Hallmark trials in ANCA-associated vasculitis (AAV) for the pediatric rheumatologist

Anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) refers to a complex group of systemic vasculitides that are characterized by primary small-to-medium sized blood vessel inflammation with the presence of autoantibodies known as ANCA. AAV diseases include Granulomatosis with Polyangiitis (GPA), Eosinophilic Granulomatosis with Polyangiitis (EGPA), and Microscopic Polyangiitis (MPA). AAVs are challenging conditions associated with high cumulative disease and treatment related morbidity and mortality. Given its rarity and the resulting paucity of pediatric-specific clinical trial evidence, pediatric rheumatologists have had to often extrapolate from adult literature for management and therapeutic decisions. The aim of this review is to provide a comprehensive overview of the important findings and overall conclusions of critical landmark clinical trials in the induction and maintenance treatments in adult AAV for the pediatric rheumatologist. This review also highlights the outcomes of recent pediatric AAV observational studies and discusses the future research priorities in pediatric AAV management.


Background
Systemic vasculitis is a challenging and complex multiorgan disease that results in primary inflammation of the blood vessel wall. Anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) is a group of systemic vasculitides that is characterized by small-to-medium sized blood vessel inflammation with the presence of autoantibodies known as ANCA. AAV diseases include Granulomatosis with Polyangiitis (GPA), Eosinophilic Granulomatosis with Polyangiitis (EGPA), and Microscopic Polyangiitis (MPA). AAVs are one of the more common types of systemic vasculitis encountered by pediatric rheumatologists. Proper treatment of this condition is critical as the mortality of untreated AAV can be up to 80% [1,2]. Given the paucity of clinical trials in pediatric AAV, pediatric rheumatologists have relied on adult AAV evidence for management. In this review, we highlight key findings of critical landmark trials in AAV for the pediatric rheumatologist.

Disease activity assessments
Standardized tools are important in measuring disease activity and damage; they also help guide treatment decisions in rheumatic diseases. Numerous instruments have been developed to measure disease activity in AAV [3][4][5]. These measurements are often used to define primary or secondary outcomes in AAV trials. Thus, it is important for pediatric rheumatologists to familiarize themselves with these tools.

Birmingham Vasculitis activity score
The Birmingham Vasculitis Activity Score (BVAS), originally published in 1994 and then revised in 1997 and 2009, is the most widely used tool in clinical practice and trials [6,7]. The BVAS is a composite score that evaluates 56 clinical features from 9 organ systems that are attributed to active vasculitis. Each item is weighted according to the severity. A score of 0 is often adopted as the definition of disease remission in studies. The revised BVAS acknowledges persistent symptoms in addition to new and worsening symptoms. A variation of the BVAS available for GPA patients is known as the BVAS/WG [8]. This score has greater disease specificity in patients with GPA but cannot be generalized to other types of systemic vasculitis.

Pediatric Vasculitis activity score
There had been no validated tool for assessment of disease activity in pediatric patients with systemic vasculitis until 2012 [9], when international collaborative efforts led to the development and validation of a pediatric vasculitis assessment tool. The Pediatric Vasculitis Activity Score (PVAS) is modified from the BVAS. In the PVAS, 22 original BVAS items were redefined and 8 new items were added, resulting in 64 clinical items grouped under 9 organ systems. Every item has an assigned score in the 'new/worse' and 'persistent' scale. This score was used in a recent study to measure the early outcomes in children with AAV [10].

Disease damage assessments
Measurement of damage is an essential component in the follow-up assessments of chronic disease. The Vasculitis Damage Index (VDI) [11] is an unweighted scoring system comprising 64 items grouped under 11 organ-based systems. Damage is defined as an irreversible change lasting for more than 3 months. The VDI is a cumulative index and can only remain static or increase over time. The damage recorded needs to occur after the vasculitis diagnosis, but do not need to be attributable to the diagnosis (e.g. might be related to treatment).
There is no validated tool to assess disease damage in children with vasculitis. However, the Pediatric Rheumatology European Society (PRES) Vasculitis Working Group and Childhood Arthritis & Rheumatology Research Alliance (CARRA) are working toward validating a formal pediatric modification of the VDI [10] -PVDI, which has been piloted in some studies.

Treatment overview
Treatment of AAV is generally categorized into two phases: induction and maintenance. Induction therapy refers to the therapies required to achieve disease remission. After achieving remission, maintenance therapy is initiated to prevent relapses. This review discusses induction therapy for severe and limited AAV separately, followed by maintenance therapy trials.
The European League Against Rheumatism (EULAR) has previously described more specific disease severity definitions by expert consensus in order to conduct trials with more homogeneous AAV patient populations, and to make trials comparable ( Table 1) [12]. For clinically relevant purposes, severe (also described as generalized) AAV is defined as presence of life-or major organ-threatening manifestations. The Canadian Vasculitis Research Network (Canvasc) recommendations on AAV management provides clinical examples describing major organ-threatening manifestations, which may include severe and progressive kidney involvement, alveolar haemorrhage resulting in severe hemoptysis, severe gastrointestinal (e.g. intestinal bleeding), cardiac (e.g. heart failure due to pericarditis or myocarditis), central nervous system (e.g. rapidly progressive neuropathy), or ocular involvement (e.g. orbital pseudotumor) [13]. Limited AAV is often defined as localized involvement without organ-threatening manifestations. Patients with constitutional symptoms are generally included. Patients with renal or pulmonary manifestations may be included with the caveat that the organ manifestations do not result in threatened organ function (e.g. pulmonary infiltrates without severe hemoptysis or mildly reduced kidney function with a Cr < 120 without evidence of casts or significant proteinuria).

Induction trials
Induction trials for severe disease Several trials have addressed induction regimens for severe AAV (Table 2).  Cyclophosphamide (CYC) is one of the most commonly used and well-studied induction therapies. All CYC trials were conducted with one shared aim: to achieve disease remission and minimize medication toxicity. CYCLOPS was an open label multicenter randomized controlled trial (RCT) that evaluated the effect of intermittent intravenous (ivCYC) versus daily oral CYC (poCYC) [14]. The primary outcome was time to remission, defined as the absence of new or worse signs of disease activity on the BVAS and no more than 1 item indicating persistent disease activity (BVAS < 1). The time to remission did not differ between groups (median 3 months). The authors concluded that the use of ivCYC in comparison to poCYC had similar efficacy in achieving remission, but ivCYC therapy had the advantage of having a lower toxicity profile with a reduced cumulative dose. However, a long-term follow-up study with a median duration of 4.3 years showed higher relapse rates with ivCYC compared to poCYC. Interestingly, antiproteinase-3 (PR3) positive patients had higher relapse rates in both groups.
Rituximab (RTX), an anti-CD20 monoclonal antibody and B-cell depleting agent, has emerged as a new therapeutic agent for AAV. RITUXVAS is a multicenter RCT in which newly diagnosed AAV patients with renal disease were randomized to receive ivCYC for 3-6 months, or RTX weekly for 4 weeks plus ivCYC with the first and third RTX infusion. All patients received 1 g of methylprednisolone followed by a tapering course of oral corticosteroids, with reduction to 5 mg by 6 months [15]. The primary outcome was rates of sustained remission (BVAS of 0 for at least 6 months) and rates of serious adverse events (AEs) at 12 months. The RTX-based induction regimen was not superior to ivCYC. The longterm follow-up study also reported no difference in disease remission, AEs, or mortality [16].
The RAVE (Rituximab in ANCA-Associated Vasculitis) trial evaluated RTX in comparison to poCYC utilizing a non-inferiority trial design [17]. This trial differed from the RITUXVAS trial in that the participants of the RAVE trial were younger with less severe renal disease. Furthermore, the comparator agent was oral as opposed to intravenous CYC. The ambitious primary outcome was to achieve disease remission (BVAS of 0) off corticosteroids by 6 months. Both groups had similar corticosteroid regimens. Sixty-four percent of patients in the RTX group had disease remission in comparison to 53% in the poCYC group. This result met the definition for non-inferiority. However, remission rates were overall lower in comparison to other studies, possibly attributable to the earlier discontinuation of corticosteroids. In subsequent analyses, subgroups of patients with either relapsing disease at baseline or those with anti-PR3 positivity achieved higher response rates with RTX when compared to CYC. A long-term follow-up of the RAVE cohort showed that in patients with severe AAV without organ failure, RTX regimen was equivalent to CYC in maintaining disease remission after 18 months [18].
Plasma exchange (PLEX) is a non-pharmacologic treatment considered in patients with severe AAV. The MEPEX trial compared the addition of either IV corticosteroids or PLEX in patients with AAV who had severe renal vasculitis (serum creatinine > 500 μmol/L or dialysis dependence). The primary outcome was renal recovery (dialysis independence) at 3 months, which was achieved in 49 and 69% in the corticosteroid and PLEX arm, respectively [19]. PLEX decreased the incidence of end stage renal disease (ESRD) from 43 to 24% at 12months, but this difference was lost after long-term follow-up [20]. PEXIVAS is an international, open-label, two-by-two factorial design study that recruited patients with new or relapsing severe AAV to investigate the use of adjunctive PLEX with standard therapy (CYC or RTX) and two different corticosteroid regimens (standard or low dose) [19]. The primary outcome was a composite measure of death from any cause or ESRD. Preliminary results from 704 patients found that the primary outcome occurred in 28% in the PLEX arm compared to 31% in the no-PLEX arm. [21]. Interestingly, varying the steroid regimens did not result in a difference to the primary outcome.

Induction trials for limited disease
There has been interest in identifying safer immunosuppressive regimens in the management of patients with limited AAV. The NORAM (Non-Renal vasculitis Alternatively treated with Methotrexate), a non-blinded RCT, hypothesized that induction with oral Methotrexate (poMTX) can spare the toxicity of poCYC in early systemic AAV [22]. The primary outcome was disease remission by 6 months, with tapering of induction therapy by 12 months. All patients were treated with the same corticosteroid regimen. The NORAM trial demonstrated that poMTX was not inferior to poCYC. However, at 18 months, relapse rates were significantly higher in the poMTX group, suggesting that when given as an alternative to poCYC, poMTX may need to be given for longer than 12 months. Long-term follow-up of the cohort (median 6 years) found no difference with respect to AEs between the two regimens and higher relapse rates in the poMTX-treated group [23].
Mycophenolate mofetil (MMF) has been studied for induction therapy for limited disease. MYCYC (MMF versus CYC for remission induction of AAV) is a randomized non-inferiority trial that compared MMF with ivCYC. The primary outcome was the proportion of patients achieving remission (BVAS of 0) by 6 months. The MMF group received doses ranging from 2 to 3 g and the ivCYC group was treated with a similar regimen used in the CYCLOPS trial. Both groups received the same corticosteroid regimen. Sixty-seven percent of patients in the MMF group achieved the primary outcome in comparison to 61% in the ivCYC group. Following remission, relapses occurred significantly more frequently with MMF (33%) compared to ivCYC (19%). The authors concluded that MMF was non-inferior to ivCYC, but that MMF may result in more relapses [24].

Maintenance trials
Given the concerns of utilizing CYC long-term, such as increased risk of malignancy and infertility, investigators have evaluated the use of less toxic immunosuppressants as alternatives (Table 3). In the CYCAZAREM (CYC versus AZA for Early REMission phase of vasculitis) trial, patients with a newly diagnosed severe AAV and a serum creatinine of < 500 μmol/L in whom remission had been achieved within 3-6 months, were randomly assigned to continue poCYC or switch to azathioprine (AZA) for 12 months. All subjects were then switched to a lower dose of AZA, which continued to the end of the study (18 months). Both groups continued a tapering course of corticosteroids. The primary outcome was major and minor relapse rate at 18 months. Relapse rates were not significantly different between groups [25]. There was no difference in severe AEs, although the study was not powered to detect differences in AE rates. Long-term follow-up (median 8.5 years) revealed a trend for worse outcomes in the AZA group in terms of relapses and development of ESRD but these were not statistically significant [26].
While the CYCAZAREM study validated AZA as a suitable alternative to poCYC for maintenance, the optimal duration of AZA treatment was not examined. The RE-MAIN (prolonged REmission-MAINtenance therapy in systemic vasculitis) trial concluded that prolonged maintenance therapy with AZA and low dose corticosteroids to 48 months from diagnosis resulted in a 3-fold reduction in the frequency of relapses compared with withdrawal of AZA and corticosteroids by 24 months [27]. Moreover, the continuation group had improved renal survival with reduced incidence of ESRD. In a RCT (AZA-ANCA trial) comparing standard and extended AZA maintenance therapy in patients with PR3-AAV, patients treated with longer treatment duration (4 years after diagnosis and tapered thereafter) had a lower relapse rate, albeit not significant, compared with those treated with standard treatment (1 year after diagnosis and tapered thereafter) [28]. However, this trial was terminated prematurely given slow patient recruitment and did not achieve an adequate sample size. Therefore, although not definitively proven, a longer duration of maintenance therapy may lead to better outcomes.
The use of MTX as an alternative maintenance agent with possibly less toxicity and equal or perhaps greater efficacy than AZA was examined in the WEGENT trial. In the WEGENT trial, AAV patients in remission were randomized to either AZA or MTX and a tapering oral steroid course [29]. The rate of AEs causing death or study withdrawal was the same between the two groups, indicating MTX was similar in toxicity with AZA. Relapse rates were also similar confirming MTX as a viable alternative to AZA. In the follow-up study, 10-year overall survival rates, total number of relapses, relapse rates and AEs did not differ significantly [30].
Other immunosuppressive agents, including MMF and leflunomide, have also been studied. The IMPROVE (International MMF Protocol to Reduce Outbreaks of Vasculitides) trial compared the efficacy of MMF versus AZA for maintenance of remission in AAV patients in whom remission had been induced with corticosteroids and CYC, with or without methylprednisolone pulses and PLEX. MMF was significantly less effective at preventing relapses when compared to AZA after a median follow-up of 39 months [31]. In a RCT comparing leflunomide and MTX, leflunomide was found to be more effective than MTX in preventing major relapses but was associated with more AEs [32].
The MAINRITSAN (Maintenance of Remission using Rituximab in Systemic ANCA-Associated Vasculitis) trial is the first RCT to evaluate RTX for maintenance therapy. After achieving remission, patients were randomly assigned to receive either RTX at 0 and 2 weeks following randomization then every 6 months until 18 months or AZA until 22 months. Both groups were treated with corticosteroids for at least 18 months. There were significantly fewer major relapses at 28 months in the RTX group compared with the AZA group (5% versus 29%) [33]. In the long-term study, RTX remained superior to AZA up to 60 months, with greater rates of relapse-free and overall survival [34]. The MAINRITSAN2 compared an individually tailored RTX regimen with fixed-schedule regimen [35]. Patients in the tailored-infusion arm received RTX at randomization and received repeat infusions based on lymphocyte counts and ANCA titers until 18 months. The fix-scheduled arm received the same regimen from the original MAINRITSAN trial. There was no significant difference between the number of relapses but the tailoredinfusion arm received fewer number of infusions overall.
The BREVAS (Belimumab in Remission of Vasculitis) is a recent RCT that evaluated the efficacy of belimumab, a monoclonal antibody against B lymphocyte stimulator, as an adjunctive therapy to a regimen of AZA with low-dose corticosteroids [36]. Overall, the addition of belimumab did not reduce relapses.

Eosinophilic granulomatosis with Polyangiitis (EGPA)
EGPA treatment recommendations are less robust due to the lack of RCTs. The treatment is often inferred  Only a few trials have studied immunosuppressive therapy in larger numbers of EGPA patients ( Table 4) [37][38][39][40]. The most commonly studied agent for EGPA has been CYC, although studies have largely looked at variations in CYC regimens as opposed to comparing the efficacy of CYC to other immunosuppressants. Currently, the EGPA Consensus task force recommends the use of CYC for patients with organ threatening disease for induction [41].
Based on a prospective cohort study, it is recommended that EGPA patients without life or organ threatening manifestations can be treated with corticosteroid monotherapy because of excellent 5-year survival rates, although relapse rates are high on a monotherapy regimen [42]. Puechal et al. performed a trial that evaluated the addition of AZA to corticosteroids for induction [43]. The study concluded no additional accrued benefits by adding AZA, with no differences in induction failures or relapses within 2 years (47% versus 49%) [41]. It is likely that a different adjunctive therapy is required for induction in non-severe EGPA patients.
Unique to EGPA is the potential role of interleukin-5 (IL-5) blockade. Wechsler et al. performed a RCT on the use of mepolizumab, an anti-IL-5 monoclonal antibody, in patients with relapsing or refractory EGPA [44]. Participants were either randomized to mepolizumab or to placebo. The two primary end points were the accrued weeks of remission at 52-weeks and the proportion of patients in remission at weeks 36 and 48. The mepolizumab group met the two primary end points with significantly more accrued weeks of remission and a higher percentage of patients in remission. However, 47% of participants did not achieve remission, and it is unclear why certain patients responded better than others. AEs were similar between groups, while serious AEs were slightly higher in the placebo group (26% versus 18%), although some of the events may be attributable to underlying EGPA activity.
RTX has not been studied by controlled trial in EGPA patients, although a few observational studies have reported successful use in both induction and maintenance [45,46]. The adjunctive use of PLEX to CYC has been evaluated in an early study, which reported no additional benefit in survival [47].

Pediatric considerations
Few studies have evaluated the outcomes of pediatric AAV (pAAV) or compared these with adult-onset AAV (aAAV). While some conflicting results exist, it appears that disease severity in pAAV is similar or slightly higher than in aAAV. Rottem et al. prospectively compared 23 pediatric GPA patients with 135 adult-onset patients and found that remission rates, relapse rates, and serious AE rates were similar [48]. Sacri et al. found that renal impairment occurred in 90% of pAAV at disease onset, which is more common than reports in aAAV which range from 10 to 20% at diagnosis to 60-80% during the disease course [49]. Iudici et al. compared 35 pAAV patients with 151 aAAV patients in a matched case-control study [50]. Both groups received similar induction therapy, most commonly corticosteroids and IV CYC. The authors report that by 5 years, pAAV patients had higher relapse rates, accumulated more damage, and were more likely to remain on corticosteroids and immunosuppressive agents than their adult counterparts. Eleven percent of pAAV and 9% of aAAV patients died.
To date, there have been no RCTs evaluating treatment regimens in pAAV. The four largest cohort studies assessing contemporary outcomes in pAAV found that the majority of patients received corticosteroids and CYC for remission induction, followed by AZA or MTX for remission maintenance ( Table 5). The definitions and rates of remission and relapse varied between studies. Sacri et al. reported disease remission in 92% of pAAV patients, and a relapse rate of 41% [51]. In a singlecentre study from Toronto [52], all 20 pAAV followed for a minimum of 6 months achieved remission; the relapse rate was 75% at a median follow-up of 10 months. Iudici et al. reported inactive disease in all pediatric EGPA and MPA patients and 68% of GPA patients at the last follow-up visit (median 96 months), however many of these patients continued to be on therapy, including many on corticosteroids [51].
The largest pAAV cohort to date, the ARChiVe (A Registry for Childhood Vasculitis: e-entry) cohort [10], evaluated the outcomes of 105 pAAV patients and found that 42% achieved remission by 12 months (remission defined as PVAS of 0 on < 0.2 mg/kg/day corticosteroids or equivalent), and 61% had inactive disease on higher doses of corticosteroids. All but 3 patients in remission were on maintenance therapy at 12 months, and 48% had discontinued corticosteroids. Rates of remission were similar for those treated aggressively (CYC and/or RTX, 43%) versus moderately (MTX, AZA or MMF, 30%); those treated aggressively had higher baseline PVAS scores. Twenty-four percent of patients experienced minor relapses after achieving inactive disease. The remission rate of 42% in this study is lower than that reported in the adult studies, which ranges from 53 to 93% [14,17,22] but direct comparison is complicated by differences in study design, remission definitions, and corticosteroid regimens. Sixtythree percent of the ARChiVe cohort had evidence of damage by 12 months, compared to 87% reported in grouped adult data [52]. Given that we have speculated greater organ reserve in children compared to adults, this    The Single Hub and Access point for Pediatric Rheumatology in Europe (SHARE) initiative recently developed consensus-based guidelines for the management of rare pediatric vasculitides. These included recommendations for pAAV (including EGPA) along with Polyarteritis Nodosa and Takayasu Arteritis [53]. Given the paucity of pediatric-specific evidence and highly variable clinical practice amongst various centers, the aims of the recommendations were to define the minimum standard of care for these patients. A large portion of the guidelines is directed at establishing the diagnosis of systemic vasculitis in a pediatric patient.
Adult-derived literature was primarily used to address pAAV treatment recommendations, as the quality of pediatric evidence was poor and largely based on descriptive studies. Induction recommendations for severe disease largely remain similar to adult guidelines, which included corticosteroids and ivCYC as primary agents. The guidelines stress the importance of IV over PO CYC use given the lower cumulative toxicity but similar efficacy. Interestingly, PLEX was defined as a "typical" initial induction agent in severe AAV patients, while adult guidelines generally describe it as an adjunctive agent with insufficient evidence to support its use as a first-line therapy. Furthermore, while adult guidelines strongly support the use of RTX as a first-line remission induction therapy, the same recommendation was not made in the SHARE guidelines, and was considered a second or third line induction agent. Unfortunately, the guidelines do not distinguish severe from limited AAV, and do not provide any specific treatment recommendations for pediatric patients with limited disease, which could potentially lead to confusion or possible over-treatment of a unique subgroup of pAAV patients. Similar agents have been recommended for maintenance therapies.

Future directions
There are questions that remain unanswered in AAV management. In the realm of induction management, LoVAS (Low-dose Glucocorticoids Plus Rituximab Versus High-dose Glucocorticoids Plus Rituximab for Remission Induction in ANCA-associated Vasculitis) is a trial currently underway to evaluate whether corticosteroid regimens can be used in lower doses when RTX is used as the induction agent [54]. The PEXIVAS study provides some preliminary suggestion that regimens utilizing lower doses of corticosteroids do not impact rates of severe outcomes (such as death or ESRD), but final results are pending [21]. The CLEAR study was a recent trial that demonstrated that Avacopan, a C5 receptor inhibitor, could potentially replace or reduce corticosteroid doses, bringing forward unique therapies with a possible steroid-sparing role [55].
With regards to maintenance, the MAINRITSAN3 is a RCT comparing RTX for 46 months compared to 18 months. The RITAZAREM trial is an ongoing study using higher dose RTX in patients with relapsing disease [56]. TAPIR (The Assessment of Prednisone In Remission Trial) is a trial comparing continuation of low-dose corticosteroid versus stopping corticosteroid entirely in GPA patients during maintenance.
Given that EGPA remains understudied in comparison to the other AAVs, further research is needed to determine the efficacy of conventional immunosuppressants and RTX in EGPA, and the optimal patient candidates and dosing regimen for mepolizumab. ANCA-negative patients, excluded by most trials to date, also remain understudied, constituting another future research priority.
Evidence-based guidelines for pAAV management remain sparse. The EULAR/EUVAS guidelines do not comment on the pediatric population. The CanVasc recommendations make 4 pediatric-specific statements, primarily suggesting that pediatric patients should be treated according to adult guidelines [35]. As previously mentioned, the SHARE guidelines for the management of pAAV have been developed in order to set the minimum standard of care when it comes to systemic vasculitis treatment [53]. The majority of treatment recommendations however, were based on low-quality pediatric evidence, expert opinion, or extrapolated from adult studies. While these recent guidelines address the pediatric rheumatology community's desire and need for more pediatric specific recommendations [57], evaluating its uptake and usefulness is warranted. Additional multi-center studies in pAAV are required to address questions around efficacy and toxicity of existing therapies in the pediatric setting, so that pediatric guidelines may incorporate higher quality evidence.

Conclusion
Significant progress has been made in our understanding of the management and outcomes of AAV over the last two decades. Future studies should be directed towards addressing the remaining unanswered questions, which include determining the optimal duration and regimen of AAV induction and maintenance therapy, improving our understanding of EGPA management, and developing evidence for pAAV to better inform pediatric-specific treatment guidelines.