Chapter 29. Goodpasture's Syndrome/Anti-Glomerular Basement Membrane Disease

• Rapidly deteriorating renal function, with or without hemoptysis and pulmonary shadowing on chest radiograph.
• Hematuria and proteinuria on urine dipstick and erythrocyte casts and/or dysmorphic erythrocytes on urine microscopy.
• Circulating antibodies directed against the glomerular basement membrane (GBM).
• Renal biopsy showing focal necrotizing glomerulonephritis with linear deposition of immunoglobulin.

Anti-glomerular basement membrane disease is a rare but well-characterized cause of glomerulonephritis, with an incidence of 1 case per million per year in white populations. All ages can be affected, but the peak incidence is in the third decade in young men and in the sixth and seventh decades in either sex. Lung hemorrhage occurs more often in younger patients, and isolated glomerulonephritis is more common in older patients. The disease is defined by the presence of pathogenic anti-GBM antibodies directed against the NC1 domain of the α3 chain of type IV collagen [α3(IV)NC1], a component of selected basement membranes including those of glomeruli and pulmonary alveoli. These cause focal necrotizing glomerulonephritis and result in widespread crescent formation on renal biopsy; clinically, the result is rapidly progressive glomerulonephritis. When this occurs in association with pulmonary hemorrhage, the condition is known as Goodpasture's syndrome.

The disease occurs in genetically susceptible individuals exposed to an environmental trigger. The HLA type strongly influences susceptibility; HLA types DR15 and DR4 predispose to disease, while HLA DR7 and DR1 are protective.

Environmental factors are involved in both triggering the disease and influencing its clinical presentation. Several reports describe clusters of cases, suggesting that an infective or other exogenous agent is involved in the pathogenesis, but no specific cause has been identified. Cigarette smoking has an important influence on the extent of lung injury, with pulmonary hemorrhage affecting almost all current smokers and being almost exclusively confined to this group. Lung hemorrhage following hydrocarbon exposure has also been described. In addition, genetic factors affect disease susceptibility.

Symptoms and Signs

Patients often have a history of malaise, arthralgia, and weight loss, but these features are generally far milder than in other causes of focal necrotizing glomerulonephritis, such as antineutrophilic cytoplasmic antibody (ANCA)-associated systemic vasculitis. Anemia is common, and may be symptomatic even in patients with minimal hemoptysis. The principal clinical symptoms relate either to pulmonary hemorrhage or to the development of renal failure. The severity of pulmonary hemorrhage varies and can range from minor hemoptysis to life-threatening hemorrhage with respiratory failure. Typically, hemoptysis is intermittent at the outset and can occur spontaneously or can be precipitated by intercurrent infections or fluid overload. There is a poor correlation between the severity of hemoptysis and the quantity of pulmonary blood loss, with other clinical signs being variable. These include inspiratory crackles and bronchial breathing, with patients often tachypnoeic and cyanosed. Historically hemoptysis has been the most common presenting feature, but its incidence is decreasing with the reduced prevalence of cigarette smoking. It now occurs in about 50% of cases. Even in those with life-threatening pulmonary hemorrhage, the lungs of patients who survive the acute illness recover completely and have no residual pulmonary symptoms of radiologic abnormalities. Histologically they are left with little residual lung pathology or fibrosis.

Renal disease can occur either alone or in association with pulmonary hemorrhage. Mild renal impairment may improve spontaneously, but once significant renal injury has occurred, improvement is rare and deterioration may be extremely rapid. Evolution from normal renal function to established renal failure is less than 12 hours in some patients. Microscopic hematuria is an early sign of renal pathology, and dysmorphic red cells and red cell casts are seen in the urine as the disease progresses. Proteinuria is present, but is usually modest (<3 g/L). Some patients present with macroscopic hematuria and severe loin pain, which are often features of very severe disease. Oliguria is a poor prognostic sign.

Without treatment anti-GBM disease almost always progresses to renal failure, but anti-GBM antibody synthesis is transient, often lasting for less than 2 months. Patients are treated with cytotoxic drugs and plasma exchange. Recurrence of anti-GBM disease is rare, but has been reported, sometimes precipitated by infection or exposure to a toxic agent. The outcome in these cases is generally better than with the original presentation, since the diagnosis is often made more quickly. The disease almost invariably recurs in renal allografts performed when circulating anti-GBM antibodies are still present, but transplantation is safe once they are no longer detectable. Exceptionally, however, renal transplantation can reinitiate anti-GBM antibody production, but this probably occurs in less than 1% of cases. Accordingly, transplantation is normally deferred for at least 6 months after the disappearance of circulating anti-GBM antibodies, and transplanted patients should be monitored closely for changes in urinary sediment or antibody titers.

Laboratory Findings

The diagnosis of Goodpasture's disease is dependent on the detection of anti-GBM antibodies either in the circulation or at renal biopsy. These antibodies are usually detected using an enzyme-linked immunosorbent assay (ELISA), and equivocal results can be confirmed by Western blotting. There is a clear correlation between the titer of anti-GBM antibodies and the severity of renal injury, but the pattern of lung disease is independent of antibody titer.

Other laboratory findings are nonspecific. Anemia is common, and is often microcytic and hypochromic. As in other types of rapidly progressive glomerulonephritis, there may be evidence of microangiopathy on the blood film, but this is mild. Abnormalities of the urinary sediment are an early sign of renal pathology, and are often followed by increases in serum urea nitrogen and creatinine as renal failure develops.

Imaging Studies

Chest radiographs can be normal or show alveolar shadowing in those with pulmonary hemorrhage. Typically this spares the bases and upper lung fields, but distinguishing pulmonary hemorrhage from infection or edema can be impossible. However, hemorrhage itself is not associated with curly B-lines and is not limited by fissures. Typically the radiologic changes appear and resolve more quickly than is the case for infection (Figure 29–1).

There are no specific morphologic abnormalities on any type of renal imaging and the appearance cannot be distinguished from other types of acute renal failure.

Special Tests

A renal biopsy is essential in the diagnosis and management of suspected anti-GBM disease (Figure 29–2). The early histologic changes include very minor mesangial expansion and hypercellularity and are followed by the development of a focal and segmental glomerulonephritis, often with marked neutrophil infiltration. This progresses to a focal necrotizing glomerulonephritis with rupture of glomerular capillaries and leakage of blood into Bowman's space. This initiates the formation of cellular crescents, composed of parietal epithelial cells and macrophages. There is some experimental evidence that crescents may also be comprised of visceral epithelial cells. Are we certain that is not the case in Goodpasture's disease? Characteristically all the crescents are at the same stage of evolution, a feature that reflects the explosive nature of the disease and distinguishes it from other forms of focal necrotizing and crescentic glomerulonephritis. Linear binding of antibody to GBM is found in all patients, regardless of the severity of the renal pathology. The antibody is almost always immunoglobulin (Ig) G, but in one-third of cases IgA or IgM is also detected; exceptionally, reports describe deposition of IgA or IgM alone. Linear C3 is seen in 60–70% of patients, and its presence does not influence the severity of the renal injury.

Goodpasture's disease must be distinguished from other causes of combined acute renal and respiratory failure, the most important of which are listed in Table 29–1. The detection of anti-GBM antibodies makes Goodpasture's disease very likely, but it is increasingly recognized that these antibodies are also found in a small proportion of patients with ANCA-associated focal necrotizing glomerulonephritis. The ANCA specificity in most of these patients is for myeloperoxidase, and their titer of anti-GBM antibodies is usually lower and more easily suppressed than in those patients with anti-GBM antibodies alone. Renal function in this group of patients may be recoverable even if they have severe renal failure at presentation.

The prognosis for Goodpasture's disease was very poor before the development of modern treatment regimens, with almost all patients dying of renal failure or pulmonary hemorrhage. The recognition that anti-GBM antibodies were pathogenic gave rise to the current treatment strategy, which is comprised of plasma exchange combined with immunosuppressive drugs. As more effective treatments were developed, it became apparent that the severity of the disease at presentation was a major determinant of the eventual outcome, especially with regard to recovery of renal function. Patients with microscopic hematuria and normal serum creatinine values often underwent spontaneous remission but remained at risk of developing devastating renal injury as long as anti-GBM antibodies remained in the circulation. In contrast, patients presenting with oligoanuric or dialysis-dependent renal failure almost never regain renal function.

Most centers now use protocols combining cyclophosphamide, prednisolone, and plasma exchange. A typical regimen is given in Table 29–2. However, the rarity of the condition makes it difficult to perform properly empowered randomized controlled clinical trials. The only controlled trial of treatment in Goodpasture's disease demonstrated a better outcome with plasma exchange than with immunosuppression and steroids alone, but used a less intensive plasma exchange regimen and lower doses of cyclosphosphamide than those in common use. A study done at Walter Reed failed to demonstrate a significantly better outcome in the plasma exchange group. Nevertheless, the uniformly better results in series of patients treated with plasma exchange combined with alkylating agents such as cyclophosphamide provide strong evidence for their effectiveness.

Pulmonary hemorrhage is usually controlled within 24–48 hours, and moderately impaired renal function can be recovered in most patients (80% of patients with a creatinine <600 μmol/L show response to therapy). Advanced renal failure, in contrast, cannot usually be reversed by any current treatment, despite control of anti-GBM antibody levels. Consequently, the risk of treating patients with a serum creatinine over 600 μmol/L and no history of pulmonary hemorrhage probably outweighs the benefits and such patients are usually left untreated.

There are, however, a number of anecdotal reports of patients who recovered renal function despite having dialysis-dependent renal failure at presentation. These patients tend to have a short history of rapidly declining renal function and a renal biopsy showing either unexpectedly mild or very recent changes with fewer crescents than anticipated. Aggressive treatment should therefore be considered despite advanced renal failure in patients such as these, and a renal biopsy plays a key role in the evaluation. The need for early renal transplantation in cases in which regular dialysis would be particularly difficult provides another reason for aggressive treatment despite renal failure; plasma exchange regimens shorten anti-GBM antibody synthesis by up to a year.

Patients being treated by immunosuppressive drugs and plasma exchange need very careful monitoring, especially in the presence of severe renal failure. The DlCO, a measure of the diffusing capacity of the lung, is markedly elevated in pulmonary hemorrhage, and can be used in diagnosis. Minor reductions in DlCO may occur long after resolution of the disease. Albumin solutions used as replacement fluid for plasma exchange have a high sodium concentration. Consequently plasma exchange is associated with sodium loading and can result in fluid accumulation and pulmonary edema. Accordingly, patients must be monitored carefully for fluid overload. Use of immunosuppressive drugs should also be monitored equally assiduously. Careful monitoring of the complete blood count is essential, and cyclophosphamide should be discontinued if the white cell count falls below 3.5×109/L. Progress can also be assessed by measuring serum creatinine, hemoglobin, and anti-GBM antibody titer, as well as by serial chest radiographs and DlCO estimation. Recent studies have documented 1-year patient survival rates of 79–93%.

Alport's syndrome is an inherited glomerular disease caused by mutations of the genes that encode the α3, α4, and α5 chains of collagen, which together form the collagen network in glomerular capillaries. Consequently the GBM lacks these collagen chains. The disease usually progresses to end-stage renal failure and after renal transplantation, patients with Alport's syndrome are at risk of developing antibodies directed against the normal α chains in the renal allograft. The transient appearance of low titers of anti-GBM antibodies is quite common, and a small number of patients (3–5%) develop a focal necrotizing glomerulonephritis with crescents in the renal allograft. The clinical cause and morphologic appearances are indistinguishable from anti-GBM disease in native kidneys. Treatment is similarly identical, but the outcome is poor. Recurrence in subsequent grafts is the norm and usually occurs within days. Importantly, the anti-GBM antibodies are directed against the mutated α chain, most commonly the α5 chain. Consequently most transplanted patients with Alport's syndrome have anti-α5 antibodies rather than anti-α3 antibodies, and these are detected poorly in standard anti-GBM ELISA assays. Loss of renal allografts due to de novo Goodpasture's syndrome in patients with Alport's syndrome is nevertheless uncommon. Recurrence of Goodpasture's syndrome is uncommon if the allograft is placed after anti-GBM titers are no longer present.