, Sedor JR: Glomerulonephritis. N Engl J Med 339:888-899, 1998 3. Jennette JC, Wilkman AS, Falk RJ: Anti-neutrophil cytoplasmic autoantibody-associated glomerulonephritis and vasculitis. Am J Pathol 135:921-930, 1989 4. Jennette JC, Falk RJ, Andrassy K, et al: Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum 37:187-192, 1994 5. Lerner RA, Glassock RJ, Dixon FJ: The role of antiglomerular basement membrane antibody in the pathogene-

ANTIBODIES IN CRESCENTIC GLOMERULONEPHRITIS

sis of human glomerulonephritis. J Exp Med 126:989-1004, 1967 6. Xiao H, Heeringa P, Hu P, et al: Antineutrophil cytoplasmic autoantibodies specific for myeloperoxidase cause glomerulonephritis and vasculitis in mice. J Clin Invest 110:955963, 2002 7. Hellmark T, Niles JL, Collins AB, McCluskey RT, Brunmark C: Comparison of anti-GBM antibodies in sera with or without ANCA. J Am Soc Nephrol 8:376-385, 1997 8. Jennette JC: Rapidly progressive crescentic glomerulonephritis. Kidney Int 63:1164-1177, 2003 9. Bosch X, Mirapeix E, Font J, et al: Prognostic implication of anti-neutrophil cytoplasmic autoantibodies with myeloperoxidase specificity in anti-glomerular basement membrane disease. Clin Nephrol 36:107-113, 1991 10. Hudson BG, Tryggvason K, Sundaramoorthy M, Neilson EG: Alport’s syndrome, Goodpasture’s syndrome, and type IV collagen. N Engl J Med 348:2543-2556, 2003 11. Heeringa P, Brouwer E, Klok PA, et al: Autoantibodies to myeloperoxidase aggravate mild anti-glomerularbasement-membrane-mediated glomerular injury in the rat. Am J Pathol 149:1695-1706, 1996 12. Tiebosch AT, Wolters J, Frederik PF, et al: Epidemiology of idiopathic glomerular disease: A prospective study. Kidney Int 32:112-116, 1987 13. Stilmant MM, Bolton WK, Sturgill BC, Schmitt GW, Couser WG: Crescentic glomerulonephritis without immune deposits: Clinicopathologic features. Kidney Int 15:184195, 1979 14. Wegener F: Uber die eigenartige Rhinogene Granulomatose mit besonderer Beteiligung des Arteriensystems un der Nieren. Beitr Pathol Anat 102:36-68, 1939 15. Former F: Uber die granulomatose Periglomerulitis. Schweiz Zeitschrift allgemeine Pathol Bakteriol 13:42-59, 1950 16. Franssen CF, Gans RO, Arends B, et al: Differences between anti-myeloperoxidase- and anti-proteinase 3-associated renal disease. Kidney Int 47:193-199, 1995 17. Boomsma MM, Damoiseaux JG, Stegeman CA, et al: Image analysis: A novel approach for the quantification of antineutrophil cytoplasmic antibody levels in patients with Wegener’s granulomatosis. J Immunol Methods 274:27-35, 2003 18. Boomsma MM, Stegeman CA, Oost-Kort WW, et al: Native and recombinant proteins to analyze auto-antibodies to myeloperoxidase in pauci-immune crescentic glomerulonephritis. J Immunol Methods 254:47-58, 2001 19. Bonsib SM, Goeken JA, Kemp JD, et al: Coexistent anti-neutrophil cytoplasmic antibody and antiglomerular basement membrane antibody associated disease—Report of six cases. Mod Pathol 6:526-530, 1993 20. Denton M, Magee C, Niles J: Rapidly progressive glomerulonephritis. Ann Intern Med 137:E375-E376, 2002 21. Jayne DR, Marshall PD, Jones SJ, Lockwood CM: Autoantibodies to GBM and neutrophil cytoplasm in rapidly progressive glomerulonephritis. Kidney Int 37:965-970, 1990 22. Levy JB, Hammad T, Coulthart A, Dougan T, Pusey CD: Clinical features and outcome of patients with both ANCA and anti-GBM antibodies. Kidney Int 66:1535-1540, 2004

261

23. Niles JL, Bottinger EP, Saurina GR, et al: The syndrome of lung hemorrhage and nephritis is usually an ANCA-associated condition. Arch Intern Med 156:440-445, 1996 24. Saxena R, Bygren P, Arvastson B, Wieslander J: Circulating autoantibodies as serological markers in the differential diagnosis of pulmonary renal syndrome. J Intern Med 238:143-152, 1995 25. Short AK, Esnault VL, Lockwood CM: Anti-neutrophil cytoplasm antibodies and anti-glomerular basement membrane antibodies: Two coexisting distinct autoreactivities detectable in patients with rapidly progressive glomerulonephritis. Am J Kidney Dis 26:439-445, 1995 26. Weber MF, Andrassy K, Pullig O, Koderisch J, Netzer K: Antineutrophil-cytoplasmic antibodies and antiglomerular basement membrane antibodies in Goodpasture’s syndrome and in Wegener’s granulomatosis. J Am Soc Nephrol 2:1227-1234, 1992 27. Segelmark M, Hellmark T, Wieslander J: The prognostic significance in Goodpasture’s disease of specificity, titre and affinity of anti-glomerular-basement-membrane antibodies. Nephron Clin Pract 94:c59-c68, 2003 28. Godman GC, Churg J: Wegener=s granulomatosis. AMA Arch Pathol 58:533-553, 1954 29. Choi HK, Lamprecht P, Niles JL, Gross WL, Merkel PA: Subacute bacterial endocarditis with positive cytoplasmic antineutrophil cytoplasmic antibodies and anti-proteinase 3 antibodies. Arthritis Rheum 43:226-231, 2000 30. Hellmich B, Ehren M, Lindstaedt M, et al: Anti-MPOANCA-positive microscopic polyangiitis following subacute bacterial endocarditis. Clin Rheumatol 20:441-443, 2001 31. Bajema IM, Hagen EC, Ferrario F, et al: Renal granulomas in systemic vasculitis. EC/BCR Project for ANCA-Assay Standardization. Clin Nephrol 48:16-21, 1997 32. Hauer HA, Bajema IM, Van Houwelingen HC, et al: Determinants of outcome in ANCA-associated glomerulonephritis: A prospective clinico-histopathological analysis of 96 patients. Kidney Int 62:1732-1742, 2002 33. Franssen CF, Stegeman CA, Oost-Kort WW, et al: Determinants of renal outcome in anti-myeloperoxidaseassociated necrotizing crescentic glomerulonephritis. J Am Soc Nephrol 9:1915-1923, 1998 34. Levy JB, Turner AN, Rees AJ, Pusey CD: Long-term outcome of anti-glomerular basement membrane antibody disease treated with plasma exchange and immunosuppression. Ann Intern Med 134:1033-1042, 2001 35. Serratrice J, Chiche L, Dussol B, et al: Sequential development of perinuclear ANCA-associated vasculitis and anti-glomerular basement membrane glomerulonephritis. Am J Kidney Dis 43:e26-e30, 2004 36. Segelmark M, Persson B, Hellmark T, Wieslander J: Binding and inhibition of myeloperoxidase (MPO): A major function of ceruloplasmin? Clin Exp Immunol 108:167-174, 1997 37. Griffin SV, Chapman PT, Lianos EA, Lockwood CM: The inhibition of myeloperoxidase by ceruloplasmin can be reversed by anti-myeloperoxidase antibodies. Kidney Int 55:917-925, 1999 38. Vissers MC, Winterbourn CC: The effect of oxidants on neutrophil-mediated degradation of glomerular basement

262

membrane collagen. Biochim Biophys Acta 889:277-286, 1986 39. Kalluri R, Gattone VH II, Noelken ME, Hudson BG: The alpha 3 chain of type IV collagen induces autoimmune Goodpasture syndrome. Proc Natl Acad Sci U S A 91:62016205, 1994 40. Kalluri R, Cantley LG, Kerjaschki D, Neilson EG: Reactive oxygen species expose cryptic epitopes associated with autoimmune Goodpasture syndrome. J Biol Chem 275:20027-20032, 2000 41. Hamano Y, Zeisberg M, Sugimoto H, et al: Physiological levels of tumstatin, a fragment of collagen IV

RUTGERS ET AL

alpha3 chain, are generated by MMP-9 proteolysis and suppress angiogenesis via alphaV beta3 integrin. Cancer Cell 3:589-601, 2003 42. Meli DN, Christen S, Leib SL: Matrix metalloproteinase-9 in pneumococcal meningitis: Activation via an oxidative pathway. J Infect Dis 187:1411-1415, 2003 43. Heeringa P, Bijl M, de Jager-Krikken A, et al: Renal expression of endothelial and inducible nitric oxide synthase, and formation of peroxynitrite-modified proteins and reactive oxygen species in Wegener’s granulomatosis. J Pathol 193:224-232, 2001