Antiglomerular basement membrane disease
Antiglomerular basement membrane disease (anti-GBM disease, also known as Goodpasture’s disease) is a rare autoimmune disease caused by pathogenic autoantibodies directed against the noncollagenous, C-terminal domain of the α-3 chain of type IV collagen (α3(IV)NC1). Immunohistology is characteristic, with linear deposition of IgG (sometimes with IgA or IgM) and complement C3 along the GBMs.
Clinical features—anti-GBM disease classically presents with pulmonary haemorrhage (in two-thirds of patients) and rapidly progressive glomerulonephritis (RPGN, with haematuria and proteinuria, urinary red cell casts, and typically severe acute renal failure). Haemoptysis can be triggered by cigarettes, inhaled toxins, fluid overload, and intercurrent infection. The condition must be distinguished from other cause of RPGN, especially antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis, in particular because there is only a small window of opportunity in which to rescue renal function in patients with Goodpasture’s disease. Key investigations are (for diagnosis) serological testing for anti-GBM antibodies and ANCA, and renal biopsy, and (for detection of pulmonary haemorrhage) chest radiology and estimation of carbon monoxide transfer factor (Kco).
Management and prognosis—untreated anti-GBM disease is usually fatal, and renal function never recovers. Immunosuppressive treatment (plasma exchange, cyclophosphamide, oral steroids) is given immediately on diagnosis to patients with serum creatinine levels lower than 600 µmol/litre at presentation and/or with active pulmonary haemorrhage. In contrast, patients with serum creatinine levels more than 600 µmol/litre at presentation rarely recover renal function, hence immunosuppressive treatment in such cases would be restricted in most centres to those with pulmonary haemorrhage or in whom the renal biopsy suggested additional mechanisms of renal damage such as acute tubular necrosis. In recent series, 1-year patient survival is 66 to 92%, 1-year renal survival is 15 to 59%, with renal recovery in patients with initial creatinine levels greater than 600 µmol/litre of 0 to 21%. Patients do not need long-term immunosuppression, and the disease rarely recurs. Transplantation is safe if performed after autoantibodies have been suppressed or naturally disappeared.
Antiglomerular basement membrane disease is an autoimmune disease in which patients develop pathogenic autoantibodies against the glomerular basement membrane (GBM). They typically present with renal failure and pulmonary haemorrhage, but isolated renal disease and more rarely isolated lung disease are well recognized. The triad of anti-GBM antibodies, rapidly progressive glomerulonephritis (RPGN), and pulmonary haemorrhage is referred to as Goodpasture’s disease in the United Kingdom, while the term Goodpasture’s syndrome describes patients with RPGN and pulmonary haemorrhage of various aetiologies.
The term ‘Goodpasture’s syndrome’ was first used in 1958 by Stanton and Tange in their report of nine patients with pulmonary–renal syndrome, in which they referred to a patient with fulminant pulmonary haemorrhage and proliferative glomerulonephritis described by Goodpasture during the influenza pandemic of 1919. In retrospect, this original patient may have had systemic vasculitis, not anti-GBM disease. In recent years much has been learnt about the immune response in Goodpasture’s disease, but despite huge advances in our understanding of the aetiopathogenesis, therapy has changed little in the last 25 years.
Aetiology and pathogenesis
The Goodpasture antigen
All patients with Goodpasture’s disease have circulating antibodies that bind a GBM antigen, the α3 chain of type IV collagen (α3(IV)), with some also binding the α-5 chain. Type IV collagen is found in all basement membranes, but the α-3, α-4, and α-5 chains are restricted in their distribution primarily to the GBM and alveolar basement membranes. The epitope for autoantibodies in Goodpasture’s disease is carried at the N-terminus of the 230-amino acid, noncollagenous, C-terminal domain of the α-3 chain (α3(IV)NC1), which is normally hidden within the collagen network. The α3(IV)NC1 is also found in the basement membranes of the choroid plexus, the cochlea, Bruch’s membrane in the eye, retinal capillaries, and the thymus. It is now clear that a very limited number of amino acids within this domain confirm antigen specificity on the structure, and that both B-cell and T-cell immune responses are driven by this protein. Furthermore, the antigenic epitopes are normally not available for driving an immune response and only become immunogenic antigens when the collagen NC1 domains undergo conformational change.
Anti-GBM antibodies and the T-cell-mediated immune response
Transfer of antibodies from patients into squirrel monkeys initially confirmed the pathogenicity of the autoantibodies. Clinical studies report a correlation between antibody levels at presentation and disease activity, and the disease recurs immediately in renal transplants when the recipient still has circulating antibodies. All patients have antibodies against α3(IV)NC1, either circulating or bound to the GBM. A small number of patients develop antibodies against other GBM components, particularly the α-1 (15% of patients) or α-4 (4% of patients) chains of type IV collagen. In some cases it can be impossible to detect circulating antibodies using conventional assays, but antibody is found bound either to glomerular or alveolar basement membranes in all. However, anti-GBM antibodies are unlikely to be the only cause of glomerular injury, and a cell-mediated immune response is also important in inducing renal damage. In animal models it is possible to induce disease by immunizing with the Goodpasture antigen in the complete absence of B lymphocytes or immunoglobulin, although the damage induced in these cases by T lymphocytes and macrophages is less severe.
Alveolar haemorrhage generally requires a second insult, either local to the lungs (e.g. cigarette smoking or pulmonary oedema), or systemic with activation of cytokines and inflammatory mediators (e.g. sepsis).
Goodpasture’s disease has been reported in four sibling pairs and two sets of identical twins, but discordant twin pairs are also documented. More striking is the association with the HLA serotype HLA DR2, which is carried by more than 85% of patients with Goodpasture’s disease, compared with 30% of controls. Molecular analysis of HLA alleles has confirmed the association with HLA DR15 (DRB1*1501 and -1502), and a weaker association with HLA DR4 (DRB1*04). A negative association with HLA DR7 (DRB1*07) has been demonstrated. These associations have been confirmed in several studies of different white and Asian populations. Thus, specific characteristics of the HLA molecules on antigen-presenting cells determine susceptibility to Goodpasture’s disease.
No specific pathogens or toxins have been identified that can initiate Goodpasture’s disease, but many case reports have documented exposure to hydrocarbons before the development of clinical manifestations, and cigarette smoking undoubtedly precipitates pulmonary haemorrhage. It seems more likely that organic solvents trigger overt pulmonary damage (and possibly renal injury) in the presence of circulating autoantibodies than that hydrocarbons are involved in the initiation of autoimmunity. Several clusters of cases have been reported, but no clear associations with influenza virus or other infectious agents have been proven. In animal models it has been demonstrated that molecular mimicry between bacterial antigens and the Goodpasture antigen might be responsible for initiating disease. Thus, a single peptide derived from a Clostridium botulinum protein which shares crucial amino acids with the T-cell epitope in Goodpasture’s disease can induce both glomerular injury and even pulmonary haemorrhage in rats, but, as stated above, evidence for specific infective causes in humans is lacking.
Anti-GBM disease is only rarely associated with other autoimmune disorders, apart from systemic vasculitides. Increasing numbers of patients (up to 30%) have been shown to have circulating antineutrophil cytoplasmic antibodies (ANCA), generally p-ANCA, in addition to anti-GBM antibodies. Conversely, only few patients with ANCA-associated vasculitis also have anti-GBM antibodies (2.5–8%). This is an important distinction since the ‘double positive’ patients tend to behave more like those with ‘pure’ Goodpasture’s disease than systemic vasculitis.
Anti-GBM disease has been reported after lithotripsy and urinary tract obstruction, and in some patients with membranous nephropathy. In all these cases it is possible that disruption of the GBM in susceptible individuals can lead to a breakdown in tolerance to the α-3 chain of type IV collagen, with the development of autoantibodies and clinical disease. Patients with HIV infection are also described with anti-GBM antibodies, but almost never with Goodpasture’s disease, these antibodies being nonpathogenic and usually of low titre. HIV is, however, associated with a variety of renal diseases, including membranous nephropathy, IgA disease, collapsing glomerulopathy, and immune complex disease.
Immunohistology is characteristic (Fig. 18.104.22.168), with linear deposition of IgG (sometimes with IgA or IgM) and complement C3 along the GBMs. Rare patients have been reported with IgM or IgA alone. Less intense linear staining with IgG can occasionally be seen in diabetes, systemic lupus erythematosus, myeloma, and transplanted kidneys. The most characteristic morphological finding is severe crescentic glomerulonephritis, with almost all the glomeruli exhibiting cellular crescents, usually at the same stage of evolution. Segmental necrosis and cellular proliferation may occur. Blood vessels are usually normal, but rarely vasculitis has been reported, even in the absence of detectable ANCA. There is often a prominent interstitial cellular infiltrate.
Histological specimens are rarely obtained from lungs, since transbronchial biopsy does not usually penetrate beyond the bronchial mucosa. Open-lung biopsy can reveal alveoli full of red blood cells, macrophages, and fibrin, interspersed between relatively normal alveoli. Immunofluorescence inconsistently reveals linear deposition of antibody.
Limited epidemiological studies suggest that Goodpasture’s disease has an incidence of 0.5 to 1 new case per million of the population per year. It is found in 1 to 2% of renal biopsies. In comparison, systemic vasculitides have an incidence of 15 to 30 new cases per million of the population per year. The disease is less common in Afro-Caribbean populations, but increasingly reported in Asian people. There is a bimodal age distribution, with peak incidence in the third and sixth decades, and a slight excess of men.
Most patients present with RPGN or lung haemorrhage, or both. Some patients have isolated lung haemorrhage and never develop renal failure (although most of these have haematuria and proteinuria), and a few have mild isolated nephritis. General malaise, fatigue, weight loss, and anaemia are the commonest systemic features, while other signs and symptoms are much rarer than in patients with systemic vasculitis.
Pulmonary haemorrhage occurs in two-thirds of patients, more commonly in young men. It usually precedes presentation with acute renal failure, and is strongly associated with cigarette smoking. Patients often complain of breathlessness and cough. Haemoptysis can be triggered by cigarettes, inhaled toxins, fluid overload, and intercurrent infection, either local (pneumonia) or systemic (sepsis), but there is a poor relationship between overt haemoptysis and the degree of alveolar haemorrhage. Clinical signs are often indistinguishable from those of pulmonary oedema or infection. The most sensitive indicator is an elevated Kco (diffusing capacity for carbon monoxide), which identifies the presence of haemoglobin in alveolar spaces by increased binding of inhaled carbon monoxide. Radiographic features are not specific, but alveolar shadowing in the central lung fields is typically seen (Fig. 22.214.171.124).
Patients can present with isolated haematuria, chronic renal failure, or mild renal insufficiency, but classically present with severe acute renal failure due to rapidly progressive glomerulonephritis. The clinical features of the nephritis are indistinguishable from any other cause of RPGN, with cellular casts in the urine, haematuria, and mild-to-moderate proteinuria (nephrotic range proteinuria is uncommon but reported). Hypertension and oliguria are late features. A few patients have relatively normal renal function at presentation, but always have abnormal urine findings and evidence of antibody deposition on renal biopsy.
It is crucial to distinguish anti-GBM disease from other causes of RPGN, and especially ANCA-associated vasculitis. There is only a small window of opportunity in which to rescue renal function in patients with anti-GBM disease, in contrast to systemic vasculitis in which renal failure can be reversed at a later stage. All patients with suspected RPGN, acute renal failure of unknown cause, or lung haemorrhage and urinary abnormalities, should have both anti-GBM antibody and ANCA assays performed urgently. Other differential diagnoses to consider include systemic lupus erythematosus, cryoglobulinaemia, haemolytic uraemia syndrome, and other causes of pulmonary–renal syndrome (see Box 126.96.36.199).
Serological testing for anti-GBM antibodies and ANCA is crucial for confirming the diagnosis, and a renal biopsy is almost always warranted. Some healthy individuals exposed to inhaled oils, hydrocarbons, or solvents may have borderline raised anti-GBM antibody levels, and anti-GBM antibody has also been detected in HIV-negative patients with pneumocystis pneumonia. Other investigations are detailed in Table 188.8.131.52. Alveolar haemorrhage is an important cause of mortality and must be identified early. All patients should have baseline Kco and chest radiology, repeated as necessary.
Table 184.108.40.206 Investigation of patients with anti-GBM disease
Dipstick shows proteinuria and haematuria
Microscopy shows numerous erythrocytes and red cell casts
Full blood count
Haemoglobin often reduced
Monitor white blood cell count during immunosuppression
Urea and electrolytes
Severe acute renal failure common
Moderately elevated (less than in vasculitis)
Normal or mildly raised
Usually positive, usually IgG, but may be only mildly raised, even in typical disease. The occasional patient has no detectable serum antibodies, but deposition in the kidney
Negative in true anti-GBM disease. If ‘double-positive’, may have a poor outcome from renal failure
ANA, dsDNA, complement, cryoglobulins, ASOT
Negative or normal
Usually bilateral airspace shadowing. Difficult to distinguish infection and fluid overload from pulmonary haemorrhage
Raised in alveolar haemorrhage. Normal or reduced in pulmonary oedema or infection
Crescentic glomerulonephritis and linear IgG deposition along the GBM
Not usually diagnostic (difficult to obtain alveoli; open biopsy more useful). Helpful in diagnosis of infection
ANA, antinuclear antibody; ANCA, antineutrophil cytoplasmic antibody; ASOT, antistreptolysin-O titre; dsDNA, double-stranded DNA; GBM, glomerular basement membrane; Kco, diffusing capacity for carbon monoxide.
Untreated anti-GBM disease is usually fatal, and renal function never recovers. In most centres immunosuppressive treatment is given immediately on diagnosis to those with a serum creatinine concentration below 600 µmol/litre at presentation and/or with active pulmonary haemorrhage. Those with a serum creatinine concentration above 600 µmol/litre and without active pulmonary haemorrhage need more careful consideration before being treated since they have a small chance of recovery of renal function (see below).
Treatment with plasma exchange, cyclophosphamide, and corticosteroids, together with dialysis when required, can allow up to 90% of patients to survive, but only around 40% of survivors will recover renal function. Daily plasma exchange removes circulating antibodies, while cyclophosphamide prevents further antibody synthesis. There has only been one controlled trial of plasma exchange, which utilized a low intensity of exchanges in a small number of patients and showed a nonsignificant trend towards an improved outcome. However, the dramatic improvement in overall mortality and renal function coincident with the introduction of a treatment regimen of the type described above has led to its widespread use. A large single centre cohort study from China has also confirmed the benefit of plasma exchange on patient and renal outcomes. Patients treated with steroids and cyclophosphamide alone showed increased survival but without renal recovery. The regimen we use is shown in Box 220.127.116.11. An alternative to plasma exchange is protein-A immunoadsorption, which has not shown any benefit over plasma exchange in the most severely affected patients. Ciclosporin has been used in occasional patients unresponsive to other therapies, but is of doubtful benefit. Case reports have suggested benefit of mycophenolate in place of cyclophosphamide, or the addition of rituximab. Long-term treatment is unnecessary, and patients can stop taking cyclophosphamide after 2 to 3 months, and withdraw prednisolone over approximately 6 months.
The outcome of patients with Goodpasture’s disease in published series is shown in Table 18.104.22.168. Most will now survive the acute illness, but pulmonary haemorrhage and infection remain important causes of death. In those with a serum creatinine concentration below 600 µmol/litre at presentation, the creatinine should begin to fall within 1 to 2 weeks of treatment, and most will recover renal function. However, patients with a creatinine concentration above 600 µmol/litre, or with oligoanuria, less commonly recover renal function. For this reason most centres would not give immunosuppressive agents to this group with the sole intention of trying to restore renal function, although they would for concurrent active pulmonary haemorrhage or if the renal biopsy suggests that tubular necrosis may be contributing to the severity of the renal failure (see above). Crescent scores over 50% and delay in diagnosis are also markers of a poor renal prognosis.
Table 22.214.171.124 Outcome of patients with Goodpasture’s disease. Data from all published series
Number of patients
1-year patient survival (%)
1-year renal survival (%)
Renal recoverya (% treated patients)
Benoit et al. (1964)
No treatment. Patients with Goodpasture’s syndrome of all causes
Wilson and Dixon (1973)
Beirne et al. (1977)
Seven patients not treated. Remainder immunosuppression only
Teague et al. (1978)
Excluded patients without lung haemorrhage
Briggs et al. (1979)
Seven patients not treated. Remainder immunosuppressed
Peters et al. (1982)
Hammersmith hospital single centre. All plasma exchanged
Simpson et al. (1982)
Excluded patients without pulmonary haemorrhage
Johnson et al. (1985)
Randomized prospective study of plasma exchange
Walker et al. (1985)
Australian single centre. All plasma exchanged
Savage et al. (1986)
Data from multiple British centres
Hammersmith hospital single centre
Williams et al. (1988)
Single British unit. Patients presenting over 13 months
Bouget et al. (1990)
French single centre
Herody et al. (1993)
French single centre. Most plasma exchanged
Merkel et al. (1994)
Survival at time of analysis. All plasma exchanged
Daly et al. (1996)
All plasma exchanged
Levy et al. (2001)
Extended Hammersmith series. All plasma exchanged. Only 8% of patients requiring dialysis recovered renal function
Segelmark et al (2003)
Li et al. (2004)
Hong Kong Chinese patients. 80% plasma exchanged
Cui et al. (2005)
Chinese series. Only 45% plasma exchanged
Cui et al. (2011)
Chinese series. Only 43% plasma exchanged
NA, not available.
a Renal recovery if initial creatinine >600 µmol/litre.
The prognosis in anti-GBM disease is in marked contrast to that of patients with a diagnosis of ANCA-associated RPGN. Renal recovery is to be expected in the latter group with immunosuppression, and around 70% of patients presenting with a creatinine concentration above 600 µmol/litre will recover renal function.
Relapses of pulmonary haemorrhage and worsening of renal function can occur early during the course of treatment in the presence of circulating autoantibodies, and can be triggered by smoking, infection, or fluid overload. True late recurrence is very unusual. Transplantation is safe once autoantibodies are no longer detectable, and is best delayed until between 6 and 12 months after the disappearance of anti-GBM antibody.
Anti-GBM disease in Alport’s syndrome
Patients with X-linked Alport’s syndrome have a mutation in the α5 chain of type IV collagen, but also have undetectable Goodpasture antigen in their kidneys despite a normal α3(IV) gene. Transplantation of a normal kidney into such recipients may allow the development of anti-GBM antibodies as a result of the exposure of the immune system to neoantigens to which tolerance has not developed. These antibodies are usually anti-α5(IV)NC1, but can also be anti-α3(IV)NC1 (classic Goodpasture autoantibodies). Most patients do not develop overt nephritis, simply depositing antibody along the GBM without recruiting a glomerular inflammatory response, but a few develop severe glomerulonephritis. In the absence of lung antigen, pulmonary haemorrhage never occurs.
Cui Z, et al. (2004). Characteristics and prognosis of Chinese patients with anti-glomerular basement membrane disease. Nephron Clin Pract, 99, 49–55.Find this resource:
Cui Z, et al. (2011). Anti-glomerular basement membrane disease. Outcomes of different therapeutic regimens in a large single center Chinese cohort study. Medicine, 90, 303–11.Find this resource:
Fisher EG, Lager DJ (2006). Anti-glomerular basement membrane glomerulonephritis. Am J Clin Pathol, 125, 445–50.Find this resource:
Herody M, et al. (1993). Anti-GBM disease: predictive value of clinical, histological and serological data. Clin Nephrol, 40, 249–55.Find this resource:
Hudson BG (2004). The molecular basis of Goodpasture and Alport syndromes: beacons for the discovery of the collagen IV family. J Am Soc Nephrol, 15, 2514–27.Find this resource:
Hudson BG, et al. (2003). Alport’s syndrome, Goodpasture’s syndrome and type IV collagen. N Eng J Med, 348, 2543–56.Find this resource:
Johnson JP, et al. (1985). Therapy of anti-glomerular basement membrane antibody disease: analysis of prognostic significance of clinical, pathological and treatment factors. Medicine, 64, 219–27.Find this resource:
Kitagawa W, et al. (2008). The HLA-DRB1*1501 allele is prevalent among Japanese patients with anti-glomerular basement membrane antibody-mediated disease. Nephrol Dial Transplant, 23, 3126–9.Find this resource:
Lerner RA, Glassock RJ, Dixon FJ (1967). The role of anti-glomerular basement membrane antibodies in the pathogenesis of human glomerulonephritis. J Exp Med, 126, 989–1004.Find this resource:
Levy JB, Pusey CD (2007). Plasmapheresis. In: Johnson R, Floege J, Feehally J (eds) Comprehensive clinical nephrology, 3rd edition, pp. 83.1–8. Mosby, London.Find this resource:
Levy JB, et al. (2001). Long-term outcome of anti-GBM antibody diseases treated with plasma exchange and immunosuppression. Ann Int Med, 134, 1033–42.Find this resource:
Levy JB, et al. (2004). Clinical features and outcome of patients with both ANCA and anti-GBM antibodies. Kidney Int, 66, 1535–40.Find this resource:
Merkel F, et al. (1994). Course and prognosis of anti-basement membrane antibody mediated disease, a report of 35 cases. Nephrol Dial Transplant, 9, 372–6.Find this resource:
Pedchenko V, et al. (2010). Molecular architecture of the Goodpasture autoantigen in anti-GBM nephritis. N Engl J Med, 363, 343–54.Find this resource:
Saus J, et al. (1988). Identification of the Goodpasture antigen as the α3(IV) chain of collagen IV. J Biol Chem, 263, 13374–80.Find this resource:
Savage COS, et al. (1986). Anti-GBM antibody mediated disease in the British Isles 1980–1984. BMJ, 292, 301–4.Find this resource:
Stanton MC, Tange JD (1958). Goodpasture’s syndrome, pulmonary haemorrhage associated with glomerulonephritis. Australas Ann Med, 7, 132–44.Find this resource: