Article Text

EULAR recommendations for the management of ANCA-associated vasculitis: 2022 update
  1. Bernhard Hellmich1,
  2. Beatriz Sanchez-Alamo2,
  3. Jan H Schirmer3,
  4. Alvise Berti4,5,
  5. Daniel Blockmans6,
  6. Maria C Cid7,
  7. Julia U Holle8,
  8. Nicole Hollinger1,
  9. Omer Karadag9,
  10. Andreas Kronbichler10,11,
  11. Mark A Little12,
  12. Raashid A Luqmani13,
  13. Alfred Mahr14,
  14. Peter A Merkel15,
  15. Aladdin J Mohammad11,16,
  16. Sara Monti17,18,
  17. Chetan B Mukhtyar19,
  18. Jacek Musial20,
  19. Fiona Price-Kuehne11,
  20. Mårten Segelmark21,
  21. Y K Onno Teng22,
  22. Benjamin Terrier23,
  23. Gunnar Tomasson24,25,
  24. Augusto Vaglio26,
  25. Dimitrios Vassilopoulos27,
  26. Peter Verhoeven28,
  27. David Jayne11
  1. 1 Klinik für Innere Medizin, Rheumatologie und Immunologie, Medius Kliniken, Akademisches Lehrkrankenhaus der Universität Tübingen, Kirchheim unter Teck, Germany
  2. 2 Nephrology Department, Hospital Universitario del Sureste, Arganda del Rey, Spain
  3. 3 Rheumatology & Clinical Immunology and Cluster of Excellence Precision Medicine in Chronic Inflammation, University Medical Center Schleswig-Holstein, Campus Kiel, Kiel, Germany
  4. 4 CIBIO, Universita degli Studi di Trento, Trento, Italy
  5. 5 Rheumatology, Santa Chiara Hospital, Trento, Italy
  6. 6 Department of Internal Medicine, University Hospital of Leuven, Leuven, Belgium
  7. 7 Department of Autoimmune Diseases, Hospital Clinic, University of Barcelona, Institut d’Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), Barcelona, Spain
  8. 8 Rheumazentrum Schleswig-Holstein Mitte, Neumuenster, Germany
  9. 9 Division of Rheumatology, Department of Internal Medicine, Vasculitis Research Center, Hacettepe University School of Medicine, Anakra, Turkey
  10. 10 Department of Internal Medicine IV, Medical University, Innsbruck, Austria
  11. 11 Department of Medicine, University of Cambridge, Cambridge, UK
  12. 12 Trinity Health Kidney Centre, Trinity College Dublin, Dublin, Ireland
  13. 13 Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Science (NDORMs), University of Oxford, Oxford, UK
  14. 14 Klinik für Rheumatologie, Kantonspital St Gallen, St Gallen, Switzerland
  15. 15 Division of Rheumatology, Department of Medicine, Division of Epidemiology, Department of Biostatistics, Epidemiology, and Informatics, University of Pennsylvania, Philadelphia, Pennsylvania, USA
  16. 16 Department of Clinical Sciences, Lund University & Department of Rheumatology, Skåne Hospital, Lund, Sweden
  17. 17 Department of Internal Medicine and Therapeutics, Università di Pavia, Pavia, Italy
  18. 18 Division of Rheumatology, Fondazione IRCCS Policlinico San Matteo, Pavia, Italy
  19. 19 Vasculitis Service, Rheumatology Department, Norfolk and Norwich University Hospital NHS Trust, Norwich, UK
  20. 20 2nd Department of Internal Medicine, Jagiellonian University Medical College, Kraków, Poland
  21. 21 Division of Nephrology, Department of Clinical Sciences, Lund University, Skane University Hospital, Lund, Sweden
  22. 22 Centre of Expertise for Lupus-, Vasculitis-, and Complement-Mediated Systemic Autoimmune Diseases (LuVaCs), Department of Internal Medicine, Section Nephrology, Leiden University Medical Centre, Leiden, The Netherlands
  23. 23 National Referral Center for Rare Systemic Autoimmune Diseases, Université Paris Descartes, Hôpital Cochin, Assistance Publique-Hôpitaux de Paris (APHP), Paris, France
  24. 24 Department of Epidemiology and Biostatistics, Faculty of Medicine, University of Iceland, Reykjavik, Iceland
  25. 25 Department of Rheumatology and Centre for Rheumatology Research, University Hospital Reykjavik, Reykjavik, Iceland
  26. 26 Nephrology Unit, Meyer Children’s Hospital, and Department of Biomedical, Experimental and Clinical Science, University of Florence, Florence, Italy
  27. 27 2nd Department of Medicine and Laboratory, Clinical Immunology-Rheumatology Unit, National and Kapodistrian University of Athens, School of Medicine, Hippokration General Hospital, Athens, Greece
  28. 28 Dutch Patient Vasculitis Organization, Haarlem, The Netherlands
  1. Correspondence to Prof Dr Bernhard Hellmich, Klinik für Innere Medizin, Rheumatologie und Immunologie, Medius Kliniken, Akademisches Lehrkrankenhaus der Universität Tübingen, Kirchheim unter Teck 73230, Germany; b.hellmich{at}


Background Since the publication of the EULAR recommendations for the management of antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) in 2016, several randomised clinical trials have been published that have the potential to change clinical care and support the need for an update.

Methods Using EULAR standardised operating procedures, the EULAR task force undertook a systematic literature review and sought opinion from 20 experts from 16 countries. We modified existing recommendations and created new recommendations.

Results Four overarching principles and 17 recommendations were formulated. We recommend biopsies and ANCA testing to assist in establishing a diagnosis of AAV. For remission induction in life-threatening or organ-threatening AAV, we recommend a combination of high-dose glucocorticoids (GCs) in combination with either rituximab or cyclophosphamide. We recommend tapering of the GC dose to a target of 5 mg prednisolone equivalent/day within 4–5 months. Avacopan may be considered as part of a strategy to reduce exposure to GC in granulomatosis with polyangiitis (GPA) or microscopic polyangiitis (MPA). Plasma exchange may be considered in patients with rapidly progressive glomerulonephritis. For remission maintenance of GPA/MPA, we recommend rituximab. In patients with relapsing or refractory eosinophilic GPA, we recommend the use of mepolizumab. Azathioprine and methotrexate are alternatives to biologics for remission maintenance in AAV.

Conclusions In the light of recent advancements, these recommendations provide updated guidance on AAV management. As substantial data gaps still exist, informed decision-making between physicians and patients remains of key relevance.

  • granulomatosis with polyangiitis
  • systemic vasculitis
  • rituximab
  • cyclophosphamide

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The antineutrophil cytoplasmic antibody (ANCA)-associated vasculitides (AAV) include granulomatosis with polyangiitis (GPA), microscopic polyangiitis (MPA) and eosinophilic GPA (EGPA).1–3 AAV represent a subgroup within the spectrum of primary systemic vasculitis defined by the Chapel Hill consensus conference nomenclature.4

In 2009, the EULAR developed its first recommendations for managing small and medium vessel vasculitis.5 An update focusing on AAV was published in 2016.6 These recommendations provided guidance to clinicians and researchers and have been widely cited. Recent landmark studies on the role of plasma exchange (PLEX), standardisation of glucocorticoid (GC) dosing, use of rituximab (RTX) for maintenance therapy, C5a receptor (C5aR)-targeted and anti-interleukin 5 (IL-5) therapy in EGPA make this an opportune time to update the 2016 guidelines.

These recommendations address the diagnosis and treatment of adult patients with AAV and are intended to give advice to clinicians, other health professionals, pharmaceutical companies and regulatory organisations.


The recommendations were drafted according to the 2014 update of the EULAR standardised operating procedures for the development of EULAR-endorsed recommendations7 and the updated version of the Appraisal of Guidelines for Research & Evaluation recommendations,8 where applicable (see online supplemental file 1 for a full description of methods). The task force consisted of 20 clinical experts including rheumatologists (MCC, BH, JH, OK, RAL, AJM, CBM, JM, PM, GT, DV), internists (AM, DB, BT) and nephrologists (AK, MAL, MS, YKOT, AV, DJ), from 15 European countries and the USA (PM), 2 methodologists (RAL; GT), convenor (BH) and co-convenor (DJ), 2 delegates of the EULAR young rheumatologists’ network EMEUNET (AB, SM), 2 fellows (BS-A, JHS), 1 health professional (NH) and 2 patient representatives (PV, FP-K).

Supplemental material

Based on results of a Delphi survey among the task force, we defined 14 key research questions addressing the management of AAV. For the update domains, the systematic literature review (SLR) was restricted to literature published from first of February 2015 (the date of the last set of recommendations) onwards. For new domains and drugs not included in the last update, the search was unrestricted. The following databases were used: PubMed, EMBASE and Cochrane Library. Each article was assigned a level of evidence according to the standards of the Oxford Centre for Evidence-Based Medicine (2009) and was systematically assessed for bias.7 The methods and results of the SLR are published separately.9 10

During a face-to-face meeting, task force members independently voted on each recommendation. Agreement on each recommendation and on the overarching principles on a scale of 0–10 (10 meaning full agreement) was given anonymously after the meeting by electronic mail. A research agenda was formulated based on controversial issues and evidence gaps. The final manuscript was approved by the EULAR Executive Committee.


General aspects

Definitions of disease activity states in AAV differed across clinical trials. For the purpose of these recommendations, we propose consensus definitions for disease activity states in AAV (table 1), which are based on the concept of activity states developed for the EULAR recommendations for conducting clinical trials in AAV11 that have been validated for use in clinical trials.12 American College of Rheumatology (ACR)/EULAR criteria for treatment response in AAV are now in development, which are expected to replace these definitions in the future.

Table 1

EULAR consensus definitions for disease activity states in AAV

Patients with AAV have previously been subdivided into those with ‘severe’ and ‘non-severe’ disease, or ‘generalised’ versus ‘non-generalised’, and some guidelines have adopted this categorisation.13–16 However, the terms ‘severe/non-severe’, ‘limited’ or ‘early-systemic’ are variably defined and misleading in clinical practice. Patients who appear to have less severe disease and may receive less intense treatment yet are at risk of developing organ-threatening or life-threatening manifestations.17 18 In most recent randomised controlled clinical trials (RCTs), this concept has been discarded and patients with different stages of disease severity were assigned the same intensity of induction treatment.19–21 As patients with ‘non-severe’ AAV are at risk of being undertreated, this task force decided not to change the categorisation of the 2016 recommendations that distinguishes patients with and without organ-threatening or life-threatening disease (table 2), instead of adapting the terminology of ‘severe’ and ‘non-severe’ AAV.

Table 2

Examples of organ/life-threatening and not organ/life-threatening manifestations in patients with AAV

Overarching principles

In line with other recent EULAR recommendations,22–24 general principles deemed fundamental for the management are now added to the AAV recommendations (table 3). These principles were consensus based and did not result directly from the SLR. Statements 1, 13, 14 and 15 of the 2016 update addressed topics based on low-quality evidence specific to AAV. Therefore, these statements have been moved into overarching principles B, C and D, while the content remains mostly unchanged.

Table 3

EULAR recommendations for the management of AAV—2022 update

Table 4

Glucocorticoid dosing (mg/day, prednisolone equivalent) with rituximab or cyclophosphamide-based regimens for remission induction in GPA or MPA according to the PEXIVAS Study93

A. Patients with AAV should be offered best care which must be based on shared decision-making between the patient and the physician considering efficacy, safety and costs.

This highlights the importance of shared decision-making between patients and physicians. Adherence to effective therapies is crucial to prevent permanent organ damage related to uncontrolled inflammation in AAV. Therefore, the committee considers efficacy, safety and tolerability as important factors in the decision-making process. This includes other factors such as kidney or liver function, fertility and pregnancy, lifestyle/smoking habits or concomitant interacting medications. Costs of treatment also need to be considered as access to expensive medication may be restricted in some countries.

B. Patients should have access to education focusing on the impact of AAV and its prognosis, key warning symptoms and treatment (including treatment-related complications).

Patients with AAV should be given a clear explanation of the nature of their disease, the treatment options, side effects of treatment, and their short-term and long-term prognosis. This statement is unchanged from the 2016 update (formerly statement no. 14) and has been moved to an overarching principle. Structured education programmes in patients with AAV increase knowledge in areas such as treatment and side effects.25 26 Patients should be informed on how to reach a vasculitis patient organisation.

C. Patients with AAV should be periodically screened for treatment-related adverse effects and comorbidities. We recommend prophylaxis and lifestyle advice to reduce treatment-related complications and other comorbidities.

Statements 11, 13 and 15 of the 2016 update have been transferred to principle C. As the use of cyclophosphamide (CYC) is associated with an increased risk of bladder cancer,27 all patients treated with CYC should have periodical urinalysis for the duration of their follow-up. In the presence of haematuria confirmed on urine microscopy that is not due to glomerulonephritis, a urology opinion must be sought. In common with other chronic inflammatory diseases, increased cardiovascular risk for patients with AAV is not explained by traditional risk factors alone and the risk of cardiovascular events is related to the burden of AAV disease activity.28 29 Additionally, as a result of damage due to AAV and its treatment, the frequency of cardiovascular risk factors such as diabetes and hypertension is increased.30 Therefore, both adequate control of vascular inflammation, and screening for and treatment of traditional cardiovascular risk factors, are important.31 Screening for and management of other treatment-related and disease-related comorbidities, such as osteoporosis or chronic kidney disease, should also be conducted. While the available evidence is insufficient to recommend an AAV-specific evaluation of comorbidities, several EULAR and other recommendations31–35 provide general guidance.

D. AAV are rare, heterogeneous, and potentially life-threatening and organ-threatening diseases and thus require multidisciplinary management by centres with, or with ready access to, specific vasculitis expertise.

This is based on statement 1 of the 2016 recommendations. Since AAV are rare, expertise in their management is more likely to be available in specialised centres. Accurate diagnosis, assessment of disease severity and differentiation between active vasculitis, infection and other complications or comorbidities can be challenging and often require rapid and low-threshold access to multidisciplinary diagnostic evaluation and treatment. In view of the limited number of formally approved therapies, access to treatment with novel drugs within clinical trials can be important, particularly in patients with relapsing or refractory AAV. Appropriately trained nurses and other healthcare providers experienced in AAV can support patients and provide education. These and other services can be bundled in dedicated vasculitis centres, such as the vasculitis centres within the European Reference Network for rare immune disorders (www.ern-rita-org). Better outcomes of patients with centre-based management compared with earlier cohorts have been reported from single-centre cohorts,36–38 but high-quality evidence on this topic is still lacking.


1. A positive biopsy is strongly supportive of a diagnosis of vasculitis and we recommend biopsies to assist in establishing a new diagnosis and for further evaluation for patients suspected of having relapsing vasculitis.

This recommendation (former statement no. 2), and additional guidance related to the role of biopsies outlined in the 2016 update,6 have not been revised, as the key evidence supporting this recommendation is unchanged. In addition to supporting a clinical diagnosis, biopsies (particularly from the kidney) can be helpful for distinguishing active disease from damage as the cause of clinical decline. A clinicopathological renal risk score gives prognostic information for end-stage kidney disease (ESKD) but histopathological subtypes are insufficient to guide treatment decisions.39–44 Repeat kidney biopsy may differentiate recurrent or refractory disease activity from damage or alternative diagnoses.45

This task force acknowledges that it may not be feasible to obtain a biopsy in every patient with suspected AAV, and initiation of treatment should not be delayed while awaiting histological information.46 Barriers to biopsies may include difficulty accessing tissue (eg, retro-orbital mass in GPA), unjustified risk of procedure (eg, patients who are on anticoagulant therapy) and anticipated low yield (eg, the diagnostic sensitivities of upper airway and transbronchial biopsies are only 30% and 12%, respectively).47 48 In patients with pulmonary lesions that cannot be clearly attributed to active AAV, thoracoscopic or open lung biopsies can be considered.49–51 When obtaining or interpreting a biopsy is challenging, surrogate markers can support a clinical diagnosis of AAV that is based on a typical clinical presentation and positive proteinase 3 (PR3)-ANCA or myeloperoxidase (MPO)-ANCA serology.52 Such surrogate parameters can be either clinical (such as mononeuritis multiplex confirmed by electrophysiological studies), laboratory data (such as red blood cell casts in the urine suggestive of glomerulonephritis) or findings on imaging.52

No studies have investigated the diagnostic accuracy of imaging compared with a definite clinical diagnosis or a positive biopsy in AAV.10 Therefore, no evidence-based recommendations on the use of imaging for the diagnosis of AAV can be made. However, imaging is recommended as an integral part of the diagnostic evaluation to detect organ involvement and to identify potential biopsy sites. CT of the chest is more sensitive than conventional radiographs and helps to distinguish disease manifestations of AAV from infection and other comorbidities,53–56 and to detect interstitial lung disease in patients with MPA.54 55 MRI can detect central nervous system lesions, pachymeningitis, retro-orbital lesions, or subglottic inflammation in GPA or cardiac disease in EGPA.57–60 F-18-fluorodeoxyglucose positron emission tomography with CT allows detection of occult sites of disease activity, concomitant malignancy and chronic infection.61–63 Endoscopy contributes to the management of certain organ-specific manifestations, such as subglottic or bronchial stenosis, or vasculitis of the gastrointestinal tract.64–66 Bronchoalveolar lavage contributes to the evaluation of pulmonary infiltrations, particularly alveolar haemorrhage or eosinophilic alveolitis, and microbiological analysis of the lower respiratory tract.

2. In patients with signs and/or symptoms raising suspicion of a diagnosis of AAV, we recommend testing for both PR3-ANCA and MPO-ANCA using a high-quality antigen-specific assay as the primary method of testing.

This recommendation was added due to the increasing relevance of ANCA for the diagnosis and classification of AAV and new data on the methodology of ANCA testing. ANCA is detectable in most patients with newly diagnosed GPA and MPA and contributes to the diagnosis. Although ANCA is a sensitive and specific tool to support a diagnosis of AAV, the diagnosis should not be made on ANCA serology alone, as ANCA can be found in other inflammatory diseases and infections, or may be drug induced.67 68 Antigen-specific immunoassays have better diagnostic accuracy than indirect immunofluorescence (IIF).69 The 2017 international consensus statement on testing of ANCA in GPA and MPA recommended high-quality immunoassays for PR3 and MPO-ANCA as the preferred screening method for diagnosis.67 If the immunoassay is negative, but the clinical suspicion for AAV is still high, a second test (either another immunoassays and/or IIF) is advised. A negative ANCA does not exclude a diagnosis of AAV, as a small proportion of patients with disease limited to the respiratory tract, or with renal-limited vasculitis, are ANCA negative.70 The 2017 international consensus statement contains detailed advice regarding other aspects of ANCA testing in GPA and MPA, such as indications for testing, the role of antibody levels and laboratory methodology. Additional testing for antibodies against glomerular basement membrane (anti-GBM) is advisable in the context of pulmonary-renal syndrome, as patients with anti-GBM/AAV overlap have a lower renal survival71 72 and may benefit from routine use of PLEX.

With a prevalence of 30% at diagnosis, ANCA is less frequent in patients with EGPA, in whom MPO-ANCA is the predominant serotype.73 74 EGPA with PR3-ANCA shares clinical features with GPA.75 A genome-wide association study reported that ANCA-positive and ANCA-negative EGPA are genetically different syndromes.73 Glomerulonephritis and neuropathy occur more frequently in ANCA-positive EGPA, while pulmonary infiltrates and cardiomyopathy are more frequent in ANCA-negative patients.73 The international consensus statement on testing of ANCA in EGPA stated that the presence of MPO-ANCA is neither sensitive nor specific enough to identify whether a patient should be subclassified as having ‘vasculitic’ or ‘eosinophilic’ EGPA.76 Furthermore, no differences in response to treatment between ANCA-positive and ANCA-negative patients were seen in two recent RCTs examining the use of RTX or mepolizumab in EGPA.77 78

ANCA serology is also relevant for the subclassification of AAV. In a large multicentre cohort study, PR3-ANCA was detected in 84%–85% of patients with GPA and 2%–27% of patients with MPA, while MPO-ANCA was found in 16% of patients with GPA and 75%–97% with MPA.79 Patients with PR3-ANCA and MPO-ANCA have distinct genetic backgrounds and differ in the frequency of some clinical manifestations, relapse rates and other clinical outcomes.79 80 The 2012 Chapel Hill consensus conference recommended adding the prefix to the name to indicate ANCA reactivity (ie, MPO-ANCA, PR3-ANCA or ANCA-negative), while the presence of PR3 or MPO-ANCA is weighted highly in the 2022 ACR/EULAR classification criteria for GPA, MPA and EGPA.1–3 Therefore, ANCA serotype is emerging as a key clinical classification criterion.

3. For induction of remission in patients with new-onset or relapsing GPA or MPA with organ-threatening or life-threatening disease, we recommend treatment with a combination of GCs and either RTX or CYC. RTX is preferred in relapsing disease.

The two major changes of this recommendation wording are the recommendation for a preferential use of RTX in relapsing GPA or MPA and the exclusion of EGPA, for which separate recommendations have been created.

Recent trials of induction therapy with CYC-based or RTX-based regimens in GPA and MPA included both new-onset and relapsing patients.19 81 In the largest trial comparing RTX and CYC for remission induction, remission rates at 6 and 12 months in relapsing patients were higher for RTX. This superiority of RTX over CYC did not extend to month 18,81 82 probably because there was no maintenance treatment in the RTX arm, whereas patients in the CYC arm were switched to receive azathioprine (AZA) for 12–15 months. Therefore, we favour treatment with RTX in relapsing patients (figure 1). The recently published data from the induction part of the RITAZAREM Study have shown that RTX can effectively restore remission in patients with relapsing AAV.20 There are limited data on use of CYC in patients relapsing after induction with RTX and the risk of malignancy increases when repeated courses of CYC are given.27

Figure 1

The 2022 EULAR algorithm for treatment of granulomatosis with polyangiitis (GPA) or microscopic polyangiitis (MPA). Dashed lines indicate supplementary action to consider. GC doses are provided as prednisolone equivalent. 1See table 2 for examples of organ/life-threatening and not organ/life-threatening manifestations. 2See table 4 and recommendation no. 5 for details and consider lower starting dose of 0.5 mg/kg/day in individual patients without organ-threatening or life-threatening manifestations. 3As part of a strategy to substantially reduce exposure to GCs (see recommendation no. 6 for details). 4Prefer RTX over CYC in relapsing disease and patients (m/f) with childbearing potential or previous exposure to CYC at an individual cumulative dosage considered to be associated with an increased risk of complications. 5In selected patients with serum creatinine >300 µmol/L due to active glomerulonephritis, plasma exchange may be considered taken into account individual risk for end-stage kidney disease and patient preferences. 6Stop avacopan after duration of treatment of 6–12 months; there are no data on use of avacopan beyond 1 year, so longer-term use cannot be recommended. AZA, azathioprine; CYC, cyclophosphamide; GC, glucocorticoid; MMF, mycophenolate mofetil; MTX, methotrexate; RPGN, rapid progressive glomerulonephritis; RTX, rituximab.

In new-onset GPA or MPA, RTX was non-inferior to CYC for induction of remission in two high-quality RCTs.81 83 Additional RCTs comparing both of these agents for induction of remission in new-onset GPA or MPA have not been published since the last update, nor have new data been released showing differences in long-term outcomes between them. There has been an increasing preference for RTX over CYC, mostly because of concerns about long-term safety of CYC.84 As CYC reduces ovarian reserve and increases the risk of premature ovarian failure and male infertility,85 86 RTX is preferable in patients who wish to preserve their reproductive potential. CYC has been associated with development of bladder cancer, bone marrow failure, myelodysplastic syndrome and other malignancies.27 87–90 The use of RTX is lowering CYC exposure and reducing the risk of malignancy in patients with AAV.91

A recent meta-analysis that included retrospective studies found that efficacy and safety outcomes do not differ between the RTX protocol used in the RAVE trial (375 mg/m2 per week for 4 weeks), which is approved for induction of remission in GPA and MPA in the European Union and the two-dose protocol (1 g in weeks 0 and 2) approved for rheumatoid arthritis.92 Recent retrospective studies found similar efficacy of RTX and RTX biosimilars in patients with AAV.93–95 Until recently, experience with RTX without concomitant CYC in patients with severe kidney failure has been limited to data from retrospective studies.96 The recent PEXIVAS Study included patients with severe renal disease and/or diffuse alveolar haemorrhage (DAH) treated with RTX and outcomes appear not to differ compared with CYC, but the study was not sufficiently powered to demonstrate non-inferiority of RTX over CYC in this subgroup.97 Although pharmacokinetics and mode of action of RTX do not suggest inferior efficacy in patients with renal failure or DAH, some task force members prefer CYC over RTX in this setting. No RCTs have assessed the benefit of RTX/CYC combination over RTX. However, the RTX/CYC combination has been shown to be CYC reducing in RITUXVAS,83 and retrospective studies98–101 have indicated the possibility of GC minimisation and improved responses that require investigation in an RCT (NCT03942887).

The MYCYC trial showed that mycophenolate mofetil (MMF) was non-inferior to CYC for remission induction in new-onset MPA or GPA.21 There was no safety benefit of MMF demonstrated and the study subjects in the MMF group who were PR3-ANCA positive had a much higher relapse rate.21 Thus, use of MMF for remission induction should be limited to situations where RTX and CYC are not tolerated or are contraindicated.

4. For induction of remission of non-organ-threatening or non-life-threatening GPA or MPA, treatment with a combination of GCs and RTX is recommended. Methotrexate (MTX) or MMF can be considered as alternatives to RTX.

In contrast to the 2016 update, this recommendation now includes RTX. Although, there are no RCTs comparing the use of RTX with other agents in patients with non-organ-threatening AAV, the RAVE trial and recent trials using RTX for induction therapy included such patients. Efficacy and safety outcomes were not inferior compared with those who had more severe disease at baseline.19–21 81

With respect to MTX and MMF, this statement refers to new-onset disease only. The MYCYC Study also included patients without organ-threatening manifestations.21 Another RCT comparing CYC and MMF in AAV found numerically lower disease-free survival rates in the MMF group at 2 and 4 years, respectively.102 Two smaller RCTs, primarily focusing on MPA, concluded equivalence of CYC and MMF for remission induction and safety.103 104 Given the probable lower long-term efficacy in patients with PR3-ANCA-positive AAV, the lack of superiority in safety and the lack of formal approval for use in AAV, there is insufficient evidence to support the routine use of MMF as a treatment of first choice for new-onset GPA or MPA over RTX or CYC. MMF can be considered as an alternative to RTX-based regimens, particularly in patients with intolerance or contraindications to RTX.

The NORAM trial comparing oral CYC versus oral MTX in new-onset GPA found no difference in remission rates and safety at 6 months.105 106 However, around 50% of patients in the MTX arm had either not attained remission or experienced a relapse by month 12 despite continued MTX and high GC exposure (starting dose 1 mg/kg, slow taper to 15 mg per day by month 3), had a shorter time to first relapse and developed additional relapse in the absence of maintenance therapy after month 12. Both NORAM and MYCYC employed GC regimens with higher doses than are currently recommended (see recommendation no. 5), which increased the chance of demonstrating non-inferiority. In contrast, recent data from the induction phase of the RITAZAREM trial showed that 66 of 69 patients with GPA or MPA without organ-threatening manifestation who were treated with RTX were in remission at month 4 despite the use of a lower-dose GC regimen with a starting dose of 30 mg prednisolone per day.20 In summary, the use of RTX over MTX or MMF should be considered in patients with GPA and MPA even without organ-threatening manifestations as RTX-based induction and remission regimens are associated with higher rates of sustained remission and lower GC exposure (see statement no. 5).

CYC is associated with long-term complications and should not be used as a first-line option in non-organ-threatening disease. It may be considered for remission induction in non-organ-threatening disease when the alternatives RTX, MTX and MMF cannot be used or are ineffective.

5. As part of regimens for induction of remission in GPA or MPA, we recommend treatment with oral GCs at a starting dose of 50–75 mg prednisolone equivalent/day, depending on body weight. We recommend stepwise reduction in GCs according to table 4 and achieving a dose of 5 mg prednisolone equivalent per day by 4–5 months.

Long-term follow-up of 535 patients with MPA or GPA and a broad spectrum of severity stages revealed an increased mortality ratio of 2.6 (95% CI 2.2 to 3.1) compared with an age-matched and sex-matched general population. The main causes of death within the first year were infection (48%) and active vasculitis (19%).107 High-dose GC contributes to the risk of infections,108–110 and patients are concerned about adverse effects of GC,111 thus reducing GC exposure in AAV without compromising control of vasculitis is a priority.

The PEXIVAS trial compared two GC taper regimens in 704 patients with GPA and MPA and active organ-threatening or life-threatening disease.97 The reduced-dose prednisone regimen (table 4) resulted in a 40% reduction in oral GC exposure in the first 6 months (figure 2); it was not inferior for the primary efficacy endpoint but led to a reduction of serious infections during the first year. We recommend tapering GC according to the PEXIVAS reduced GC regimen.112

Figure 2

Protocol target glucocorticoid (GC) doses in AAV induction trials81 106 113 221–226 (black line), illustrating how these compare with the reduced GC group from the PEXIVAS trial (red line). The line and error bars represent the mean and 95% CIs across a range of weights, genders and ages. AAV, antineutrophil cytoplasmic antibody-associated vasculitis.

The RITAZAREM trial of 190 patients with GPA/MPA at relapse permitted physician selection of either 0.5 mg/kg/day or 1.0 mg/kg/day starting dose of GC in conjunction with RTX.20 Although the GC dosing regimens were non-randomised, when patients were stratified for ‘major’ or ‘minor’ relapse, no differences in efficacy were seen for either severity subgroup between the two doses.20 A recent randomised, open-label multicentre trial in patients with predominantly MPA excluding severe kidney disease and/or alveolar haemorrhage compared a reduced GC starting dose (0.5 mg/kg) with a standard starting dose (1 mg/kg) for induction of remission in combination with RTX. At 6 months, remission rates were similar in both groups, but serious adverse events and infections occurred less frequently with the reduced dose. It is premature to give a general recommendation to use lower GC starting doses of 0.5 mg/kg for remission induction in all patients with active AAV. However, these data encourage further research of lower GC starting doses in cohorts with a broader spectrum of risk factors for unfavourable disease outcomes. For now, lower GC starting doses of 0.5 mg/kg/day may be considered on an individual basis in selected patients without life-threatening or organ-threatening disease.

Administration of intravenous methylprednisolone (MP) pulses in doses of 1000–3000 mg has been used in induction protocols including the RAVE, MEPEX and PEXIVAS trials81 97 113 and is common practice in many institutions, without an evidence base. No head-to-head trials have studied the role of MP pulses in AAV, the best available evidence being derived from indirect comparison across different trials. Observational studies have reported no efficacy benefit but increased rates of infections with the use of higher initial doses of GC, including MP pulses.108 109 114 115 Taken together, there is no compelling evidence to support the routine use of MP pulse therapy in addition to oral GC induction therapy, and there is a need for further research on this topic. In view of this limitation, and based on the evidence from PEXIVAS trial, MP pulse therapy should be limited to treatment of severe organ-threatening manifestations, particularly either active renal involvement with a documented estimated glomerular filtration rate (eGFR) of <50 mL/min/1.73/m2, or DAH.

6. Avacopan, in combination with RTX or CYC, may be considered for induction of remission in GPA or MPA as part of a strategy to substantially reduce exposure to GCs.

This is a new recommendation based on results of the ADVOCATE RCT in 331 patients with newly diagnosed or relapsing MPA or GPA that compared the use of the oral C5aR inhibitor avacopan (30 mg two times per day) with a GC regimen tapering from 1 mg/kg/day to 0 by 21 weeks (a GC withdrawal time similar to the RAVE trial81) as part of a standard induction protocol (RTX or CYC).19 The primary endpoint (remission at week 26) was reached at similar rates with avacopan (72.3%) and GCs (70.1%). Patients with active glomerulonephritis at baseline had greater recovery of kidney function compared with patients treated with GCs. The cumulative GC dose in the avacopan group over 1 year was 2.3 g lower than in the prednisone group, and GC-induced toxic effects measured by the Glucocorticoid Toxicity Index at week 26 were lower in the avacopan compared with the prednisone group. The incidence of adverse events, severe adverse events and infections was not different between groups. There are no data on use of avacopan beyond 1 year, so longer-term use cannot be recommended. We recommend consideration of avacopan in those subgroups that are likely to have enhanced benefit compared with GC therapy, that is, patients at risk of development or worsening of GC-related adverse effects and complications or patients with active glomerulonephritis and rapidly deteriorating kidney function who had better recovery of kidney function with avacopan.19

In ADVOCATE, remission sustained until week 52 (the second primary endpoint) was reached at a higher rate in the avacopan (65.7%) compared with the GC treatment groups (54.9%). Thus, avacopan appears to have efficacy for maintenance of remission. Future studies are needed to evaluate the role of avacopan for this purpose beyond 1 year, for patients presenting with a GFR <15 mL/min/1.73 m2, and for those with refractory disease, and whether avacopan can be stopped when RTX is given for maintenance of remission.

7. PLEX may be considered as part of therapy to induce remission in GPA or MPA for those with a serum creatinine >300 µmol/L due to active glomerulonephritis. Routine use of PLEX to treat alveolar haemorrhage in GPA and MPA is not recommended.

Compared with the 2016 update, the strength of recommendation supporting the use of PLEX for patients with active glomerulonephritis has been reduced (‘may be considered’ compared with ‘we recommend’). While the cut-off point serum creatinine qualifying for PLEX has been lowered from 500 to 300 µmol/L (3.41–5.68 mg/dL), the routine use of PLEX to treat alveolar haemorrhage in GPA and MPA is not routinely recommended.

The 2016 statement was based on the results of the MEPEX trial113 that included only patients with severe glomerulonephritis defined by a serum creatinine >500 µmol/L. In view of a meta-analysis,116 which suggested that the evidence supporting PLEX was not robust, but there may be benefit in less severe presentations, the PEXIVAS trial was conducted to evaluate the efficacy of PLEX as an adjunct to standard induction therapy in patients with newly diagnosed or relapsing MPA or GPA, with positive PR3 or MPO-ANCA who had active kidney involvement with an eGFR <50 mL/min/1.73 m2 or DAH.97 After a median follow-up of 2.9 years, no difference for the primary composite endpoint of death of any cause or ESKD was found between patients randomised to PLEX (28%) compared with those randomised to no PLEX (31%).

A meta-analysis of nine RCTs97 113 117–124 confirmed that PLEX had no effect on all-cause mortality.125 Outcome data for ESKD were reported in 999 patients of which 597 came from PEXIVAS, and a meta-analysis revealed that PLEX reduced the risk of ESKD at 12 months (relative risk 0.62 (95% CI 0.39 to 0.98)). As baseline serum creatinine predicts ESKD risk, subgroups based on baseline creatinine with low risk (≤200 µmol/L), low to moderate risk (>200–300 µmol/L), moderate to high risk (>300–500 µmol/L) and high risk (>500 µmol/L) were analysed. While little absolute risk reduction of ESKD was observed following use of PLEX in the low-risk and low moderate-risk groups, a 4.6% absolute reduction of ESKD at 12 months was estimated for the moderate high-risk group and 16.0% for the high-risk group.125 126 This translates into a number of patients to treat with PLEX of 21.7 for the moderate to high-risk group and 6.25 for the high-risk group to prevent one case of ESKD at 12 months. The impact of PLEX on ESKD risk diminished over a 3-year follow-up (relative risk 0.79 (95% CI 0.58 to 1.08)). PLEX increased the risk of serious infections at 12 months by 8.5% in the moderate high-risk group, and 13.5% in the high-risk group or patients, and no effect on quality of life was found.125 Thus, treating 14 patients with PLEX will result in one serious infection. Two retrospective studies involving 251 and 188 patients with AAV and severe kidney disease, respectively, found no efficacy of PLEX on death or ESKD.127 128

Thus, PLEX may reduce the risk of ESKD 12 months but may increase the risk of severe infection. This benefit declines over longer follow-up, suggesting that PLEX might prolong the time to dialysis. Balancing the reported benefit in a subgroup of patients at high risk of ESKD against the risk of severe infection, the cost and risks of the procedure, PLEX may be considered as an adjunctive treatment of GPA and MPA for selected cases with a serum creatinine >300 µmol/L, after discussion of the risks and benefits with the patient.

The SLR and recent meta-analysis revealed no evidence for a clinically relevant benefit of PLEX in patients with AAV and DAH.125 A small open-label study reported survival in 19 of 20 patients with DAH of whom 9 had severe DAH,129 while an observational study of 73 patients with DAH of which 34 required mechanical ventilation did not find a benefit of PLEX on mortality or other outcomes.130 In PEXIVAS, DAH was present in 191 patients and was associated with hypoxia in 61. No significant effect of PLEX on the combined endpoint of death from any cause or development of ESKD was found in these patients even after adjustment for the severity of DAH, but this substudy was underpowered for this endpoint, and further analysis of the impact of PLEX on DAH mortality is ongoing. As isolated DAH is rare in AAV, the driver for PLEX in those with DAH is usually the degree of associated renal impairment, and there is insufficient evidence to make a recommendation for or against PLEX in isolated DAH.

PLEX is recommended for those patients with AAV also positive for anti-GBM antibodies.131 Although high-quality evidence for this small subgroup is lacking, most clinicians follow management recommendations for both anti-GBM disease and AAV in their initial treatment of these dual-positive patients.132 There is low level of evidence derived from a prospective randomised trial that included 62 patients with either EGPA or polyarteritis nodosa for a lack of short-term and long-term efficacy of PLEX on remission and mortality in patients with EGPA.133 134

8. For patients with GPA or MPA with disease refractory to therapy to induce remission, we recommend a thorough reassessment of disease status and comorbidities and consider options for use of additional or different treatment. These patients should be managed in close conjunction with, or referred to, a centre with expertise in vasculitis.

Given new options, it is too narrow to limit the recommendation for refractory disease to switching from CYC to RTX or vice versa, as stated in 2016. Time to a treatment response varies individually in the early treatment phase (weeks 0–4). Raising the GC dose for some time can be reasonable strategy, particularly if only minor symptoms persist. The combination of RTX and CYC is used in patients with refractory organ-threatening or life-threatening disease by many centres, but data on this approach in true refractory AAV are lacking. Adding intravenous immunoglobulins can be an option for persistent disease manifestations, particularly in patients with increased risk of infection.135 No controlled studies on the management of refractory GPA or MPA have been published since the last update. Refractory disease is rare, and management should include review of the diagnosis and careful assessment of disease activity. Refractory AAV needs to be distinguished from infections, other comorbidities and alternative diagnoses. Therefore, patients with suspected refractory severe AAV should be managed at centres of expertise.

9. For maintenance of remission of GPA and MPA, after induction of remission with either RTX or CYC, we recommend treatment with RTX. AZA or MTX may be considered as alternatives.

This recommendation was changed towards favouring RTX in view of consistent results from two high-quality RCTs confirming a higher efficacy of RTX compared with AZA136 137 and other recent prospective trials on the use of RTX for maintenance of remission.138 139

In the MAINRITSAN trial, patients who attained remission after induction therapy with GC and CYC, repeat-dose RTX (500 mg two times at 6 months then 500 mg every 6 months three times) over 2 years was associated with a lower relapse rate than treatment with AZA, with comparable safety.136 Long-term data of this trial showed that the rate of sustained remission remained superior over 60 months with repeat-dose RTX, with better overall survival.140 First results (abstract) of another multicentre RCT (RITAZAREM) now confirm the higher efficacy of RTX (1 g every 4 months five times) compared with AZA for patients receiving RTX induction therapy for relapsing disease.137 Results of MAINRITSAN and RITAZAREM complemented each other with similar findings despite methodological differences (ie, type of induction therapy, duration and dose of AZA, inclusion of relapsing patients, RTX dose and dosing interval).9 RTX is considered cost-effective by preventing costs associated with the occurrence of relapses, particularly since RTX biosimilars have become available.141 In an RCT, ‘tailored’ RTX maintenance treatment based on biomarkers (rise of ANCA concentration, switch from negative to positive ANCA or repopulation of CD19+ lymphocytes) was associated with a higher but not statistically different relapse rate (17.3%) compared with the approved fixed regimen (9.9%).139 As the anticipated and observed relapse rates differed substantially, the trial was considered underpowered to exclude inferiority of the biomarker-triggered regimen. In view of these uncertainties, this task force favours the use of the 500 mg every 6 months RTX maintenance regimen. The higher dose of 1 g or shorter dosing interval of 4 months or both may be considered for patients who relapse on the 500 mg every 6 months regimen.

RTX impairs humoral responses to vaccination142–144 and there is increasing information concerning the risks of secondary immunodeficiency in patients with GPA/MPA receiving RTX.145 146 Patients should be counselled about the risk of hypogammaglobulinaemia (see also recommendation no. 14) and further research is required into the safety, duration and dosing of repeated RTX in this disease.

Evidence regarding the use of conventional immunosuppressive agents has not changed substantially since the last update. AZA and MTX are similarly effective maintenance agents in AAV147 and can be used if RTX is contraindicated (eg, previous allergic reaction to RTX) or appears inappropriate (eg, urgent need for vaccination, severe hypogammaglobulinaemia). Doses lower than those recommended for AZA and MTX (online supplemental table 1) have been associated with higher relapse rates.136 147 148 MTX can be continued in patients in whom it was used to induce remission. MMF was associated with a higher relapse rate compared with AZA in the only phase III RCT149 and can be considered in patients with intolerance or contraindications to RTX, AZA or MTX. In patients with GPA, leflunomide can be considered in patients with intolerance to all the above-mentioned drugs.150 Results of two recent meta-analyses revealed that trimethoprim–sulfamethoxazole (T/S) does not reduce relapse risk in patients with GPA.151 152 The addition of belimumab to an AZA-based maintenance regimen did not improve relapse-free survival in an RCT, which was stopped early due to slow recruitment and had a low relapse rate in the placebo group, making a positive result with belimumab unlikely to be detected.153

Since there is little evidence to guide low-dose GC therapy during remission in AAV,154 duration and dosage need to be individualised on a shared decision basis, taking into account the patient’s individual disease course, risk for or presence of GC-related comorbidities and patient preferences. There is lower-quality evidence that GC withdrawal increases relapse risk,154 but high-quality prospective studies on the role of GC are yet lacking. Regular screening for GC-related comorbidities during continued low-dose GC therapy is recommended according to EULAR recommendations for monitoring adverse events of low-dose GC therapy.34

10. We recommend that therapy to maintain remission for GPA and MPA be continued for 24–48 months following induction of remission of new-onset disease. Longer duration of therapy should be considered in relapsing patients or those with an increased risk of relapse, but should be balanced against patient preferences and risks of continuing immunosuppression.

For the 2016 update, this statement was based on low-quality evidence derived from observational studies or long-term follow-up of RCTs.105 155 156 Since then, three trials138 157 158 have directly compared the duration of maintenance regimens and this recommendation has therefore been changed accordingly. In the REMAIN Study, 117 patients received AZA for a total of 24 months after induction therapy with CYC. Patients were then randomised to withdrawal of AZA/GC or continued dosing for an additional 24 months. Those treated for 4 years had fewer relapses (22%) than those treated for 2 years (63%).157 Four patients in the withdrawal group developed ESKD versus none in the group treated for another 2 years (p=0.012). A meta-analysis of REMAIN and the smaller AZA-ANCA trial concluded that prolonged administration of AZA reduced the risk of relapse.152 Results of the open-label randomised MAINRITSAN-3 trial showed that more patients remain relapse free after an additional 18 months of RTX maintenance therapy than after treatment for only 18 months.138 Prolonged therapy with RTX was not associated with an excess of serious adverse events or infections.

The clinical disease type (GPA vs MPA), ANCA serotype (PR3-ANCA vs MPO-ANCA) and ANCA status (positive vs negative) have all been associated with the risk of relapse. In several studies, a higher risk of relapse was observed in patients with GPA than in patients with MPA.159–161 Regardless of the clinical phenotype, patients with positive PR3-ANCA at diagnosis are at higher risk of relapse than those who are MPO-ANCA positive.80 140 156 162 163 Persistent ANCA positivity despite clinical remission,164–166 or seroconversion from negative to positive ANCA,164 is also each associated with an increased risk of relapse. B cell repopulation within 12 months of RTX164 and persistent haematuria167 have been identified as risk factors for relapse in individual studies. The intensity of induction therapy also impacts the risk of relapse. Lower cumulative CYC doses or induction therapy with MTX or MMF instead of CYC have been associated with an increased risk of relapse.21 105 156 168

There is low-level to moderate-level evidence that patients with renal-limited vasculitis and patients positive for MPO-ANCA have a lower relapse risk compared with PR3-ANCA-positive patients and patients with respiratory tract involvement.169 170 In a series of 228 patients with ESKD, the proportion of patients with AAV in remission off immunosuppression had increased time spent on dialysis and patients were far less likely to relapse from their vasculitis than to display serious infectious or cardiovascular events.171 Therefore, the benefit of relapse prevention should be weighed against the risk of complications resulting from immunosuppressive therapy in patients with renal-limited MPO-ANCA-associated vasculitis and some task force members do not routinely use maintenance therapy in these patients. Patients with drug-induced AAV rarely relapse and do not require routine immunosuppressive therapy after remission is achieved and the implicated drug is discontinued.172

In summary, there is no consistent high-quality evidence available to guide decisions about the duration of maintenance therapy based on biomarkers such as ANCA or other factors alone (see also recommendation no. 14).9 While there is now consistent evidence from two RCTs that extending maintenance therapy for longer than 24 months reduces relapse risk, we recommend considering individual risk factors for relapse and damage as well as patient preferences for decisions about the length of maintenance treatment.

11. For induction of remission in new-onset or relapsing EGPA with organ-threatening or life-threatening manifestations, we recommend treatment with a combination of high-dose GCs and CYC. A combination of high-dose GCs and RTX may be considered as an alternative.

Since the last update of these recommendations, results of three RCTs enrolling only or mainly patients with EGPA have been reported,77 78 173 which allowed the development of separate recommendations for EGPA for this update (figure 3).

Figure 3

The 2022 EULAR algorithm for treatment of eosinophilic granulomatosis with polyangiitis (EGPA). 1See table 2 for examples of organ/life-threatening and not organ/life-threatening manifestations. 2See table 4 for an example of GC dosing (note: validated in MPA and GPA only). 3Consider use of RTX over CYC in patients (m/f) with childbearing potential or previous exposure to CYC at an individual cumulative dosage considered to be associated with an increased risk of complications. 4Individualised duration of maintenance treatment. 5In patients with relapsing or refractory EGPA without organ-threatening manifestations at the time of relapse, MEPO is preferred for maintenance of remission, and AZA, MTX or RTX can be used as alternatives if MEPO is not tolerated or ineffective. AZA, azathioprine; CYC, cyclophosphamide; GC, glucocorticoid; GPA, granulomatosis with polyangiitis; MEPO, mepolizumab; MPA, microscopic polyangiitis; MTX, methotrexate; RTX, rituximab.

The Five Factor Score (FFS) is used for prognostic assessment of EGPA.174 Particularly, cardiac involvement has been associated with increased mortality in EGPA.36 175 176 However, under optimised management in centres of expertise, the prognosis of cardiac involvement appears to be better than previously reported. This may reflect more frequent diagnosis of milder forms of cardiac disease through use of cardiac MRI and greater awareness among physicians regarding cardiac disease in EGPA.177 In a recent series from the French Vasculitis Study Group of 70 patients with EGPA with cardiac manifestations treated with high-dose GC, mostly along with CYC, no patient died as a consequence of cardiac involvement during a 10-year observation period.178 With the aim of preventing permanent organ damage due to EGPA, patients with severe involvement of the kidneys, central and peripheral nervous system, or gastrointestinal tract are also considered to be candidates for treatment with CYC (table 2).179 In a randomised, open-label trial in patients with EGPA and poor prognosis (FFS ≥1), 12 compared with 6 pulsed doses of CYC were associated with a lower rate of minor relapses but did not improve response rate or reduce severe relapses.180 Therefore, we recommend treatment be switched to a less intensive remission maintenance therapy after six pulses of CYC if remission is achieved, and the GC dose is reduced by then to approximately 7.5 mg per day (see online supplemental table 2 for protocols).

An RCT examining the use of RTX in EGPA (REOVAS) included 105 patients with new-onset or relapsing EGPA of whom 42 had life-threatening or organ-threatening disease (FFS ≥1) (abstract).77 Patients with FFS ≥1 received high-dose GCs plus either 2×1 g RTX (days 1 and 15) or nine pulses of CYC over 13 weeks. The primary endpoint of on-treatment remission was reached at similar frequencies at days 180 and 360 in both groups, but the limited number of patients, the superiority design and the lack of fully published results do not allow for strong conclusions regarding non-inferiority. Adverse events, cumulative prednisone doses and quality of life were not different between groups. Results were similar in both newly diagnosed and relapsing disease. In contrast to earlier observational studies,181 182 the response to RTX was not higher in MPO-ANCA-positive patients compared with ANCA-negative patients, consistent with consensus recommendations on ANCA testing that treatment decisions in EGPA should not be influenced solely by ANCA status.76 Keeping in mind that the results of the REOVAS trial have not been fully published yet, the data reported so far are deemed sufficiently strong to consider RTX as an alternative to CYC, particularly in patients in which exposure to CYC needs to be avoided, and are consistent with earlier observational reports (see recommendation no. 3).

In contrast to GPA and MPA, no studies have compared different GC tapering strategies in the treatment of EGPA. In the absence of data to support an evidence-based recommendation on GC tapering in EGPA, recommendations made for GPA and MPA (statement no. 4) can be used as an orientation. However, asthma and ear, nose and throat (ENT) exacerbation increase the GC requirement in patients with EGPA, leading to prolonged tapering.183 Therefore, interdisciplinary management involving pulmonologists and/or otorhinolaryngologists aimed at optimising treatment (including topical agents) of asthma, polyposis and sinusitis is recommended.

12. For induction of remission in new-onset or relapsing EGPA without organ-threatening or life-threatening manifestations, we recommend treatment with GCs.

Patients with EGPA without adverse prognostic factors (FFS=0) treated with GC only achieve remission >90% of the time, but relapses are common once GCs are tapered.174 184 Therefore, clinicians frequently combine GCs with other immunosuppressants or biologics. However, the SLR revealed that evidence supporting GC-sparing therapy in newly diagnosed patients with EGPA without organ-threatening or life-threatening manifestations is low.10 A prospective placebo-controlled study showed that therapy with AZA for 1 year in addition to GC had no effect on the risk of relapse, cumulative GC requirement, or the rate of asthma and sinusitis exacerbation compared with GC monotherapy in EGPA without poor prognostic factors (FFS=0).173 Recent long-term study data also showed that, within 5 years, 48% of all patients experienced vasculitis relapses, and prior therapy with AZA did not reduce this risk.184 The REOVAS trial included 63 new-onset or relapsing patients with EGPA with an FFS of 0 who were randomised to receive high-dose GCs together with either 2×1 g RTX (days 1 and 15) or placebo. Efficacy and safety outcomes after 180 and 360 days were not different between RTX and placebo group. Because the trial was not designed as a non-inferiority trial and since it has been reported only in abstract format so far, the data preclude from making strong conclusions but provide no support for use of RTX for remission induction in this subgroup of patients. No RCTs are available on the use of MTX, MMF or leflunomide in EGPA. Small observational studies on the use of MTX or MMF did not include control groups and carry a high risk of bias.185 186 As the evidence to support immunosuppression beyond GCs in new-onset EGPA without risk factors for worse outcome is low,10 decisions on the use of GC-sparing therapy in this subset of patients may be made on an individual basis considering risk factors of GC-related morbidity.

13. For induction of remission in patients with relapsing or refractory EGPA without active organ-threatening or life-threatening disease, we recommend the use of mepolizumab.

The IL-5 inhibitor mepolizumab was evaluated in a randomised double-blind placebo-controlled phase III study (MIRRA) that included 136 patients with relapsing or refractory EGPA with disease duration of at least 6 months.78 The study protocol allowed the inclusion of patients without vasculitic manifestations, while patients with active life-threatening or organ-threatening manifestations were excluded. After treatment with prednisolone at a stable dose ≥7.5 mg per day prior to baseline, 54% of patients in the mepolizumab arm and 71% of patients in the control arm had active disease at randomisation with a Birmingham Vasculitis Activity Score (BVAS) >0. Patients were randomised 1:1 to continuation of standard therapy (including other conventional immunosuppressive agents in more than 50% of patients in each arm) or standard therapy plus mepolizumab at a dose of 300 mg subcutaneously every 4 weeks. Both co-primary endpoints (the number of weeks in remission on a prednisolone dose reduced to 4 mg and the proportion of patients in remission at weeks 36 and 48) were met in favour of mepolizumab.78 A post-hoc analysis showed that treatment with mepolizumab was associated with additional clinically relevant endpoints, such as remission and GC reduction of >50%, in over half of the patients treated with mepolizumab.187 Based on its association with clinically meaningful improvement of disease control and reduction of GC demand, and good safety profile compared with conventional immunosuppressants, this task force recommends the use of mepolizumab with relapsing or refractory, non-organ-threatening or life-threatening EGPA. Mepolizumab has recently been approved for this indication in many European countries. Data from studies using mepolizumab for treatment of life-threatening or organ-threatening or new-onset EGPA are currently lacking.10

Other IL-5 or IL-5 receptor inhibitors (reslizumab, benralizumab) showed efficacy in small open-label pilot studies in EGPA,188 189 but data from RCTs using these agents are not yet available. In a retrospective multicentre series, anti-IgE-targeted therapy with omalizumab appeared to be less effective than mepolizumab.190 As discussed above, data showing improved outcomes with other biological or conventional drugs for non-severe relapsing or refractory EGPA are lacking. In patients for whom mepolizumab is not effective or not tolerated, AZA, MTX, MMF or RTX can be considered on an individual basis.184–186 190–192

A true refractory course of EGPA with life-threatening and organ-threatening manifestations is rare if patients are treated with high-dose GC and a CYC-based or RTX-based induction regimen.77 180 Data guiding treatment decisions in this small subgroup of patients are scarce, and true refractory severe EGPA needs to be carefully distinguished from infections and comorbidities. Therefore, patients with suspected refractory severe EGPA should be managed at centres of expertise.

14. For maintenance of remission of relapsing EGPA after induction of remission for non-organ-threatening or life-threatening manifestations at the time of relapse, we recommend treatment with mepolizumab. For maintenance of remission of EGPA after induction of remission for organ-threatening or life-threatening disease, treatment with MTX, AZA, mepolizumab or RTX should be considered.

In view of the relapsing nature of EGPA requiring long-term use of GCs in most patients, other agents for maintenance of remission are commonly prescribed in an attempt to be GC sparing. In an RCT (MIRRA, see recommendation no. 13 for details) that enrolled patients with EGPA who had relapsing or refractory disease, rates of severe and non-severe relapses were significantly lower and the median GC dose throughout the study was lower in the mepolizumab group compared with the control group.78 Adverse events occurred at similar rates in the mepolizumab group and the placebo group, while serious adverse events were somewhat more common in the placebo group (18% vs 26%). In view of its efficacy and good safety profile, the use of mepolizumab after induction of remission for non-organ-threatening or life-threatening manifestations at the time of relapse is recommended. In an RCT that enrolled 51 patients with EGPA and no organ-threatening or life-threatening manifestations, AZA for 1 year in addition to GC had no effect on the risk of relapse, cumulative GC requirement, or the rate of asthma and sinusitis exacerbation compared with GC monotherapy.173 There is little evidence to recommend routine use of other immunomodulatory agents for maintenance of remission in EGPA without organ-threatening or life-threatening manifestations.10

The SLR identified only one prospective study addressing remission maintenance strategies in patients with EGPA who attained remission after treatment for life-threatening or organ-threatening disease.10 A single-centre prospective randomised trial compared oral CYC with MTX for 1 year after remission induction with CYC in different subtypes of AAV.193 In the subgroup of 30 patients with EGPA who had either an FFS >1 or peripheral neuropathy, no difference in relapse rates between the two treatment arms was observed. Although no excess in adverse events was found in this study, we do not recommend CYC for remission maintenance in view of its toxicity. However, this study provides a rationale for use of MTX for maintenance of remission in EGPA, although the small sample size of patients with EGPA precludes a strong recommendation. As observational studies reported favourable outcomes on the use of AZA, mepolizumab and RTX for maintenance of remission,36 190 these agents can also be considered for remission maintenance in EGPA after induction of remission for organ-threatening or life-threatening manifestations. In view of its efficacy in eosinophilic asthma, mepolizumab should also be considered for patients with EGPA with residual GC-dependent asthma who achieved remission of major organ involvement.

15. In the management of patients with AAV, we recommend that structured clinical assessment, rather than ANCA and/or CD19+ B cell testing alone, should inform decisions on changes in treatment.

This recommendation (former statement no. 10) has been amended to include CD19+ B cell testing but is otherwise unchanged as data from recent RCTs have confirmed earlier studies on this topic. Although ANCA status is associated with relapse,165 194–196 prospective trials on maintenance of remission showed conflicting results in ANCA status or CD19+ B cell counts to predict future relapses at a level deemed insufficient to guide treatment decisions for individual patients.138 139 Administration of RTX for maintenance of remission based on changes of ANCA status and/or B cell counts was associated with a non-significantly higher rate of relapses compared with regular treatment at a 6-month interval, while adverse events occurred at a similar rate.139 Regarding the role of ANCA measurements for monitoring in AAV, we also refer readers to the recent international consensus statements on ANCA testing.67 76 Further prospective studies are clearly necessary to identify predictive markers of relapse.

As AAV involves multiple organs and relapses are frequent, a structured clinical assessment during follow-up at regular intervals is recommended. The BVAS197 has been used in different variants in the majority of RCTs in AAV and can be helpful in clinical practice to document response to treatment in a systematic fashion. Damage resulting from AAV or its treatment needs to be distinguished from active disease to avoid unnecessarily escalating treatment. The Vasculitis Damage Index198 is a validated instrument to record damage in AAV and provides definitions that help distinguish damage from active disease.

16. In patients with AAV receiving RTX, we recommend measurement of serum immunoglobulin concentrations prior to each course of RTX to detect secondary immunodeficiency.

The SLR revealed no data published since the last update that suggested a change of this recommendation,10 but wording has been rephrased to highlight the purpose of immunoglobulin measurement. Results of the MAINRITSAN-3 Study have shown that long-term treatment of patients with GPA or MPA with RTX over 36 months was associated with the development of hypogammaglobulinaemia (IgG <5 g/L) in 21% of patients, confirming earlier reports on decreased IgG levels following treatment with RTX or CYC in AAV.199 200 For further details on risk factors for secondary immunodeficiency after RTX, monitoring, indications, dosage and discontinuation of immunoglobulin replacement therapy, we refer readers to evidence-based consensus recommendations.201

17. For patients with AAV receiving RTX, CYC and/or high doses of GCs, we recommend the use of T/S as prophylaxis against Pneumocystis jirovecii pneumonia (PJP) and other infections.

This is a new recommendation based on results of an observational study in 192 patients with AAV treated with RTX, showing that the prophylactic use of T/S was associated with a lower frequency of severe infections (HR 0.30, 95% CI 0.13 to 0.69).202 In an earlier RCT investigating the role of T/S in therapeutic dosage (960 mg two times per day for 2 years), a reduction in respiratory tract infections and a trend towards fewer non-respiratory tract infections compared with placebo had been observed.203 Thus, available evidence suggests that T/S not only reduces the risk of PJP, but is also associated with a reduction of the overall risk of infection. T/S also reduced the 1-year incidence of PJP and related mortality in a cohort of 1092 patients with various rheumatic diseases treated with ≥30 prednisolone mg/day for ≥4 weeks.204 For patients treated with ≥15–<30 mg/day of prednisone for ≥4 weeks, the risk of PJP and benefit of T/S are lower, but in a subgroup of patients with lymphopenia at baseline and those receiving GC pulse treatment, the number needed to treat to prevent one PJP was lower than the number needed to harm by serious adverse events.205 As infections are the leading cause of death within the first year of induction therapy in patients with AAV,206 infection prophylaxis with T/S (800/160 mg on alternate days or 400/80 mg daily) is recommended for all patients with AAV receiving CYC or RTX and patients where treatment with GCs at a dose of ≥30 mg/day for 4 weeks or longer is envisioned, irrespective of other concomitant immunosuppressants. Although there have been concerns about synergistic toxicities to T/S given at therapeutic doses and MTX, recent studies found no evidence for an interaction between MTX and T/S at prophylactic doses,207 but data on the safety of this combination in patients with rheumatic diseases are lacking. While there are insufficient data to guide the total duration of prophylaxis with T/S, it seems reasonable to continue this drug for the estimated duration of the biological effect of CYC and RTX of around 3 and 6 months after the last dose or B cell reconstitution, respectively. For patients treated with GCs in combination with immunosuppressants other than CYC or RTX, T/S may be stopped once GC doses have been tapered to 15 mg/day, but that strong consideration should be given to continuing it until lower doses are achieved if other risk factors such as pulmonary disease or hypogammaglobulinaemia are present.208 Patients who develop adverse reactions often tolerate re-introduction of T/S if the dose is gradually increased according to published regimens.209 210 Alternatives for patients who cannot tolerate T/S are dapsone,211 atovaquone212 213 or aerosolised pentamidine.214

Vaccinations are an integral part of infection prophylaxis in patients with autoimmune diseases receiving immunosuppressive therapy. Since the approach to vaccination in patients with AAV does not differ from other rheumatic diseases treated with a similar intensity of immunosuppression, we refer to the 2019 EULAR recommendations for vaccination in adult patients with autoimmune inflammatory rheumatic diseases and local guidelines.215


Since the publication of the 2016 EULAR recommendations on the management of AAV, several high-quality RCTs have expanded our knowledge about these complex diseases and allowed updates of previous recommendations. We have made substantial alterations, including the introduction of overarching principles and new recommendations on ANCA testing, therapy with GC, use of agents with novel modes of action (C5aR inhibition, IL-5 blockade) and prophylaxis against infections. While most of the original recommendations addressed AAV in general, new data allowed us to devise separate recommendations for GPA/MPA and EGPA for some management principles.

Given the complexity and variability of multiorgan involvement in AAV, we emphasise that these recommendations are not intended to propose a ‘one-size-fits-all’ strategy. Comorbidities, the individual patient’s history, toxicities, local availability and costs of medication, and patient preferences should all be considered in the process of informed decision-making. High-quality evidence of management of AAV in pregnancy is lacking and we refer readers to EULAR recommendations for the management of family planning, assisted reproduction, pregnancy, and menopause in patients with systemic lupus erythematosus and other rheumatic diseases.216 217 In addition, many of the organ manifestations such as severe kidney disease, MPO-ANCA-associated interstitial lung disease, bronchial/subglottic stenosis, orbital mass, severe ENT manifestations, cardiac involvement in EGPA and central nervous system disease or vasculitic neuropathy may require specific pharmacological and non-pharmacological interventions. Recognition and supportive management of organ damage is another important aspect. However, the level of evidence guiding management of these organ manifestations or organ damage in AAV is mostly low. Thus, it was beyond the scope and format of these EULAR recommendations to specifically address these important areas of management.

The COVID-19 pandemic has had a major impact on patients with AAV and influences their management.218 In the light of changing virus variants, availability of vaccinations and antiviral treatments, conditions affecting the management of patients with AAV in the pandemic change rapidly. Therefore, specific recommendations for management of patients with AAV in the pandemic are beyond the scope of this project, since these would be outdated at the time these recommendations are published. Instead, we refer to the most recent national guidelines and EULAR points to consider on the use of immunomodulatory therapies and vaccinations in COVID-19.219 220

Given the relative rarity of AAV and the limitations of the published studies, particularly in terms of outcome assessment and long-term follow-up, important questions remain unanswered. We have listed key issues in a research agenda (box 1) and encourage investigators to use them as a basis for conducting future high-quality research in the field of AAV.

Box 1

Research agenda

A. Diagnosis and classification

  • Develop data-driven diagnostic criteria for AAV.

  • Develop data-driven definitions for disease activity states (remission, response, relapse) and standardisation of outcome measures, including patient-reported outcomes, for use in trials in AAV.

  • Develop data-driven definitions of disease subtypes of importance.

  • Identify reliable biomarkers and risk factors for relapsing disease and damage.

  • Identify reliable biomarkers including imaging and biopsies to assess subclinical disease activity and monitor treatment response.

B. Treatment

  • Evaluate benefits and harms of higher-dose GC therapy (eg, intravenous MP) compared with standard starting dosing (1 mg/kg/day) for patients with different subtypes, severity stages and risk factors for adverse outcomes.

  • Evaluate benefits and harms of reduced starting doses of GC (eg, 0.5 mg/day) compared with standard starting doses (1 mg/kg/day) in patients with different ANCA subtypes, severity stages and risk factors for adverse outcomes.

  • Investigate optimal duration of therapy with GC.

  • Study benefit of the combination of RTX and CYC versus RTX only.

  • Investigate the safety and optimum schedule of repeat-dose RTX maintenance therapy.

  • Investigate the effect of immunomodulators with novel modes of action (eg, JAK inhibitors).

  • Study long-term outcomes after induction therapy with avacopan and its efficacy for maintenance therapy, in combination with and/or compared with standard therapy.

  • Further study the potential of C5a blockade to fully replace GCs for induction of remission and for extended use.

  • Study efficacy and safety of IL-5 inhibitors in newly diagnosed patients with EGPA and patients with organ-threatening manifestations compared with other types of induction therapy (CYC, RTX).

  • Study maintenance therapies for EGPA.

  • Study the optimal duration and dosage of T/S or other agents for prophylaxis against infection.

  • Management of AAV during conception and pregnancy and potential impact on fertility.

C. Long-term outcome and biomarkers

  • Identify biomarkers to predict drug toxicity.

  • Identify predictors for good response, remission or relapse.

  • Define and validate use of patient-reported outcomes for management of AAV in clinical practice.

  • Study the impact of long-term GC therapy on GC-related adverse effects and comorbidities.

AAV, ANCA-associated vasculitis; ANCA, antineutrophil cytoplasmic antibody; CYC, cyclophosphamide; EGPA, eosinophilic granulomatosis with polyangiitis; GC, glucocorticoid; IL-5, interleukin 5; MP, methylprednisolone; RTX, rituximab; T/S, trimethoprim–sulfamethoxazole.

In conclusion, we substantially revised the recommendations for the management of AAV. Despite progress over the past 10 years, we acknowledge that some recommendations had to be made based on low-quality evidence. Nevertheless, the level of agreement for each recommendation was consistently high among the task force members. We encourage clinicians to implement these recommendations into their clinical practice to effectively manage AAV and to improve the patients’ quality of care.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.


The authors wish to thank the librarian Oliver Weiner (Medical Department of the Kiel University Library, Kiel, Germany) for advice and assistance during the SLR. DJ was supported by the NIHR Cambridge Biomedical Research Centre.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Handling editor Josef S Smolen

  • Twitter @AKronbichler, @thkc1, @cmukhtyar

  • Contributors The task force consisted of 20 clinical experts including rheumatologists (MCC, BH, JH, OK, RAL, AM, CBM, JM, PM, GT, DV), internists (AJM, DB, BT) and nephrologists (AK, MAL, MS, YKOT, AV, DJ), from 15 European countries and the USA (PM), 2 methodologists (RAL, GT), convenor (BH) and co-convenor (DJ), 2 delegates of the EULAR young rheumatologists’ network EMEUNET (AB, SM), 2 fellows (BS-A, JHS), 1 health professional (NH) and 2 patient representatives (PV, FP-K). All task force members were involved in preparing the project outline. BH drafted the first version of the manuscript. DJ, the methodologist and other steering committee members revised the manuscript before it was sent to the task force. All task force members made edits and comments and approved the final version.

  • Funding Funding was provided by the EULAR.

  • Competing interests AB received honoraria for consulting from GSK. AJM received speaker fees and/or consultancies from Amgen, Lilly, Vifor, Roche and GSK. AJM received speaker fees and/or consultancies from Amgen, Celgene, Chugai, Novartis, Roche and Vifor. AK received speaker fees and/or consultancies from Catalyst Biosciences, Walden Bioscience, Delta4, Otsuka, UriSalt and Vifor. BH received speaker fees and/or consultancies from AbbVie, Amgen, AstraZeneca, BMS, Boehringer, Chugai, GSK, InflaRx, Janssen, MSD, Pfizer, Novartis, Phadia, Roche and Vifor. BT received consulting fees from AstraZeneca, GlaxoSmithKline, Vifor and Pharma. DB received consultancies from Roche. DJ received speaker fees and/or consultancies from Amgen, AstraZeneca, BMS, Boehringer, Chemocentryx, Chugai, GSK, Novartis, Roche, Takeda and Vifor. DV received speaker fees and/or consultancies and/or grants from AbbVie, Genesis-Pharma, Janssen, Lilly, MSD, Novartis, Pfizer, Roche and UCB. The work is supported by the Greek Rheumatology Society and Professional Association of Rheumatologists (ERE-EPERE) and the Special Account for Research Grants (SARG), National and Kapodistrian University of Athens, Greece. JHS received research grants from the John Grube Foundation and the Deutsche Gesellschaft für Rheumatologie (German Society for Rheumatology). MAL received unrestricted research funding and consultancy fees from Vifor with Chemocentryx. MCC received consulting fees from GSK, Vifor, AbbVie and Janssen, and a research grant from Kiniksa Pharmaceuticals. MS has received consultancy fees from Hansa Biopharma, Vifor, AstraZeneca, Toleranzia and Chemocentryx. PM reports receiving funds for the following activities in the past 2 years: consulting: AbbVie, AstraZeneca, Boehringer-Ingelheim, Bristol-Myers Squibb, Chemocentryx, CSL Behring, Dynacure, EMDSerono, Forbius, Genentech/Roche, Genzyme/Sanofi, GlaxoSmithKline, Immagene, InflaRx, Janssen, Kiniksa, Kyverna, Magenta, MiroBio, Mitsubishi, Neutrolis, Novartis, NS Pharma, Pfizer, Regeneron, Sparrow, Takeda and Talaris; research support: AbbVie, AstraZeneca, Boehringer-Ingelheim, Bristol-Myers Squibb, Chemocentryx, Eicos, Electra, Forbius, Genentech/Roche, Genzyme/Sanofi, GlaxoSmithKline, InflaRx, Sanofi and Takeda; stock options: Kyverna. OK reports receiving speaking fees and/or consultancies from Amgen, AbbVie, Lilly, UCB-Pharma, Novartis, Celltrion; and research support from AbbVie, Viela-Bio, Roche and Novartis. RAL received speaker fees and/or consultancies and/or grants from AbbVie, BMS/Celgene, Chemocentryx, Chugai, GSK, InflaRx, Pfizer Global, Roche and Vifor. The LUMC received on behalf of YKOT an unrestricted research grant from CSL Vifor, GlaxoSmithKline, Aurinia Pharmaceuticals. The LUMC received consulting fees from Aurinia Pharmaceuticals, Novartis, GSK, KezarBio, Vifor Pharma and Otsuka Pharmaceuticals on consultancies delivered by YKOT. The work of YKOT is supported by the Dutch Kidney Foundation (17OKG04) and by the Arthritis Research and Collaboration Hub (ARCH) foundation. ARCH is funded by Dutch Arthritis Foundation (ReumaNederland). AV, BS-A, CBM, FP-K, GT, JH, JM, NH, PV and SM reported no conflicts of interest.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.