Pathology of Antineutrophil Cytoplasmic Antibody–Associated Pulmonary and Renal Disease Open Access

In 2012 the nomenclature of vasculitis was updated.¹⁰  The previous consensus nomenclature system¹¹  was developed prior to widespread ANCA serologic testing and did not differentiate sufficiently the different forms of small vessel vasculitis. One of the main areas of revision included the description of 2 pathophysiologically different groups of small vessel vasculitis including immune complex vasculitis and AAV (Table 1).¹²  Immune complex vasculitis included cryoglobulinemic, immunoglobulin A, hypocomplementemic urticarial, anti-C1q, and anti-GBM vasculitis. Three distinct forms of AAV were defined, including MPA, GPA, and EGPA. It is important to note that this nomenclature system was not intended as a classification or diagnostic system. As a group, the AAVs were defined broadly as “a necrotizing vasculitis, with few or no immune deposits, predominantly affecting small vessels, and associated with MPO-ANCA or PR3-ANCA.” ¹⁰ 

The presence or absence of immune complex, either as granular deposits or as a linear coating of the basement membranes, is an important distinguisher in the classification system of small vessel vasculitis. This is done on a daily basis in renal pathology laboratories with immunofluorescence (IF) studies on kidney biopsies, looking mostly at the glomerular basement membranes. As such, renal pathologists develop much experience in the interpretation of IF studies that may be challenging. Although similar patterns of immune complex disease can be seen in IF studies performed on fresh-frozen lung tissue,¹³  there are several limitations. First, routine harvest of fresh lung tissue for IF is not standard practice at most laboratories. Second, renal biopsies are far more common than surgical lung biopsies and there is not as much opportunity to develop expertise in the interpretation. Finally, some believe there is more challenging artifact in the interpretation of the IF of the capillary network within the pulmonary interstitium compared with the kidney. For these reasons, if given the option, IF is best performed on renal biopsy samples.

The spectrum of pulmonary pathology seen in AAV is quite broad, ranging from acute injury to pulmonary fibrosis. The acute injury seen in AAV is most often in the setting of DAH.¹⁴  Criteria for DAH include a constellation of findings including acute injury (eg, fibrin, hyaline membranes) and the presence of alveolar hemorrhage (Figure 1, A). Alveolar hemorrhage is diagnosed most easily when there are alveolar siderophages indicative of past bleeding into the alveoli with subsequent macrophage processing of the red blood cells. One should use caution when only fresh red blood cells are present in the alveolar spaces; the vast majority of the time, this is an artifact of the tissue harvest/processing. The exception is when there is corroborating clinical history (hemoptysis) and fibrin intimately associated with the red blood cells. After arriving at a diagnosis of DAH, a careful search for the presence of capillaritis is essential, as this impacts the differential diagnosis (Table 2).²  Capillaritis can be very focal and often missed on large surgical lung biopsies. The presence of alveolar fibrin can be a useful indicator of where to search more carefully for capillaritis. A diagnosis of “acute and organizing alveolar hemorrhage syndrome, with (or without) capillaritis” is often sufficient for the pathologist, as a more specific diagnosis requires clinical history and serologic testing. If capillaritis is identified, the diagnosis should be communicated promptly, as these patients are at risk for massive hemoptysis and mortality.^15,16 

Organizing pneumonia is the common repair mechanism following acute injury. Therefore, all AAV may show organizing pneumonia (Figure 1, B). Usually the organizing pneumonia is seen in the setting of other histologic findings suggestive of an AAV etiology (DAH, capillaritis, necrotizing granulomas); however, especially on smaller transbronchial biopsies, organizing pneumonia may be seen in isolation.

Both GPA and EGPA may present with nodules, either by imaging studies or by surgical lung biopsy. The nodules of GPA and EGPA consist of irregular geographic regions of necrosis surrounded by histiocytes, sparse multinucleated giant cells, mixed inflammation, fibrin, and acute lung injury (Figure 1, C).^9,17  In GPA the necrosis is classically basophilic and contains abundant neutrophils, whereas EGPA shows eosinophilic necrosis with abundant eosinophils.

Acute eosinophilic pneumonia is a frequent pattern of injury seen especially in EGPA.¹⁸  Acute eosinophilic pneumonia shows intra-alveolar fibrin with embedded eosinophils, markedly reactive type II pneumocytes, and mild interstitial edema (Figure 1, D). There is a broad differential diagnosis for acute eosinophilic pneumonia, including adverse drug reaction, infection, changes in smoking habits, collagen vascular disease, EGPA, and idiopathic disease. Inclusion of EGPA in the differential diagnosis of acute eosinophilic pneumonia will help direct ANCA studies, especially in a patient with other clinical symptoms of EGPA (eg, asthma and heart disease).

Finally, the granulomas of GPA and EGPA may heal with variable degrees of fibrosis that may complicate the radiographic picture and histologic specimens. Patients going for surgical lung biopsy may show a spectrum of injury ranging from acute lung injury with siderophages to fibrosis in the same lobe. The presence of fibrosis with associated acute injury helps to confirm recurrent injury over time that is common in AAV.

The renal pathology associated with AAV is far less complex than that seen in the lung. The predominant histologic finding is a focal segmental necrotizing glomerulonephritis (Figure 2, A).^19–21  A variable number of crescents may be present (Figure 2, B). In the majority of cases, the glomeruli without crescents will appear completely normal with open capillary loops, thin delicate glomerular basement membranes, and no significant hypercellularity. In subacute and chronic disease there may be a component of glomerulosclerosis. Occasionally, glomeruli will show endocapillary proliferation of neutrophils reminiscent of the capillaritis seen in the lung. Special stains like Jones silver and periodic acid–Schiff can help to identify rupture of the GBM in glomeruli with segmental necrosis and/or crescents (Figure 2, C). The differential diagnosis includes immune complex–mediated disease, anti-GBM disease, and AAV. Although all AAVs have been associated with this pattern of renal injury, MPA and GPA more commonly involve the kidney compared with EGPA. To further clarify the diagnosis, IF studies are required.²²  Three basic IF patterns are encountered in this setting: positive immunoglobulin and complement staining in a granular distribution suggestive of immune complex–mediated disease (systemic lupus erythematosus, immunoglobulin A, postinfectious); linear immunoglobulin G staining along the GBM (anti-GBM disease); and few or no IF deposits (AAV), which is referred to as pauci-immune glomerulonephritis in the renal literature. Electron microscopy studies will show the crescentic disease and disruptions in the glomerular basement membrane, but electron-dense immune complex deposits are not identified.

Tubulointerstitial inflammation and fibrosis are usually commensurate with the degree of active glomerular lesions and glomerulosclerosis, respectively (Figure 2, B). Periglomerular inflammation can occur and may be vaguely granulomatous. True interstitial granulomas, although uncommonly encountered, are quite suggestive of AAV.²³  True necrotizing vascular lesions are identified in the minority of biopsy samples (Figure 2, D).²⁴ 

There are 2 laboratory methods to test for ANCA, screening indirect IF (IIF) assays and more specific enzyme-linked immunosorbent assays (ELISAs). Recommendations are to first perform IIF and, if it is positive, to then perform ELISA testing. Testing for specific ANCA, in the form of myeloperoxidase (MPO) and proteinase 3 (PR3) in the absence of positive IIF, rarely leads to the identification of patients with active vasculitis.²⁵  In one study of 55 patients with AAV-like symptoms who ended up being diagnosed as not having AAV, 9 had a positive IIF screen with negative ELISA and 1 had a positive ELISA in the setting of a negative IIF.²⁶  Based on meta-analysis data, the sensitivity of combined positive IIF and ELISA testing showed a sensitivity of 82% and a specificity of 99% for AAV. Although it would be very useful to know the positive and negative predictive values for these tests, the positive and negative predictive values depend on the prevalence of disease within the group of patients being tested. In a single institution the positive and negative predictive values for IIF and ELISA testing varied widely by department, depending on how selective the department was in ordering the test.²⁶ 

Screening IIF studies are performed by obtaining an ethanol-fixed buffy coat from a control sample and mixing with patient serum. Three possible patterns of ANCA reactivity are observed under IIF: cytoplasmic ANCA, perinuclear ANCA, and atypical ANCA. In most laboratories, IIF testing is performed manually and has a component of subjectivity. Distinguishing perinuclear ANCA from antinuclear antibodies can be a pitfall in the interpretation. Automated reading systems have been developed but still require improvement prior to widespread implementation.²⁷ 

In the setting of a positive IIF study, more specific ELISA testing for MPO and PR3 is recommended.²⁸  Myeloperoxidase and PR3 are proteases found in the granules of neutrophils and play a role in the respiratory burst. Several different enzyme-based immunoassays have been developed to identify and titer antibodies developed against these antigens.²⁷  There is conflicting literature on how closely antibody titer correlates with disease activity and treatment response.^20,21  Dual reactivity for MPO and PR3 antibodies is suggestive of drug-induced ANCA-mediated disease.²⁹ 

Table 3 compares and contrasts the 3 major AAV diseases: MPA, GPA, and EGPA.

The 2012 revised vasculitis nomenclature defines MPA as a necrotizing vasculitis with few or no immune complex deposits predominantly affecting small vessels with frequent necrotizing glomerulonephritis and pulmonary capillaritis. Granulomatous inflammation is absent.¹⁰ 

Microscopic polyangiitis affects approximately 10 000 people in the United States, with about 900 new cases diagnosed each year.³⁰  It occurs in both sexes equally, and although it can occur at any age, it is more commonly a disease of the elderly.^31,32  Microscopic polyangiitis may involve the kidney, skin, lung, gastrointestinal tract, neurologic system, and head and neck sites.³¹  The clinical presentation includes generalized symptoms of fever, weight loss, weakness, arthralgia, and myalgia as well as site-specific symptoms like hematuria, hemoptysis, and rash.

Patients with lung involvement typically show bilateral infiltrates on chest x-ray and bilateral ground-glass opacification by high-resolution computed tomography.³³  As with most vascular diseases, there is often a predilection for the lower lobes. Bronchoalveolar lavage may show red blood cells or hemosiderin-laden macrophages.

The urinalysis in patients with renal involvement is positive for hematuria and variable proteinuria. Examination of the urine sediment reveals red blood cell casts and variable white blood cells. Creatinine and blood urea nitrogen are elevated compared with baseline. Eighty percent of patients are positive for perinuclear ANCA with an MPO specificity and 19% of patients are positive for cytoplasmic ANCA with PR3 specificity.³⁴  It is very rare for patients to have negative ANCA studies. Other autoantibodies that may be identified include RF and ANA. Acute-phase reactants such as erythrocyte sedimentation rate and C-reactive protein are often elevated.

Diffuse alveolar hemorrhage with capillaritis is the most common pulmonary manifestation of MPA (Figure 1, A).¹⁴  There may be a variable degree of organizing pneumonia and the presence of hyaline membranes of diffuse alveolar damage. The pathologic approach to patients with DAH has been nicely reviewed.³  The differential diagnosis for this injury pattern is shown in Table 2 and depends on the presence or absence of capillaritis. Final diagnosis usually requires clinical and serologic correlation.

Kidney biopsies early in the disease show focal and segmental necrotizing glomerular lesions with crescents (Figure 2, A through C). On the Jones silver stain, disruptions in the glomerular basement membrane can be identified (Figure 2, C). Endocapillary proliferation with neutrophils and macrophages may be seen but mesangial hypercellularity is quite rare. There is a variable degree of glomerulosclerosis and fibrous/fibrocellular crescents depending on the degree of chronicity in any given case. With severe glomerular disease and crescents, rupture of Bowman capsule may occur, generating a vaguely granulomatous inflammatory cell infiltrate. True tubulointerstitial granulomas are not seen in MPA. The tubulointerstitial inflammation generally correlates with the degree of glomerular involvement. Smaller arteries may show necrotizing vasculitis, often with relatively little associated inflammation (Figure 2, D).²⁴  Immunofluorescence and electron microscopy studies typically show few to no immune complex deposits. This lesion is referred to as pauci-immune glomerulonephritis and it is indistinguishable from other forms of AAV that involve the kidney. A prognostic classification schema has been introduced that classifies cases into 4 different classes based on the prevalence of crescents and glomerulosclerosis: sclerotic, focal, crescentic, and mixed.¹⁹ 

The prognoses of MPA and GPA are similar. Untreated patients have a mortality of 90% at 2 years whereas treated patients show 34% mortality at 1 year.¹⁶  Patients are more likely to die from infectious complications of treatment than they are of the disease.¹⁶  There is a correlation with the degree of renal involvement and need for hemodialysis with renal failure and overall mortality.^15,19 

The 2012 revised vasculitis nomenclature defines GPA as necrotizing granulomatous inflammation usually involving the upper and lower respiratory tract with necrotizing vasculitis involving small- and medium-sized vessels and frequent necrotizing glomerulonephritis.¹⁰  Granulomatosis with polyangiitis was originally described by Wegener in 1938 and was referred to as Wegener granulomatosis until multi-organization recommendations for a name change were issued in 2011.

The prevalence of GPA is 30 000 patients in the United States, with an estimated 2600 new cases diagnosed each year.³⁰  It may occur at any age but is generally a disease of the elderly and affects both sexes equally.^31,32  Granulomatosis with polyangiitis may involve a variety of organs including the head and neck, lung, and renal, and musculoskeletal system.^9,31  Less frequent sites of involvement include the nervous, gastrointestinal, and cutaneous systems. Similar to MPA, patients often show systemic symptoms of fatigue, fever, arthralgia, and myalgia. Other symptoms are related specifically to the organs involved (dark urine with renal involvement and shortness of breath with lung involvement). Surprisingly, despite quite impressive radiologic changes, many patients do not usually appear critically ill.

Imaging studies in patients with lung involvement may show a variety of findings including bilateral nodules, infiltrates, ground-glass opacities, and pleural opacities.³³  Cavitary lesions may be encountered. The degree of radiologic abnormalities is often out of proportion to the severity of clinical presentation. If the patient presents with hemoptysis, bronchoalveolar lavage specimens may be useful in the confirmation of alveolar hemorrhage by the presence of abundant siderophages.

Approximately 77% of patients will present with renal disease.³¹  This is most often in the form of a rapidly progressive glomerulonephritis. Patients have hematuria and variable proteinuria on urinalysis. Creatinine is elevated and glomerular filtration rate is decreased. Complement levels (C3 and C4) are usually normal. Ninety percent of patients have a positive PR3-ANCA whereas 7% have a positive MPO-ANCA and 3% are negative for ANCA studies.³⁴ 

The pulmonary pathology of GPA has been well reviewed.⁹  Classic pulmonary lesions of GPA include large geographic necrotizing granulomas that show a basophilic appearance (Figure 1, C). At higher power, the basophilic necrosis appears to consist mostly of necrotic neutrophils and other debris. At the interface between the necrosis and the viable tissue are palisading histiocytes with rare scattered multinucleated giant cells (Figure 3, A).²  The multinucleated giant cells of GPA are distinctive with angulated edges and hyperchromatic syncytial-like nuclei that are often arranged in a ring or horseshoe at the periphery of the giant cell. In some cases the granulomas have a predilection for the centrilobular region. Surrounding the granulomas are usually variable degrees of organizing pneumonia and fibrin. Scattered microabscesses may be seen as well (Figure 3, B).⁹ 

The major pathologic differential diagnosis for these necrotizing lesions is infection. Necrotizing granulomas of infection have more eosinophilic central necrosis with a robust surrounding lymphoplasmacytic infiltrate (pink center with a blue rim, opposite the blue center and pink rim characteristic of GPA). In addition, they show more numerous giant cells with a more epithelioid appearance. Special stains for infectious organisms are required in these cases and it is necessary to follow culture studies. The identification of granulomatous inflammation in the arteries is not helpful as this is often seen in infection as well. Correlation with the clinical history can be helpful too, as patients with bilateral necrotizing lesions of infection are usually quite ill, whereas patients with GPA may have only minor symptoms.

In some cases there are areas of frank vasculitis of medium-sized vessels characterized by transmural necrotizing inflammation with variable giant cells (Figure 3, C). The inflammatory disruption of the internal and external elastic lamina can be confirmed with elastic tissue stains (Figure 3, D). Other cases may show DAH with capillaritis histologically indistinguishable from MPA (Figure 1, A). Numerous other so-called minor manifestations of GPA have been described, including interstitial fibrosis, lipoid pneumonia, chronic bronchiolitis, follicular bronchiolitis, and bronchocentric granulomatosis.³⁵  Surgical lung biopsies may not show all the required histologic features, and correlation with clinical and serologic data is often needed to arrive at a definitive diagnosis of GPA.⁸  In the correct clinical setting, the diagnosis of GPA can be made on transbronchial biopsy.⁷ 

The renal histologic manifestations of GPA can be nearly indistinguishable from what is seen in MPA (see above) and distinction often requires other clinical and serologic data (Figure 2). Rarely, cases of GPA may show granulomatous tubulointerstitial inflammation and/or granulomatous arteritis, both features suggestive of GPA or EGPA over MPA.

The prognosis of GPA is similar to that seen in MPA (see above).

The 2012 revised vasculitis nomenclature defines EGPA as eosinophil-rich granulomatous inflammation often involving the respiratory tract, associated with asthma and peripheral eosinophilia, and showing necrotizing vasculitis affecting small- and medium-sized vessels ¹⁰ . The presence of associated glomerulonephritis makes ANCA more frequently identified. This disease was previously known as Churg-Strauss syndrome.

Eosinophilic GPA is an exceedingly rare disease, affecting only 3 persons per million.³⁶  Patients with EGPA present with a fairly characteristic constellation of symptoms including asthma, peripheral eosinophilia, neuropathy, and cardiac disease.³⁷  Asthma may precede other symptoms for years. Eosinophilic GPA is a multisystem disorder that may involve the neurologic system, skin, gastrointestinal tract, head and neck, and musculoskeletal system, in addition to the lung and kidney. Patients often suffer from fever, cough, and shortness of breath. According to the traditional format criteria suggested by the American College of Rheumatology, the presence of 4 or more of the following is 85% sensitive and 99% specific for the diagnosis of EGPA: asthma, greater than 10% peripheral eosinophilia, neuropathy, nonfixed pulmonary infiltrates, and a biopsy showing extravascular eosinophils.³⁸  Note that the pathology needed to support a diagnosis of EGPA is relatively nonspecific and can be demonstrated in any organ.

Bilateral pulmonary infiltrates that vary in both size and location over time (so-called nonfixed pulmonary infiltrates) are characteristic of EGPA.³⁹  Bronchoalveolar lavage specimens may show a marked increase in the number of eosinophils on the differential count. If there is renal involvement, the urinalysis may show hematuria and proteinuria, and there may be variable elevation of creatinine and decrease in GFR (glomerular filtration rate). Approximately two-thirds of patients will have a positive MPO-ANCA. The frequency of a positive PR3-ANCA and ANCA negativity are variable depending on the study, but ANCA negativity is likely rare.³⁴  Serum levels of immunoglobulin E are often elevated.³⁷ 

Although there have been a variety of pathologic findings described in the setting of EGPA, in multiple organs, the pathologist must remember that their role in supporting a diagnosis lies only in the identification of extravascular eosinophils.³⁸  Lung biopsies may show a variety of findings, including large geographic necrotizing granulomas with eosinophilic necrosis (Figure 4, A), eosinophilic vasculitis, eosinophilic capillaritis in the setting of DAH (Figure 4, B),⁴⁰  and acute eosinophilic pneumonia.³⁷  The eosinophilic pneumonia seen in EGPA is not distinguishable from the numerous other etiologies and is characterized by airspace fibrin, embedded eosinophils, and reactive type II pneumocytes (Figure 1, D).

Although patients with EGPA typically have less impressive renal impairment when compared with GPA and MPA, if there is renal involvement it is histologically identical to the focal segmental necrotizing glomerulonephritis seen in MPA (see above) but perhaps with increased eosinophils.⁴¹ 

Cardiac involvement in EGPA is the leading cause of morbidity and mortality.^37,39  The 5-year survival is between 62% and 79%, with heart disease and gastrointestinal tract involvement leading the causes of mortality.^39,42 

Most treatment trials have enrolled only MPA and GPA patients and have excluded EGPA. The optimal treatment of MPA and GPA is based on several randomized controlled trials and involves both induction and maintenance phases.^43–45  At the acute presentation, treatment consists of cyclophosphamide and high-dose steroids. Rituximab has been used for induction in both newly diagnosed patients and patients experiencing relapse.^46,47  It shows similar response rates to cyclophosphamide. Plasma exchange has been shown to improve renal survival but not overall mortality.^45,48  Maintenance phases consist of low-dose steroids and methotrexate/azathioprine.^49,50 

The treatment of EGPA has not been as thoroughly studied as that of MPA and GPA, but strategies are typically similar.

The marked variation in the histology of AAV, ranging from the neutrophilic capillaritis and segmental glomerular necrosis of MPA and GPA to the larger necrotizing lesions of GPA and EGPA, has led many to propose different pathophysiologic mechanisms.

The characterization of the injury at the capillary level has been fairly well worked out. In this model, neutrophils are primed by an event related to bacterial wall exposure, tumor necrosis factor, or complement factor C5a.⁵¹  The priming of the neutrophils involves moving MPO and PR3 antigens close to the surface of the neutrophil, along with the surface expression of adhesion molecules. The ANCA antibodies in the serum bind to the MPO or PR3 antigens and activate the respiratory burst of the neutrophil, including degranulation.⁵¹  This leads to increased chemokine activity and resulting white blood cell adhesion, edema, endothelial damage, and microvascular injury. This is the lesion of necrotizing glomerulonephritis and DAH with capillaritis.

There is more controversy surrounding the pathogenesis of the necrotizing granulomatous lesions. Some believe the large geographic granulomas are simply an extension of the above process, but occurring within tissue as opposed to the vascular space.²¹  Neutrophils are again primed within the interstitium, and the respiratory burst is activated by ANCA; the result is microabscess formation. Indeed, PR3 ANCAs have been demonstrated to cause direct tissue pathology in vivo.⁵  Then an exaggerated granulomatous response occurs to protect the surrounding tissues from the microabscess in an antigen-independent fashion.²¹  Others propose a type IV–mediated inflammatory pathway that is directly antigen dependent and involves complicated dysregulation of the immune system.⁵² 

The AAV diseases enter into the differential diagnosis for a variety of different lung pathologic presentations of disease ranging from acute lung injury, alveolar hemorrhage, granulomas, nodules, and vasculitis. Conversely, focal segmental necrotizing glomerular lesions with or without crescents are the dominant renal pathology manifestations of AAV. Recognition of the acute presentations including acute alveolar hemorrhage with capillaritis and necrotizing glomerulonephritis with crescents and prompt communication with the clinical services are important in rapid initiation of an appropriate serologic workup and/or treatment. In both the lung and the kidney, there are no specific pathology features, and definitive classification often requires correlation with the clinical presentation, radiology studies, and serologic studies, including ANCA studies.