Clinical manifestations

Presentation and course

In this article, we focus on antineutrophil cytoplasmic antibody-associated vasculitis, predominantly affecting small vessels. Vasculitis means inflammation of blood vessel walls. All intraparenchymal vessels are small vessels; therefore, ANCA-associated vasculitis affects organs (43). When first manifesting, ANCA-associated vasculitis can affect only one or multiple organs. The presentation varies widely, from a limited disease to a life-threatening generalized systemic vasculitis. It is important to differentiate between acute disease activity, organ damage due to the disease, and complications from treatment, as all impact management.

The disease can persist for many years in a limited form, and spontaneous remissions are possible. Transition to the generalized phase may occur at any time. Frequently affected tissues include the upper and lower respiratory tract, kidneys, skin, and peripheral nerves. Acute exacerbations may be accompanied by nonspecific inflammatory symptoms, such as arthralgia, fever, and myalgia. Symptoms depend on which vessels are involved: involvement of dermal venules results in purpura; inflammation of the capillaries results in pulmonary hemorrhage or glomerulonephritis; affected vessels in the epineurium result in peripheral neuropathy; and small vessels in the eye and orbit result in ocular inflammation (44). Some ANCA vasculitides (GPA and EGPA) have extravascular granulomatous inflammation, which seems separate from vasculitic inflammation, mostly affecting the respiratory tract (44).

Systemic ANCA-associated vasculitides are further categorized depending on the clinical presentation and pathology (44):

Further classification can be made based on ANCA type. The two best-described subtypes are myeloperoxidase and proteinase 3. Because patients with different subtypes differ in outcomes and HLA association, it is useful to account for the specific subtype in the name, such as MPO-ANCA MPA or PR3-ANCA GPA (44). Although this helps to classify a disease once the diagnosis of vasculitis has been made, there are currently no diagnostic criteria for ANCA-associated vasculitis (49). The “Diagnostic and Classification Criteria in Vasculitis Study” (DCVAS) is a study conducted by the American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) to be completed in December 2015. The study aims at establishing diagnostic criteria and to further refine the classification of diseases.

Granulomatosis with polyangiitis and microscopic polyangiitis. There is substantial overlap in many of the clinical features of the ANCA-associated vasculitides, mostly between granulomatosis with polyangiitis and microscopic polyangiitis. Differentiation based on clinical features is difficult but can be confirmed by pathology (eg, absent granulomatous inflammation suggests microscopic polyangiitis) and supported by ANCA types: microscopic polyangiitis typically directed against myeloperoxidase and in granulomatosis with polyangiitis against proteinase 3. Treatment recommendations do not differ between the two diseases at this time.

The upper airways are affected in 85% to 90% of patients with granulomatosis with polyangiitis, with sinusitis, serous otitis media, sensorineural hearing loss, and subglottic stenosis (97). Lung disease occurs in more than 80% and may involve both the airways and parenchyma. Airways are typically affected by stenoses and ulcerations, whereas parenchymal disease manifests with nodules, cavitary lesions, consolidations, and alveolar hemorrhage. Renal disease is recognized by the presence of microscopic hematuria, proteinuria, and rising serum creatinine, with pauci-immune necrotizing glomerulonephritis seen on biopsy. A review of 735 patients enrolled in several treatment studies and followed for extent of organ damage at 6 months revealed that more patients with microscopic polyangiitis (with MPO ANCA) had proteinuria (MPA:GPA: 50.7% vs. 31.6%) and GFR<50 mL/min (46.5% vs. 22.8%), while otolaryngologic damage was more frequent in patients with granulomatosis with polyangiitis (along with PR3 ANCA), such as hearing loss (MPA:GPA: 4.2% vs. 22.5%) and nasal crusting or blockage (3.7% vs. 29.1%). These differences largely remained stable at long-term follow up at 7 years, with progression to end-stage renal disease more common in microscopic polyangiitis (20% vs. 9%) (77).

The musculoskeletal system is involved in approximately 60% of patients. Arthralgias and myalgia are often seen (83). Orbital masses (pseudotumors, more common in GPA) can cause proptosis, diplopia, or vision loss. Other ocular manifestations include scleritis, keratitis, conjunctivitis, episcleritis, and uveitis. Nasolacrimal duct obstruction is most typical of granulomatosis with polyangiitis (83).

Granulomatosis with polyangiitis or microscopic polyangiitis can be limited to one organ before evidence of systemic vasculitis. However, it is recommended to include them in the ANCA-associated vasculitis category if the patients are ANCA-positive and have typical clinical manifestations and granulomatous disease on pathology (then GPA), eg, of the respiratory tract.

Neurologic manifestations. Both the peripheral and central nervous systems may be affected in granulomatosis with polyangiitis (40). Between 10% and 45% of patients will develop peripheral nerve disease, most commonly either a sensorimotor polyneuropathy or mononeuritis multiplex due to vasculitis and ischemia of the vasa nervorum. In a study of 128 consecutive patients, 56 developed a polyneuropathy, 31 had a distal symmetric polyneuropathy, and 25 had a mononeuritis multiplex (17). The majority of patients develop neuropathy within the first 2 years of disease, with males and the elderly being more frequently affected. Although 35% of neuropathies will resolve within the first 6 months, 15% continue to be symptomatic over 5 years (93).

The central nervous system is affected in about 10% of patients, typically wit cerebral vasculitis and pachymeningitis. Other potential manifestations include cranial neuropathies (cranial nerve I, II, VI, and VII usually), cerebrovascular accidents, intracerebral hemorrhage, cerebritis, Horner syndrome, myelopathy, and cerebral vein thrombosis (67; 84). Diabetes insipidus has been reported as a result of granulomatous involvement of the posterior pituitary.

In a large French cohort group, peripheral neuropathy was observed at 6 months (MPA:GPA 7.9% versus 15%). After 7 years, more patients with microscopic polyangiitis had developed neuropathy as well: 14.1% versus 22.2% (77).

Eosinophilic granulomatosis with polyangiitis. EGPA is a disease characterized by nongranulomatous eosinophilic inflammation, necrotizing granulomatous inflammation often involving the respiratory tract, and necrotizing vasculitis; it is associated with asthma and blood eosinophilia. Typically, the disease develops in 3 stages, starting out in the second or third decade with prodromes of asthma (present in 96% to 100% of patients), sinusitis, nasal polyps, rhinitis, otitis media, or sensorineural hearing loss (27; 29). The disease progresses to the eosinophilic phase with blood eosinophilia and lung involvement, manifesting as pulmonary infiltrates, heart and gastrointestinal involvement (mostly involving small bowel), otherwise unexplained abdominal pain, bleeding, and, on occasion, intestinal perforation. The disease then can progress to a third, vasculitic, phase (29).

Three studies showed that ANCA-positive patients differed from ANCA-negative patients in their clinical manifestations. Compared to ANCA- negative patients, ANCA-positive patients more frequently had a vasculitic pattern with peripheral neuropathy, renal involvement, or purpura (likely ANCA-mediated) but less frequently had cardiomyopathy (78; 85; 11). ANCA antibodies are positive in approximately 40% of these patients (78; 85; 99; 11; 27; 29); they are mostly characterized by a perinuclear pattern (pANCA) at their immunofluorescence assay and are mostly directed against the neutrophil myeloperoxidase. Common laboratory findings in EGPA patients beyond ANCA are marked hypereosinophilia (at least > 1500 eosinophils/µL or > 10% of the total white blood cells) as well as an increase in nonspecific inflammatory markers (ESR, CRP) (27). In 116 patients with EGPA, 26.7% showed signs of renal dysfunction, and 13.8% had rapidly progressive renal insufficiency (85). ANCA was more common in patients with nephropathy than without (75.0% vs. 25.7%), and all patients with necrotizing crescentic glomerulonephritis were ANCA-positive. Skin involvement encompasses purpura (most common), pseudo-urticarial rash, hives, subcutaneous nodules, livedo reticularis, and gangrenous necrotic lesions (11).

A retrospective study of 383 patients with EGPA in the French Vasculitis Study Group cohort revealed that at diagnosis 91.1% of the patients had asthma, and the remainder of patients developed it within 6 months after diagnosis. The most common EGPA manifestations at diagnosis included weight loss (49.3%), mononeuritis multiplex (46.0%), peripheral neuropathy (51.4%), nonerosive sinusitis or polyposis (41.8%), skin lesions (39.7%), and lung infiltrates (38.6%). Furthermore, patients presented with cardiovascular involvement (27.4%), such as pericarditis (15.1%), Raynaud phenomenon, or deep venous thrombosis, and 16.4% presented with cardiomyopathy, gastrointestinal tract involvement (23.2%), and kidney involvement (21.7%). Less common manifestations were alveolar hemorrhage (4.2%), cranial neuropathies (3.1%), or CNS manifestations (5.2%), including ischemic stroke (2.1%) or hemorrhagic stroke (0.5%). Of the patients tested, 31% tested were ANCA positive, most (66%) with a perinuclear pattern but some with cytoplasmic or mixed pattern; 62% had myeloperoxidase specificity, 3.7% had proteinase 3 specificity on ELISA, and few had a mixed pattern. Interestingly, none of the patients who were initially ANCA negative subsequently tested positive (11).

Neurologic manifestations. The central nervous system can be involved, with ischemic stroke or hemorrhagic stroke (66) or both (01), and this can be the initial clinical presentation. One case has been published with ischemic stroke post tPA administration with hemorrhagic transformation. Concern was raised about safety of tPA in this population given the presumed increased risk of bleeding (21). Wolf and colleagues followed newly diagnosed patients with EGPA and monitored neurologic involvement at diagnosis and follow up prospectively: 12 of 14 patients had neurologic involvement as acute or subacute mononeuropathy multiplex (most common), axonal polyneuropathy, cranial neuropathy (II and VII), ischemic infarcts in basal ganglia, and myositis. Commonly affected nerves were the sural nerve (n = 11), common peroneal nerve (n = 8), tibial nerve (n = 8), and median nerve (n = 8). Nerve biopsy results were not reported, except one combined muscle and nerve biopsy that showed interstitial myositis. The onset of both types of neuropathies (mononeuritis multiplex and axonal polyneuropathy) was rather rapid--within a few days or weeks. New symptoms were rare with appropriate treatment. There was no correlation with inflammatory markers. CSF was normal in most cases (6/8). It was observed that 6 of 8 patients with a multiplex mononeuropathy and 2 of 3 patients with an axonal polyneuropathy initially suffered from severe neuropathic pain that improved with immunosuppressive therapy (105).

Prognosis and complications

Studying complications and prognosis has been challenging. Based on the most recent new definitions of diseases, the disease may have not been defined the same way in previous studies (for example microscopic polyangiitis may have been diagnosed as granulomatosis with polyangiitis). Also there is no generalized definition of what should be considered a true relapse. Retrospective studies are mostly performed on study populations, which may be nonrepresentative samples as they excluded mild cases, not eligible to enter trials (24). Also, many studies have been restricted to European populations. Here we will discuss general prognostic concepts whereas more treatment-specific outcomes will be discussed in the management outcomes section of this article.

Assessment tools have been established to predict prognosis and allow for comparability between studies. The Five-Factor Score (FFS) was established in 1996 to evaluate prognosis at diagnosis for systemic necrotizing vasculitides (for panarteritis nodosa, microscopic polyangiitis, and EGPA) and was extended and revised as 2009 FFS to account for granulomatosis with polyangiitis (33). A total of 1108 patients registered in the French Vasculitis Study Group database were analyzed until 2005. Overall mortality over 5 years was 19.8% (219/1108); mortality for each disease was as follows: MPA (60/218, 27.5%), GPA (41/311, 13.2%) and EGPA (32/230, 13.9%). Factors increasing mortality were age older than 65 years, cardiac symptoms, gastrointestinal involvement, and renal insufficiency (stabilized peak serum creatinine ≥150 μmol/L). The presence of each was accorded +1 point. Otolaryngologic symptoms affecting patients with GPA and EGPA were associated with a lower relative risk of death, and their absence was scored +1 point. According to the 2009 FFS, 5-year mortality rates for scores of 0, 1, and 2 or higher were 9%, 21%, and 40% respectively. The score implicates visceral involvement as a major influence on outcome (33).

Another tool used to assess disease activity is the Birmingham Vasculitis Activity Score (BVAS, further outlined in management outcomes section), a clinical checklist of relevant symptoms, signs, and features of active disease. The BVAS allows for comparison between vasculitic diseases, follow up in clinical trials, and guiding clinical decision making. This is separate from the vasculitis damage index (VDI), which represents chronic damage either from vasculitis or from treatment, which is not responsive to further therapy and which should not prompt escalation of treatment (94; www.vasculitis.org).

Granulomatosis with polyangiitis and microscopic polyangiitis. Prior to recognition by Fauci and colleagues that cyclophosphamide and prednisone could induce remission in a majority of patients, prognosis in granulomatosis with polyangiitis was dismal, with a mean patient survival of 5 months (22). With better therapies, however, survival has dramatically improved, with over 90% of patients achieving remission. ANCA-associated vasculitides are now considered chronic diseases with relapses. Cumulative survival for patients with ANCA-associated vasculitis (observed in a large cohort with 52% GPA and 48% MPA) is estimated to be 88% at 1 year, 85% at 2 years, and 78% at 5 years (24). Flossman found no difference in survival in patients with either granulomatosis with polyangiitis or microscopic polyangiitis or patients with different ANCA specificities. It was observed that 44% of the deaths occurred within the first year after enrollment and were due to infection (48%) or active vasculitis (19%). For patients who died after 1 year of enrollment, the causes were mostly cardiovascular disease (26%), malignancy (22%), or infection (20%). Patients with granulomatosis with polyangiitis have a significantly increased risk of myocardial infarction, especially early after the diagnosis of granulomatosis with polyangiitis (02). Infection also plays a major role in mortality both in induction and maintenance phase. Standardized mortality ratio remains elevated compared to the general population at 2.6 (p< 0.0001). Advanced renal impairment and higher disease activity were predictors of mortality, as measured by the Birmingham Vasculitis Activity Score (BVAS), which categorizes organ involvement and disease activity in each organ system (64; 24). Advanced renal failure and higher age were independent risk factors for treatment-related toxicity, including infections (24). Further risk factors for infections are level of aggressiveness of induction therapy, the cumulative exposure to steroids, and dialysis dependence (51).

ANCA-associated vasculitis with pulmonary involvement and with presence of large cavities predisposes to fungal infections. Antibiotic coverage is recommended in the early induction phase with trimethoprim-sulfamethoxazole with all types of immunosuppression. More individualized preventive strategies are recommended depending on laboratory markers such as serum creatinine or the lymphocyte counts including CD4+ T-cell count (51).

In addition to specific organ-related damage, patients with granulomatosis with polyangiitis are significantly more likely to suffer from depression (22% vs. 7%) and fibromyalgia symptoms as compared to the general population. Fatigue is also extremely common, significantly limiting activities in nearly half of these patients, and sleep disorders are reported in up to 29%. Interestingly, these symptoms often persist during periods of remission and do not correlate well with disease duration, activity level, or damage (34). Patient perspectives on disease impact have revealed that fatigue and low energy account for the greatest burden of disease for patients with granulomatosis with polyangiitis, eclipsing even certain severe manifestations, such as the requirement for dialysis (36).

Eosinophilic granulomatosis with polyangiitis. Samson and colleagues prospectively evaluated 118 patients with EGPA with a mean followup of 81.3 months who were enrolled in 2 prospective, randomized, clinical trials. Overall survival 1-, 3-, 5- and 7-year rates were 98%, 94%, 92%, and 90%, respectively, regardless of baseline severity of disease, number of relapses, ANCA-positivity or eosinophil counts. Long-term remission was achieved in 29% of patients. Age 65 years or older was the only factor significantly associated with higher mortality. Patients with cardiomyopathy had a lower survival rate (64; 11; 81). Long-term organ damage was common, usually chronic asthma and peripheral neuropathy. The risk of relapse was higher for patients with MPO ANCA and lower for those with greater than 3000 eosinophils/mm3 (11; 81). Compared to ANCA-negative patients, ANCA-positive patients more frequently had clinical vasculitic manifestations, such as peripheral neuropathy or renal involvement, but less frequently had cardiomyopathy. Their mortality rate was lower, despite the slightly higher risk of relapses (11).

Sinico and colleagues examined retrospectively 116 patients and found a similar mortality rate (85). Five-year mortality rates were 11.7% in patients with nephropathy. Five of the 10 patients who died had renal involvement, and the authors considered that nephropathy might be associated with a poor prognosis.

A retrospective literature analysis in 2008 revealed a survival in EGPA of 93% to 94% at 1 year and 60% to 97% at 5 years (64). The absence of any of the 5 manifestations, incorporated in the Five Factor Score (FFS), (proteinuria, renal failure, gastrointestinal involvement, cardiomyopathy, and concurrent neurologic manifestations were part of the score) was associated with a good prognosis, while the presence of 2 or more of the factors increased mortality in EGPA.

Clinical vignette

A 30-year-old woman presented with insidious-onset sinus congestion and facial pain not responsive to multiple courses of antibiotics. Over the following 2 months, she developed progressive fatigue and weight loss, along with dyspnea and 2 discrete episodes of hemoptysis. She was hospitalized but continued to worsen, developing bilateral hearing loss, migratory arthralgias, and sudden, new-onset numbness and weakness in the right hand.

Physical examination was notable for saddle-nose deformity, nasal crusting, right tympanic membrane perforation, and sensory loss and motor weakness in the distribution of the right ulnar nerve. Chest x-ray, followed by chest CT revealed multiple cavitary lung nodules. Bronchoscopy with bronchoalveolar lavage was negative for bacterial, mycobacterial, or fungal organisms. CBC revealed stable counts; creatinine was preserved, and urinalysis was negative for protein or blood. ESR was markedly elevated at 100 mm/hr. C-ANCA and p-ANCA were negative; however, CT-guided lung biopsy of a nodule revealed granulomatous inflammation and necrotizing vasculitis.

She was diagnosed with severe, generalized granulomatosis with polyangiitis and was treated with oral prednisone 60 mg once daily and rituximab (4 weekly infusions). Trimethoprim/sulfamethoxazole 3 times weekly was added, along with daily calcium 1000 mg and vitamin D 1000 IU. The hemoptysis resolved, and she was discharged home. As an outpatient, prednisone was tapered, and 3 months later, azathioprine was added at 150 mg once daily. The systemic features, joint pain, sinus symptoms, weakness, and dyspnea resolved. Follow-up chest CT 3 months later showed significant improvement.

Biological basis

Etiology and pathogenesis

The exact etiology of ANCA-associated vasculitis is unknown. It is believed to result from a combination of genetic predisposition, epigenetic factors, and environmental exposure, which lead to the production of pathogenic antineutrophil cytoplasmic autoantibodies (ANCAs) that activate neutrophils and monocytes, resulting in necrotizing vascular inflammation.

In granulomatosis with polyangiitis, approximately 90% of patients will produce ANCA directed against proteinase 3, and another 6% will have ANCA against myeloperoxidase. Microscopic polyangiitis mostly has more ANCA directed against myeloperoxidase (78%) than proteinase 3 (18%) (49), and EGPA has antibodies against myeloperoxidase in 40%. Proteinase 3 and myeloperoxidase are intracellular granules found within neutrophils. For autoantibody-antigen interactions to occur, neutrophils must first be “primed” by circulating inflammatory cytokines such as TNF-alpha, IL-1, and IL-18 to express these antigens on the cell surface. Once accessible, ANCA binds to proteinase 3 or myeloperoxidase on the cell surface via the Fab portion of the autoantibody, while the Fc portion activates receptors on the same neutrophil, nearby neutrophils, or endothelial cells, resulting in adherence of neutrophils to the endothelium and transmigration through the vessel wall. Neutrophils bound by ANCA degranulate and undergo a respiratory burst, releasing lytic enzymes and reactive oxygen species that cause vascular necrosis and inflammation. There is concomitant activation of the alternative complement pathway and recruitment of additional leukocytes to perpetuate the damage (50).

The pathogenesis of extravascular granulomatous inflammation remains less well understood. It is hypothesized that it results from neutrophil migration into the tissues of the upper and lower respiratory tract in response to a stimulus, possibly infection. ANCAs present in the tissues bind to neutrophils and stimulate activation and degranulation via a similar process as described earlier. Acutely activated neutrophils cause tissue necrosis and fibrin deposition. This inflammatory episode serves to recruit monocytes to the area, which then convert to macrophages and form multinucleated giant cells (44). In addition to humoral mechanisms of autoimmunity, there is also evidence for T-cell dysregulation in ANCA-associated vasculitis. Normally, regulatory T cells help downregulate lymphocyte activation and, therefore, inhibit B-cell production of autoantibodies. In ANCA-associated vasculitis, regulatory T cells are less effective and fewer in number as compared to controls, and this correlates with a higher risk of relapse (44). B-cell regulation is also impaired. Patients with active ANCA-associated disease have a reduced percentage of circulating CD5+ B cells, whereas patients in remission have a normal percentage. With rituximab therapy, the degree of normalization of circulating CD5+ B cells may correlate with more persistent remission, but this requires further investigation (44).

A genome-wide association study in ANCA-associated vasculitis (GPA and MPA) demonstrated a strong genetic association with the antigenic specificity of ANCA and not with the clinical syndrome: PR3-ANCA was associated with HLA-DP and genes encoding for alpha 1-antitryspin and proteinase 3, whereas MPO-ANCA was associated with HLA-DQ (58). Therefore, antigen recognition is important in ANCA-associated disease. That the initial antigen recognized by antigen-presenting cells is different in PR3-ANCA-associated disease as opposed to MPO-ANCA-associated disease argues that they are genetically distinctive diseases (44).

In some studies, PR3-ANCA is associated with a higher rate of relapses compared to MPO-ANCA, suggesting a genetic basis (102; 59). PR3-ANCA-associated vasculitis (PR3-AAV) may differ from MPO-ANCA-associated vasculitis (MPO-AAV) with regard to treatment response, relapse rate, and outcome. Further trials should separate study patients with PR3-AAV and MPO-AAV and will provide clarification whether “PR3-AAV” and “MPO-AAV” are more accurate terms for disease entities than “GPA” and “MPA” (49).

Little is known about patients with ANCA-associated vasculitis who test positive for ANCAs (microscopic polyangiitis or granulomatosis with polyangiitis). It is unclear whether ANCAs cannot be detected with the current available techniques or whether they have different pathomechanisms (43). These patients are oftentimes excluded from treatment trials. A new autoantigen specificity for lysosomal-associated membrane protein 2 (LAMP-2) remains controversial as the results were not reproduced in 2 later studies (44). ANCAs also occur in patients with inflammatory bowel disease, but they typically do not have a specificity for proteinase 3 or myeloperoxidase (44).

In EGPA, the presence or absence of ANCA is associated with 2 different clinical phenotypes, supporting the role of ANCA in disease pathomechanism. Presence of ANCA is associated with renal involvement, peripheral neuropathy, and biopsy-proven vasculitis, whereas ANCA absence is associated with heart disease (78; 99). ANCAs were found in 75.0% of patients with nephropathy and in only 25.7% of patients without nephropathy. All patients with necrotizing crescentic glomerulonephritis were ANCA-positive (85). The ANCA status in EGPA also has an association with prognosis: the risk of relapse is significantly higher for MPO-ANCA-positive patients at diagnosis (11; 81), but the mortality is lower (11), which may be related to the fact that cardiomyopathy is a major risk factor for death and more common in ANCA-negative patients.

The development of granulomatosis with polyangiitis may require exposure to an environmental trigger, resulting in disruption of autoimmune homeostasis; however, this has not been proven. Infection with Staphylococcus aureus is frequently implicated based on the high prevalence of nasal carriage of S aureus in patients with ANCA-associated vasculitis and evidence for prevention of relapse with high-dose trimethoprim-sulfamethoxazole therapy (89). With respect to pathogenesis, S aureus secretes superantigens that may stimulate T- and B-cell proliferation and release of inflammatory cytokines to “prime” neutrophils for ANCA binding (07). In addition to infection, silica-containing compounds and medications, including hydralazine and propylthiouracil, and levamisole-adulterated cocaine induce a similar small-vessel vasculitis.

In EGPA, leukotriene receptor antagonists are related to disease onset, but the role in triggering EGPA is still uncertain. They may have unmasked previous existing disease by their steroid-sparing effect and enabling steroid tapers without flare-up of disease (99; 27).

Histopathology. Pathologically, GPA and EGPA are characterized by both a necrotizing small- to medium-vessel vasculitis affecting the capillaries, arterioles, venules, arteries, and veins (04) and granulomatous inflammation. The typical histopathological lesion in granulomatosis with polyangiitis consists of areas of necrosis of variable size with a surrounding granulomatous reaction, including giant cells. Vasculitis, often with luminal occlusion, occurs adjacent to necrotic foci. Deposits of immune complexes and complement are absent. Granulomas in granulomatosis with polyangiitis contain CD4+CD28- T cells (catalysts of inflammation), monocyte-derived tissue macrophages, giant cells, and large numbers of neutrophils surrounding a necrotic center (65).

EGPA was originally identified on autopsy as a pathological triad consisting of eosinophilic infiltration, necrotizing vasculitis, and extravascular granuloma formation (09). Because glucocorticoids are used commonly and early, pathology-based identification of granulomas has become infrequent. Typically, they would show a core of necrotic eosinophilic material surrounded by palisading lymphocytes and epithelioid and multinucleated giant cells (29). Renal biopsy can reveal necrotizing crescentic glomerulonephritis but also eosinophilic interstitial nephritis, mesangial glomerulonephritis, and focal sclerosis (85).

Epidemiology

Epidemiological studies in ANCA-associated vasculitis are challenging. These are relatively rare diseases with variation in geographical distribution and ethnicity of patients and overlapping features.

ANCA-associated vasculitis predominantly affects Caucasians, and both men and women are equally affected (12). The annual incidence of granulomatosis with polyangiitis ranges from 5 cases per million in Spain (103) to 11.3 per million in the United Kingdom (104) and appears to be increasing over time, perhaps due to increased recognition. Microscopic polyangiitis has a reported annual incidence of 5.9 per million in the United Kingdom (104). Although there was no major difference in ANCA-associated vasculitis incidence between Japan and the United Kingdom, microscopic polyangiitis and MPO-ANCA was more common in Japan compared to granulomatosis with polyangiitis (83% of patients with ANCA-associated vasculitis met criteria for microscopic polyangiitis) (25).

It may affect any age group; however, the majority of new cases are diagnosed between 65 to 74 years of age (104). Although considered rare in childhood, a study from Southern Alberta, Canada, indicated an average annual incidence in pediatric patients of 2.75 cases per million (30), which is not considerably different from that seen in adults. The prevalence of granulomatosis with polyangiitis varies from 30 per million population in a hospital-based United States study (12), to 60 per million in a population-based German study (75) and 148 per million in a British population-based study (104). The prevalence of microscopic polyangiitis in a United Kingdom-based study was 63.1 per million (104). The incidence of granulomatosis with polyangiitis shows no seasonal variation but has periodicity, with peaks occurring every 7.6 years, supporting a possible infectious trigger for this disease (104).

Although nonwhite populations are less commonly affected by granulomatosis with polyangiitis, the HLA-DRB1*15 allele is associated in African Americans with higher risk of ANCA-associated vasculitis with positive PR3-ANCA, typically corresponding to the granulomatosis with polyangiitis phenotype (06).

EGPA can manifest at any age, but mean age at onset has been described between 40 and 60 years (85; 27; 29). The estimated annual incidence is 0.11 to 2.66 per 1 million people, with an overall prevalence of 10.7 to 14 per 1 million adults. There is no gender predominance or ethnic predisposition in EGPA (72).

Prevention

ANCA-associated vasculitis is an autoimmune disease. There are no known preventable measures to date. In EGPA, triggering factors such as treatment with leukotriene antagonists remain highly controversial as they may help unmask the disease, but may not play a role in disease etiology.

Differential diagnosis

When assessing features of small-vessel vasculitis, it is important to consider both primary and secondary etiologies. In the primary vasculitides, there is considerable overlap among the ANCA-associated vasculitides, but distinctions remain. Microscopic polyangiitis may be differentiated from granulomatosis with polyangiitis by the absence of upper respiratory tract manifestations and granulomatous inflammation, and ANCAs are usually directed against myeloperoxidase. EGPA is distinguished by characteristic allergic rhinitis, nasal polyposis, asthma, and peripheral eosinophilia. In contrast to GPA, only 40% of these patients will have a positive ANCA, usually directed against myeloperoxidase.

In the absence of clinical or histologic evidence of vasculitis or absence of ANCA, it may be difficult to differentiate between EGPA and other systemic eosinophilic disorders, such as chronic eosinophilic pneumonia, allergic bronchopulmonary aspergillosis, or hypereosinophilic syndrome. However, response to treatment is usually better and faster in EGPA than in primary systemic eosinophilic disorders; therefore, treatment response can help to re-evaluate the initial diagnosis (11).

Panarteritis nodosa is one of the main elements of differential diagnosis. It is a necrotizing arteritis of the medium and small arteries and needs to be distinguished from the ANCA-associated vasculitis. Although it can appear similar clinically and pathologically, it is not associated with ANCA, granulomatous inflammation, or glomerulonephritis. A typical finding in panarteritis nodosa are medium artery aneurysms, which are not present in ANCA-associated vasculitis (43).

Secondary causes of small-vessel vasculitis should also be considered, including those related to rheumatic diseases, such as systemic lupus erythematosus, rheumatoid arthritis, Sjögren syndrome, cryoglobulinemia, and sarcoidosis. Infection-associated vasculitides include infectious endocarditis, syphilis, mycobacteria, leprosy, hepatitis B, hepatitis C, and HIV, and in some cases, fungal infections. Numerous medications induce small-vessel vasculitis; notable among these are propylthiouracil and hydralazine. In addition, the adulterant levamisole, found in up to 70% of all cocaine seized in North America, is now a well-recognized cause of thrombotic microangiopathy with necrotic cutaneous lesions and retiform purpura that may mimic granulomatosis with polyangiitis. In these patients, neutropenia (uncommon in granulomatosis with polyangiitis) and positive p-ANCA, with specificity against human neutrophil elastase, are frequently reported (28). Malignancy-associated vasculitides should be ruled out when appropriate.

With respect to saddle nose or midline destructive lesions, the differential diagnosis includes relapsing polychondritis; congenital syphilis; cocaine-induced destruction; or midline lethal granuloma, now identified as NK/T-cell lymphoma.

In its most severe form, granulomatosis with polyangiitis may present with alveolar hemorrhage and rapidly progressive glomerulonephritis. In this case, additional causes of pulmonary-renal syndrome should be considered, such as microscopic polyangiitis, anti-GBM disease, systemic lupus, and catastrophic antiphospholipid antibody syndrome.

Diagnostic workup

The workup for ANCA-associated vasculitis includes laboratory testing, imaging, and, if possible, a biopsy of affected tissue as directed by the clinical symptoms and signs.

Laboratory workup. Nonspecific laboratory abnormalities seen in granulomatosis with polyangiitis or microscopic polyangiitis include leukocytosis, normocytic anemia, and thrombocytosis. EGPA is characterized by marked peripheral eosinophilia (usually > 1500 cells/μl or > 10% of total leucocyte count) (29). Renal function may be impaired, and inflammatory markers such as ESR or CRP may be significantly elevated. All patients with suspected ANCA-associated vasculitis should have a urinalysis performed for detection of hematuria and proteinuria, and, if present, microscopy to evaluate for dysmorphic RBCs and RBC casts. Select testing to rule out other diseases should be performed, including antinuclear antibody, complements, anti-GBM antibodies, cryoglobulins, viral serology, and blood cultures.

For EGPA, serologic testing for toxocariasis, aspergillus, and human immunodeficiency virus is indicated. A peripheral blood smear (looking for dysplastic eosinophils or blasts) is useful (31). Paraneoplastic eosinophilia may be considered in atypical cases. For further details of work up in atypical cases, we recommend review of the EGPA Consensus Task Force recommendations (31).

Antineutrophil cytoplasmic antibody testing. ANCA testing is important in supporting the diagnosis of ANCA-associated vasculitis and to help differentiate the types of ANCA-associated vasculitis. It is still recommended that ANCA first be measured by immunofluorescence to detect a cytoplasmic (c) or perinuclear (p) pattern, followed with antigen-specific testing by ELISA in order to maximize sensitivity and specificity. With low antibody titers, immunofluorescence can be negative, whereas antigen-specific testing is positive (14). If clinical suspicion is high, both tests should be ordered. In granulomatosis with polyangiitis, c-ANCA pattern is detected in 80% to 95% of patients, with specificity against proteinase 3, whereas p-ANCA pattern with specificity against myeloperoxidase may be found in 5% to 20% (39). Microscopic polyangiitis mostly has ANCA directed against myeloperoxidase (76%) but also EGPA against myeloperoxidase (40%). In granulomatosis with polyangiitis, sensitivity of ANCA testing varies with extent of disease, exceeding 90% in the acute generalized phase, but is found in only 50% to 78% of those with limited disease (92). Thus, the absence of ANCA antibodies does not rule out ANCA-associated vasculitis, particularly if disease is limited, and it is useful to recheck during the course of the disease. An atypical ANCA (where ANCA is positive by immunofluorescence, but negative for specificity against proteinase 3 or myeloperoxidase) should raise suspicion for the presence of another autoimmune disease, such as ulcerative colitis, primary sclerosing cholangitis, autoimmune hepatitis, connective tissue diseases, or rheumatoid arthritis. Levamisole may cause dual c- and p-ANCA positivity (74).

Imaging procedures. Imaging of the chest should be performed in patients with suspected ANCA-associated vasculitis. Up to 34% of patients may have asymptomatic disease in granulomatosis with polyangiitis (38). Chest CT may reveal nodules, cavitary lesions, infiltrates, and ground glass opacities suggestive of hemorrhage. CT of the sinuses and MRI of the brain may also be useful to evaluate the sinuses, orbits, and brain. In EGPA, CT is more sensitive in detecting abnormalities (31).

Histological confirmation. Definitive diagnosis of ANCA-associated vasculitis requires histologic confirmation. Demonstration of necrosis and pauci-immune vasculitis is characteristic. Granulomatous inflammation is diagnostic for granulomatosis with polyangiitis (but not present in microscopic polyangiitis) and is seen in EGPA along with eosinophilia. Biopsy should be taken from an involved organ site as “blind” biopsies rarely produce useful findings. Skin biopsy of a fresh vasculitic lesion is the least invasive approach; however, it often reveals only nonspecific leukocytoclastic vasculitis. Yield of ear-nose-throat biopsies vary, with the highest success rate in patients with macroscopic abnormalities in the nasal cavity (75). Renal biopsy should be performed if there is evidence of kidney involvement to establish the diagnosis and to assess for disease activity and chronicity. This is confirmed by a pauci-immune segmental necrotizing glomerulonephritis. Transbronchial biopsies are rarely diagnostic due to inadequate tissue sampling; however, open lung biopsy or video-assisted thorascopic biopsy of affected lung tissue has a high yield. Cultures and special stains should be performed on tissue samples to exclude infection.

Other tests. Bronchoalveolar lavage may be useful in ruling out infectious causes of lung infiltrates or nodules. It is instrumental in diagnosing alveolar hemorrhage when serial aliquots instilled into the pulmonary tree are increasingly hemorrhagic.

Assessing activity and damage. Regular assessments will determine disease activity in patients with ANCA-associated vasculitis (75). Instruments combining scores from each of the potentially involved organ systems have proven useful in monitoring disease course and therapy; of these, the Birmingham Vasculitis Activity Score is most commonly used. This tool measures signs of vasculitic activity that have existed for a maximum of 4 weeks (90; 94). Each symptom is then graded according to severity and organ system affected. Damage is measured by the Vasculitis Damage Index, which measures the accumulation of irreversible injury caused either by the disease or its therapy over time (20; 94).

In EGPA, differentiating disease activity from worsening of underlying asthma and sinusitis is challenging but important for treatment decisions. An attempt has been made to find biomarkers indicating disease activity. In studies of patients with EGPA with active disease and prior to treatment, eotaxin-3, TARC/CCL17, and IgG4 were identified as potential biomarkers of disease activity (73; 13; 100). However, in a prospective, longitudinal EGPA cohort study, these serum biomarkers (as well as ESR, CRP, and eosinophil count) did not discriminate between periods of remission and active disease within individual patients (19). Given that this disease is rare, numbers are small and further studies are needed.

Management

Treatment of ANCA-associated vasculitis, microscopic polyangiitis, or granulomatosis with polyangiitis is organized into 2 phases: induction and maintenance therapy. Medications are selected according to severity of disease (limited vs. generalized). Generalized disease can be further characterized as mild or severe (organ- or life-threatening). Induction therapy is usually for 3 to 6 months, with a goal of achieving disease remission. Patients are then switched to maintenance therapy for prevention of relapse. Limited disease is neither organ- nor life-threatening (eg, sinusitis), whereas generalized disease refers to organ involvement that is likely to result in serious morbidity and mortality if not treated aggressively (eg, pulmonary hemorrhage). The majority of therapeutic procedures discussed below are empirical and unsupported by controlled studies as ANCA-associated vasculitis has only been the subject of multi- or single-center controlled studies since the mid-1990s. Typical induction and maintenance protocols are summarized and reviewed in Table 1 (26; 52; 56; 49; 57).

Table 1. Treatment of Patients with GPA or MPA Based on Stage of Treatment and Disease Severity

Induction therapy

Granulomatosis with polyangiitis and microscopic polyangiitis. The intensity of the initial therapeutic approach must be adjusted for each patient and for the type and seriousness of the ANCA-associated vasculitis in order to avoid 2 pitfalls: (1) excessive treatment associated with a significant risk of side effects or (2) insufficient treatment with a risk of failure or early relapse (11). In cases with severe generalized manifestations of ANCA-associated vasculitis, such as pulmonary hemorrhage, mononeuritis multiplex, and rapidly progressive glomerulonephritis, standard therapy is still as defined by Fauci and colleagues in 1983: daily oral cyclophosphamide 2 mg/kg plus oral prednisone 1 mg/kg per day with gradual taper (22). However, because prolonged use of cyclophosphamide was found to lead to unacceptable toxicities (including 26% to 46% rates of infection, 12% to 43% risk of hemorrhagic cystitis, and 57% risk of infertility), the original protocol has been modified. Cyclophosphamide is now used for a finite period, usually for 3 to 6 months (41). In a further attempt to minimize cyclophosphamide exposure, intravenous cyclophosphamide therapy (15 mg/kg every 2 to 3 weeks) was compared directly with daily oral dosing (2 mg/kg/day) for induction of remission in ANCA-associated vasculitis by de Groot and colleagues (15). Despite receiving half the cumulative dose of drug as compared to the daily oral group (8.2 g vs. 15.9 g), those treated with intravenous cyclophosphamide were equally likely to achieve remission. Long-term follow-up of these patients demonstrated that, indeed, intravenous treatment had a significantly higher risk of relapse over time (39.5% in intravenous arm vs. 20.8% in oral arm), although mortality and renal function remained similar (35).

Next, the NORAM (NonRenal ANCA-associated Vasculitis Alternatively Treated with Methotrexate) trial demonstrated that in patients with limited disease and without seriously impaired renal function, methotrexate is equivalent to oral cyclophosphamide for inducing remission but, again, carries a greater risk of relapse (16). Based on this study, methotrexate remains a useful induction agent in those with mild disease without renal involvement who wish to avoid the adverse effects associated with cyclophosphamide.

Most recently, studies have demonstrated that rituximab, an anti-CD20 monoclonal antibody that depletes B cells, is equally effective to cyclophosphamide for inducing remission in patients with generalized disease, including those with major renal involvement, with a similar if not improved adverse event profile (91). RITUXVAS was a 24-month, randomized controlled trial involving 44 patients with newly diagnosed ANCA-associated renal vasculitis. Patients received glucocorticoids plus either rituximab (375 mg/m2/week×4) with 2 intravenous cyclophosphamide pulses or intravenous cyclophosphamide (15 mg/kg every 2 weeks for 3 doses followed by infusions every 3 weeks) for 3 to 6 months followed by azathioprine. There was no difference in adverse events. Both groups were similar in the frequency of the composite outcome of death, ESRD. Relapses in rituximab patients were observed after B cell return and usually after ANCA return, suggesting that although the rituximab induction regimen offers a potential alternative to cyclophosphamide/azathioprine over 24 months, it did not induce long-term remission or cure. The authors concluded that after treatment with rituximab, vigilance for relapse is warranted particularly in patients with detectable peripheral blood B cells and ANCA (47). In relapsing cases, an additional dose of rituximab could be considered.

An NIH funded multicenter, randomized, double-blind, double-dummy, noninferiority trial compared rituximab (375 mg/m2 weekly for 4 weeks) followed by placebo with cyclophosphamide (2 mg/kg per day) administered for 3 to 6 months followed by azathioprine for 12 to 15 months in 197 patients (88). All patients received 1 to 3 doses of methylprednisolone (1000 mg) followed by prednisone (1 mg/kg per day). Patients had granulomatosis with polyangiitis (75% of enrolled patients) or microscopic polyangiitis (25%), all were ANCA-positive and had a BVAS score higher than 3. Rituximab was noninferior to cyclophosphamide in inducing remission by 6 months (64% vs 53%). Among the 101 patients who had relapsing disease at baseline, rituximab was superior to conventional immunosuppression at 6 months (P = 0.01) and at 12 months (P = 0.009). Therefore, in patients with severe ANCA-associated vasculitis, a single course of rituximab was as effective as continuous conventional immunosuppressive therapy for the induction and maintenance of remissions over the course of 18 months. There was no significant between-group difference in adverse events.

Overall, in cases with contraindication to cyclophosphamide or refusal because of concerns about cancer or fertility, rituximab is the drug of choice because in these 2 randomized trials, it was as effective as cyclophosphamide in inducing remission among patients with newly diagnosed or relapsing granulomatosis with polyangiitis or microscopic polyangiitis. In a post-hoc analysis of the RAVE study, patients with PR3 ANCA-associated vasculitis responded better to rituximab than to cyclophosphamide/azathioprine (98). An ANCA type-based classification may guide immunosuppression in ANCA-associated vasculitis. It should also be considered in patients who have already received cyclophosphamide and had a relapse (11).

For some authors, rituximab has replaced cyclophosphamide in the management of granulomatosis with polyangiitis and microscopic polyangiitis. Despite this, Specks considers cyclophosphamide in compliant, newly diagnosed, MPO-ANCA-positive patients without cancer history and without fertility concerns as well as in patients with lack of response to rituximab (87).

With induction treatment, it is recommended that ANCA-associated vasculitis be treated with a systemic corticosteroid along with the immunosuppressant in combination. Oral prednisone is recommended at a daily starting dose of 1 mg/kg. There is no recommended tapering scheme, but generally prednisone taper is started after 3 to 4 weeks of oral prednisone treatment and should continue during maintenance phase, typically with duration of 6 to 8 months until discontinuation.

Eosinophilic granulomatosis with polyangiitis. Unlike the other ANCA-associated vasculitides, there are no randomized-controlled treatment trials for EGPA. The following recommendations are mostly based on recommendations developed by a consensus task force indicating different levels of evidence (31). Glucocorticoids are the cornerstone of therapy for EGPA and are often sufficient to achieve remission. Methylprednisolone intravenously (7.5 to 15 mg/kg/day) should be given for life-threatening symptoms along with additional immunosuppression followed by oral prednisone. Otherwise, the induction dose prescribed should be 1 mg/kg/day prednisone for patients with organ- or life-threatening manifestations for 2 to 3 weeks, followed by gradual tapering over 6 months to the minimal effective dose or to discontinuation if tolerated (31). However, in a retrospective study, approximately 85% of EGPA patients required long-term prednisone (mean dose 12.9 ± 12.5 mg/day) (11), raising the question whether steroid sparing agents are needed.

Patients with life- or organ-threatening disease manifestations should be treated with another immunosuppressant (eg, cyclophosphamide). Adjunctive cytotoxic drugs are recommended to treat patients with a Five-Factor Score (FFS) ≥ 1. The FFS is a prognostic tool that consists of 5 items, 4 of which are associated with poor prognosis (age older than 65 years, heart and gastrointestinal peak serum creatinine > 150 μmol/L), while the fifth (otolaryngologic manifestations) is associated with better outcomes, and its absence is scored+1.

Cyclophosphamide may be equally effective when given as continuous oral therapy (2mg/kg/day) or as intravenous pulses (15 mg/kg or 0.6 g/m2, first 3 infusions every 2 weeks with a maximum of 1.2 g per infusion, followed by pulses 3 to 6 at 15 mg/kg or 0.7 g/m2 every 3 weeks) (31).

Maintenance therapy

Granulomatosis with polyangiitis and microscopic polyangiitis. Once remission is attained, treatment is switched to a more tolerable maintenance regimen, typically for 12 to 24 months. A number of medications are suitable for maintenance of remission (see Table 1). Data from the CYCAZAREM trial suggest that cyclophosphamide can be replaced by azathioprine during the maintenance phase, with no increased relapse rate and improved safety (41). Methotrexate is also effective for maintenance therapy; extended 32-month follow-up shows no increased risk of relapse when compared to extended use of cyclophosphamide (54; 55). In a direct comparison, the Wegener Granulomatosis Entretien Trial (WEGENT) demonstrated no difference in either efficacy or safety between methotrexate and azathioprine for maintenance therapy (76).

Mycophenolate mofetil remains an option in those who fail or cannot tolerate azathioprine or methotrexate or rituximab. Although effective, a comparison study showed it was associated with a higher rate of relapse (55% vs. 38%) when compared to azathioprine (37). Leflunomide was studied in a prospective randomized controlled clinical trial against oral methotrexate. The study was discontinued early due to high relapse rates in the oral methotrexate group. The incidence of adverse reactions was concerning in the leflunomide group (34 adverse events in 55 patients) and resulted in discontinuation of the drug in 4 patients (61).

Retrospective studies indicate the efficacy of rituximab as a maintenance therapy (07; 86). The MAINRITSAN study was the first randomized study to compare rituximab with azathioprine as maintenance of remission therapy (32). The study showed that 500 mg rituximab infusions every 6 months were superior to azathioprine as maintenance therapy for PR3-ANCA vasculitis. Patients with newly diagnosed or relapsing granulomatosis with polyangiitis, microscopic polyangiitis, or renal-limited ANCA-associated vasculitis in complete remission after a cyclophosphamide-glucocorticoid regimen were randomly assigned to receive either 500 mg of rituximab on days 0 and 14 and at months 6, 12, and 18 after study entry or daily azathioprine until month 22. The primary outcome was a major relapse. One hundred fifteen enrolled patients (87 with GPA, 23 with MPA, and 5 with renal-limited ANCA-associated vasculitis) were all ANCA-positive, with PR3 70%, MPO-ANCA 23%. A major relapse occurred in 17 patients in the azathioprine group (29%) and in 3 patients in the rituximab group (5%) (HR for relapse, 6.61). The frequency of severe adverse events was similar in the two groups. The authors concluded that more patients with ANCA-associated vasculitides had sustained remission at month 28 with rituximab than with azathioprine. The limitations of the study were the predominance of proteinase 3 and low number of myeloperoxidase ANCA and microscopic polyangiitis cases; therefore, it is unclear whether the results can be generalized to myeloperoxidase ANCA vasculitis. Further studies are needed, and it is possible that treatment recommendations in the future will vary based on ANCA subtype. Also, the patients remained on steroids, which could account for the generally low relapse rates in both groups, discontinuation was deferred to the local investigator, and azathioprine was tapered, which has been a point of critique. However, more relapses occurred in the azathioprine group prior to the taper (32). (A lower dose of rituximab was used to account for previous induction therapy.)

Overall, existing trial data suggest that azathioprine and methotrexate are both options for maintenance treatment in granulomatosis with polyangiitis and microscopic polyangiitis. For patients with PR3-ANCA, rituximab may be superior.

MPO-ANCA-positive patients with complete remission may be considered for careful clinical monitoring given that relapses are less common in this group, but more studies are needed. Specific markers allowing reliable identification of low relapse risk have to be studied further (18).

Further studies are ongoing. The BREVas study is investigating the efficacy and safety of belimumab (a monoclonal IgG1-lambda antibody that inhibits B-cell activating factor, also known as B-lymphocyte stimulator), in combination with azathioprine, for the maintenance of remission following a standard induction regimen in patients with granulomatosis with polyangiitis or microscopic polyangiitis. (Completion expected February 2017, NCT01663623.)

Eosinophilic granulomatosis with polyangiitis. In a prospective study, 48 patients with severe EGPA were treated with oral glucocorticoids plus 12 or 6 intravenous cyclophosphamide pulses as maintenance therapy. The difference in overall relapses was not significant between the 12-pulse and the 6-pulse regimens. However, when including only the number of mild relapses, the 12 cyclophosphamide pulses were better able to control severe Churg-Strauss syndrome than a 6-pulse regimen. The best long-term duration of treatment remains uncertain. Interestingly, oral glucocorticoids could be stopped for only 5 (11.4%) of 44 patients (10). This suggests that EGPA patients would also benefit from maintenance cyclophosphamide therapy to avoid relapses and also allow glucocorticoid tapering (31). The CYCAZAREM trial showed that azathioprine (2 mg/kg/day) was as effective as cyclophosphamide for preventing relapses in ANCA-associated vasculitis (though excluding EGPA) (41). Cyclophosphamide remains an important agent for induction but not for remission maintenance. For patients with relapsing disease involving life- or organ-threatening disease, maintenance therapy with azathioprine or methotrexate is recommended (31). There is no prospective randomized study to date comparing the two in EGPA. The optimal duration of maintenance therapy after achieving remission also remains unknown (31).

ANCA-positivity, cutaneous manifestations, and a low eosinophil count at the time of EGPA diagnosis predict relapses (11).

The CHUSPAN study, a prospective, randomized, open-label study of newly diagnosed EGPA without poor prognosis factors (FFS = 0, 72 patients), involved initial treatment with corticosteroids alone, and in case of failure or relapse, patients were randomized to receive 6 months of oral azathioprine or 6 pulses of cyclophosphamide; 93% achieved remission with corticosteroids therapy alone, and 35% relapsed during corticosteroid dosage tapering or after termination of corticosteroid therapy (especially those with peripheral neuropathy). The 19 patients randomized to additional immunosuppression achieved remission, with no significant difference between the cyclophosphamide and azathioprine group. At the end of follow up, 79% of the patients whose disease was in remission still required low-dose corticosteroid therapy, mainly to control respiratory disease (76). The inability to wean off steroids and the high rate of relapse suggests that patients might also benefit from early additional immunosuppression (31). A follow-up analysis suggests that the initial mononeuritis multiplex subpopulation within the group was a distinct prognosis group compared to other FFS=0 patients, with the same overall survival but a higher risk of relapse, corticosteroid failure, and corticosteroid-dependence (80).

It is unclear whether immunosuppressive drugs should be added for patients unable to taper prednisone for asthma or otolaryngologic manifestations (31). The CHUSPAN 2 trial (NCT00647166, active but not recruiting, completion expected July 2015, results not published) is currently evaluating the effectiveness of adjunctive azathioprine for FFS= 0 EGPA patients.

Therapy in refractory cases. Seven percent to 10% of patients will not achieve remission after 6 months of daily oral prednisone and cyclophosphamide. In these patients, rituximab may be a more effective agent (48; 91). For patients with severe renal involvement at presentation (creatinine > 500 μmol/L), the addition of plasmapheresis to traditional immunosuppression provides a greater likelihood of recovering renal function (42). Plasmapheresis is often used empirically in cases of severe alveolar hemorrhage, although evidence to support this approach is lacking (08). A prospective, randomized controlled trial (PEXIVAS) to address this question is currently in progress (still recruiting, completion April 2018, NCT00987389).

Within the MAINRITSAN trial, 20% of patients had prior relapses (n=23). Five of them experienced relapses during the study, and all were in the azathioprine treatment group (32). In a large retrospective trial with maintenance therapy with rituximab (1000 mg every 4 months), the rate of relapse did not differ between the patients with new diagnosis or previous relapses (71). In another retrospective study, scheduled rituximab retreatment over 2 years reduced relapse rates during the retreatment period and a more prolonged period of remission during subsequent follow-up (86).

A new study RITAZAREM, is a randomized controlled trial comparing rituximab with azathioprine as maintenance therapy in relapsing ANCA-associated vasculitis. Trial completion is expected in December 2018 (NCT01697267). A randomized trial comparing continuous-scheduled rituximab, with dosing based on rate of ANCA and CD19 lymphocytes, is underway (MAINRITSAN 2; NCT01731561); completion is expected October 2017. MAINRITSAN 3 investigates the best duration of maintenance treatment and is a randomized placebo-controlled trial comparing conventional maintenance treatment (18 months) against long-term rituximab maintenance treatment (46 months) (completion expected January 2019, NCT02433522).

Eosinophilic granulomatosis with polyangiitis. The prospective trial comparing rituximab against azathioprine as maintenance therapy did not include patients with EGPA. Data on rituximab use for EGPA, limited to case reports or series (79; 95; 63), suggest that it is effective and safe. Responses are reported in both ANCA-negative and -positive patients, though it may be more effective in the ANCA-positive patients (63). Two patients had severe bronchospasms during the first infusion (05). Rituximab can be considered for selected ANCA-positive patients with renal involvement or refractory disease or with contraindications to other agents (31).

Intravenous immunoglobulins can be considered a second-line therapy for patients on glucocorticoids (or other immunosuppressants) with EGPA flares refractory to other treatments or during pregnancy (31).

Mepolizumab, a humanized monoclonal antibody targeting IL-5, effective against eosinophilic asthma, was effective in a pilot study in refractory EGPA. An international randomized placebo-controlled trial on EGPA is ongoing (ClinicalTrials NCT02020889, recruiting, estimated completion September 2016). The study is enrolling both ANCA-positive and -negative patients. It will be of interest to compare treatment responses between the groups as it is still debatable whether they represent the same disease or may require distinct therapeutic strategies (31).

Local therapy. Subglottic stenosis is a potentially life-threatening manifestation that is often refractory to standard therapy. For these lesions, local dilatation combined with injection of glucocorticoids is the treatment of choice (53).

Follow-up. In light of the chronicity of granulomatosis with polyangiitis, its tendency to relapse, and the toxicity of treatment, long-term follow-up combined with patient education is advisable. Relapses are often foreshadowed by constitutional symptoms, sinusitis, arthralgias, skin lesions, or hematuria. Increase in c-ANCA titers may portend relapse (03); however, other published studies have found no correlation between ANCA level and disease activity (23). Treatment should not be guided by ANCA level alone. However, caution is advised when ANCA-titers change from negative to positive or rise substantially between visits (18).

Outcome measures

Disease outcome is influenced by disease activity, permanent damage from the disease (scarring), and treatment side effects (infections, malignancy, etc). To be able to measure disease activity, evaluate for treatment indications across studies, and compare outcomes, validated assessment instruments are necessary. The Birmingham Vasculitis Activity Score (BVAS) is the most effective validated tool to document disease activity and to define remission and response to therapy (60; 49; 57). Several versions all appear to be of similar validity (60). It includes a list of typical symptoms related to vasculitis based on history or clinical examination findings affecting several organ systems. The Vasculitis Damage Index (VDI) is used to assess permanent damage as a result of vasculitis, but also includes treatment effects (such as osteoporosis, diabetes, cataracts, atrophy and weakness, malignancy, gonadal failure, marrow failure, chemical cystitis, and avascular necrosis) (20; 77). The VDI is correlated with mortality (57). The French Vasculitis Study Group has introduced 5 prognostic factors as the Five-Factor Score (FFS) in patients with necrotizing vasculitis, which is often used in EGPA. It includes: (1) elevated serum creatinine levels (> 1.58 mg/dl), (2) proteinuria (> 1 g/day), (3) gastrointestinal tract involvement, (4) cardiomyopathy, and (5) otolaryngologic symptoms (60; 29). In the absence of reliable biomarkers, these clinical tools are used to guide treatment and compare study results.

Granulomatosis with polyangiitis and microscopic polyangiitis. Despite therapy-induced clinical remission and ongoing maintenance therapy, between 25% and 59% of patients with ANCA-associated vasculitis will relapse within 5 years. Risk factors for relapse include PR3 ANCA positivity (sHR 1.62) and cardiovascular disease (sHR1.59), whereas elevated creatinine at presentation (particularly above 200 μmol/l or 2.26 mg/dl) appears protective (64; 102).

Patients previously treated within the RAVE study (randomized, double-blind, placebo-controlled trial comparing rituximab to cyclophosphamide followed by azathioprine for induction of remission in patients with granulomatosis with polyangiitis and microscopic polyangiitis) were followed for nonsevere relapses between months 1 and 18. The patients developing relapses were more likely to have PR3-ANCAs, granulomatosis with polyangiitis, and previous relapses. An increase in prednisone dose led to remission in 80% of patients. However, only 30% of patients were able to maintain second remissions through a mean of 12.5 months, whereas 31 patients (70%) had a second disease relapse, nearly half of them with severe disease. The mean time to second relapse was 9.4 months. Therefore, increase of corticosteroids for nonsevere relapses in ANCA-associated vasculitis achieved only temporary remission, and recurrent relapses were common (62).

No markers predict relapses reliably. In the RAVE trial, increases in ANCA titer did not predict relapses; neither did B cell counts. However, as long as both B cells and ANCA were undetectable, the risk of relapse was low. In a long-term follow-up study, 38% of patients experienced a relapse (102). Patients who were anti-PR3 positive, had cardiovascular involvement, and did not have renal impairment at the time of diagnosis were at the highest risk of relapse.

In most studies, patients who were ANCA-negative or with alveolar hemorrhage on ventilatory support were excluded; therefore, data for these patient populations are sparse (88; 77; 87).

To evaluate long-term organ damage, data from 6 European Vasculitis Study group trials were combined (CYCAZAREM, MEPEX, CYCLOPS, IMPROVE, RITUXVAS, and NORAM). The extent of damage was assessed via the Vasculitis Damage Index (VDI). Renal, otolaryngologic, and treatment-related (cardiovascular, disease, diabetes, osteoporosis, and malignancy) damage increases over time, with around one third of patients having 5 or more items of damage at a mean of 7 years postdiagnosis. The most common disease-related symptoms were proteinuria, impaired glomerular filtration rate, hypertension, nasal crusting, hearing loss, and peripheral neuropathy. The most common treatment-related damage was hypertension, osteoporosis, malignancy, and diabetes (77). Lifelong monitoring and patient education are of paramount importance for positive long-term outcomes.

Complications and mortality in the rituximab studies. In the MAINRITSAN study, severe infections developed in 8 patients in the azathioprine group (14%) and in 11 patients in the rituximab group (19%); cancer developed in 2 patients in the azathioprine group and 1 in the rituximab group, and 2 patients in the azathioprine group died (1 from sepsis and 1 from pancreatic cancer) (32).

In the RITUXVAS trial, mortality rates were high (20% overall), and most deaths occurred within the first 3 months when patients were most vulnerable due to active disease and treatment exposure, including corticosteroids. Older age and more severe renal function likely contributed to the higher mortality rate (47). Infections continue to be a major concern contributing to morbidity and mortality, despite introduction of new therapies such as rituximab. In the early induction phase, where most adverse events occur, treatment with trimethoprim-sulfamethoxazole is recommended, independent of the type of immunosuppression, while on corticosteroids (51).

Treatment and outcome in the elderly. Because age is a negative prognostic factor and infections play a major role, studies have been investigating optimal treatment in the elderly, which may differ from younger populations. In an open-label, randomized, controlled trial including 104 patients older than 65 years of age who were newly diagnosed with ANCA-associated vasculitis, including EGPA, and panarteritis nodosa, a rapid corticosteroid dose tapering and limited cyclophosphamide exposure was investigated against conventional dosing (70). The experimental treatment group received corticosteroids for 9 months and a maximum of six 500 mg fixed-dose intravenous cyclophosphamide pulses every 2 to 3 weeks, followed by maintenance azathioprine or methotrexate. The control treatment included 26 months of corticosteroids with 500 mg/m2 intravenous cyclophosphamide pulses every 2 to 3 weeks until remission, then followed by maintenance.

The induction regimen limiting corticosteroid exposure and with fixed low-dose intravenous cyclophosphamide pulses reduced severe adverse events (79), mostly infections (32 [60%] versus 40 [78%] had > 1 SAE [P = 0.04]), in comparison to conventional therapy; there was no significant effect on remission rate. Three-year relapse rates and mortality were high for both groups. (Mortality was 6.7% during induction and reached 20% at 3 years.) SAEs occurred mainly during the first year and the causes of deaths were more likely due to serious adverse events than vasculitis (70).

Safety of rituximab in the elderly population was the goal of a retrospective review of 31 patients with ANCA-associated vasculitis with mean age older than 60 years (19 granulomatosis with polyangiitis, 12 microscopic polyangiitis, and 42% with relapsing disease) (96). Nearly all patients achieved remission (n=30), with a mean time to remission of 57 days. The mean prednisone dose at 6 months was 5.6 +/- 4 mg. Remission maintenance therapy was started within 12 months of rituximab in 6 patients (PR3 ANCA-positive and relapsing disease, 4 with rituximab, 1 with azathioprine, and 1 with mycophenolate mofetil). One-year survival among the 25 patients with available follow-up data was 100%. There were no episodes of infusion reaction or leukopenia. This suggests that rituximab is effective for remission induction in elderly patients with ANCA-associated vasculitis, though there was a high incidence of infectious complications (16.1%). Further prospective studies are needed to evaluate efficacy and safety (96).

Eosinophilic granulomatosis with polyangiitis. Compared to granulomatosis with polyangiitis and microscopic polyangiitis, EGPA has been considered a vasculitis with lower mortality but with remission rates similar to granulomatosis with polyangiitis and higher than that of microscopic polyangiitis (64). Based on data obtained from 2 prospective, randomized, clinical trials, survival rates at 1, 3, 5 and 7 years were: 98%, 94%, 92%, and 90%, respectively (81). As for granulomatosis with polyangiitis and microscopic polyangiitis, there is no reliable biomarker to measure EGPA activity (including ESR, CRP, eosinophilia) (19; 31).

Per the 2015 task force (31), disease remission is defined as the absence of clinical symptoms and biological abnormalities in patients on small doses of prednisone (no cutoff provided) or immunosuppressant doses. Ear-nose-throat manifestations or asthma flares are monitored separately as they may not necessarily reflect vasculitic activity. A relapse is defined as new, recurrent, or worsening of clinical EGPA manifestations (excluding asthma or ear-nose-throat manifestations), resulting in dose adjustments of the immunosuppressive regimen (31).

Among the patients with EGPA, ANCA-positive patients had a lower mortality (11) but more relapses (81). Age was significantly associated with mortality, and patients with cardiomyopathy seemed to have poorer survival rates (both included in FFS) (11). However, the differences between phenotypes based on presence or absence of ANCA has not been studied prospectively for a large cohort of EGPA patients (81). Relapse rates have been described around 41%, mostly affecting the peripheral nervous system, with 57% of relapses occurred when corticosteroid-tapering reached less than 10 mg/day. With treatment modification, remissions were achieved in more than 90% of the patients, but 38% continued to relapse.

At the last follow up after 7 years, most patients (more than 80%) were still treated with corticosteroids, resulting in treatment-related damage. Furthermore, 45% had peripheral neuropathy with disabling symptoms (81).

Special considerations

Pregnancy

Pagnoux described 8 pregnancies in 4 patients with granulomatosis with polyangiitis, 2 in a patient with microscopic polyangiitis, and 6 in 3 patients with EGPA (68). All mothers survived and live birth children were healthy. There was one first trimester miscarriage in each disease group. There was 1 therapeutic abortion at 8 weeks (granulomatosis with polyangiitis, receiving cyclophosphamide) and 1 at 5 weeks (EGPA, who had just completed an eventful pregnancy before). Four pregnancies (3 patients with granulomatosis with polyangiitis and 1 with microscopic polyangiitis) were uneventful from an obstetrical perspective; 2 had premature rupture of membranes with preterm births (one GPA and one EGPA). A twin pregnancy of an EGPA patient, with heart involvement and an ejection fraction of 53% at pregnancy onset, was complicated by sudden maternal cardiac failure at 32 weeks, leading to caesarean section. The newborns were healthy, and the mother recovered after 5 days of supportive care. Rescue caesarean was also required at 32 weeks for a patient with granulomatosis with polyangiitis with impaired renal function due to thrombotic microangiopathy (68). Medications such as azathioprine and corticosteroids were continued, and cyclophosphamide or methotrexate were discontinued (68). The risk of new vasculitis relapse during pregnancy is estimated as less than 10% to 40% with granulomatosis with polyangiitis in remission at conception (up to 100% in patients with active disease), 25% to 50% in EGPA (less if in remission at conception), and less than 50% in microscopic polyangiitis (but few data are available) (69).

Pregnancy outcome in patients with systemic vasculitis was compared to a group of women matched for age and ethnicity: 9 patients with GPA and 3 with EGPA (82).

The only significant finding was that the median gestational age was significantly lower in the vasculitis group (data reaching significance was also including other types of vasculitis): 36 weeks versus 40 weeks in the control group (P< 0.03). Three patients with GPA and 1 with EGPA had a relapse during pregnancy, which was mild to moderate in nature; 3 with GPA and 1 with EGPA relapsed after pregnancy. These findings suggest a higher risk of preterm birth correlate with the findings of Pagnoux (68) though there was no significant increase of miscarriages compared to the control group in this study, as shown by Pagnoux (82).

Overall, it is recommended that patients be well-controlled and in remission at the time of conception. Patients should be monitored closely during pregnancy by a high-risk obstetrician (68). Most maternal complications during pregnancy seem to be due to vasculitis damage (subglottic or bronchial stenoses) and hypertension in granulomatosis with polyangiitis, renal insufficiency in microscopic polyangiitis, and asthma or cardiac damage in EGPA.

Cyclophosphamide and azathioprine are both associated with pregnancy risk factor D. Azathioprine and cyclophosphamide can cause congenital anomalies. Cyclophosphamide may cause sterility in males and females (which may be irreversible), and women of childbearing potential should avoid pregnancy while receiving cyclophosphamide and for up to 1 year after completion of treatment. In young women or men with childbearing plans, rituximab should be considered as induction therapy to preserve ovarian (or testicular) function (18; 87).

Anesthesia

Given the frequency of respiratory involvement, including subglottic stenosis, general anesthesia should not be administered without prior pulmonary function tests, ear-nose-throat consultation, and chest radiograph.