Crescentic glomerulonephritis

(proliferative extracapillary)

Anti-GBM disease - Pauci-immune glomerulonephritis


Proliferative extracapillary glomerulonephritis (GN) or crescentic GN is not a specific disease, but a histologic manifestation of severe glomerular damage. The term “extracapillary proliferation” is used to designate the cellular and/or fibrous proliferation that occupies the Bowman’s space, arising from its capsule. Extracapillary indicate that proliferation occur outside of capillary tuft. There is no universal agreement on the percentage of involved glomeruli to diagnose crescentic GN, some authors or texts say that more than 50%, others more than 30%, but that percentage of glomeruli with crescents, to call it extracapillary, depends on the clinical context. In ANCA associated GN it is possible to say that it is an extracapillary GN with only one crescent. In immunocomplexes mediated GN there is more controversy, for example, in IgA nephropathy, some authors propose more than 50% and others more than 30%. More important than that, is to quantify the crescents in all cases, of course, the higher the percentage of crescents, more aggressive the clinical behavior is expected to be.

The definition of crescent is the presence of at least two layers of cells that are filling totally (circumferential) or partially (circumscribed) the Bowman’s space.

This alteration is also known like rapidly progressive GN, subacute GN, and malignant GN. It is histologically characterized by crescents and, clinically, by loss of kidney function in weeks to months. Several glomerular diseases can produce this clinicopathologic picture and many of the glomerulopathies seen in other chapters can produce crescents. Nevertheless, in this chapter we will see glomerular diseases in which crescents constitute the main histologic feature.

Crescentic GN with rapidly progressive disease can be seen in postinfectious GN, membranoproliferative GN, IgA nephropathy, dense deposit disease, and others, but in these cases it is better to call it according to the primary glomerulopathy, adding the commentary “… with extensive crescents, in X% of glomeruli”.

There is a classification of crescentic GN according to immunopathologic features:

Type I: Produced by anti-glomerular basement membrane (GBM) antibodies.

Type II: Due to immune complexes deposited in glomeruli.

Type III: Without deposits of immunoglobulins or complement in glomeruli: pauci-immune. This group has been subdivided in: associated to antineutrophil cytoplasmic antibodies (ANCA): microscopic polyangiitis, granulomatosis with polyangiitis (Wegener), eosinophilic granulomatosis with polyangiitis (Churg–Strauss), or renal-limited vasculitis; and not associated to ANCA (idiopathic crescentic GN).

A fourth type has been proposed for those cases in which coexistence of anti-GBM disease and ANCA-associated GN is documented. These types clearly are differentiated in all texts of the matter, nevertheless, the denomination: type I, type II and type III is not very used in histopathological diagnosis.

In 2010 an international working group of renal pathologists proposed a new histopathological classification that is based on glomerular pathology:

Classification schema for ANCA-associated glomerulonephritis (Berden AE, et al. Histopathologic classification of ANCA-associated glomerulonephritis. J Am Soc Nephrol. 2010;21:1628-36. [PubMed link] [Free full text]):

Focal: ≥50% normal glomeruli.

Crescentic: ≥50% glomeruli with cellular crescents.

Mixed: <50% normal, <50% crescentic, <50% globally sclerotic glomeruli.

Sclerotic: ≥50% globally sclerotic glomeruli.

This classification is of aid in the prognostication of patients at the time of diagnosis and facilitates uniform reporting between centers.

Crescentic GN mediated by immune complexes can be seen in lupus nephritis, Henoch-Schönlein purpura, IgA nephropathy, postinfectious GN, membranoproliferative GN, cryoglobulinemia, hepatitis B, endocarditis and, in general, in any GN mediated by immune complexes. Each one of these diseases will be presented in other chapters. In the present one we will present anti-GBM disease and puci-immune crescentic GN.

Crescentic GN corresponds to <10% of all the biopsies with GN diagnosis. According to the three defined types the distribution is approximately thus: anti-GBM disease: 20%; immune complexes mediated: 40%, and pauci-immune: 40%.

Origin of extracapillary proliferation. Crescents are formed by a variable proportion of epithelial cells and monocytes (macrophages); according to the evolution of crescent the predominant cells are monocytes (earlier phase) or epithelial cells (later phases). Also other inflammatory cells may be identified, mainly lymphocytes and fibroblasts in fibroepithelial phases. In cases with segmental necrosis and/or rupture of the Bowman’s capsule usually there are more monocytes, and in cases of immune complexes disease there are more epithelial cells. Crescents are associated to rupture of capillary walls and fibrin in the urinary space. This fibrin and other proteins of the plasma seem to have the capacity to stimulate epithelial cells so that they proliferate, in addition, the mediators released by monocytes and platelets would contribute to the cellular proliferation and, probably, to posterior fibrosis. Crescents may evolve to fibrosis (fibrous crescent) or disappear (by apoptosis). Fibrosis is mediated by infiltration of fibroblasts from the periglomerular interstitium through spaces in the Bowman’s capsule.


Anti-GBM disease

This disease frequently appears like a pulmonary-renal syndrome, characterized by severe hemoptysis, pulmonary alterations, and rapidly progressive GN (RPGN). Although many decades ago patients with these clinical characteristics have been described, only in 1964 and 1965 was described linear deposits of immunoglobulins in the GBM in patients with the disease. Later Lerner et al established that the anti-GBM antibodies are the cause of pulmonary and glomerular lesions (J Exp Med 126:989-1004, 1967 [Full text link]).

From the initial description of Ernest Goodpasture of a patient with renal and pulmonary disease, and necrotizing vasculitis of spleen and intestine, the understanding and nomenclature of the disease has changed very much. The eponym “Goodpasture disease” (or Goodpasture syndrome) has been used in a variable way in the literature, but at the moment it is used to denominate the pulmonary-renal syndrome produced by anti-GBM antibodies. Nevertheless, all the patients with anti-GBM antibodies do not present Goodpasture disease: in some patients there is not pulmonary involvement.

The anti-GBM antibodies are directed against the alpha-3 chain in the C-terminal non-collagenous domain of type IV collagen (alpha-3 [IV] chain of NC1 domain). In patients with Alport’s syndrome (without formation of type IV collagen of adult type: without alpha-3, alpha-4 nor alpha-5 chains) the disease is not developed. The target antigen is in kidney, lung, and other tissues, such as choroids plexus and choclea. The antibody recognizes the same antigen in patients with Goodpasture syndrome, isolated GN, or exclusive pulmonary involvement, without GN. Therefore, it is considered that other factors must exist, additional to the presence of the antibody, so that the disease is developed.

Two epitopes have been identified in the alpha-3 (IV) NC1 domain (EA and EB); both are inaccessible for antibody binding unless dissociation of the hexamer occurs. The EA and EB epitopes are located in close proximity to each other, near the triple helical junction, and they are sequestered at the interface between NC1 domains within a triple helical protomer, rather than between adjacent protomers. It is presumed that an environmental factor, such as exposure to hydrocarbons or tobacco smoke, is required in order to reveal cryptic epitopes to the immune system. Endogenous oxidants can open this privileged site, as can certain subpopulations of anti–glomerular basement membrane antibodies. (Hudson BG, et al, N Engl J Med; 348:2543-56, 2003 [PubMed link][Free full text]).

When the antibodies bind to the GBM activate complement and proteases, this activation produces damage of the membrane and proteins liberation to the urinary space facilitating the formation of crescents. CD4 and CD8 lymphocytes induce migration of macrophages and neutrophils. Interleukin 12 and gamma-interferon contribute to crescent formation. There are also genetic factors related to disease development. People with HLA-DRB1*1501 and DRB1*1502 have more susceptibility, while HLA-DR7 and DR1 are protective (Kalluri R, et al, J Clin Invest 100:2263-2275, 1997 [PubMed link] [Free full text]).

Anti-GBM disease is rare, affects people of any age, but it is commonest in 20-40 years of age. There is greater frequency in men than in women.

Clinical features: It is usually manifested as RPGN, often accompanied by pulmonary hemorrhage. Although many patients present with severe pulmonary and renal disease, the range is ample and the disease rarely presents with hematuria and proteinuria in the absence of renal insufficiency or pulmonary hemorrhage. Sometimes the disease is preceded by a catarrhal episode, although it is not known if a viral infection has relevant roll in the pathogenesis. In approximately one third of the cases renal disease is accompanied by pulmonary hemorrhage with hemoptysis, dyspnea, and ronchi, although in follow-up this percentage increase approximately until 50%. Most patients with pulmonary hemorrhage have pallor because of anemia.

Renal disease is characterized by oliguric or anuric renal failure, macro or microhematuria, and other unspecific general symptoms. According to sodium retention and volume expansion, hypertension may be detected.

Treatment is based on high doses of steroids, other immunosuppressors and plasma exchange. Anti-GBM disease has worse prognosis than immune complexes GN and pauci-immune GN. Around half of the patients will develop terminal renal failure. After a complete remission recurrence of active disease is uncommon but may occur even many years later. The disease also can recur in renal allografts, but this is uncommon if transplantation is delayed until anti-GBM antibodies are undetectable.

See Case 93 of our case series: Anti-GBM disease

Histopathology, immunofluorescence and electron microscopy: See below.


Pauci-immune proliferative extracapillary GN

Crescentic GN without immunoglobulins or complement deposition is called pauci-immune. The disease may be part of a systemic vasculitis and, since affects glomerular capillaries, it would be called small vessel vasculitis. When it is not part of a systemic disease and there is only renal involvement, it is also known with the names idiopathic crescentic GN, primary crescentic GN, or vasculitis limited to the kidney.

Approximately 90% of patients with pauci-immune crescentic GN have ANCAs. Cases of idiopathic crescentic GN not associated to ANCAs, immune complexes nor anti-GBM antibodies are very rare (approximately 5%). Pauci-immune crescentic GN is the most frequent cause of rapidly progressive GN and the pulmonary-renal syndrome. The differential diagnosis in crescentic pauci-immune GN is: Wegener’s granulomatosis, microscopic polyangiitis, Churg-Strauss syndrome, and idiopathic crescentic GN.

The diagnosis of granulomatosis with polyangiitis (Wegener) is based on the demonstration of granulomatous vasculitis. Granulomas are more frequently demonstrated in respiratory tract, they are very unusual in renal tissue, but, when they are identified in this organ must be separated of glomeruli with Bowman’s capsule rupture, since when this happens it is very frequent formation of granulomas as a reaction to this destruction. We diagnosed eosinophilic granulomatosis with polyangiitis (Churg–Strauss) when there is a history of asthma and eosinophilia and there is pulmonary granulomas; and we diagnosed microscopic polyangiitis when there is a systemic vasculitis without granulomas or history of asthma. There is clinical or morphologic evidence of renal involvement in 90% of patients with microscopic polyangiitis, 80% of patients with Wegener and 45% of patients with Churg-Strauss.

The roll of ANCAs in the pathogenesis of these diseases is not completely understood. ANCAs can be directed against myeloperoxidase (P-ANCAs: perinuclear pattern of staining) or against proteinase 3 (C-ANCAs: cytoplasmic pattern of staining). The cytoplasmic pattern of staining refers to the appearance when indirect immunofluorescence in neutrophils alcohol-fixed is performed; perinuclear pattern seems to be an artifact for precipitation of the substrate. In Wegener’s granulomatosis usually there are C-ANCAs; in microscopic polyangiitis there are more frequency of P-ANCAs; and in patients with crescentic GN limited to the kidney and patients with Churg-Strauss syndrome there is predominantly P-ANCAs, although there is variability.

We know that levels of ANCA antigens correlate with the degree of activity of the disease. ANCAs can activate the neutrophils and produce degranulation and liberation of free radicals. The activation mechanism seems to imply union to Fc receptors and union of Fab'2 in the cellular surface (Kettritz R, et al, J Am Soc Nephrol; 8:386, 1997 [PubMed link]). Neutrophils activated by ANCAs injure the endothelium causing vessels damage (vasculitis); endothelial necrosis seems to be an initial event in the pathogenesis. In agreement with some hypotheses, in patients with circulating ANCAs an inflammatory event, like a respiratory infection, prepares (or initiates) neutrophils and monocytes to be activated by ANCAs; approximately 90% of patients relate a catarrhal episode before the beginning of the symptoms.

Clinical features: As in the case of anti-GBM disease, it usually appears with a rapidly progressive GN. The initial presentation usually is with oliguria, creatinine increase, variable proteinuria, macro or microhematuria, and arterial hypertension. Occasionally there is nephrotic syndrome. In some patients the serum creatinine can become stabilized in high levels, but in others renal insufficiency progresses until chronic terminal disease.

Extrarenal presentation includes fever, arthralgias, myalgias, purpura, peripheral neuropathy, respiratory alterations, and signs of vasculitis in digestive tract. There are cutaneous lesions (purpura and/or nodules) in near 40% of the cases. In Churg-Strauss syndrome there is extrarenal involvement by vasculitis in a high percentage of the cases. In Wegener there is greater frequency of superior respiratory tract involvement by necrotizing granulomas or vasculitis (80-90%). Pulmonary cavities and nodules can be seen (necrotizing granulomas); alveolar hemorrhage, arteritis and capillaritis are frequent (pulmonary involvement in near 90%). In microscopic polyangiitis there is superior respiratory tract involvement in approximately 30% of the cases, and pulmonary involvement in 50%. The clinical presentation of the Wegener’s granulomatosis and microscopic polyangiitis does not allow making a precise differential diagnosis; this is possible only demonstrating presence of granulomas. In many cases without granulomas these two diseases can not be distinguished with accuracy; nevertheless, relevance for treatment is not very great since it is similar.

Approximately 90% of patients respond to aggressive immunosuppressor treatment (usually cyclophosphamide and steroids) with complete remission rate near 75%. Many patients present relapses that respond to treatment in 70%. In severe cases plasmapheresis can be useful but there is controversy with respect to the results.

Laboratory findings: In anti-GBM disease and in pauci-immune crescentic GN we can find similar findings in serum and urine analyses: hematuria (macro or microscopic), erythrocyte cast, variable proteinuria, and creatinine and BUN increase. In most of patients with anti-GBM disease serum antibodies (anti-GBM) can be detected, but it is a difficult technique in which there are false negatives. ANCAs are detected in 80-90% of patients with Wegener’s granulomatosis and microscopic polyangiitis, and in 60% of patients with Churg-Strauss syndrome. In approximately 30% of patients with anti-GBM disease ANCAs are detected, but it is difficult to know their pathogenic roll in this disease; these patients seem to have better prognosis than those with anti-GBM antibodies without ANCAs. Complement levels usually are normal.


Histopathological presentation in renal tissue is similar for anti-GBM and pauci-immune crescentic GN. There is glomerular inflammation with necrotizing lesions; GN is usually focal and segmental, but it may be global and diffuse (Figure 5). Necrotizing lesions are invariably accompanied by crescents, although it is possible to see crescents in glomeruli without tuft necrosis. Crescents are associated to capillary walls rupture (easier to demonstrate with the methenamine-silver staining) (Figure 2) and fibrin deposition in Bowman’s space. Crescents are formed by proliferating epithelial cells and monocytes; sometimes there are lymphocytes and polymorphous. When the crescents contain only cells, without collagen, is called: epithelial crescent (Figures 1 and 2). When advancing the process fibroblasts migrating to the Bowman’s space synthesize collagen that replaces progressively the crescent (easier to demonstrate with trichrome or methenamine-silver stains); when cellular components are mixed with collagen the lesion is called: fibroepithelial crescent (Figure 3); in advanced stages, without epithelial cells and only fibrous tissue, the lesion is called fibrous crescent (figure 4). These three stages indicate active or chronic lesions and therefore give an idea of the response to treatment.

Figure 1. This case corresponds to a 58-years-old patient with rapidly progressive GN. See proliferation of cells occupying the entire Bowman’s space (green arrows) and compressing the tuft (red arrows). The Bowman’s capsule is indicated by the blue arrows. This is the characteristic aspect of an epithelial crescent, in this case circumferential. Cells forming the crescent may be epithelial, monocytes or other inflammatory cells. (Masson’s trichrome, X400).

Figure 2. In epithelial crescents the proliferating cells are not accompanied by fibrous or collagen tissue. Methenamine-silver stain emphasizes the compressed tuft and rupture of capillary walls (arrows). This capillary damage is an important pathogenic phenomenon in the generation of the extracapillary proliferation. (Methenamine-silver, X400).

Figure 3. In fibroepithelial crescents we can see a mixture of large cells, with cleared nucleus, prominent nucleolus, and ample cytoplasm, that are mixed with fibrous or collagen tissue that is highlighted with the silver stains (like in this photo) and with the trichrome stains. The blue arrows indicate the tuft compressed and distorted; the red arrows indicate large, active, proliferating cells, and green arrows mark the fibrosis replacing many areas of this crescent; this combination of lesions indicates a process that evolves to chronicity. (Methenamine-silver, X400).

Figure 4. When advancing the destructive process of glomeruli the crescent becomes more fibrous (or scarred) and the cells composing it diminish progressively until disappearing, being replaced by fibroblasts. Stains for fibrous tissue, like trichrome and silver, emphasize this component (green arrows), in this case without epithelial component; the blue arrows indicate remains of glomerular tuft. (Methenamine-silver, X400).

In the tuft (in addition to necrosis) capillary collapse, mesangial matrix increase, and karyorrhexis can be identified. Frequently there is segmental or extensive destruction of Bowman’s capsule (Figure 5); in these cases it is usual to find granulomas and multinucleated cells surrounding the glomerulus, these granulomas do not indicate Wegener’s granulomatosis.

Figure 5. In this glomerulus we can see the tuft deformed, with fibrin in a segment, and fragmentation of nuclei (karyorrhexis) (green arrows) that indicate segmental necrosis. There is an epithelial circumferential crescent with an area evolving to fibroepithelial (left superior). Red arrows mark the capillary tuft collapsed. Mononuclear inflammatory infiltrate is prominent around this glomerulus; this severe infiltrate is associated with rupture of the Bowman’s capsule. In some cases we will be able to demonstrate periglomerular granulomas that do not indicate Wegener’s granulomatosis. (Masson’s trichrome, X400).

In some cases vessel can show transmural vasculitis, with fibrinoid necrosis, lymphocyte infiltrates and fragmented polymorphous (leucocytoclastic vasculitis) (Figure 6). Any vessel can be involved, being the arterial lesions easiest to demonstrate. In interstitium it is usual to find inflammatory mononuclear infiltrates, edema and, in advanced stages, fibrosis. Tubules may have acute changes: tubular necrosis, epithelial degeneration, and tubulitis; or they may have chronic lesion: atrophy. In anti-GBM disease the antibodies may also recognize antigens in tubular basement membranes presenting greater tubular damage and interstitial inflammation. Interstitial granulomas not associated to glomeruli indicate Wegener’s granulomatosis, but it is a very unusual feature in kidney; habitually they have a necrotic center and they are surrounded by epithelioid histiocytes and multinucleated giant cells.

Figure 6. The case of this microphotography corresponds to a 62-year-old man with microscopic polyangiitis. In an intrarenal artery we demonstrated this vascular lesion. The arterial wall is pointed with green arrows, and the lumen with asterisks. In the central zone we can see an area with fibrinoid necrosis of the wall, polymorphous and cellular detritus (blue arrows). These lesions are not very frequent in intrarenal arteries, but its presence helps much in the vasculitis diagnosis. (Gomori’s trichrome, X400)..

In the tubules erythrocytes or cylinders can be identified, but its absence does not imply that the patient has not had hematuria. There is tubular epithelial damage in until 57% of biopsies: cellular desquamation, brush border loss, cytoplasmic vacuolation, and tubular dilatation; these changes are found mainly in patients with acute renal failure. Other signs of chronic damage can be also demonstrated: tubular atrophy and interstitial fibrosis. In 5% of cases there is severe acute tubulointerstitial inflammation. There is intimal fibrosis of arteries, medial hypertrophy, and hyaline arteriolar deposits in a variable proportion of cases.

In lung there is extensive alveolar hemorrhage, capillaritis and variable interstitial inflammation (in both, anti-GBM disease and systemic vasculitis). Involvement in other organs is characterized by small vessel vasculitis; accompanied by granulomas in Wegener and eosinophilic infiltrates in Churg-Strauss.

Histopathological classification for ANCA-associated glomerulonephritis: Berden AE, et al. Histopathologic classification of ANCA-associated glomerulonephritis. J Am Soc Nephrol. 2010;21(10):1628-36. [PubMed link] [Free full text]


In anti-GBM disease there are linear, global and diffuse deposits of IgG (Figure 7), usually accompanied by smaller amounts of linear C3. Occasionally there is linear staining for IgM or IgA in smaller intensity. Very unusual is the presence of IgA without IgG. The predominant IgG is IgG1 subclass. There is linear staining of IgG in tubular basement membranes in more than 50% of cases. There is also linear staining in the basement membranes of alveolar capillaries, but the staining is more irregular, with less intensity, more background, and more difficult for interpretation than renal staining.

Figure 7. Immunofluorescence for IgG in a case of Goodpasture’ syndrome (a29-year-old man). See the linear staining in the capillary walls. The immunostaining for C3 may show a similar aspect, but usually it is less intense. (Direct immunofluorescence for IgG, antibodies anti-IgG human marked with fluorescein, X400).

In the pauci-immune crescentic GN (any of the three vasculitic diseases or in idiopathic crescentic GN) there are not deposits of immunoglobulins or complement. In some cases weak immunostaining can be identified, but if strong staining is found we must think in anti-GBM disease (if linear deposits) or in immune complexes disease (if granular staining). Necrotizing segments can have unspecific binding of IgM and/or C3.

The crescent show positivity for fibrinogen.

Electron microscopy

In anti-GBM disease and pauci-immune GN the findings are very similar: rupture of GBM and Bowman’s capsule, focal effacement of podocyte foot processes, fibrin in urinary space and tuft, and fibrinoid necrosis.

It must not have electron-dense deposits indicating immune complexes disease.

Prognostic factors

There are no histopathologic findings that specifically indicate better or worse prognosis, but, like in all the glomerulopathies, chronic lesions correlate with irreversible damage and its extension should be quantified or semiquantified: percentage of glomerulosclerosis, fibroepithelial and fibrous crescents, interstitial degree of fibrosis, tubular atrophy, and intimal arterial fibrosis. In addition, severity of the glomerular changes should be also quantified: percentage of glomeruli with crescents and necrosis. But outcome in each case is the truly prognostic determinant. The histopathologic classification is useful for prognosis (Berden AE, et al.Histopathologic classification of ANCA-associated glomerulonephritis. J Am Soc Nephrol. 2010;21(10):1628-36. [PubMed link] [Free full text])


Suggested articles: