evidence-based management of anca vasculitis

Best Practice & Research Clinical Rheumatology 23 (2009) 367–378

Contents lists available at ScienceDirect

Best Practice & Research ClinicalRheumatology

journal homepage: www.elsevierheal th.com/berh

5

Evidence-based management of ANCA vasculitis

David Carruthers, MBBS, PhD, FRCP, Consultant Rheumatologist a,*,Jonathan Sherlock, MB BChir, MRCP(UK), Academic ClinicalFellow in Rheumatology b

a Department of Rheumatology, City Hospital, Sandwell and West Birmingham Hospitals NHS Trust, Birmingham, UKb Sandwell and West Birmingham Hospitals NHS Trust, and Department of Rheumatology, Division of Immunity and Infection,University of Birmingham, Birmingham, UK

Keywords:ANCA-associated vasculitisWegener’s granulomatosiscyclophosphamidemethotrexateremission inductionmaintenance therapy

* Corresponding author. Tel.: þ44 121 507 4853;E-mail address: [email protected]

1521-6942/$ – see front matter � 2008 Elsevier Ltdoi:10.1016/j.berh.2008.12.003

The vasculitides associated with antineutrophil cytoplasmic anti-bodies (ANCAs) present to and are managed by a wide spectrum ofphysicians, reflecting the multi-organ nature of the conditions.Treatment strategies for these primary inflammatory vasculardiseases have varied based on the outcomes of different clinical trialsand practice reviews. The individual drugs used and their route ofadministration, dose, and duration of therapy have varied and havebeen the source of much debate. Advances in our understanding ofdisease immunopathogenesis, clinical assessment and outcomehave formed the basis for several recent good-quality clinical trials.Now, with the results of these large-scale multicentre collaborativestudies, there is a firmer evidence base to guide management deci-sions for individual patients. This evidence base, reviewed here, hasled to the publication of treatment guidelines which importantlyencompass many of the broader aspects of disease management.

� 2008 Elsevier Ltd. All rights reserved.

Cyclophosphamide (Cyc), in combination with steroids, has long been accepted as the gold standardfor the management of patients with systemic necrotizing vasculitis (SNV). Lieb’s publication in 1979showed that in a mixed group of patients with SNV receiving no specific therapy the 5-year mortalitywas 90%. The addition of steroids improved the outcome to 50% survival at 5 years, but it was with theaddition of a range of immunosuppressive therapies, which included azathioprine (Aza) and Cyc, to thesteroids that the 5-year prognosis improved to 80% [1]. In 1992 the National Institutes of Health (NIH)published a review of 158 patients with Wegener’s granulomatosis (WG) in which continuous oral Cyc

Fax: þ44 121 507 5451.(D. Carruthers).

d. All rights reserved.

mailto:[email protected]

www.sciencedirect.com/science/journal/15216942

http://www.elsevierhealth.com/berh

D. Carruthers, J. Sherlock / Best Practice & Research Clinical Rheumatology 23 (2009) 367–378368

was advocated for treatment of this condition [2]. The treatment regimen consisted of continuous oralCyc at a dose of 2 mg/kg/day until remission, followed by a further 12 months of treatment. Oralprednisolone in a dose tapering over 1 year was an important part of the treatment regimen. Severalother case series supported this approach [3–5]. However, these earlier studies were limited by beinguncontrolled and often retrospective reviews of a heterogeneous group of patients with varyingdiagnoses where the toxicity from Cyc and steroids was high. There were also poorly defined outcomedefinitions for remission and relapse, making interpretation and comparison between studies difficult.

Since these early studies the concept for treatment has changed from one drug for all patients toa phased approach to therapy modified for the individual patient. This change in practice has comeabout for many reasons, including the development of criteria for classification [6,7] and nomenclature[8] that allow the differentiation of one form of vasculitis from another, an important factor given thedifferent outcomes of the specific vasculitides. In addition there are now validated clinical tools for theassessment of disease activity (Birmingham Vasculitis Activity Score, BVAS) [9] and organ damage(Vasculitis Damage Index, VDI) [10] which have been modified and improved since their first intro-duction and are now widely used in clinical studies of vasculitis [11]. Far more is also understood aboutthe immunopathogenesis and in particular the role of antineutrophil cytoplasmic antibodies (ANCAs)[12]. All these factors have contributed to effective collaboration between interested clinicians acrossEurope and America. This unified approach has allowed development and completion of severaltreatment studies on ANCA-associated vasculitis (AAV), particularly under the auspices of the EuropeanVasculitis Study Group (EUVAS), which now allow an evidence-based approach to the therapy of thisgroup of conditions.

This review will synthesize the evidence from large-cohort studies and randomized clinical trials inAAV and provide the clinician with a practical, evidence-based approach towards management of thisdifficult group of patients.

EUVAS collaborative studies

The EUVAS studies have examined the therapeutic approach to patients with AAV based on theirdisease stage and severity. NORAM compared Cyc against methotrexate (Mtx) for induction ofremission in early systemic AAV [13]. CYCLOPS examined the role of pulse versus continuous oral Cycfor induction of remission in AAV [14]. MEPEX studied the role of plasma exchange in patients pre-senting with rapidly progressive renal failure [15], and CYCAZERAM compared Cyc with Aza formaintenance therapy of AAV [16]. Table 1 summarizes the diagnoses and numbers of patients enteredinto these studies, illustrating that a collaborative approach between centres (built on sound diseaseand assessment definitions) can lead to successful large-scale studies even in rare diseases.

Table 1European Vasculitis Study Group (EUVAS) trials summaries.

Trial Number of patients Diagnosis Stage Presentation

CYCAZERAM 155 WG (95)MPA (60)

Generalized New

NORAM 100 WG (94)MPA (6)

Limited New

CYCLOPS 160 WG (61)MPA (99)

Generalized New

MEPEX 137 WG (42)MPA (95)

Severe renal impairment New

TOTAL 552 WG (292)MPA (260)

Generalized refers to patients presenting with AAV and active renal vasculitis and/or other immediate vital organ threateningdisease activity. Limited refers to patients presenting with AAV and one or more involved organ systems plus constitutionalsymptoms with no immediately vital-organ-threatening disease activity. Severe renal impairment indicates evidence of biopsy-proven necrotizing/crescentic glomerulonephritis and a creatinine of >500 mmol/L or significant oliguria at presentation. WG,Wegener’s granulomatosis; MPA, microscopic polyangiitis.

D. Carruthers, J. Sherlock / Best Practice & Research Clinical Rheumatology 23 (2009) 367–378 369

Stages of therapy

This review will look at the different phases of therapy and the important therapeutic questionsraised for each stage.

Remission induction

When considering how effective oral Cyc is at inducing remission for the individual patient, it maybe important to consider whether the disease is in a limited or generalized stage, whether there shouldbe any difference in induction regimen used based on a specific diagnosis of microscopic polyarteritis(MPA) or WG, and whether the patient presenting with active disease has new or relapsing disease.

Three of the EUVAS studies used 2 mg/kg/day Cyc for remission induction in at least one arm of thetrials [13,14,16]. Three additional randomized controlled studies have also used this regimen as part oftheir treatment strategy [17–19]. Although there are other studies reporting the use of continuous oralCyc for remission induction [5], these are the main randomized controlled trials from which compa-rable data can be obtained. From these six studies, 340 patients out of a total of 566 who were enrolledreceived induction therapy with 2 mg/kg/day of Cyc in combination with oral prednisolone (Pred),generally at a dose of 1 mg/kg/day initially, with reduction to a dose of 12.5 mg/day by 3 months. Therate of remission induction with oral Cyc in this pooled group of patients was 76% at 3 months and 87%(range 68–93%) at 6 months. In the EUVAS studies, remission was defined as the absence of new orworse clinical disease activity (BVAS1 score of 0), but allowed persistent activity (BVAS2) in one itemwith a score of �2. This is similar to the 90% remission rate at the same time point reported forcontinuous oral Cyc in several of the large uncontrolled series [2,4,20]. Interestingly, the rate ofremission induction was identical in those patients with WG to those with MPA, as were the remissionrates in patients with generalized versus limited or ‘early systemic disease’ (defined as one or moreorgan systems involved with constitutional symptoms and a creatinine level <150 mmol/L) from theNORAM study. However, these studies do not allow comparison of rates of remission inductionbetween patients with new and relapsing disease, as all recruited patients were at first presentation.This question can be addressed in part by examining data from the control arm of the WGET studywhich examined the role of etanercept in the treatment of WG [21]. Those patients entered into thestudy with relapsing disease (n¼ 42/160 patients) achieved 90% remission rates at some point in thestudy. These patients, who had both limited and systemic disease, received standard immunosup-pressive therapy (Mtx or Cyc respectively) with Pred but without etanercept. This suggests that the rateof remission induction with continuous oral Cyc may be similar in patients with new-onset or relapsingdisease.

The overall mortality over the duration of the six studies referred to above was 9%. However, it isimportant to note that all these studies excluded patients with immediately life-threatening disease atpresentation, a group which would be considered to have a high BVAS at presentation and a worseprognosis [9].

One such poor prognostic group was examined in the MEPEX study which looked at patients whohad more severe disease at presentation, with a creatinine level >500 mmol/L. The mean BVAS at studyentry was 21 (95% CI: 12–41). All patients were treated with a reducing dose of oral Pred and oral Cyc(2 mg/kg/day) for 6 months, and then switched to Pred and Aza. Patients were randomized to receiveadditional therapy with either intravenous methylprednisolone 1 g/day for 3 days or plasma exchange(60 mL/kg) on seven occasions. At 3 months there was a significant difference between those patientswho were alive and dialysis-independent favouring plasma exchange (69%) as against the methyl-prednisolone group (49%) (P¼ 0.02). At 12 months, fewer patients receiving plasma exchange were inend-stage renal failure (19% versus 43%, P¼ 0.008), but the overall 1-year survival was the same (73%for plasma exchange versus 76% for methylprednisolone), whilst serious adverse events approached50% in both groups. Based on the improved renal outcome, the use of plasma exchange is supported inthis group of patients with rapidly progressive renal failure, and may also be indicated in patients withpulmonary haemorrhage.

Having looked at the effect of continuous oral Cyc in remission induction, we should now considerwhether the route of Cyc administration makes a difference to the outcome. Within the literature there


are many uncontrolled retrospective reviews of practice using different Cyc and steroid regimens.Patient disease status, the dose of Cyc per pulse, and the pulse interval vary widely in these papers,making comparison between regimens difficult. A meta-analysis in 2001 [22] showed that pulse Cycwas less likely to fail in remission induction compared with continuous oral Cyc (OR 0.29, 95% CI: 0.12–0.73), but there appeared to be no difference in mortality rate within this meta-analysis (14 out of 69patients receiving pulse Cyc and 16 out of 73 receiving continuous Cyc).

The CYCLOPS study was designed to address the issue of remission induction comparing continuousoral to pulse Cyc; 160 patients with systemic disease, a mean creatinine of 185 (67–681) mmol/L anda BVAS of 20 (10–41) at entry were randomized to receive either 15 mg/kg intermittent pulse Cyc or2 mg/kg/day continuous oral Cyc for at least 3 months or until remission induction. This was followedby a further 3 months’ consolidation Cyc therapy, either at a reduced-pulse frequency or at a lower oraldose, as shown in Fig. 1. All patients then switched to maintenance therapy with Aza and continued ona reducing course of Pred (initial dose of 60 mg/day reducing to 12.5 mg at 3 months). Preliminary datafrom the study (which at the time of writing has still not been published in full) showed that at both 3and 6 months the rate of remission induction was equivalent for pulse compared to continuous oralCyc. For pulse versus oral Cyc at 3 months the remission rate was 70% versus 65%, and at 6 months 92%versus 86% respectively. Patient survival was not significantly different between the two arms, sup-porting the hypothesis that pulse Cyc at a dose of 15 mg/kg and an initial 2-weekly pulse interval (incombination with oral steroid), increasing to 3-weekly, is as effective as continuous oral Cyc in inducingremission.

However the question of whether Cyc is necessary for remission induction in all patients with AAVwas examined in the NORAM study where patients with limited systemic disease (creatinine<150 mmol/L at presentation and �1 organ system involved) were randomized to receive continuousoral Cyc (2 mg/kg/day) or Mtx (20–25 mg/week) as well as a titrating dose of steroids (starting dose of60 mg/day Pred). The median BVAS at trial entry was 15 (2–40) compared with a BVAS of 20 (10–41) inpatients in the CYCLOPS study, reflecting slightly milder disease and a better prognostic group ofpatients. Most of the NORAM patients had a diagnosis of WG and not MPA (Table 1). The remission rateat 6 months was not inferior in the Mtx-treated compared with the Cyc-treated patients (90% for Mtx,93% for oral Cyc, P¼ 0.041), but the median time to remission was shorter with oral Cyc (a median of 2months with a range of 1–5 months versus 3 months for Mtx with a range of 1–9 months). Mortalitywas equal (n¼ 2) in both groups. There was, however, delayed remission in the Mtx-treated group forthose patients who had pulmonary disease (P¼ 0.03) or a high disease extent index (DEI score)(P¼ 0.03) [23]. There was no difference in other outcome measurements of the VDI or Short Form 36

Azathioprine 2 mg/kg/day15 mg/kg/pulse

0 3 (–6) 12 186 (–9)

Induction Maintenance

Month

Consolidation

Prednisolone (mg/day)60 25 12.5 7.5

Cyc 2mg/kg/day → 1.5 mg/kg

A

B

9

Fig. 1. Treatment regimen used in the CYCLOPS study. The first three cyclophosphamide (Cyc) pulses are given at 2-weekintervals, increasing to 3-weekly. A switch to consolidation therapy was made after remission induction and between 3 and 6months.


quality-of-life assessment between the two treatment groups. The role of Mtx in remission induction inWG patients has been studied by several other groups where remission rates are reported at between35–88% [24–27]. Disease stage (extent and new or relapsing disease) and steroid dose are two variablesthat confound comparison between these studies but generally add support to the use of Mtx withsteroids in remission induction in limited or early systemic WG, although the rate at which remission isinduced may be delayed in some subgroups of patients.

Remission maintenance

There are several key questions about what happens after remission induction, and these will beexamined here. (1) Is prolonged Cyc needed? (2) Is remission maintenance influenced by the routeof Cyc used for remission induction? (3) Is remission maintained after induction with Mtx? (4) Isthere an optimum agent for maintenance therapy? And finally, (5) are there any predictors ofrelapse?

Historically, Cyc has been maintained for 1 year after remission induction. This prolonged use ofcontinuous oral Cyc in particular added to the considerable toxicity of treatment regimens. TheCYCAZARAM study aimed to address whether prolonged Cyc was actually needed to maintainremission [16]. In this study, patients with WG (n¼ 95), MPA (n¼ 60), and a median BVAS of 19 (17.5–20.4) had remission induced with 3–6 months of continuous oral Cyc (2 mg/kg/day). Remission withoral Cyc was reached in 77% of patients by 3 months, with a further 16% (25 patients) achievingremission between 3 and 6 months, at which time randomization took place to either Aza or Cyc asillustrated in Fig. 2. After a total of 12 months from study entry, both groups of patients weremaintained on Aza for a further 6 months. The rate of relapse during the initial maintenance phase upto 12 months was equal in the Cyc and Aza maintenance arms of about 10%. Between 12 and 18 months,when both groups were on Aza and Pred, there was again no difference in relapse rate, with 85% of eachgroup maintaining remission at 18 months. In addition there was no difference between groups whenalternative outcome measures, the VDI or SF36, were looked at. These data support the conclusion thatup to 18 months after remission induction with continuous oral Cyc, Aza in combination with pred is noless effective than Cyc at remission maintenance.

The relapse rate was higher within the WG cohort (18% versus 8% for MPA, P¼ 0.03). This supportsother studies showing a higher relapse rate for WG [2,28], and is complemented by studies lookingat ANCA status which show that disease-free survival is better in those patients who are myeloper-oxidase- (MPO-)positive and therefore more likely to have MPA than those who are proteinase-3-(PR3-)positive and more likely to have WG (RR 3.2; 95% CI 1.4–4.4, P¼ 0.002) [29].

Cyc 2mg/kg/day

Aza 2mg/kg

Cyc 1.5mg/kg

Aza 1.5mg/kg

0 3 (–6) 12 189

Induction Maintenance

Month

Prednisolone (mg/d)60 25 12.5 7.5

Fig. 2. Treatment regimen used in the CYCAZERAM study. Aza, azathioprine; Cyc, cyclophosphamide.


Relapse after remission induction

Data from the meta-analysis of pulse versus continuous oral Cyc [22] suggested a slightly greater riskof relapse in those patients who had received pulse compared with continuous oral Cyc (27/64 versus15/56 patients, OR 1.79, 95% CI: 0.85–3.75, P¼ 0.12). This higher relapse rate after pulse Cyc was mostmarked in Guillevin’s study, where 59.2% of WG patients treated with pulse Cyc relapsed as compared to13% treated with continuous oral Cyc [18]. About half of these relapses in each group occurred duringwhat would now be considered an overly prolonged 2-year treatment phase with Cyc, the other halfoccurring after all therapy had been stopped (4-year follow-up). This is in contrast to the shorterCYCLOPS study where there is no reported difference in relapse rate after remission induction with pulsecompared to oral Cyc (P¼ 0.12) [14]. However, there was a trend similar to that shown in the meta-analysis towards a higher relapse rate in those patients who received pulse compared with continuousoral Cyc (14 relapses for the pulse-treated cohort, seven of which were major, versus six relapses, three ofwhich were major, for the oral Cyc group). In contrast to Guillevin’s study, the majority of these patientsreceived only 6 months of Cyc treatment before switching to maintenance therapy for the remaining 12months of the trial, and most relapses occurred after the cessation of Cyc therapy.

For those patients in the NORAM study, the time to relapse after remission induction was shorter forthose patients who had received Mtx: 13 (2–17) months versus 15 (4–17) months for the Cyc group(P¼ 0.023). In the Cyc group there were 20 relapses, nine of which were major (defined as recurrence ofnew vasculitis activity that threatened vital organ function) compared with 32, of which 14 were major,in the Mtx group. However, between 12 and 18 months the rate of relapse in both groups was equal andgreatest, but this was after all therapy (including steroids) had been withdrawn at 12 months. Thesedata therefore also suggest that in AAV maintenance therapy does need to be continued beyond 1 yearafter diagnosis to reduce the risk of relapse. Another factor to take into consideration when consideringa switch to maintenance therapy is ANCA status. Relapse appears to be more frequent in those patientswho are persistently PR3-ANCA-positive at the time of switching from remission induction to main-tenance therapy with Aza (disease-free survival at 4 years¼ 65% for PR3-negative compared with 15%for PR3-ANCA-positive at switch to Aza maintenance) [30].

Unanswered questions

The data from these studies do not allow us to address the optimum duration of maintenancetherapy. Open studies show that 20–53% of WG patients will relapse despite being on maintenancetherapy at a median time interval of 15–19 months after remission induction [24,31,32]. The compo-sition of any maintenance regimen is likely to include prednisolone, but it is also unclear how longsteroid therapy should be continued as part of the regimen, and there are conflicting data on this. TheNORAM study showed that when steroids, as well as immunosuppressive therapies, were withdrawn at12 months there was a high subsequent relapse rate [13]. The WGET study had steroid withdrawal at 6months, but Mtx was continued for 12 months after remission was induced and there was still a 74%relapse rate in the control (non-etanercept) arm,12% of which were severe, suggesting that steroids maybe an important component of a maintenance regimen [21]. However, other studies have suggested thatremission maintenance with Mtx alone is equal to that achieved with Pred plus Mtx [33]. The EUVASstudy group is looking at the optimum duration of maintenance therapy after remission inductioncomparing a 2-year to a 4-year maintenance period with low-dose Pred and Aza in the REMAIN study.

Alternative maintenance agents have been tried, and a higher rate of remission maintenance wasseen in patients with generalized WG who received Mtx (86%) versus cotrimoxazole (58%) [33]. Inanother study, cotrimoxazole was compared with placebo, and remission maintenance was looked at in81 patients at 2 years; a superior benefit was shown with cotrimoxazole (80%) versus placebo (60%) [34].

Toxicity of pulse versus oral Cyc

When considering the beneficial effects of any treatment it is also important to balance this withpotential toxicity. This subject is covered in detail in the chapter by Harper (Adverse Events and TheirPrevention in Vasculitis Therapy). However, the meta-analysis of pulse versus continuous Cyc suggested


that there was a higher instance of leukopenia (P¼ 0.009) and infections (P¼ 0.02) in those patientsreceiving continuous oral Cyc compared with pulse Cyc. In the CYCLOPS study, this pattern wasrepeated, with fewer episodes of leukopenia in the pulse group (9/73 versus 26/71 patients with mildto moderate leukopenia and 1 versus 6 patients with severe leukopenia). Severe infections were alsoless frequent in the pulse Cyc group (3/73 versus 9/71 patients). There were very few patients whodeveloped haemorrhagic cystitis, amenorrhoea, or new diabetes during the course of the study,although hypersensitivity reactions were three times more frequent with pulse compared to daily oralCyc, but these were mild to moderate (9 versus 3 events). Overall there were 19 severe and life-threatening events with daily oral Cyc compared with six with pulse Cyc, supporting pulse Cyc asa safer alternative to continuous oral Cyc.

Guidelines for management of ANCA-associated vasculitis and consensus regimens

Guidelines for the treatment of patients with AAV have been published recently by the BritishSociety of Rheumatology (BSR) [35] and by the European League against Rheumatism (EULAR) [36].The major messages from the guidelines are reviewed here and summarized in Fig. 3.

Diagnosis

A firm diagnosis of vasculitis is essential before beginning therapy. Whilst classification criteria areoften used in diagnosis, there has been no consensus on definitive diagnostic criteria. Both the BSR andEULAR guidelines insist on the necessity of symptoms and signs of vasculitis in the presence of directhistological or serological evidence (positive ANCA) or specific indirect evidence such as imaging,mononeuritis multiplex or mononeuropathy on neurophysiological testing. In addition, clinicalfeatures must be shown not to be accounted for by malignancy, infections (especially bacterialendocarditis), drugs, other forms of vasculitis, or mimics of vasculitis.

Remission induction

GlucocorticoidsWhilst glucocorticoid therapy alone extends the 5-month mean survival of untreated vasculitis [37]

by only a few months [1], the addition of immunosuppressive agents to glucocorticoids is highly

Remission induction ConsolidationConsolidation Remission maintenanceRemission maintenance

Pulse Cyc 15mg/kg Aza or Mtx

0 3 (–6) 12 18 6 (–9) 9Month

Mtx

Prednisolone

Cyc 2mg/kg/dayAza or Mtx

A

B

C

Fig. 3. Guideline summary for the management of ANCA-associated vasculitis (AAV). Remission can be achieved with either (A)pulse cyclophosphamide (Cyc) or (B) continuous oral Cyc. Limited disease can be treated with methotrexate (Mtx) (C). All receivea reducing course of steroids and are switched to maintenance therapy at 3–6 months.


beneficial. Both the BSR and EULAR guidelines endorse this addition of immunosuppressive agents toglucocorticoids, the BSR basing their dosing regimen on the CYCAZAREM trial [16]. Whilst bothguidelines suggest starting Pred at 1 mg/kg/day, the EULAR guidelines suggest that the initial high dosebe maintained for 1 month, whereas a tapering regimen (Table 2) is advised in the BSR guidelines.Where pulse Cyc is used as the induction regimen, intravenous (IV) methylprednisolone is advocatedfor the first and second Cyc pulses only, at a dose of 500 mg and 250 mg respectively. Subsequent IV ororal Cyc pulses are given without additional steroid, while maintaining a titrating dose of daily oralPred. Methylprednisolone is widely used in severe vasculitis and rapidly progressive glomerulone-phritis, at total cumulative doses of 500–3000 mg (daily infusions of 500–1000 mg for 3 days), and maybe combined with plasma exchange for these indications [15]. Whilst the initial oral Pred dosage maybe reduced if high-dose IV methylprednisolone is used, no quantitative recommendations are madewithin the guidelines.

CyclophosphamideBoth guidelines recommend the use of Cyc in all categories of disease severity, the only exception

being in localized or early disease where Mtx can be used. Whilst highlighting the recent evidence thatpulsed IV Cyc is as effective as continuous oral therapy in terms of induction and subsequent main-tenance of remission, both continuous oral and IV regimens are supported, although the lower rates oftoxicity with IV therapy are emphasized. If oral Cyc is used, it should be prescribed at 2 mg/kg/day(maximum 200 mg/day) for 3 months. The dose of oral Cyc should be reduced by 25% in those over 60years and by 50% in patients older than 75 years. Remission is usually achieved by 3 months, but oralCyc may be continued beyond this point (to 6 months) if remission is not attained. Upon induction ofremission, the dose of Cyc can be reduced to 1.5 mg/kg/day and continued so that a total of no morethan 6 months of the drug is given. Pulsed IV Cyc is recommended at 15 mg/kg (maximum 1.5 g), withappropriate adjustments made for age and renal function (Table 3). Pre-hydration with 1 L 0.9% salineand oral hydration with 3 L oral fluid/24 hours is recommended. Although not mentioned in the BSRguidelines, it may be possible to use oral pulses of Cyc after the first three have been given intrave-nously. The total Cyc dose and pulse interval remain unchanged, but the dose of Cyc is spread over 3days (equivalent to 5 mg/kg/day for 3 days). Details of haematological and biochemical monitoring aredescribed in the guidelines to optimize therapy and reduce complications. The advised pulse interval is2-weekly for the first three pulses and then 3-weekly for the next three. Based on data from theCYCLOPS study a further 3-month period of consolidation therapy with pulse Cyc seems appropriate,given at 3-week intervals. Transfer of patients to maintenance therapy after 3 months of oral Cyc andbetween 3 and 6 months after pulse therapy is advised.

MethotrexateLocalized or early systemic disease without threatened vital organ involvement may be treated with

Mtx and glucocorticoids, based on the dosing regimen used in the NORAM study [13], since thisappears as effective as Cyc and avoids the toxicity of the latter. Both BSR and EULAR guidelines

Table 2Oral steroid dose for ANCA-associated vasculitis (AAV) in combination with immunosuppressants.

Time from commencement Prednisolone dose

0 1 mg/kg/day1 week 0.75 mg/kg/day2 weeks 0.5 mg/kg/day3 weeks 0.4 mg/kg/day4 weeks 0.4 mg/kg/day6 weeks 0.33 mg/kg/day8 weeks 0.25 mg/kg/day12 weeks 15 mg/day16 weeks 12.5 mg/day6 months 10 mg/day12–15 months 7.5 mg/day15–18 months 5 mg/day

Table 3Pulse cyclophosphamide dose reductions.

Age (years) Creatinine 150–300 mmol/L Creatinine 300–500 mmol/L

<60 15 mg/kg/pulse 12.5 mg/kg/pulse>60 and <70 12.5 mg/kg/pulse 10 mg/kg/pulse>70 10 mg/kg/pulse 7.5 mg/kg/pulse


recommend that methotrexate is begun at 15 mg/week and increased to a maximum of 20–25 mg/week by week 12. Some patients may tolerate parenteral therapy better. Folic acid may be prescribed.Oral Pred is recommended at the same dosing regimen as described with Cyc (Table 2). However, iflocal or early systemic disease has potentially severe local consequences – for example in localizedretro-orbital disease – Cyc is preferred. Close follow-up is essential because of the potential for morerelapses if therapy is withdrawn at 12 months, and because of the increased risk of progressionassociated with Mtx use, in comparison with Cyc. However, recommendations for maintenancetherapy discussed below do not support withdrawal of therapy at this early stage.

Plasma exchangeIn severe or life-threatening disease, plasma exchange is advised in addition to Cyc and steroids (at

the doses already described). The addition of plasma exchange has been shown in the MEPEX study toimprove renal but not overall survival in acute renal failure secondary to ANCA-associated glomeru-lonephritis [15]. Seven exchanges of 4 L are recommended over 2 weeks. Plasma exchange can beconsidered in other life-threatening manifestations such as pulmonary haemorrhage.

Remission maintenance

AzathioprineOf all the therapies used in maintenance, Aza has the highest level of evidence. The CYCAZAREM trial

demonstrated that continuous oral Cyc can be safely substituted with Aza after remission, and theguidelines suggest that this switch to maintenance therapy can be made after 3 months of oral Cyc or after3–6 months of IV administration, thus limiting Cyc exposure to the active phase of disease. Given the highrelapse rates when therapy is withdrawn at 12 or even 18 months [13], at least 2 years of therapy isrecommended following remission. However, since WG has a higher relapse rate than MPA [16], untilother evidence is available it is recommended that patients with WG should continue for up to 5 years, asshould those who remain ANCA-positive. Azathioprine is prescribed at 2 mg/kg/day from remissioninduction to 12 months and 1.5 mg/kg thereafter. Detailed monitoring guidance must be adhered to.

MethotrexateSince methotrexate has efficacy in preventing relapse similar to that of Aza, with relapse rates in one

study of 33% for Mtx and 41% for Aza at 3 years [38], it may also be used for maintaining remission. It isrecommended to commence Mtx at 0.3 mg/kg (maximum 15 mg) per week and escalate by 2.5 mgeach week to a maximum of 20–25 mg/week. The BSR guidelines emphasize, however, that theevidence for methotrexate is not as strong as for azathioprine which has been studied in a randomizedcontrolled trial [16].

SteroidsThe role of steroids in maintenance therapy is unclear, and no strong guidance is given beyond 18

months of therapy. In the absence of evidence, individual clinician preference as to the dose andduration of therapy as part of a maintenance regimen is advocated.

Relapsing disease

Consensus opinion advocates increasing the dose of prednisolone to 30 mg/day to treat minorrelapses (defined as relapse in the absence of threatened vital organ involvement), with optimization


of concurrent immunosuppression. In the presence of a major relapse (threatened vital organinvolvement), the addition of Cyc at the doses described for remission induction is advised. Plasmaexchange and methylprednisolone may also be considered for more severe flares. When minor relapsesoccur on Mtx or Aza, mycophenolate mofetil or leflunomide may be able to re-induce and maintainremission. Evidence is highlighted that trimethoprim/sulphamethoxazole is useful in maintainingremission, possibly by decreasing Staphylococcus aureus nasal carriage.

Refractory disease

Refractory disease is rare, and so it is emphasized that infections or malignancies which may rendervasculitis resistant to treatment are excluded. Placebo-controlled and open-label trials have demon-strated partial disease control with high-dose intravenous immunoglobulin (IVIg) in refractoryvasculitis [39,40], and this therapy may be considered particularly in refractory disease accompaniedby infection, or in pregnancy. Tumour necrosis factor (TNF) inhibition with infliximab may have a rolein refractory disease though not in routine induction regimens. T-cell depletion with anti-thymocyteglobulin [41], B-cell depletion with rituximab [42], and pan-lymphocyte depletion with CAMPATH 1H[43] have all been associated with sustained remission in refractory disease. There are insufficient datato recommend their routine use, although it is hoped that the results of ongoing clinical trials willclarify the place of rituximab in management of vasculitis. Use of such biological agents is recom-mended to be confined to centres of expertise (BSR), and is the subject of a separate review in thechapter by Jayne and Hiemstra (Newer Therapies in Vasculitis).

Clinical follow-up

Monthly assessments for the first 3 months followed by 3–6-monthly assessments are advised byexpert consensus, with future intervals no longer than 6 months. The VDI and initial BVAS scores help topredict the long-term outcome and can be useful clinical tools in the routine clinic. It is advised thattreatment should not be altered solely on the basis of a rise in ANCA titre, since only approximately half ofsuch rises are followed by relapse, and around half of relapses occur in the absence of rises in ANCA titre.

Minimization of side-effects

The efficacy of treatments for vasculitis has meant that whilst the disease can frequently be wellcontrolled, major morbidity and mortality can arise from the therapies used and the complications ofimmunosuppression. Specific recommendations in the guidelines are made to try to reduce this,particularly by limiting exposure to Cyc. In addition, monitoring for the bladder malignancy associatedwith Cyc is suggested, as is prophylaxis and monitoring for many conditions, including Pneumocystisjiroveci infection, fungal infections, cervical intraepithelial neoplasia, infertility, osteoporosis, myco-bacterial infection, vaccinations and cardiovascular risk management. This topic is covered in depth inthe chapter by Dr Lorraine Harper (Adverse Events and Their Prevention in Vasculitis Therapy).

Summary

Large-scale collaborative studies into management of the different phases of AAV have started toprovide solid evidence upon which the clinician can base management decisions. With an increasedunderstanding of the prognosis and outcome of the different vasculitic syndromes, a balance betweenclinical efficacy of any treatment strategy and the risk of drug toxicity can be made. A careful clinicalassessment of the individual patient will allow tailored therapeutic decisions to be made usinga phased approach to therapy. Clinical assessment tools (e.g. BVAS and VDI) will assist in this process.The evidence from the EUVAS controlled clinical trials has been used in conjunction with data frommany large-cohort and other clinical studies to produce guidelines that make recommendations for themanagement of AAV. Importantly, these guidelines include broader aspects of patient care which gosome way to minimize the wide-reaching effects of the disease and its therapy. There are still many


unanswered questions relating to optimum therapy using current drugs, but the new range of agentscurrently in clinical trials will hopefully further improve the outcome of this patient group.

Research agenda

� investigation of the role of biologic therapies in increasing the remission rate, reducingrelapse rates, and minimizing patient exposure to high cumulative doses of steroids andcyclophosphamide and other immunosuppressive drugs� it is unclear what the optimum drug combination and duration of maintenance therapy is

with steroids and Aza or Mtx, but this is currently being addressed in clinical studies

Practice points

� remission induction can be induced with continuous oral Cyc in 87% of AAV patients by6 months, with many of them in remission at 3 months� pulse Cyc has equivalent efficacy to but is less toxic than continuous oral Cyc� in patients with limited disease, rates of remission induction with Mtx are the same as with

Cyc, but remission occurs more slowly� relapse may be more frequent after remission induction with pulse therapy� prolonged Cyc is not necessary for remission maintenance therapy� there is no clear advantage of Aza over Mtx for maintenance therapy

Conflict of interest

Neither author has any conflict of interest (personal or financial) to declare.

References

[1] Lieb ES, Restivo C, Paulus HE. Immunosuppressive and corticosteroid therapy of polyarteritis nodosa. Am J Med 1979;67:941–7.

[2] Hoffman GS, Kerr GS, Leavitt RY, et al. Wegener granulomatosis - an analysis of 158 patients. Ann Intern Med 1992;116:488–98.

[3] Fauci AS, Wolff SM. Wegener’s Granulomatosis: studies in eighteen patients and a review of the literature. Medicine1973;52:535–61.

[4] Fauci AS, Haynes BF, Katz P, Wolff SM. Wegeners granulomatosis - prospective clinical and therapeutic experience with85 patients for 21 years. Ann Intern Med 1983;98:76–85.

[5] Langford CA, Talar-Williams C, Barron KS, Sneller MC. A staged approach to the treatment of Wegener’s granlomatosis.Arthritis Rheum 1999;42:2666–73.

[6] Fries JF, Hunder GG, Bloch DA, et al. The American College of Rheumatology 1990 criteria for the classification of vasculitis -summary. Arthritis Rheum 1990;33:1135–6.

[7] Leavitt RY, Fauci AS, Bloch DA, et al. The American College of Rheumatology 1990 criteria for the classification ofWegener’s granulomatosis. Arthritis Rheum 1990;33:1101–7.

[8] Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides - proposal of an international consensusconference. Arthritis Rheum 1994;37:187–92.

[9] Luqmani RA, Bacon PA, Moots RJ, et al. Birmingham Vasculitis Activity Score (BVAS) in systemic necrotizing vasculitis.QJM 1994;87:671–8.

[10] Exley AR, Bacon PA, Luqmani RA, et al. Development and initial validation of the vasculitis damage index for thestandardized clinical assessment of damage in the systemic vasculitides. Arthritis Rheum 1997;40:371–80.

[11] Bacon PA, Moots RJ, Exley A, et al. VITAL assessment of vasculitis. Clin Exp Rheumatol 1995;13:275–8.*[12] Morgan MD, Harper L, Williams J, Savage C. Anti-neutrophil cytoplasm-associated vasculitis. J Am Soc Nephrol 2006;17:

1124–34.*[13] de Groot K, Rasmussen N, Bacon PA, et al. Randomized trial of cyclophosphamide versus methotrexate for induction of

remission in early systemic antineutrophil cytoplasmic antibody-associated vasculitis. Arthritis Rheum 2005;52:2461–9.[14] de Groot K, Muhler M, Reinhold Keller E, et al. Randomised controlled trial of daily oral versus pulsed cyclophosphamide

for induction of remission in ANCA associated systemic vasculitis (abstract). Kidney Blood Pressure Res 2005;28:195.


*[15] Jayne DRW, Gaskin G, Rasmussen N, et al. Randomized trial of plasma exchange or high dose methylprednisolone asadjunctive therapy for severe renal vasculitis. J Am Soc Nephrol 2007;18:2180–8.

*[16] Jayne D, Rasmussen N, Andrassy C, et al. A randomised trial of maintenance therapy for vasculitis associated withantineutrophil cytoplasmic antibodies. N Engl J Med 2003;349:36–44.

[17] Gordon M, Luqmani RA, Adu D, et al. Relapses in patients with a systemic vasculitis. Q J Med 1993;86:779–89.*[18] Guillevin L, Cordier JF, Lhote F, et al. A prospective, multicenter, randomized trial comparing steroids and pulse cyclo-

phosphamide versus steroids and oral cyclophosphamide in the treatment of generalized Wegener’s granulomatosis.Arthritis Rheum 1997;40:2187–98.

[19] Haubitz M, Schellong S, Gobel U, et al. Intravenous pulse administration of cyclophosphamide versus daily oral treatmentin patients with antineutrophil cytoplasmic antibody-associated vasculitis and renal involvement: a prospective,randomized study. Arthritis Rheum 1998;41:1835–44.

[20] Reinhold-Keller E, Beuge N, Latza U, et al. An interdisciplinary approach to patients with Wegeners granulomatosis:long-term outcome in 155 patients. Arthritis Rheum 2000;43:1021–32.

*[21] The Wegener’s Granulomatosis Etanercept Trial (WGET) Research Group. Etanercept plus standard therapy for Wegener’sgranulomatosis. The N Engl J Med 2005;352:351–61.

*[22] de Groot K, Adu D, Savage COS. The value of pulse cyclophosphamide in ANCA-associated vasculitis: meta-analysis andcritical review. Nephrol Dial Transplant 2001;16:2018–27.

[23] de Groot K, Gross WL, Herlyn K, Reinhold-Keller E. Development and validation of a disease extent index for Wegener’sgranulomatosis. Clin Nephrol 2001;55:31–8.

[24] Hoffman GS, Leavitt RY, Kerr GS, Fauci AS. The treatment of Wegener’s granulomatosis with glucocorticosteroids andmethotrexate. Arthritis Rheum 1992;35:1322–9.

[25] Sneller MC, Hoffman GS, Talar-Williams C, et al. An analysis of forty-two Wegener’s granulomatosis patients treated withmethotrexate and prednisone. Arthritis Rheum 1995;38:608–13.

[26] de Groot K, Muhler M, Reinhold-Keller E, et al. Induction of remission in Wegener’s granulomatosis with low dosemethotrexate. J Rheumatol 1995;38:608–13.

[27] Stone JH, Tun W, Hellman DB. Treatment of non-life threatening Wegener’s granulomatosis with methotrexate and dailyprednisone as the initial therapy of choice. J Rheumatol 1999;26:1134–9.

[28] Guillevin L, Gasselin B, Cevallos R, et al. Microscopic polyangiitis - clinical and laboratory findings in eighty-five patients.Arthritis Rheum 1999;42:421–30.

[29] Hogan SL, Falk R, Chin H, et al. Predictors of relapse and treatment resistance in antineutrophil cyctopplasmic antibody-associated small-vessel vasculitis. Annals Intern Med 2005;143:621–32.

*[30] Slot MC, Tervaert CJW, Boomsma MM, Stegeman CA. Positive classic antineutrophil cytoplasmic antibody (c-ANCA) titerat switch to azathioprine therapy associated with relapse in proteinase 3-related vasculitis. Arthritis Rheum 2004;51:269–73.

[31] Langford CA, Talar-Williams C, Barron KS, Sneller MC. Use of a cyclophosphamide-induction methotrexate-maintenanceregimen for the treatment of Wegener’s granulomatosis: extended follow-up and rate of relapse. Am J Med 2003;114:463–9.

[32] Specks U. Methotrexate for Wegener’s granulomatosis: what is the evidence? Arthritis Rheum 2005;52:2237–42.[33] de Groot K, Reinhold-Keller E, Tatsis E, et al. Therapy for the maintenance of remission in sixty-five patients with

generalized Wegener’s granulomatosis. Methotrexate versus trimethoprim/sulfamethoxazole. Arthritis Rheum 1996;39:2052–61.

[34] Stegeman CA, Cohen Tervaert JW, de Jong PE, Kallenberg CG. Trimethoprim-sulfamethoxazole (co-trimoxazole) for theprevention of relapses of Wegener’s granulomatosis. Dutch Co-Trimoxazole Wegener Study Group. N Engl J Med 1996;335:16–20.

*[35] Lapraik C, Watts RW, Bacon P, et al. BSR and BHPR guidelines for the management of adults with ANCA-associatedvasculitis. Rheumatology 2007;46:1–11.

*[36] Mukhtyar C, Guillevin L, Cid M, et al. EULAR recommendations for the management of primary small and medium vesselvasculitis. Ann Rheum Dis 2008. e publication: April 15.

[37] Walton EW. Giant cell granuloma of the respiratory tract (Wegener’s Granulomatosis). BMJ 1957;2:265–70.[38] Mahr A, Pagnoux C, Cohen P. Treatment of ANCA associated vasculitides; corticosteroids and pulse cyclophopshamide

followed by maintenance therapy with methotrexate or azathioprine a prospective multi-centre randomised trial(abstract). Kidney Blood Pressure Res 2005;28:194.

[39] Richter C, Schnabel A, Csernok E, et al. Treatment of antineutrophil cytoplasmic antibody (ANCA)-associated systemicvasculitis with high-dose intravenous immunoglobulin. Clin Exp Immunol 1995;101:2–7.

[40] Jayne DRW, Chapel H, Adu D, et al. Intravenous immunoglobulin for ANCA-associated systemic vascultis with persistentdisease activity. QJM 2000;93:433–9.

[41] Schmitt WH, Hagen EC, Neumann I, et al. Treatment of refractory Wegener’s granulomatosis with anti-thymocyteglobulin (ATG): an open study in 15 patients. Kidney Int 2004;65:1440–8.

[42] Keogh KA, Wylam ME, Stone JH, Specks U. Induction of Remission by B lymphocyte depletion in eleven patients withrefractory antineutrophil cytoplasmic antibody-associated vasculitis. Arthritis Rheum 2005;52:262–8.

[43] Lockwood CM, Thiru S, Stewart S, et al. Treatment of refractory Wegener’s granulomatosis with humanized monoclonalantibodies. QJM 1996;89:903–12.

evidence-based management of anca vasculitis

Documents