Abstract

Background. We reported previously that in renal disease in relation to antineutrophil cytoplasm auto‐antibodies (ANCA)‐associated vasculitis, renal outcome correlates better with the percentage of normal glomeruli than with separate active lesions. This may imply that glomeruli, once affected by necrotizing and crescentic lesions, are irreversibly damaged. We quantified and evaluated the course of renal lesions in the present study.

Methods. We retrospectively analysed 31 patients with renal disease in relation to ANCA‐associated vasculitis, all treated with immunosuppressive drugs. In all patients, a renal biopsy was performed at diagnosis. A follow‐up biopsy was performed in all patients on the indication of a suspected renal relapse, after a mean interval of 31 months.

Results. The mean percentage of normal glomeruli in the renal biopsy did not change over time (29% in the initial and 30% in the follow‐up biopsy). The mean percentage of glomeruli with crescents, however, significantly decreased from 57 to 30% (P<0.001). The percentage of glomerulosclerosis significantly increased from 12 to 39% (P<0.001). The data were independent of diagnosis, gender, age, time interval between the biopsies, and treatment.

Conclusions. This is the first study to quantify glomerular changes between two time points in patients with renal vasculitis. Our results suggest that, on average, no new glomeruli are recruited into the active disease process. The sum of the percentage of crescentic and sclerotic glomeruli in the initial biopsies is larger than the percentage of sclerotic glomeruli in the follow‐up biopsies. Thus, therapy seems not only to prevent normal glomeruli from being recruited into the active disease process for a certain time, but seems also to allow part of the active lesions to revert into a normal phenotype, although another part of the active lesions will be transformed to a chronic phenotype.

Introduction

In antineutrophil cytoplasm auto‐antibodies (ANCA)‐associated systemic vasculitis with kidney involvement, the main renal lesions are glomerular extracapillary proliferation (crescents) and fibrinoid necrosis [13]. The renal biopsy primarily serves as a diagnostic tool in combination with clinical findings and the results of ANCA serology [4]. To a lesser extent, histological parameters in the renal biopsy may be used to predict renal outcome after therapy [5].

In a previous study, we investigated the correlation between the findings in the initial renal biopsy and renal function, both at the time of biopsy and during follow‐up [6]. In this clinical‐pathological analysis, the renal biopsy findings at diagnosis from 157 patients with ANCA‐associated systemic vasculitis were related to serum creatinine values after at least 1 year of follow‐up. The percentage of normal glomeruli in the biopsy was found to be the histological parameter with the best predictive value for renal function [6]. This seems to suggest that only glomeruli that are unaffected by the disease are relevant for the re‐establishment of renal function. Another hypothesis suggests that lesions considered to represent the acute phase of the disease, such as extracapillary proliferation, are to some extent reversible although there is a certain point of no return, for instance when crescents become fibrous [7]. This has implications for the effects that may be expected from therapy.

Insight into the validity of these hypotheses can only be gained in human research by studying follow‐up biopsies. We present a retrospective clinical‐pathological analysis of 31 patients with ANCA‐associated vasculitis of whom follow‐up renal biopsies were available; in these biopsies, the evolvement of renal lesions was quantified and evaluated.

Subjects and methods

Patients and renal biopsies

Clinical data and renal biopsies were available from 12 patients included in the ‘EC/BCR Project for ANCA Assay Standardization’ [8]. In addition, patients were selected who were not studied in the mentioned project. Altogether, 31 patients were included in the present study from the Leiden University Medical Center (n=7), the Erasmus University Medical Center Rotterdam (n=3), The Netherlands, Hôpital Necker, Paris, France (n=7), Ospedale San Carlo Borromeo, Milan, Italy (n=9), University Clinics of Heidelberg, Germany (n=3), Addenbrooke's Hospital, Cambridge (n=1), and Hammersmith Hospital, London (n=1) UK.

Initial biopsies were performed before the onset of treatment. In the original pathology report on the initial renal biopsy, pauci‐immune crescentic necrotizing glomerulonephritis was concluded in all but three cases. These were three patients with histologically proven Wegener's granulomatosis and active extra‐renal disease manifestations. Two of them presented with erythrocyturia, but without elevated serum creatinine or histopathological renal lesions; the third patient presented with chronic renal lesions only. In all 31 patients, follow‐up renal biopsies were performed on the indication of a suspected renal relapse, 3–72 months (mean 31 months) after the initial biopsy. In one case the follow‐up biopsy was performed after 161 months.

Relevant clinical data at the moment of the initial biopsy and of the follow‐up biopsy are listed in Table 1, grouped by diagnosis. Patient's age at the initial biopsy ranged from 13 to 70 years (mean 49). Included were 14 females and 17 males. ANCA testing by IIF (indirect immunofluorescence test) was performed in 26 patients: 16 patients displayed a C‐ANCA (cytoplasmic) staining pattern, nine a P‐ANCA (perinuclear) staining pattern, and one had a negative IIF test. As some patients in this study were recruited before ANCA testing was generally in use, ANCA‐test results were not available of five patients. Most patients received immunosuppressive treatment after the initial biopsy, in most cases consisting of corticosteroids in combination with either cyclophosphamide or azathioprine (Table 1). One patient (no. 18, Table 1) refused to receive immunosuppressive treatment. This patient proceeded to end‐stage renal failure in approximately 6 months. Of another patient (no. 22, Table 1), it remained uncertain whether he received immunosuppressive drugs or not.

Table 1. 

Patient characteristics

Patient Agea Genderb Diagnosisc Initial biopsy
 

 

 

 

 
Time intervali Treatmentj Follow‐up biopsy
 
 

 

 

 

 
Creatd
 
ANCA‐IIFe
 
PR3‐ANCAf
 
MPO‐ANCAg
 
U/Ah
 

 

 
Creatd
 
U/Ah
 
 1 40 MPA 1041 n.d. n.d. n.d.  35 P 9g  312 
 2 53 MPA  400 − 161 P 45g+MP 23g  200 
 3 35 MPA  184 n.d. n.d. n.d.  71 P 19g  990 
 4 66 MPA 1184 − n.d. n.d.  57 P 10g+MP 6g  320 
             +Cyc 27g   
 5 53 MPA  462 − −   4 P 7g+Cyc 36g 
             +Aza 14g   
 6 54 MPA  650 −  29 P 1g+MP 3g  388 
             +Cyc 0.6g   
             +Aza 30g   
 7 34 MPA  198 n.d. n.d.  12 P 9g+Cyc 42g  333 
 8 70 MPA  253 −  28 P 4g+Cyc 10g  744 
             +Aza 10g   
 9 41 MPA  138   6 P 13g+MP 3g  103 
             +Cyc 10g   
10 33 MPA  213 −  29 P 8g+MP 3g   89 − 
             +Cyc 3g   
11 60 MPA  167  35 P 4g+Aza 53g  179 
12 58 RLV  186 −  45 P 8g  486 
13 56 RLV  211 −   9 P 8g+MP 3g  150 
             +Cyc 3g   
14 45 RLV  988 − −  11 P 8g+MP 3g  498 
             +Cyc 14g   
15 61 RLV  466 −  32 P 11g+MP 3g  423 
             +Cyc 6g   
16 31 WG −  55 P 5g+Cyc 27g 
17 48 WG  303 n.d. n.d.  38 MP 8g+Cyc 49g  104 
18 60 WG  410 n.d. n.d.   6 Patient refused 1329 
             treatment   
19 52 WG  328 −  40 P 13g+Cyc 11g  226 
             +Aza 116g   
             +ATG 1g   
20 25 WG   78 −  32 Cyc 60g   82 
21 56 WG  125 n.d. n.d. −  37 P 15g+Cyc 57g  253 
22 43 WG  28 Unknown  655 
23 65 WG   89 n.d. n.d.  21 P 7g+MP 3g   89 
             +Cyc 4g   
24 63 WG  594 −   3 P 1g+Cyc 1g  342 
25 35 WG  534 n.d. n.d.  24 P 11g+MP 3g  107 
             +Cyc 5g   
26 58 WG 1607 −  24 P 8g+Cyc 28g  209 
27 49 WG  120 −  26 P 12g+Cyc 1g  160 − 
28 66 WG  250  12 P 8g+MP 2g  177 
29 70 WG  180 −  42 P 10g+Cyc 13g  128 
             +Aza 39g   
30 49 WG  205 −  16 P 8g+MP 3g  125 
             +Cyc 26g   
31 13 WG  107 −  15 P 7g   71 − 
Patient Agea Genderb Diagnosisc Initial biopsy
 

 

 

 

 
Time intervali Treatmentj Follow‐up biopsy
 
 

 

 

 

 
Creatd
 
ANCA‐IIFe
 
PR3‐ANCAf
 
MPO‐ANCAg
 
U/Ah
 

 

 
Creatd
 
U/Ah
 
 1 40 MPA 1041 n.d. n.d. n.d.  35 P 9g  312 
 2 53 MPA  400 − 161 P 45g+MP 23g  200 
 3 35 MPA  184 n.d. n.d. n.d.  71 P 19g  990 
 4 66 MPA 1184 − n.d. n.d.  57 P 10g+MP 6g  320 
             +Cyc 27g   
 5 53 MPA  462 − −   4 P 7g+Cyc 36g 
             +Aza 14g   
 6 54 MPA  650 −  29 P 1g+MP 3g  388 
             +Cyc 0.6g   
             +Aza 30g   
 7 34 MPA  198 n.d. n.d.  12 P 9g+Cyc 42g  333 
 8 70 MPA  253 −  28 P 4g+Cyc 10g  744 
             +Aza 10g   
 9 41 MPA  138   6 P 13g+MP 3g  103 
             +Cyc 10g   
10 33 MPA  213 −  29 P 8g+MP 3g   89 − 
             +Cyc 3g   
11 60 MPA  167  35 P 4g+Aza 53g  179 
12 58 RLV  186 −  45 P 8g  486 
13 56 RLV  211 −   9 P 8g+MP 3g  150 
             +Cyc 3g   
14 45 RLV  988 − −  11 P 8g+MP 3g  498 
             +Cyc 14g   
15 61 RLV  466 −  32 P 11g+MP 3g  423 
             +Cyc 6g   
16 31 WG −  55 P 5g+Cyc 27g 
17 48 WG  303 n.d. n.d.  38 MP 8g+Cyc 49g  104 
18 60 WG  410 n.d. n.d.   6 Patient refused 1329 
             treatment   
19 52 WG  328 −  40 P 13g+Cyc 11g  226 
             +Aza 116g   
             +ATG 1g   
20 25 WG   78 −  32 Cyc 60g   82 
21 56 WG  125 n.d. n.d. −  37 P 15g+Cyc 57g  253 
22 43 WG  28 Unknown  655 
23 65 WG   89 n.d. n.d.  21 P 7g+MP 3g   89 
             +Cyc 4g   
24 63 WG  594 −   3 P 1g+Cyc 1g  342 
25 35 WG  534 n.d. n.d.  24 P 11g+MP 3g  107 
             +Cyc 5g   
26 58 WG 1607 −  24 P 8g+Cyc 28g  209 
27 49 WG  120 −  26 P 12g+Cyc 1g  160 − 
28 66 WG  250  12 P 8g+MP 2g  177 
29 70 WG  180 −  42 P 10g+Cyc 13g  128 
             +Aza 39g   
30 49 WG  205 −  16 P 8g+MP 3g  125 
             +Cyc 26g   
31 13 WG  107 −  15 P 7g   71 − 

aAge (years) at time point of initial biopsy. bF, Female; M, Male. cMPA, microscopic polyangiitis; RLV, renal limited vasculitis; WG, Wegener's granulomatosis. dSerum creatinine (μmol/l); #, not determined at same day as biopsy date or data were not available. eIndirect immunofluorescence test for antineutrophil cytoplasm auto‐antibodies; C, cytoplasmic staining pattern; P, perinuclear staining pattern; −, negative; n.d., not determined; @, data were not available. fAntiproteinase‐3 antineutrophil cytoplasm antibodies. gAntimyeloperoxidase antineutrophil cytoplasm antibodies. hUrinary sediment activity (positive or negative). iTime interval between initial and follow‐up biopsy (months). jTreatment in the time interval between initial and follow‐up biopsy. P, prednisolone; MP, methylprednisolone pulses; Cyc, cyclophosphamide; Aza, azathioprine; ATG, antithymocyte globulin.

Patient classification

Disease definitions were based on the ‘Chapel Hill consensus conference on the nomenclature of systemic vasculitis’ [9]. Patients were categorized in the following three groups:

  • (1) Wegener's granulomatosis (n=16). Histologically proven granulomatous inflammation of the respiratory tract together with necrotizing vasculitis affecting small to medium‐sized vessels.

  • (2) Microscopic polyangiitis (n=11). Histologically proven non‐granulomatous necrotizing vasculitis predominantly affecting small vessels without airway symptoms compatible with Wegener's granulomatosis.

  • (3) Renal limited vasculitis (n=4). Isolated histologically proven pauci‐immune crescentic necrotizing glomerulonephritis.

Scoring of the biopsies

Paraffin sections were stained with silver, periodic acid‐Schiff, haematoxylin and eosin, and trichrome, and were forwarded to two out of five participating nephropathologists (I.M.B., L.H.N., F.F., R.W., J.A.B). Both pathologists scored the biopsies separately, blinded to patient data and to the original pathology reports, and according to a previously standardized protocol for scoring renal biopsies of patients with ANCA‐associated systemic vasculitis, which has been elaborately described in previous publications [6,10]. In short, each glomerulus had to be scored separately for the presence of fibrinoid necrosis, crescents, global glomerulosclerosis, or any other lesion. The number of glomeruli affected by these lesions was expressed as the percentage of the total number of glomeruli in the biopsy. The percentage of normal glomeruli was scored as well. The total percentage of glomeruli affected by the disease was calculated as the percentage of glomeruli with extracapillary proliferation plus the percentage of glomerulosclerosis. We neglected the possible influence of physiological processes of ageing on the induction of glomerulosclerosis. Interstitial infiltrates (−/+/++/+++), interstitial fibrosis (−/+/++), and tubular atrophy (−/+/++) were scored semi‐quantitatively. Tubular necrosis was scored as absent or present. The arteries and arterioles were evaluated for the presence of vasculitis. The scores were entered into a central database (MSAccess) and major discrepancies between the observers were solved by conference during central reviews, achieving consensus for each biopsy.

Statistical analysis

Differences between glomerular parameters in the initial and follow‐up biopsies, and renal function in terms of natural logarithm of serum creatinine at the time of these two biopsies were analysed with a paired t‐test. For the interstitial parameters, this analysis was performed with the Sign test. The correlation of the histopathological differences with the differences in renal function was calculated with Spearman's rho correlation test. The influence of diagnosis, gender, and treatment (corticosteroids only vs corticosteroids and additional drugs such as cyclophosphamide and azathioprine) on our results was tested with the t‐test (quantitative parameters) and with the Mann–Whitney U test (ordinal parameters). The influence of age and time interval between the biopsies was tested by Spearman's rho correlation test. The reported P‐values are two‐sided. The level of significance used was 0.05. Reported results are based on analyses in which no minimal number of glomeruli per sample was defined for inclusion of biopsies. The influence of a small number of glomeruli on the outcome of our results was tested, but it was found that both a minimum of five and ten glomeruli per biopsy as inclusion criteria did not influence these results.

Results

Results on the changes in renal histology and serum creatinine were not influenced by differences in diagnosis, time interval between the initial and follow‐up biopsy, gender, patient's age, and treatment, which is shown in Table 2. The distribution of the number of glomeruli per biopsy is illustrated in Figure 1.

Normal glomeruli

The mean percentages of normal glomeruli in the initial (29%) and follow‐up (30%) biopsy were not significantly different (Table 3 and Figure 2a). In most individual patients, the percentage of normal glomeruli in the initial and the follow‐up biopsy was similar, although there were some patients with large differences as well.

Fig. 1. 

Distribution of sample size in terms of the number of glomeruli per biopsy in the initial biopsies (a) and in the follow‐up biopsies (b). The influence of a small number of glomeruli on the outcome of our results was tested, but it was found that both a minimum of five and 10 glomeruli per biopsy as inclusion criteria did not influence these results.

Fig. 1. 

Distribution of sample size in terms of the number of glomeruli per biopsy in the initial biopsies (a) and in the follow‐up biopsies (b). The influence of a small number of glomeruli on the outcome of our results was tested, but it was found that both a minimum of five and 10 glomeruli per biopsy as inclusion criteria did not influence these results.

Fig. 2. 

The percentage of normal glomeruli (a), glomeruli with fibrinoid necrosis (b), extracapillary proliferation (c), global sclerosis (d), and the calculated diseased glomeruli (e) in initial and follow‐up biopsies paired for each individual. Each line counts for one patient. However, lines may overlap, in most cases between zero in the initial biopsy and zero in the follow‐up biopsy. The mean percentages of glomerular lesions per biopsy with standard errors are added for both initial and follow‐up biopsies.

Fig. 2. 

The percentage of normal glomeruli (a), glomeruli with fibrinoid necrosis (b), extracapillary proliferation (c), global sclerosis (d), and the calculated diseased glomeruli (e) in initial and follow‐up biopsies paired for each individual. Each line counts for one patient. However, lines may overlap, in most cases between zero in the initial biopsy and zero in the follow‐up biopsy. The mean percentages of glomerular lesions per biopsy with standard errors are added for both initial and follow‐up biopsies.

Table 2. 

Influence of diagnosis, gender, age, time interval between the biopsies, and treatment on differences in renal histology, and serum creatinine

Parametera
 
MPA vs RLVb
 
MPA vs WGb
 
RLV vs WGb
 
Gender
 
Age
 
Time interval
 
Treatmentc
 
Normal glomeruli 0.548d 0.608 0.519 0.104 0.125 0.135 0.386 
Fibrinoid necrosis 0.410 0.239 0.951 0.462 0.625 0.848 0.372 
Extracapillary proliferation 0.872 0.327 0.341 0.599 0.554 0.190 0.562 
Glomerulosclerosis 0.373 0.611 0.114 0.368 0.177 0.915 0.366 
Diseased glomeruli 0.371 0.424 0.867 0.131 0.067 0.105 0.799 
Interstitial infiltrates 0.661 0.904 0.494 0.215 0.963 0.813 0.581 
Interstitial fibrosis 0.851 0.865 0.963 0.444 0.199 0.826 0.618 
Tubular atrophy 0.343 0.110 0.750 0.173 0.279 0.471 0.158 
Tubular necrosis 0.104 0.422 0.290 0.597 0.177 0.573 0.694 
Serum creatinine 0.829 0.590 0.493 0.141 0.939 0.849 0.593 
Parametera
 
MPA vs RLVb
 
MPA vs WGb
 
RLV vs WGb
 
Gender
 
Age
 
Time interval
 
Treatmentc
 
Normal glomeruli 0.548d 0.608 0.519 0.104 0.125 0.135 0.386 
Fibrinoid necrosis 0.410 0.239 0.951 0.462 0.625 0.848 0.372 
Extracapillary proliferation 0.872 0.327 0.341 0.599 0.554 0.190 0.562 
Glomerulosclerosis 0.373 0.611 0.114 0.368 0.177 0.915 0.366 
Diseased glomeruli 0.371 0.424 0.867 0.131 0.067 0.105 0.799 
Interstitial infiltrates 0.661 0.904 0.494 0.215 0.963 0.813 0.581 
Interstitial fibrosis 0.851 0.865 0.963 0.444 0.199 0.826 0.618 
Tubular atrophy 0.343 0.110 0.750 0.173 0.279 0.471 0.158 
Tubular necrosis 0.104 0.422 0.290 0.597 0.177 0.573 0.694 
Serum creatinine 0.829 0.590 0.493 0.141 0.939 0.849 0.593 

aThe difference of the parameter between the initial and the follow‐up biopsy. bMPA, microscopic polyangiitis; RLV, renal limited vascultis; WG, Wegener's granulomatosis. cCorticosteroids only vs corticosteroids and additional drugs such as azathioprine and cyclophosphamide. dP‐value. The tests that were used are described in the methods section.

Table 3. 

Glomerular lesions in initial and follow‐up biopsies

Histologic parameter Initial biopsy
 

 

 
Follow‐up biopsy
 

 

 
95% Confidence interval of the difference
 

 
Significancea (P‐value) 

 
Mean (%)
 
Range (%)
 
SDb
 
Mean (%)
 
Range (%)
 
SD
 
Lower
 
Upper
 

 
Normal glomeruli 29 0−100 27 30 0−100 27 −11   9 n.s.d 
Fibrinoid necrosis 22 0−73 23  8 0−50 14   3  25 0.014 
Extracapillary proliferation 57 0−96 29 30 0−88 26  14  39 <0.001 
Glomerulosclerosis 12 0−50 15 39 0−88 25 −37 −18 <0.001 
Glomeruli affected 69 0−100 27 69 0−100 29 −11  10 n.s.d 
   by diseasec          
Histologic parameter Initial biopsy
 

 

 
Follow‐up biopsy
 

 

 
95% Confidence interval of the difference
 

 
Significancea (P‐value) 

 
Mean (%)
 
Range (%)
 
SDb
 
Mean (%)
 
Range (%)
 
SD
 
Lower
 
Upper
 

 
Normal glomeruli 29 0−100 27 30 0−100 27 −11   9 n.s.d 
Fibrinoid necrosis 22 0−73 23  8 0−50 14   3  25 0.014 
Extracapillary proliferation 57 0−96 29 30 0−88 26  14  39 <0.001 
Glomerulosclerosis 12 0−50 15 39 0−88 25 −37 −18 <0.001 
Glomeruli affected 69 0−100 27 69 0−100 29 −11  10 n.s.d 
   by diseasec          

aSignificance of the difference in presence of glomerular lesions in the initial and follow‐up biopsy. bStandard deviation. cGlomeruli with extracapillary proliferation plus glomeruli with global sclerosis. dNot significant.

Fibrinoid necrosis

In the initial biopsy, fibrinoid necrosis was present in 22% of all glomeruli (Table 3 and Figure 2b) and only if extracapillary proliferation was present in the same glomerulus as well. In the follow‐up biopsy, the percentage of glomeruli with fibrinoid necrosis had decreased to a mean of 8%. This difference was considered statistically significant (P=0.014).

Extracapillary proliferation

Extracapillary proliferation was present in 57% of glomeruli in the initial biopsy (Table 3). This percentage decreased to 30% in the follow‐up biopsy (P<0.001). Figure 2c shows that the presence of extracapillary proliferation decreased between the initial and follow‐up biopsy in most cases. In only a few cases, the presence of the lesion was markedly increased. Despite this significant decrease in the percentage of glomeruli affected by extracapillary proliferation, the percentage of follow‐up biopsies in which crescentic glomerulonephritis (i.e. the occurrence of at least one crescent) was present did not change significantly (94% in the initial biopsy vs 87% in the follow‐up biopsy).

Glomerulosclerosis

Glomerulosclerosis was present in 12% of the glomeruli in the initial biopsy (Table 3). In the follow‐up biopsy, this percentage had significantly increased to 39% (P<0.001), which is depicted in Figure 2d. In four cases, less glomerulosclerosis was found in the follow‐up biopsy than in the initial biopsy. A percentage of 100% glomerulosclerosis was not found in either the initial or in the follow‐up biopsy (maximal 88%).

Diseased glomeruli

The calculated percentage of glomeruli affected by the disease appeared not to be the same as 100% minus the percentage normal glomeruli, because a few glomerular lesions not typically present in vasculitis—such as ischaemic changes—were found. In both the initial and the follow‐up biopsy, the percentage of diseased glomeruli was 69% (Table 3 and Figure 2e).

Interstitial infiltrates

Figure 3a illustrates the results on the severity of interstitial infiltrates in the initial and follow‐up biopsies, paired for each individual patient. Predominantly mild interstitial infiltrates were found in both initial and follow‐up biopsies. Results on the interstitial lesions are summarized in Table 4. Positive and negative changes in the presence of interstitial infiltrates occurred to approximately the same extent and no significant change between the initial and the follow‐up biopsy was found.

Fig. 3. 

The severity of interstitial infiltrates (a), interstitial fibrosis (b), tubular atrophy (c), and tubular necrosis (d) in initial (diamond) and follow‐up (square) biopsies paired for each individual patient. The lesions were scored on a semi‐quantitative scale. Major differences of opinion were averaged.

Fig. 3. 

The severity of interstitial infiltrates (a), interstitial fibrosis (b), tubular atrophy (c), and tubular necrosis (d) in initial (diamond) and follow‐up (square) biopsies paired for each individual patient. The lesions were scored on a semi‐quantitative scale. Major differences of opinion were averaged.

Table 4. 

Differences in occurrence of interstitial lesions between initial and follow‐up biopsies


 
Interstitial infiltrates
 
Interstitial fibrosis
 
Tubular atrophy
 
Tubular necrosis
 
Initial<follow‐up 11 17 15  3 
Initial=follow‐up 10  9 11 18 
Initial>follow‐up 10  5  5 10 
Sign test (P‐value) n.s.a  0.017  0.041  0.092 

 
Interstitial infiltrates
 
Interstitial fibrosis
 
Tubular atrophy
 
Tubular necrosis
 
Initial<follow‐up 11 17 15  3 
Initial=follow‐up 10  9 11 18 
Initial>follow‐up 10  5  5 10 
Sign test (P‐value) n.s.a  0.017  0.041  0.092 

aNot significant.

Interstitial fibrosis

Figure 3b illustrates that mild interstitial fibrosis was present in most initial biopsies, although absence of this lesion was frequently observed as well. In the follow‐up biopsies, interstitial fibrosis was present in more cases, and it was more severe. This difference was considered significant (P=0.017) as is shown in Table 4.

Tubular atrophy

Figure 3c shows that tubular atrophy in all levels of severity was found in the initial biopsies. In the follow‐up biopsies, the severity level was higher in most patients (Table 4, P=0.041).

Tubular necrosis

Tubular necrosis was present in half of the initial biopsies (Figure 3d). In the follow‐up biopsies, this lesion was absent in most cases where it was present in the initial biopsy. In a few cases, tubular necrosis was found in the follow‐up biopsy, whereas it was absent in the initial biopsy. These changes appeared not to be significant (Table 4).

Interstitial vasculitis

The arteries (if present) and arterioles did not show vasculitic changes in any of the patients, neither in the initial, nor in the follow‐up biopsy (data not shown).

Serum creatinine

Mean serum creatinine levels (Table 1 and Figure 4) were elevated at the time of both the initial and the follow‐up biopsy (400 and 308 mmol/l, respectively), but without a significant difference between these two time points. In the time interval between the initial and follow‐up biopsy, serum creatinine levels either lowered or stabilized (data not shown).

Fig. 4. 

Serum creatinine levels at the time points of initial and follow‐up biopsies. The mean levels of serum creatinine with standard errors are added for both time points.

Fig. 4. 

Serum creatinine levels at the time points of initial and follow‐up biopsies. The mean levels of serum creatinine with standard errors are added for both time points.

Clinical‐pathological correlation

The correlations between the change in renal function (in terms of the natural logarithm of serum creatinine levels) and the changes in histopathological features are shown in Table 5. Changes in renal function significantly correlated both with changes in the percentage of normal glomeruli and with changes in the percentage of diseased glomeruli (Figure 5).

Fig. 5. 

Correlations of the change in the percentage of normal glomeruli (a) and the change in the percentage of diseased glomeruli (b) with the change in renal function (in terms of natural logarithm of serum creatinine) between the initial and the follow‐up biopsy. The changes were calculated as the value at the time of the initial biopsy minus the value at the time of the follow‐up biopsy.

Fig. 5. 

Correlations of the change in the percentage of normal glomeruli (a) and the change in the percentage of diseased glomeruli (b) with the change in renal function (in terms of natural logarithm of serum creatinine) between the initial and the follow‐up biopsy. The changes were calculated as the value at the time of the initial biopsy minus the value at the time of the follow‐up biopsy.

Table 5. 

Correlation of differences in renal histology and difference in renal function (in terms of the natural logarithm of serum creatinine)

Lesion
 
Correlation coefficient
 
Significance (P‐value)
 
Normal glomeruli −0.484 0.009 
Fibrinoid necrosis  0.281 0.148 
Extracapillary  0.354 0.064 
   proliferation    
Glomerulosclerosis  0.052 0.794 
Diseased glomeruli  0.513 0.005 
Interstitial infiltrates  0.280 0.149 
Interstitial fibrosis  0.347 0.071 
Tubular atrophy  0.245 0.208 
Tubular necrosis  0.086 0.662 
Lesion
 
Correlation coefficient
 
Significance (P‐value)
 
Normal glomeruli −0.484 0.009 
Fibrinoid necrosis  0.281 0.148 
Extracapillary  0.354 0.064 
   proliferation    
Glomerulosclerosis  0.052 0.794 
Diseased glomeruli  0.513 0.005 
Interstitial infiltrates  0.280 0.149 
Interstitial fibrosis  0.347 0.071 
Tubular atrophy  0.245 0.208 
Tubular necrosis  0.086 0.662 

The changes were calculated as the value at the time of the initial biopsy minus the value at the time of the follow‐up biopsy. The three patients, for whom no data for initial and/or follow‐up serum creatinine were available, were excluded from this analysis.

The course of glomerular lesions

Our results on the presence of glomerular lesions in the initial and follow‐up renal biopsies were used to speculate on the evolvement of these lesions during remission. This is illustrated in Figure 6, which is a hypothesis on the course of the presence of normal glomeruli, crescents, and glomerulosclerosis in vasculitis.

Fig. 6. 

Hypothesis on the evolvement of normal glomeruli, extracapillary proliferation, and glomerulosclerosis during the time interval between the two biopsies. At presentation of the disease, 12% of glomeruli in the initial renal biopsy were globally sclerosed. Assuming an analogous course of events during relapse, 12% of glomeruli that were not sclerotic during remission will be sclerotic in the follow‐up biopsy. Subsequently, the percentage newly sclerosed glomeruli in the follow‐up biopsy (i.e. developed during relapse) can be calculated as follows: new sclerosis (follow‐up)=12/100 of the not sclerotic glomeruli during remission. The ‘not sclerotic glomeruli during remission’ can be substituted: new sclerosis (follow‐up)=12/100 (normal (follow‐up)+crescents (follow‐up)+new sclerosis (follow‐up)+other (follow‐up)). This formula can be converted to: new sclerosis (follow‐up)=12/88 (normal (follow‐up)+crescents (follow‐up)+other (follow‐up))=12/88 (30+30+1)=8%. For the percentage sclerosis, present on average during remission, we can subsequently calculate: sclerosis (remission)=sclerosis (follow‐up)−new sclerosis (follow‐up)=39−8=31%. Assuming that crescents will not persist during remission without consequence on renal function and that glomerular lesions stated as ‘other’ are not influenced by immunosuppressive treatment, we can calculate the percentage normal glomeruli during remission: normal (remission)=100−sclerosis (remission)−other (initial)=100−31−2=67%.

Fig. 6. 

Hypothesis on the evolvement of normal glomeruli, extracapillary proliferation, and glomerulosclerosis during the time interval between the two biopsies. At presentation of the disease, 12% of glomeruli in the initial renal biopsy were globally sclerosed. Assuming an analogous course of events during relapse, 12% of glomeruli that were not sclerotic during remission will be sclerotic in the follow‐up biopsy. Subsequently, the percentage newly sclerosed glomeruli in the follow‐up biopsy (i.e. developed during relapse) can be calculated as follows: new sclerosis (follow‐up)=12/100 of the not sclerotic glomeruli during remission. The ‘not sclerotic glomeruli during remission’ can be substituted: new sclerosis (follow‐up)=12/100 (normal (follow‐up)+crescents (follow‐up)+new sclerosis (follow‐up)+other (follow‐up)). This formula can be converted to: new sclerosis (follow‐up)=12/88 (normal (follow‐up)+crescents (follow‐up)+other (follow‐up))=12/88 (30+30+1)=8%. For the percentage sclerosis, present on average during remission, we can subsequently calculate: sclerosis (remission)=sclerosis (follow‐up)−new sclerosis (follow‐up)=39−8=31%. Assuming that crescents will not persist during remission without consequence on renal function and that glomerular lesions stated as ‘other’ are not influenced by immunosuppressive treatment, we can calculate the percentage normal glomeruli during remission: normal (remission)=100−sclerosis (remission)−other (initial)=100−31−2=67%.

Discussion

Renal involvement is a common and usually severe feature of ANCA‐associated vasculitis, which is characterized histopathologically by a pauci‐immune crescentic necrotizing glomerulonephritis. How the renal histopathology changes on a long‐term course is largely unknown. Insight into this issue can only be gained by studying follow‐up biopsies. Only a few studies have been published discussing renal histology in follow‐up biopsies [1120], and only two of these included patients with vasculitis. Aasarød et al. described 14 patients with Wegener's granulomatosis of which follow‐up renal biopsies were available [20]. In 13 patients, chronicity scores had significantly increased in the follow‐up biopsies as compared with the initial biopsies, but crescents were still present in eight of the cases. In a series of six patients with crescentic glomerulonephritis (of which four were pauci‐immune) and follow‐up biopsies (time points not stated), Kunis et al. observed persistent extracapillary proliferation in only one follow‐up biopsy whereas it had been present in all initial biopsies [19]. Furthermore, the percentage of sclerotic glomeruli in the follow‐up biopsy was approximately the same as the sum of the percentage of crescentic and sclerotic glomeruli in the initial biopsy. Their results suggest that treatment prevents new crescent formation, but that it cannot prevent glomerulosclerosis of the initially affected glomeruli. This hypothesis seems in accordance with the model of crescentic glomerulonephritis in the rabbit, where collagen deposition in the glomerulus starts at the onset of extracapillary proliferation [21]. On the other hand, in this rabbit model, 90% of the glomeruli show extracapillary proliferation at the beginning of the experiment, and despite an early onset of collagen deposition, follow‐up biopsies after 45 days show approximately 50% histologically normal glomeruli and global sclerosis of the other 50%. In the human situation, spontaneous improvement of crescentic glomerulonephritis sometimes occurs [7,22], although to our knowledge, this was never reported in ANCA‐associated systemic vasculitis.

We present a retrospective clinical‐pathological analysis of 31 patients with ANCA‐associated vasculitis in which we studied the evolvement of renal histology. Initial biopsies were performed before the onset of treatment. From this time point onwards, renal function either stabilized or improved. Follow‐up renal biopsies were performed for a suspected renal relapse after an interval of 3–72 months. Our results on the changes in renal pathology and serum creatinine were not influenced by diagnosis, gender, time interval between the biopsies, patient's age or applied treatment (corticosteroids only vs corticosteroids and additional drugs such as cyclophosphamide and azathioprine), although we should keep in mind that follow‐up biopsies were performed at the indication of a suspected relapse, which may have affected the influence of these variables on our results.

The percentage of normal glomeruli in the initial and the follow‐up biopsy was similar. This means that, on average, no glomeruli are recruited into the active disease process in any particular time interval in patients who receive immunosuppressive treatment and who are clinically free of relapses between both episodes. We found that the percentage of glomeruli with extracapillary proliferation was significantly lower in the follow‐up biopsies, although the lesion was still present in an average of 30% of the glomeruli and in 87% of all biopsies. This lower percentage of crescents per biopsy may be due to a relatively early follow‐up biopsy, with the diagnosis already established. Fibrinoid necrosis occurred in our patient group in combination with extracapillary proliferation only, which confirms our previously reported observations on this issue [23]. This means that in 13% of biopsies, no active glomerular lesions were present. The mean percentage of glomeruli affected by global sclerosis increased from 12 to 39%. Accordingly, the sum of the percentage of crescentic and sclerotic glomeruli in the initial biopsies was larger than the percentage of sclerotic glomeruli in the follow‐up biopsies, which is in contrast with the results that were reported previously by Kunis et al. [19]. None of the 13% of biopsies without active lesions showed 100% glomerulosclerosis (maximal 88%), which can be explained by the fact that renal biopsies are not performed in patients with small end‐stage kidneys. The current opinion on the role of immunosuppressive therapy in the treatment of renal lesions is that this type of medication has beneficial effects on at least some of the crescentic glomeruli, but not on glomerulosclerosis. Our results show that in one out of ten patients with a suspected renal relapse, no lesions or lesions insensitive to immunosuppressive drugs are found.

Figure 6 represents a hypothesis on the evolvement of normal glomeruli, extracapillary proliferation, and glomerulosclerosis during the time interval between diagnosis and the follow‐up biopsy. It suggests that during therapy, a certain percentage of the glomeruli with extracapillary proliferation recover to histologically normal glomeruli and other glomeruli are doomed to develop glomerulosclerosis. This is analogous to the aforementioned rabbit model [21]. There may be a certain point of no return, after which a crescentic glomerulus loses its potential of recovering to a histologically normal glomerulus, for instance when crescents become fibrous [7]. In addition, during and after this remission induction, therapy seems to prevent normal glomeruli from being recruited into the active disease process for a certain time. We suggest that these processes result in a quiescent phase of the disease with a stable or a relatively low serum creatinine, in which no glomeruli with extracapillary proliferation are present. At the time of relapse, recovered glomeruli or glomeruli that never have been affected by the disease—at least histologically—may develop extracapillary proliferation and glomerulosclerosis during that stage. Alternatively, it is possible that crescents still exist during remission without largely affecting renal function.

Although interstitial fibrosis was observed in both initial and follow‐up biopsies, its presence was more pronounced in the follow‐up biopsies (Table 4). For tubular atrophy, a similar pattern was observed. It is generally accepted that tubulo‐interstitial disease accompanies glomerulonephritis, especially with longer duration of the disease. Various hypotheses, reviewed by Pichler et al. [24], have been suggested for the pathogenesis of tubulo‐interstitial involvement in glomerulonephritis. According to these hypotheses, it is likely that the persistence of glomerular inflammation and the occurrence of glomerulosclerosis promote interstitial fibrosis and tubular atrophy.

We also investigated whether there was a correlation between the changes in renal histology and the change in renal function (Table 5). Our results suggest that increased serum creatinine levels can be predicted neither by acute inflammation nor by chronic damage.

The main drawback of this study consists of the problem that, in current clinical practice, patients with renal disease in relation to vasculitis will not undergo a follow‐up renal biopsy without medical indication, which makes prospective studies to address the evolvement of histopathological lesions on a larger scale virtually impossible. Our study comprises patients with a suspected relapse of the disease but not patients with aggressive rapidly progressive disease or patients with ‘grumbling’ disease. This means that one should be cautious about drawing clinical conclusions from the present study. Future studies are needed to address this issue in depth. Furthermore, a decreased incidence of lesions that are related to irreversible injury (glomerulosclerosis, tubular atrophy, and interstitial fibrosis) was observed in the follow‐up biopsies of a few patients. As the decreased incidence of these lesions did not coincide with each other, sampling error seems the most plausible explanation for this phenomenon. Finally, the retrospective character of this study makes it impossible to account for the influence of antihypertensive drugs on the histological changes between the initial and follow‐up biopsy for at least three reasons. First, the year in which the initial biopsy was performed ranged from 1972 to 1996. The development in the field of antihypertensive drugs has been enormous during this period. Secondly, the time interval between the biopsies varied between 3 and 161 months. More than one antihypertensive drug will probably have been administered to patients with a long time interval between the biopsies. Finally, some patients received antihypertensive drugs not only during the interval between the biopsies, but also before systemic vasculitis had been diagnosed.

In summary, we present a study of follow‐up renal biopsies in 31 patients with ANCA‐associated vasculitis with renal involvement, who experienced a suspected relapse after initial treatment. This is the first study in which renal histopathological differences between biopsies performed at the moment of diagnosis and biopsies performed for a suspected renal relapse were quantified. The histological changes were independent of gender, diagnostic classification, time interval between the biopsies, patient's age, and treatment. Our results suggest that, in our patient group, on average, no glomeruli are recruited into the active disease process. The sum of the percentage of crescentic and sclerotic glomeruli in the initial biopsies is larger than the percentage of sclerotic glomeruli in the follow‐up biopsies. Thus, therapy seems not only to prevent normal glomeruli from being recruited into the active disease process for a certain time, but seems also to allow part of the active lesions to revert into a normal phenotype, although another part of the active lesions will be transformed to a chronic phenotype.

We would like to thank David Jayne for kindly providing material and information of one patient in this study. Twelve of our patients were included in the European Commission/Bureau Communautaire de Référence (EC/BCR) Project for ANCA‐Assay Standardisation:

Study Coordinators: E. C. Hagen, F. J. van der Woude, M. R. Daha (Leiden University Medical Center, the Netherlands). Steering Committee: G. Gaskin and C. D. Pusey (Hammersmith Hospital, London, United Kingdom), K. Andrassy (University of Heidelberg, Heidelberg, Germany), N. Rasmussen (Rigshospitalet, Copenhagen, Denmark), A. Wiik (Statens Serum Institut, Copenhagen, Denmark), F. Ferrario and R. A. Sinico (San Carlo Borromeo Hospital, Milan, Italy), Z. Heigl (Karolinska Institute, Stockholm, Sweden), D. Jayne, C. M. Lockwood (Addenbrooke's Hospital, Cambridge, UK), C. G. M. Kallenberg, J. W. Cohen Tervaert (University Hospital Groningen, the Netherlands), Ph. Lesavre (Hôpital Necker, Paris, France), J. Lüdemann (Utecht and Lüdemann, Klausdorf, Germany), F. Mascart‐Lemone (Hôpital Erasme, Brussels, Belgium), E. Mirapeix (Hospital Clinic I Provincial, Barcelona, Spain), A. Tzioufas (National University of Athens, Athens, Greece), J. Wieslander (Wieslab AB, Lund, Sweden), K. de Groot, W. L. Gross (University of Lübeck, Bad Bramstedt, Germany). Pathology Review:Renal biopsies: I. M. Bajema, J. A. Bruijn (Leiden University Medical Center, the Netherlands), L. H. Noël (Hôpital Necker, Paris, France), R. Waldherr (University of Heidelberg, Heidelberg, Germany), F. Ferrario (San Carlo Borromeo Hospital, Milan, Italy). Respiratory tract biopsies: B. Ravn Juhl, C. B. Andersen (Rigshospitalet, Copenhagen, Denmark). Statistical Analysis: J. Hermans, B. E. Hansen (Leiden University Medical Center, the Netherlands). Data Management: M. J. K. Mallat (Leiden University Medical Center, the Netherlands). Participating investigators: E. Csernok (Rheumaklinik Bad Bramstedt, University of Lübeck, Bad Bramstedt, Germany), M. de Waele (Academisch Ziekenhuis Vrije Universiteit, Brussels, Belgium), W. Szpirt, J. Petersen (Rigshospitalet, Copenhagen, Denmark), C. Geffriaud (Hôpital Necker, Paris, France), G. Gregorini (Ple Spedali Civili, Brescia, Italy), M. Quarenghi (Ospedale S. Anna, Como, Italy), A. Lopez Soto (Hospital Clinic I Provincial, Barcelona, Spain), E. Pettersson (Huddinge University Hospital, Huddinge, Sweden), J. Berglund, T. Zweig, S. Jacobson (Karolinska Institute, Stockholm, Sweden), P. Chapman (Addenbrooke's Hospital, Cambridge, UK). Other Support:technical assistance: E. Heemskerk (Leiden University Medical Center, the Netherlands), A. Radice (San Carlo Borromeo Hospital, Milan, Italy), J. M. Flodman (Karolinska Institute, Stockholm, Sweden), A. Coulthart (Hammersmith Hospital, London, UK). Providing biopsy material: M. Thompson (Hammersmith Hospital, London, UK), S. Thiru (Addenbrooke's Hospital, Cambridge, UK).

References

1
Pinching AJ, Lockwood CM, Pussell BA et al. Wegener's granulomatosis: observations on 18 patients with severe renal disease.
Q J Med
 
1983
;
208
:
435
–460
2
Ronco P, Verroust P, Mignon F et al. Immunopathological studies of polyarteritis nodosa and Wegener's granulomatosis: a report of 43 patients with 51 renal biopsies.
Q J Med
 
1983
;
206
:
212
–223
3
Godman GC, Churg J. Wegener's granulomatosis. Pathology and review of the literature.
Arch Pathol
 
1954
;
58
:
533
–553
4
Savage COS, Harper L, Adu D. Primary systemic vasculitis.
Lancet
 
1997
;
349
:
553
–558
5
Luqmani RA, Bacon PA, Beaman M et al. Classical versus non‐renal Wegener's granulomatosis.
Q J Med
 
1994
;
87
:
161
–167
6
Bajema IM, Hagen EC, Hermans J et al. Kidney biopsy as a predictor for renal outcome in ANCA‐associated necrotizing glomerulonephritis.
Kidney Int
 
1999
;
56
:
1751
–1758
7
Atkins RC, Nikolic‐Paterson DJ, Song Q, Lan HY. Modulators of crescentic glomerulonephritis.
J Am Soc Nephrol
 
1996
;
7
:
2271
–2278
8
Hagen EC, Daha MR, Hermans J et al. Diagnostic value of standardized assays for anti‐neutrophil cytoplasmic antibodies in idiopathic systemic vasculitis.
Kidney Int
 
1998
;
53
:
743
–753
9
Jennette JC, Falk RJ, Andrassy K et al. Nomenclature of systemic vasculitides. Proposal of an international consensus conference.
Arthritis Rheum
 
1994
;
37
:
187
–192
10
Bajema IM, Hagen EC, Hansen BE et al. The renal histopathology in systemic vasculitis: an international survey study of inter‐ and intra‐observer agreement.
Nephrol Dial Transplant
 
1996
;
11
:
1989
–1995
11
Esdaile JM, Joseph L, MacKenzie T, Kashgarian M, Hayslett JP. The pathogenesis and prognosis of lupus nephritis: information from repeat renal biopsy.
Semin Arthritis Rheum
 
1993
;
23
:
135
–148
12
Moroni G, Pasquali S, Quaglini S et al. Clinical and prognostic value of serial renal biopsies in lupus nephritis.
Am J Kidney Dis
 
1999
;
34
:
530
–539
13
Villa M, Fogazzi GB, Ambroso GC. Crescentic glomerulonephritis with normal renal function after 28 years of follow‐up.
Nephrol Dial Transplant
 
1998
;
13
:
2671
–2673
14
Briggs WA, Tanawattanacharoen S, Choi MJ, Scheel PJ, Nadasdy T, Racusen L. Clinicopathologic correlates of prednisone treatment of human immunodeficiency virus‐associated nephropathy.
Am J Kidney Dis
 
1996
;
28
:
618
–621
15
Meyrier A, Noël LH, Auriche P, Callard P. Long‐term renal tolerance of cyclosporin A treatment in adult idiopathic nephrotic syndrome. Collaborative Group of the Societe de Nephrologie.
Kidney Int
 
1994
;
45
:
1446
–1456
16
Fukuda Y, Ohtomo Y, Kaneko K, Yabuta K. Pathologic and laboratory dynamics following the removal of the shunt in shunt nephritis.
Am J Nephrol
 
1993
;
13
:
78
–82
17
Myers BD, Chagnac A, Golbetz H, Newton L, Strober S, Sibley RK. Extent of glomerular injury in active and resolving lupus nephritis: a theoretical analysis.
Am J Physiol
 
1991
;
260
:
F717
–F727
18
Pillemer SR, Austin HA3, Tsokos GC, Balow JE. Lupus nephritis: association between serology and renal biopsy measures.
J Rheumatol
 
1988
;
15
:
284
–288
19
Kunis CL, Kiss B, Williams G, D'Agati V, Appel GB. Intravenous ‘pulse’ cyclophosphamide therapy of crescentic glomerulonephritis.
Clin Nephrol
 
1992
;
37
:
1
–7
20
Aasarød K, Bostad L, Hammerstrom J, Jorstad S, Iversen BM. Renal histopathology and clinical course in 94 patients with Wegener's granulomatosis.
Nephrol Dial Transplant
 
2001
;
16
:
953
–960
21
Downer G, Phan SH, Wiggins RC. Analysis of renal fibrosis in a rabbit model of crescentic nephritis.
J Clin Invest
 
1988
;
82
:
998
–1006
22
Couser WG. Rapidly progressive glomerulonephritis: classification, pathogenetic mechanisms and therapy.
Am J Kidney Dis
 
1988
;
11
:
449
–464
23
Hauer HA, Bajema IM, de Heer E, Hermans J, Hagen EC, Bruijn JA. Distribution of renal lesions in idiopathic systemic vasculitis: a three‐dimensional analysis of 87 glomeruli.
Am J Kidney Dis
 
2000
;
36
:
257
–265
24
Pichler R, Giachelli C, Young B et al. The pathogenesis of tubulointerstitial disease associated with glomerulonephritis: the glomerular cytokine theory.
Miner Electrolyte Metab
 
1995
;
21
:
317
–327

Comments

0 Comments