Background

Segmental glomerular necrosis has been described in the biopsy material in a minority of patients with idiopathic IgA nephropathy in the oldest studies on this disease, but this marker of active capillaritis has received little attention in the subsequent literature, and its significance and relevance for the clinical outcome is still unknown.

Methods

Thirty-five out of 340 patients (10.3%) biopsied in our division at the San Carlo Hospital since 1974 showed active segmental necrotizing lesions. The morphological features and the natural history of this group of patients were compared with those of a control group of 229 patients who had comparable serum creatinine and extent of glomerular sclerosis, but who lacked active segmental necrosis.

Results

Patients with the necrotic variant showed a significantly more marked extracapillary proliferation and interstitial accumulation of monocytes and T lymphocytes and, in the segmental areas of necrotizing and extracapillary lesions, infiltration of monocytes, deposition of fibrinogen, and expression of the adhesion molecule vascular cell adhesion molecule-1. No difference was found in the presenting clinical syndrome. The clinical course was frequently characterized by acute flareups, and the progression to end-stage renal failure was more frequent, although actuarial renal survival was not significantly worse (P = 0.07). The aggressive treatment with steroids and cyclophosphamide, carried out in 20 of the 35 patients, has probably been beneficial, justifying the multicenter controlled trial that recently has been initiated.

Conclusions

Vasculitic lesions of the glomerular capillaries, with histologic and immunohistological features similar to those of Henoch-Schönlein purpura and antineutrophil cytoplasmic antibody-positive renal vasculitis, were found in 10% of patients with idiopathic IgAN. Clinical features at presentation did not differ from those of the other patients with IgAN, and despite of the more frequent occurrences of recurrent acute flareups, rapid progression to end-stage renal failure was a rare phenomenon, even in untreated patients.

METHODS

Patients

From a review of our files, all cases of IgAN that had been observed and biopsied in our division from 1974 and met each of the following criteria were identified: (1) light microscopy slides available for review containing a minimum of six glomeruli; (2) immunofluorescence (IF) studies showing the prevalent mesangial deposition of IgA; and (3) no signs of systemic involvement reminiscent of Henoch-Schönlein purpura, systemic lupus erythematosus, or chronic liver disease, either at the time of the biopsy or during the complete course of the disease, before or after the biopsy.

Three hundred forty patients met all of these criteria. For each of these cases, the case folder, including all of the follow-up observations, and all available microscopic slides were reviewed (P.N. and F.F.) to select the histologic subset characterized by the presence, at least in one glomerulus, of segmental glomerular necrosis, defined as a segmental destructive lesion of the capillary loop, with fibrin deposition within and around the capillaries, usually surrounded by segmental areas of extracapillary proliferation. Of the 340 cases, 35 (10.3%) showed these features and represent the cohort of patients included in this study. In 33 of them, segmental glomerular necrosis was found in the first biopsy, while in the remaining two, a repeat biopsy, carried out because of an unfavorable clinical course, revealed the presence of the previously described characterizing histologic features that were absent at the first biopsy.

Since none of these 35 patients had a serum creatinine higher than 3.5 mg/dL and a global glomerular sclerosis in more than 50% of glomeruli at the time of the biopsy, to compare the morphological and clinical features among the 305 patients without necrotizing lesions, we arbitrarily selected only the 229 with a serum creatinine 3.5 mg/dL and a percentage of obsolescent glomeruli 50% at the time of biopsy.

Frozen biopsy tissue for the specific immunohistological studies with monoclonal antibodies was available for only 26 of the 35 patients with necrotizing lesions whose histologic material was obtained and processed in the last 10 years. As a control population to evaluate these immunohistological features, we selected 28 patients without necrotizing lesions who were consecutively biopsied during the same period of time and seven cadaver donors whose kidneys could not be grafted because of vascular abnormalities.

Histologic methods

Kidney tissue was obtained in all patients from routine biopsies. All specimens were processed routinely for light microscopy by the standard method. For each biopsy specimen, the degree of mesangial proliferation, global and segmental glomerular sclerosis, and interstitial fibrosis (graded from 0 to +++) were evaluated independently by two investigators, as previously described³⁶.

Immunoperoxidase labeling

For immunoperoxidase labeling, we used an avidin-biotin technique in which a biotinylated secondary antibody reacts with several peroxidase-conjugated streptavidin molecules.

Briefly, after incubation with 0.5% avidin (Sigma Chimica, Gallarate, Milan, Italy) and 0.01% biotin (Sigma) to suppress endogenous avidin-binding activity, tissue sections were fixed in a methanol-H₂O₂ solution (to block endogenous peroxidase) and then incubated with the primary antibody. The specificities of anti-human monoclonal antibodies used were as follows: CD45, mouse anti-human leukocyte common antigen (Dako, S.p.A., Milan, Italy); CD3, rabbit anti-human T lymphocyte (Dako); CD68, mouse anti-human monocyte/macrophage (Dako); CD54, mouse anti-human intercellular adhesion molecule-1 (ICAM-1; Serotec, Oxford, UK); CD106, mouse anti-human vascular cell adhesion molecule-1 (VCAM-1; Serotec). After washing, the sections were sequentially incubated with the secondary biotinylated antibody (Dako) and with the peroxidase-labeled streptavidin (Dako). Peroxidase activity was detected with 3,5-diaminobenzidine (DAB; Dako); sections were then counterstained with Harry's hematoxylin (BDH, Poole, UK), dehydrated, and mounted in Entellan (Merck, Darmstadt, Germany).

Specificity of labeling was proved by the lack of staining after substitution of phosphate-buffered saline (PBS) for the primary antibody.

Evaluations of immunostained sections

Immunostained sections, which contained at least six glomeruli, were evaluated independently by two investigators without any clinical and pathological information, as previously described³⁷.

Intraglomerular-labeled cells were counted and expressed as mean cells SD per glomerular cross section, and cells within crescents were included in glomerular counts.

Labeled interstitial cells were counted in a sequence of 10 consecutive 400 high-power fields using an eyepiece graticule (Periplan 6F 12.5 MF mounted in a Leitz Dialux 20 microscope; Wetzlar, Germany), avoiding glomeruli, large vessels and focal infiltrates. The results were expressed as mean cells SD/mm².

Periglomerular infiltrates were evaluated separately with a semiquantitative method (grading from 0 to ++).

Intraglomerular and tubular staining for adhesion molecules was scored semiquantitatively using a four-point score: 0 = no staining; + = focal staining; ++ = diffuse staining; +++ = massive staining.

Antineutrophil cytoplasmic antibodies

Twenty-two of the 35 patients were measured both by indirect immunofluorescence (IF) and by enzyme-linked immunosorbent assay (ELISA) according to the standardized procedures already described^38,39.

Statistical methods

Results are expressed as mean value SD. The smallest and the largest values are reported for frequency data. Data were analyzed by two-tailed Student t test for paired or unpaired data when required. Differences among proportions were tested using chi-square test. Survival curves were constructed according to the Kaplan–Meier method, and the log-rank test was applied to determine survival. P values 0.05 were considered significant. Statistical analyses were performed using statistical procedures of SAS system, version 6.04 (SAS Institute, Cary, NC, USA).

RESULTS

Histological and immunohistological features

As previously indicated, the characterizing feature in all the 35 patients selected for the study was the presence, at least in one glomerulus, of segmental glomerular necrosis, defined as a destructive lesion of the capillary loop, with fibrin deposition within and around the capillaries Figure 1a. The necrotizing lesions involved a variable percentage of glomeruli ranging from 3 to 25% (mean 8.6), and were usually associated with adjacent segmental areas of extracapillary cellular proliferation and also frequently with adhesions of the glomerular tuft to Bowman's capsule Figure 1a. As indicated in Table 1, this extracapillary proliferation involved a percentage of glomeruli ranging between 8 and 53% (mean 24.5), significantly more marked than in the control population with IgAN and absence of glomerular necrotizing lesions (mean 4.5). In this latter population, extracapillary proliferation that was frequently fibrocellular or fibrotic was present in 29% of the 229 patients, and in only 13 of them it involved more than 10% of glomeruli (10 to 37%).

Figure 1.

(A) The glomerulus shows a segmental area of tuft necrosis associated with a small cellular crescent. A periglomerular leukocyte infiltration is also present (Trichrome; 250). (B) Immunofluorescence. Deposits of IgA located mainly in the mesangial stalk region (anti-IgA antiserum; 250). (C) Immunofluorescence. Fibrinogen is strongly positive only in well-delineated areas of the glomerular tuft, corresponding to areas of segmental intracapillary necrosis seen at light microscopy (antifibrinogen antiserum; 250). (D) The glomerulus shows monocyte infiltration restricted to a segmental area of the tuft (immunoperoxidase conjugated anti-CD68 antibody; 250). (E) Intense periglomerular leukocyte infiltration in a patient with necrotizing lesions is evident (immunoperoxidase-conjugated anti-CD45 antibody; 250). (F) Strong positivity of adhesion molecule VCAM-1 is present in segmental area of the glomerular tuft corresponding to necrotizing crescentic lesion seen by light microscopy (immunoperoxidase conjugated anti–VCAM-1 antibody; 250). (G) The adhesion molecule ICAM-1 is significantly overexpressed and extended to the whole mesangial area in a glomerulus from a patient with IgAN. (H) In a repeat biopsy, a well-delineated segmental area of glomerular sclerosis with Bowman's capsule adhesion is present, as a consequence of previous (first biopsy) localized necrotizing-extracapillary lesion (Trichrome; 250).

Full figure and legend (125K)

Table 1 - Histological and immunohistological lesions in 35 patients with necrotizing IgAN and in 229 patients with non-necrotizing IgAN.

Full table

Another characterizing feature by light microscopy was the presence of interstitial infiltration of diffuse and/or segmental leukocytes, frequently with a periglomerular accentuation Figure 1a. As shown in Table 1, both focal and diffuse interstitial infiltration were found in a significantly higher number of biopsy specimens in comparison with the control population of patients with IgAN and absence of necrotizing lesions, while the extent of glomerular sclerosis and interstitial fibrosis did not differ in the two groups of patients.

By IF, there was no difference in the frequency and intensity of deposition of IgA and IgG Figure 1b, and IgM, nor in the incidence of the combined mesangial and capillary staining for IgA between patients with and without necrotizing lesion Table 1. However, segmental areas of fibrinogen deposition were present in only the subgroup of patients with necrotizing lesions Figure 1c.

Immunochemical features

Table 2 summarizes the results of the immunoperoxidase labeling of infiltrating cells in the glomeruli and in the interstitium of the fresh frozen biopsy material taken from 26 patients with segmental necrotizing IgAN compared with that obtained from 28 patients without such lesions who were consecutively biopsied during the same period of time and with the normal kidney tissue taken from cadaver donors. The two populations of IgAN patients with and without necrotizing lesions did not differ in mean age (32 12 vs. 35 15 years), male:female ratio (18:8 vs. 21:7), mean serum creatinine levels (1.3 0.5 vs. 1.2 0.3 mg/dL), and frequency of arterial hypertension (42 vs. 43% of patients), while the mean daily proteinuria was significantly higher in patients with necrotizing lesions (2.3 2.0 vs. 0.7 0.8 g/24 h; P < 0.01). The number of total intraglomerular leukocytes (CD45+ cells), monocytes (CD68+ cells), and T lymphocytes (CD3+ cells) was not significantly higher in patients with non-necrotizing IgAN than in normal controls. In patients with necrotizing IgAN, the average number of intraglomerular total leukocytes and monocytes was significantly higher than in patients with non-necrotizing lesions. The cells infiltrating the glomeruli in patients with necrotizing IgAN were concentrated in segmental areas corresponding to the areas of necrosis shown by traditional stain in the same glomeruli Figure 1d. Interstitial infiltration of total leukocytes and monocytes was not significantly more marked in patients with non-necrotizing IgAN than in normal controls, while the average number of T lymphocytes was significantly higher Table 2. All infiltrating interstitial cells were significantly more numerous in the necrotizing variant of IgAN in comparison with non-necrotizing IgAN. The cells infiltrating the interstitium had a prevalent periglomerular localization in two thirds of the patients with necrotizing IgAN Figure 1d, e.

Table 2 - Glomerular and interstitial expression of markers of leukocyte infiltration in patients with necrotizing IgAN and non-necrotizing IgAN.

Full table

Table 3 summarizes data on glomerular expression of adhesion molecules VCAM-1 and ICAM-1. VCAM-1, which was weakly expressed only on epithelial cells of Bowman's capsule in glomeruli from normal kidney tissue and from biopsy tissue of patients with non-necrotizing IgAN, was intensely expressed in segmental and focal, well-delineated areas of the tuft in some glomeruli in patients with necrotizing IgAN Figure 1f. These areas corresponded exactly, as site and percentage of glomeruli involved, to the areas of segmental necrotizing and crescentic involvement observed with thrichrome staining, as well as to the areas with monocyte infiltration described previously in this article.

Table 3 - Glomerular expression of adhesion molecules in 26 patients with necrotizing IgAN and in 28 patients with non-necrotizing IgAN.

Full table

Intercellular adhesion molecule-1 (which in normal kidneys was expressed only on the endothelial cells of the glomerular capillary wall) was significantly overexpressed in glomeruli from patients with both types of IgAN, extending to the entire mesangial area Figure 1g. The intensity of this expression was concordant with the extent of mesangial damage (proliferation and/or matrix expansion).

Repeated biopsies

In 2 of the 35 patients, the histological and immunohistological features described previously in this article were found upon a repeat biopsy four and five years after the first biopsy, respectively, which were carried out because of an unfavorable clinical course (increase of proteinuria and/or serum creatinine). Features of glomerular segmental necrosis were absent at the first biopsy in both cases. Segmental extracapillary proliferation and intraglomerular infiltration of monocytes were found in both cases only at the second biopsy, together with a marked increase of interstitial infiltration of leukocytes, both monocytes and T lymphocytes.

A second biopsy was also carried out in six additional patients from 5 to 17 years after a first biopsy showing the existence of segmental necrotizing lesions. In one of them, the new biopsy was performed because after seven years there was persistent severe proteinuria and a mild impairment of renal function. The biopsy documented that the very mild segmental necrotic involvement of the glomerular capillaries was unchanged. Unchanged as well was the extent of the extracapillary proliferation and of the intraglomerular monocyte infiltration, while the interstitial infiltration of total leukocytes and of T lymphocytes had increased from 129 to 437 mm² and from 68 to 368 mm², respectively. In the remaining five patients, the new biopsy was performed in the absence of clinical signs suggesting an active disease, to document the fate of the morphological signs of segmental capillary necrosis. In all cases, necrotic lesions were no longer present, while segmental areas of sclerosis with glomerulocapsular synechiae were found Figure 1h. Extracapillary proliferation had also completely disappeared in four of them, while it had increased in the remaining patient. Intraglomerular infiltration of monocytes had decreased in all patients to less than one cell per glomerulus, and interstitial infiltration of monocytes and T-lymphocytes was less marked in all patients, while interstitial fibrosis was moderately increased.

Clinical features and natural history

Asymptomatic urinary abnormalities characterized by microscopic hematuria and proteinuria in the non-nephrotic range were the only clinical signs at the time of biopsy in 28 (80%) of the 35 patients with necrotizing lesions. An episode of macroscopic hematuria had occurred in three (9%) patients, and nephrotic syndrome was found in one (3%). Two patients had chronic renal failure, and one patient experienced a rapid decline of renal function.

Table 4 compares demographic and clinical parameters at the time of biopsy in the 35 patients with necrotizing lesions and in 229 patients without necrotizing lesions selected as previously indicated. Both groups had a similar age and male predominance. Mean serum creatinine levels were not significantly different. Gross hematuria at the time of biopsy or a previous history of recurrent macroscopic hematuria were reported in similar percentages of patients. Frequency of arterial hypertension was not significantly different in the two groups. The mean duration of the prebiopsy follow-up was longer in patients without signs of necrotizing vasculitis than in patients with necrotizing lesions: 60 months (range 3 to 240) versus 33 months (range 4 to 191). Only the average amount of urinary daily loss of proteins at the time of biopsy was significantly different in the two groups of patients: It was higher in patients with segmental necrotic lesions.

Table 4 - Demographic and clinical features at the time of renal biopsy in 35 patients with necrotizing IgAN and in 229 patients with non-necrotizing IgAN.

Full table

Although we searched for IgG ANCAs in 22 of the 35 patients with necrotizing lesions using IF and ELISA, no patient had them.

The study of the natural history of patients with necrotizing lesions has been complicated by the interfering variables of anti-inflammatory and immunosuppressive therapies instituted in some of them during the last 10 years, in a nonrandomized fashion, because of some signs of a more progressive course (rapid, usually moderate increase of proteinuria, and of serum creatinine). In treated as well as in nontreated patients, the clinical course was frequently characterized by recurring episodes of more acute disease, as indicated by a sudden increase in urinary protein loss and slight elevation of serum creatinine levels. However, no systemic signs of Henoch-Schönlein purpura manifested during the follow-up.

The treatment was identical to that in use in our unit for the patients with rapidly progressive crescentic glomerulonephritis caused by vasculitis. It consisted of steroids (3 intravenous pulses of 0.75 to 1.0 g of methylprednisolone, followed by oral prednisone for 6 months at progressively tapering doses, starting with 0.5 mg/kg/day) and cyclophosphamide (2 mg/kg/day for 16 weeks). It was started in all patients immediately after biopsy. This specific therapy was added to the usual therapy with variable doses of angiotensin-converting enzyme (ACE) inhibitors, which has been given to all of our patients with IgAN since 1993, including those without necrotizing lesions.

The extent of necrotic involvement and of extracapillary proliferation was comparable in the groups of treated and untreated patients (9 vs. 8.3% of glomeruli and 28.7 vs. 21.4% of glomeruli, respectively), and the difference in initial serum creatinine levels was not significantly different. The duration of postbiopsy follow-up was quite variable in the different patients with necrotizing lesions, ranging from 12 to 204 months (mean 63 months). Since immunosuppressive treatment was instituted in the more recent years, the duration was shorter in the 15 treated patients (mean 50.4 months; range 12 to 108 months) than in the 20 untreated patients (mean 74 months; range 12 to 204 months). During this period, only one of the treated patients progressed to end-stage renal failure (ESRF) requiring dialysis, this occurring 48 months after biopsy. By comparison, 7 of the 20 patients who did not receive immunosuppressive treatment progressed to ESRF, with a mean interval from the renal biopsy of 70.2 months (range 36 to 96 months). However, because of the limited number of patients and of the different duration of the follow-ups, it was not possible to compare renal survival in the two subgroups of patients. A comparative analysis of the variations of mean serum creatinine and daily proteinuria levels was performed on the 20 untreated and the 15 treated patients, arbitrarily using one year after biopsy as the end point, since none of the patients had reached ESRF at that time. As shown in Table 5, mean proteinuria was significantly lower than at the time of biopsy in treated patients, while no significant differences were found in the mean levels of serum creatinine. Figure 2 illustrates the clinical course of one treated patient who had received three courses of treatment over the previous five years, with a reduction of urinary protein loss and stabilization of serum creatinine to levels approximately 2 mg/dL.

Figure 2.

Flow chart of the values of urinary protein excretion and serum creatinine in one patient with necrotizing IgAN receiving three courses of six months of treatment with steroid and cyclophosphamide during a follow-up period of more than seven years.

Full figure and legend (38K)

Table 5 - Serum creatinine and proteinuria in treated and untreated patients with necrotizing IgAN at the time of biopsy and one year later.

Full table

A Kaplan–Meier estimate of renal survival in the 229 patients without necrotizing lesions selected as the control group was compared with that of the subgroup of 15 patients with necrotizing IgAN who received treatment with steroids and cyclophosphamide and of the subgroup of 20 patients who did not receive treatment (these patients, as those with non-necrotizing IgAN, since 1993 have been treated with ACE inhibitors; Figure 3). Despite the trend to a more rapid loss of renal function in untreated patients with necrotizing IgAN, the difference in the rate of progression did not reach statistical significance (P = 0.07).

Figure 3.

Life-table analysis of renal survival in 229 patients with non-necrotizing IgAN (solid line), in 15 treated (dotted line) and in 20 untreated (dashed line) patients with necrotizing IgAN.

Full figure and legend (11K)

DISCUSSION

While segmental necrosis of the glomerular tuft usually associated with extracapillary proliferation has always been considered a characterizing histologic lesion in the glomerulonephritis of Henoch-Schönlein purpura, the less frequently documented presence of such a lesion even in some patients with idiopathic IgAN has received much less attention in the last 20 years, after its accurate description in many of the oldest morphological studies on this disease^{1,2,3,4,5,6,7,8,9}. This fact is hard to understand if one considers that it points to a pathogenetic mechanism different from that which is universally accepted as preeminent in this disease (that is, the damaging effect of deposited macromolecular IgA1 on the mesangial cells⁴⁰) and suggests a concomitant direct injury on the endothelial cells of the glomerular capillary wall, with its potential disruption that induces extracapillary proliferation.

The histological and immunohistological features of our 35 patients show a strict similarity with the necrotizing capillaritic lesion described in Henoch-Schönlein purpura⁴¹ and in ANCA-positive systemic vasculitis^42,43. Characteristics found in these two diseases were (1) a variable degree of extracapillary proliferation, usually segmental and adjacent to the areas of necrosis; (2) the intraglomerular influx of infiltrating cells, mainly monocytes, restricted to the areas of necrosis; (3) a significantly more marked interstitial accumulation of infiltrating cells (both monocytes and T lymphocytes) than in non-necrotic IgAN, with a frequent periglomerular accentuation; and (4) the intense deposition of fibrinogen in areas of tuft necrosis seen by IF.

Infiltrating glomerular monocytes were a characterizing feature of our subgroup of patients, since we did not find increased numbers of these cells in patients with non-necrotizing IgAN, including those with intense mesangial proliferation. This finding is in agreement with that of Li et al, who described intraglomerular infiltration of monocytes only in patients with "crescentic" IgAN (the concomitant presence of necrotizing lesions has not been looked for by these investigators), but not in those without extracapillary proliferation⁴⁴. Nagata et al in early biopsies of pediatric patients found a variable infiltration of glomerular monocytes/macrophages closely associated with endocapillary proliferative lesions and mesangial proliferation, but also with tuft necrosis, and with the "sites of formation of small cellular crescents"⁴⁵. A significant correlation between the extent of intraglomerular infiltration of monocytes-macrophages and degree of extracapillary proliferation in IgAN, without any mention of the concomitant presence or absence of necrotizing lesions, has also been reported by Arima et al⁴⁶. Even the more marked accumulation of monocytes and T cells in the interstitium, which we found in the necrotizing variant, had been already described by Li et al in their IgAN patients with crescents⁴⁴.

The study of ICAM-1 and VCAM-1 expression in the biopsies of our patients deserves some commentary. In eight patients with IgAN and absence of crescents, Brujin and Dinklo reported an up-regulation of ICAM-1—but not of VCAM-1—in the mesangial area, similar to that which we found in our patients, independently of the presence of necrotizing lesions⁴⁷. However, in the segmental areas of necrotizing and extracapillary lesions, we also found an abnormal expression of the adhesion molecule VCAM-1 that we had already described as a distinctive marker of the glomerular necrotizing lesions in ANCA-positive systemic vasculitis⁴³.

These data suggest that while ICAM-1 probably plays a role in the mesangial damage present in all patients with IgAN, VCAM-1 overexpression is the marker of a more specific lesion, the segmental necrotizing lesion present in a limited subgroup of patients with IgAN, either idiopathic or systemic (Henoch-Schönlein purpura), and in all patients with ANCA-positive systemic vasculitis. Since no IgG-ANCAs could be detected in any of our tested patients with necrotizing lesions, we hypothesize that the activation of VCAM-1 is the consequence of the endothelial inflammatory damage, independently of the immunologic mechanism responsible for the direct injury of the endothelial cells. This negative serum ANCA determination in our tested patients is worth emphasizing in view of the recent description of a few cases of rapidly progressive crescentic glomerulonephritis with clinical signs of systemic vasculitis, high titers of IgG-antimyeloperoxidase antibodies, and mesangial deposition of IgA shown by IF^48,49,50. As we have said, systemic signs were absent at presentation and during the clinical course in all of our patients with necrotizing IgAN. We believe that the isolated cases described in the recent literature refer to associations of IgAN and ANCA-positive systemic vasculitis, while there is agreement that in the absence of such an association between two different diseases, IgG-ANCAs are constantly negative in both idiopathic IgAN and Henoch-Schönlein purpura^51,52,53.

It is important to note that the extracapillary proliferation, although on average more marked than in the control cohort of patients without active signs of necrosis Table 1, was not widespread and circumferential in our patients with necrotizing lesions; rather, the extracapillary proliferation was characteristically segmental and located in close vicinity to the areas of necrosis in the glomeruli. Frequently, it did not involve more than 10 to 15% of glomeruli and slightly exceeded half of them in only one patient. In other words, necrotizing lesions in idiopathic IgAN are not always associated with massive crescent formation, and a mild noncircumferential extracapillary proliferation was sometimes the only sign of an extracapillary epithelial damage. On the other hand, we found signs of segmental extracapillary proliferation even in some of the patients of the group arbitrarily selected as the control group because of the absence of signs of active segmental glomerular biopsy in the biopsy specimen. Considering that in such a control group extracapillary proliferation was frequently in a fibrotic stage and that it was sometimes associated with a mild intraglomerular monocyte infiltration of leukocytes, we cannot rule out the possibility that it represents the indirect sign of a necrotic lesion, either healed or present in an active stage in a glomerular area not included in the stained section, as suggested by Shouno et al³⁵. If such is the case and the control group includes cases of "necrotizing" IgAN without obvious signs of segmental necrosis in the limited available specimen, the differences between the two groups of patients compared in Tables 1 and 4 become even more significant. Moreover, it justifies the policy of processing many more additional sections when the histologic material obtained routinely does not show intraglomerular segmental necrosis but documents an extracapillary proliferation exceeding 20% of glomeruli, especially when conventional staining shows marked interstitial infiltration of leukocytes, with a periglomerular accentuation.

At variance with what is found in patients with ANCA-positive systemic vasculitis, our patients with necrotizing IgAN showed a marked and diffuse mesangial proliferation as frequently as in the cohort of patients without necrotizing lesions Table 1. This suggests that the damaging effect on the mesangial cell coexists with that directly exerted on the endothelium of the capillary wall. Considering that the biopsy revealing active necrotizing lesions has been performed in our 35 patients many years after the apparent onset [the average interval was 33 months (range of 4 to 192)], it can be hypothesized that these lesions may occur in a patient with the typical mesangial pathology in a later stage of the disease, triggered by some as yet unknown superimposed immunologic event. This hypothesis is confirmed by the absence of active necrosis in the first biopsy in our two patients who were rebiopsied because of a rather sudden change in clinical conditions.

The postbiopsy follow-up of our patients as well as the data emerging from rebiopsy in some patients suggest that the necrotic lesions and the concomitant extracapillary proliferation may heal, leaving segmental areas of glomerular sclerosis and fibrous crescents, with a clinical course characterized by stages of quiescence (especially after treatment) and possible subsequent exacerbations (new areas of active segmental necrosis?), similar to that which is frequently documented for the glomerular involvement of ANCA-positive systemic vasculitis. The flow chart depicted in Figure 2 illustrates this clinical behavior.

If we accept the proposed hypothesis that tissue deposition of an abnormal IgA1 produced by the immunocompetent cells is responsible for the morphological lesions characterizing both idiopathic IgAN and Henoch-Schönlein purpura^40,54,55, the difference between these two diseases being that in Henoch-Schönlein purpura the damage produced by such deposition is more widespread, also involving extrarenal tissues, then we can hypothesize that the spectrum of the lesions inducible in different patients is a continuum. This continuum has the lesions restricted to the mesangium at one end and the lesions involving the mesangium and the wall of small vessels (to give a systemic vasculitis with cutaneous purpura) at the other end, but includes an intermediate range of lesions, restricted to the kidney but not to the mesangium, to give a glomerular capillaritis concomitant with the mesangial lesions. Obviously, if this hypothesis is accepted, all of the histologic classifications of IgAN based on a grading of overall severity of the glomerular and tubulointerstitial lesions that have been proposed and used in the last 15 years are inadequate and misleading.

Montinaro et al have recently described a model of IgAN obtained by the passive administration in mice of IgA/IgA immune complexes⁵⁶. By changing the type and amount of antigen to be fixed in situ to such IgA-immune complexes, they were able to reproduce glomerular lesions ranging from mild mesangial hypercellularity and expansion to severe mesangial involvement associated with segmental necrosis. The necrotizing lesions described by these investigators were comparable to those found in our 35 patients, being characterized by the intraglomerular influx and periglomerular deposition of monocytes, a more marked staining of fibrinogen by IF, and a swelling of the endothelial cells of the capillary loops by electron microscopy.

We hypothesize that in a few patients with idiopathic IgAN and in a larger number with Henoch-Schönlein purpura, IgA immune complexes of a different size and/or affinity for the endothelium of the peripheral capillary wall may form and deposit from time to time in the glomeruli, to give the necrotizing, renal-limited vasculitic lesions described in the current study. Rebiopsies show that these lesions develop during the course of the disease and that, once developed, may heal, suggesting that they may represent an acute reversible phenomenon.

This might explain why the clinical course of our patients showed that the superimposition of such necrotizing lesion on the typical mesangial injury was not necessarily an aggravating factor in the long run. At presentation, the features typical of an acute or rapidly progressive disease were not more frequent than in the control population of patients without demonstrable necrotizing lesions. In particular, only one patient exhibited at presentation signs of acute renal failure and another one a nephrotic syndrome, while recurrent macroscopic hematuria was reported in only 20% of the patients with necrotizing IgAN versus 31% of patients with non-necrotizing IgAN. Even arterial hypertension was not more frequent at the time of biopsy in patients with necrotizing lesions. In other words, necrotizing IgAN does not have characterizing clinical features at presentation and can be found casually in biopsy specimens taken from patients with an apparently indolent chronic clinical course. The retrospective analysis of the postbiopsy history suggests that even when only ACE inhibitors are administered to these patients, a rapidly progressive course is by no means the rule, although we suspect that chronic progression to ESRF may occur slightly more rapidly or frequently than in patients without necrotizing lesions.

That being said, is the aggressive treatment with steroids and immunosuppressive agents used in other types of necrotizing capillaritic lesions justified in patients with necrotizing IgAN nephropathy? Our personal experience does not allow us to answer this question, although in some patients with more active clinical signs of acute disease, such therapeutic approaches appear to have been beneficial. We have recently undertaken a multicenter controlled trial to define better the efficacy of steroids plus cyclophosphamide in this variant of IgAN.

References

1. Droz, D: Natural history of primary glomerulonephritis with mesangial deposits of IgA, in IgA Mesangial Nephropathy, Contrib Nephrol, 1976edited by D'amico G, Minetti L, Ponticelli C, Basel, Karger, pp 150–157
2. Clarkson, AR, Seymour, AE & Thompson, AJ, et al: IgA nephropathy: A syndrome of uniform morphology, diverse clinical features and uncertain prognosis. Clin Nephrol 19778: 459–471,  | PubMed | ISI | ChemPort |
3. Hood, SA, Velosa, JA & Holley, KE, et al: IgA-IgG nephropathy: Predictive indices of progressive disease. Clin Nephrol 198116: 55–62,  | PubMed | ISI | ChemPort |
4. Katz, A, Underdown, BJ & Minta, JO, et al: Glomerulonephritis with mesangial deposits of IgA unassociated with systemic disease. Can Med Assoc J 1976114: 209–215,  | PubMed | ISI | ChemPort |
5. Lomax-Smith, JD, Woodroffe, AJ & Clarkson, AR, et al: IgA nephropathy: Accumulated experience and current concepts. Pathology 198517: 219–224,  | PubMed | ChemPort |
6. Shigematsu, H, Kobayashi, Y & Tateno, S, et al: Ultrastructure of acute glomerular injury in IgA nephritis. Arch Pathol Lab Med 1980104: 303–307,  | PubMed | ISI | ChemPort |
7. Burkholder, PM, Zimmerman, SW, Moorthy, AV: A clinico-pathology study of the natural history of mesangial IgA nephropathy, in Glomerulonephritis Proceedings of the International Symposium, Glomerulonephritis, Progression and Regression, 1979, edited by Yoshitoshi LY, Baltimore, University Park Press, pp 143–166
8. Schmekel, B, Svalander, C & Bucht, H, et al: Mesangial IgA glomerulonephritis in adults. Acta Med Scand 1981210: 363–372,  | PubMed | ISI | ChemPort |
9. Croker, BP, Dawson, DV & Sanfilippo, F: IgA nephropathy: Correlation of clinical and histologic features. Lab Invest 198348: 19–24,  | PubMed | ISI | ChemPort |
10. Beukof, JR, Kardaun, O & Schaafsma, W, et al: Towards individual prognosis of IgA nephropathy in primary IgA. Kidney Int 198629: 549–556,  | PubMed |
11. D'amico, G, Minetti, L & Ponticelli, C, et al: Prognostic indicators in idiopathic IgA mesangial nephropathy. Q J Med 1986228: 363–378,
12. Nicholls, KM, Failey, KF & Dowling, JP, et al: The clinical course of mesangial IgA associated nephropathy in adults. Q J Med 1984210: 227–250,
13. Noel, LH, Droz, D & Gascon, M, et al: Primary IgA: From the first described cases to the present. Semin Nephrol 19877: 351–354,  | PubMed | ISI | ChemPort |
14. Bogenschütz, O, Bohle, A & Batz, C: IgA nephritis: On the importance of morphological and clinical parameters in the long-term prognosis of 239 patients. Am J Nephrol 199010: 137–147,  | PubMed |
15. Rekola, S, Bergstrand, A & Bucht, H: Development of hypertension in IgA nephropathy as a marker of a poor prognosis. Am J Nephrol 199010: 290–295,  | PubMed | ISI | ChemPort |
16. Johnston, PA, Brown, JS & Braumholtz, DA, et al: Clinico-pathological correlations and long-term follow-up of 253 United Kingdom patients with IgA nephropathy: A report from the MRC Glomerulonephritis Registry. Q J Med 199284: 619–627,  | PubMed | ISI | ChemPort |
17. Woo, KT, Edmondson, RPS & Wu, AYT, et al: The natural history of IgA nephritis in Singapore. Clin Nephrol 198625: 15–21,  | PubMed | ISI | ChemPort |
18. Kusumoto, Y, Takebayashi, S & Taguchi, T, et al: Long-term prognosis and prognostic indices of IgA nephropathy in juvenile and in adult Japanese. Clin Nephrol 198728: 118–124,  | PubMed | ISI | ChemPort |
19. Ibels, LS & Györy, AZ: IgA nephropathy: Analysis of the natural history, important factors in the progression of renal disease, and a review of the literature. Medicine (Baltimore) 199473: 79–102,  | PubMed | ChemPort |
20. Donadio, JV, Bergstralh, EJ & Offord, KP, et al: Clinical and histopathologic associations with impaired renal function in IgA nephropathy. Clin Nephrol 199441: 65–71,  | PubMed | ISI | ChemPort |
21. Alamartine, E, Sabatier, JC & Guerin, C, et al: Prognostic factors in mesangial IgA glomerulonephritis: An extensive study with univariate and multivariate analyses. Am J Kidney Dis 199118: 12–19,  | PubMed | ISI | ChemPort |
22. Okada, H, Suzuki, H & Konishi, K, et al: Histological alterations in renal specimens as indicators of prognosis of IgA nephropathy. Clin Nephrol 199237: 235–238,  | PubMed | ISI | ChemPort |
23. Radford, MG, Donadio, JV & Bergstralh, EJ, et al: Predicting renal outcome in IgA nephropathy. J Am Soc Nephrol 19978: 199–207,  | PubMed | ISI |
24. Katafuchi, R, Kiyoshi, Y & Oh, Y, et al: Glomerular score as a prognosticator in IgA nephropathy: Its usefulness and limitation. Clin Nephrol 199849: 1–8,  | PubMed | ISI | ChemPort |
25. Boyce, NW, Holdsworth, SR & Thomson, NM, et al: Clinicopathological associations in mesangial IgA nephropathy. Am J Nephrol 19866: 246–252,  | PubMed | ISI | ChemPort |
26. Frimat, L, Briançon, S & Hestin, D, et al: IgA nephropathy: Prognostic classification of end-stage renal failure. Nephrol Dial Transplant 199712: 2569–2575, 10.1093/ndt/12.12.2569 | PubMed | ISI | ChemPort |
27. Haas, M: Histologic subclassification of IgA nephropathy: A clinicopathologic study of 244 cases. Am J Kidney Dis 199729: 829–842,  | PubMed | ISI | ChemPort |
28. Freese, P, Nordén, G & Nyberg, G: Morphologic high-risk factors in IgA nephropathy. Nephron 199879: 420–425,  | Article | PubMed | ISI | ChemPort |
29. Yagame, M, Suzuki, D & Jinde, K, et al: Value of pathological grading in prediction of renal survival in IgA nephropathy. Nephrology 19962: 107–117,  | ISI |
30. Kobayashi, Y, Kokubo, T & Horii, A, et al: Prognostic prediction of long-term clinical courses in individual IgA nephropathy patients. Nephrology 19973: 35–40,  | ISI |
31. Lee, SMK, Rao, VM & Franklin, WA, et al: IgA nephropathy: morphologic predictors of progressive renal disease. Hum Pathol 198213: 314–322,  | PubMed | ISI | ChemPort |
32. Abuelo, JG, Esparza, AR & Matarese, RA, et al: Crescentic IgA nephropathy. Medicine (Baltimore) 198463: 396–406,  | PubMed | ChemPort |
33. Abe, T, Kida, H & Yoshimura, M, et al: Participation of extracapillary lesions (ECL) in progression of IgA nephropathy. Clin Nephrol 198625: 37–41,  | PubMed | ISI | ChemPort |
34. Ferrario, F, Napodano, P & Rastaldi, MP, et al: Capillaritis in IgA nephropathy. Contrib Nephrol 1995111: 8–12,  | PubMed | ChemPort |
35. Shouno, Y, Shimokama, T & Sakemi, T, et al: Segmental glomerular sclerosis as an active index of IgA nephropathy: A study with 100 serial sections of 128 cases. Acta Pathol Jpn 199343: 723–729,  | PubMed | ChemPort |
36. D'amico, G, Imbasciati, E & Belgioioso, GB, et al: Idiopathic IgA mesangial nephropathy: Clinical and histological study of 374 patients. Medicine 198564: 49–60,  | PubMed | ChemPort |
37. Rastaldi, MP, Ferrario, F & Yang, L, et al: Adhesion molecules expression in noncrescentic acute post-streptococcal glomerulonephritis. J Am Soc Nephrol 19967: 2419–2427,  | PubMed | ISI | ChemPort |
38. Sinico, RA, Radice, A & Pozzi, C, et al: Diagnostic significance and antigen specificity of antineutrophil cytoplasmic antibodies in renal diseases. A prospective multicentre study. Nephrol Dial Transplant 19949: 505–510,  | PubMed | ISI | ChemPort |
39. Hagen, EC, Andrassy, K & Csernok, E, et al: Development and standardization of solid phase assays for the detection of anti-neutrophil cytoplasmic antibodies (ANCA): A report on the second phase of an international cooperative study on the standardization of ANCA assays. J Immunol Methods 1996196: 1–15, 10.1016/0022-1759(96)00111-1 | Article | PubMed | ISI | ChemPort |
40. D'amico, G: Pathogenesis of immunoglobulin A nephropathy. Curr Opin Nephrol Hypertens 19987: 247–250,  | Article | PubMed | ChemPort |
41. Habib, R & Niaudet, P: Renal involvement in Schöenlein-Henoch purpura, in Renal Pathology, 1989, edited by Tisher CC, Brenner BM, Philadelphia, Lippincott Co., pp 409–433
42. Jennette, JC & Falk, RJ: The pathology of vasculitis involving the kidney. Am J Kidney Dis 199424: 130–141,  | PubMed | ISI | ChemPort |
43. Rastaldi, MP, Ferrario, F & Tunesi, S, et al: Intraglomerular and interstitial leukocyte infiltration, adhesion molecules, and interleukin-1 expression in 15 cases of antineutrophil cytoplasmic autoantibody-associated renal vasculitis. Am J Kidney Dis 199627: 48–57,  | PubMed | ISI | ChemPort |
44. Li, HL, Hancock, WW & Hooke, DH, et al: Mononuclear cell activation and decreased renal function in IgA nephropathy with crescents. Kidney Int 199037: 1552–1556,  | PubMed | ISI | ChemPort |
45. Nagata, M, Akioka, Y & Tsunoda, Y, et al: Macrophages in childhood IgA nephropathy. Kidney Int 199548: 527–535,  | PubMed | ISI | ChemPort |
46. Arima, S, Nakayama, M & Naito, M, et al: Significance of mononuclear phagocytes in IgA nephropathy. Kidney Int 199139: 684–692,  | PubMed | ISI | ChemPort |
47. Brujin, JA & Dinklo, NJ: Distinct patterns of expression of intercellular adhesion molecule-1, vascular cell adhesion molecule-1, and endothelial-leukocyte adhesion molecule-1 in renal disease. Lab Invest 199369: 329–335,  | PubMed |
48. Allmaras, E, Nowack, R & Andrassy, K, et al: Rapidly progressive IgA nephropathy with anti-myeloperoxidase antibodies benefits from immunosuppression. Clin Nephrol 199748: 269–273,  | PubMed | ISI | ChemPort |
49. Ramirez, SB, Rosen, S & Niles, J, et al: IgA antineutrophil cytoplasmic antibodies in IgA nephropathy: A clinical variant? Am J Kidney Dis 199831: 341–344,  | PubMed | ISI | ChemPort |
50. Van der Woude, F & Ferrario, F: Renal involvement in ANCA-associated systemic vasculitis. J Nephrol 199912: 55–78,
51. Savage, COS, Winearls, CG & Jones, S, et al: Prospective study of radioimmunoassay for antibodies against neutrophil cytoplasm in diagnosis of systemic vasculitis. Lancet 19871: 1390–1393,
52. Bygren, P, Rasmussen, N & Wieslander, J: ANCA in glomerulonephritis. APMIS 198997(Suppl 6):43,  | PubMed |
53. MacIsaac, AI, Moran, JE & Davies, DJ, et al: Antineutrophil cytoplasm antibody (ANCA) associated vasculitis. Clin Nephrol 199034: 5–8,  | PubMed | ISI | ChemPort |
54. Feehally, J & Allen, AC: Structural features of IgA molecules which contribute to IgA nephropathy. J Nephrol 199912: 9–15,  | PubMed |
55. Allen, AC, Willis, FR & Beattie, TJ, et al: Abnormal IgA glycosylation in Henoch-Schönlein purpura restricted to patients with clinical nephritis. Nephrol Dial Transplant 199813: 930–934, 10.1093/ndt/13.4.930 | Article | PubMed | ISI | ChemPort |
56. Montinaro, V, Esparza, AR & Cavallo, T, et al: Antigen as mediator of glomerular injury in experimental IgA nephropathy. Lab Invest 199164: 508–519,  | PubMed | ISI | ChemPort |

Acknowledgments

This work was done on behalf of EU Concerted Action "Chronic inflammation leading to scarring, the major mechanism for kidney failure" (http://www.kidney-euract.org).