Classification

The ACR classification criteria for vasculitis and GPA and the subsequent Chapel Hill Consensus Conference (CHCC) disease definitions, were derived primarily from adult data. Because of the poor performance of ACR criteria in classifying children with vasculitis, notably GPA, a pediatric-specific adaptation of ACR criteria, the EULAR/PRINTO/PRES criteria, were developed using pediatric data (see Chapter 32 ). The EULAR/PRINTO/PRES criteria and the ACR criteria are compared in Table 36-1 . The pediatric criteria for GPA had improved sensitivity compared to the adult-based ACR criteria (93% vs 83%), but there are some limitations to their application. Neither pediatric nor adult systems have criteria for MPA, and in one report, pediatric patients classified as having GPA according to ACR criteria could be described as having MPA according to the CHCC definitions. When both the ACR and pediatric-specific criteria were applied to a cohort of 155 pediatric patients with AAV and unclassified vasculitis, the sensitivities of the systems when applied to GPA patients were 62% and 77%, respectively. These differences in sensitivities might be because this last study challenged the ability of the EULAR/PRINTO/PRES criteria to distinguish GPA from MPA, in comparison with the validation cohort where there were extremely few MPA patients among a much wider spectrum of vasculitides.

Epidemiology

Recent population studies suggest a stable incidence of 0.8 to 1.2 per 100,000 persons in some regions, and a 23-year UK study demonstrates a cyclical pattern with peaks of incidence in 1996-1998 and 2005. The rising incidence between the 1970s and 1990s, described for some regions in earlier studies, may only partly be explained by improved case recognition after the introduction of ANCA testing in the 1980s. Regional differences of GPA have been described with higher incidences in both Norway and areas of the UK compared to Spain and Japan. A study in New Zealand similarly describes a higher incidence in higher latitudes.

Much less is known about the incidence of GPA in children and adolescents. American and Norwegian population studies from the 1980s and 1990s, respectively, estimated incidences of 0.6 and 0.8 per 100,000 persons per year. In these studies 3.3% and 7% of patients had disease onset before 20 years or 16 years of age, respectively, for calculated annual incidences of 0.02 and 0.06. A recent Canadian study demonstrated an increase in the incidence of childhood GPA from 0.28 to 0.64 per 100.000 per year in southern Alberta between 2003 and 2008.

In the general population, the peak incidence of GPA is in the sixth decade, and males outnumber females 1.6 : 1. In the pediatric population, the disease occurs in the second decade of life, with a female preponderance.

Pathogenesis

The cause of GPA is unknown, although it is likely a multifactorial disease. As with many other polygenic systemic autoimmune diseases, GPA is probably the result of interactions between genetic factors predisposing to loss of self-tolerance and triggering environmental exposures. Epidemiological studies describing Caucasian predisposition suggest genetic factors play some role. However, the infrequent reported familial occurrences, relatively late mean age at onset of GPA in the general population, and the variations in incidence related to season or latitude, for example, argue for the importance of environmental and nongenetic factors. The vast majority of genetic-association studies of AAV have been in relation to SNPs, and the variants most strongly and reproducibly associated with GPA have been found in human leukocyte antigen (HLA), protein tyrosine phosphatase nonreceptor type 22 (PTPN22), and cytotoxic T-lymphocyte antigen 4 (CTLA4) genes.

Since GPA was first described researchers have unsuccessfully looked for exogenous agents that might stimulate granuloma formation in the airway. An association between primary systemic vasculitis or GPA and crystalline silica exposure and farming has been shown by some but not by others. A role for an infectious trigger has been proposed by many and supported by a number of observations: increased rates of chronic nasal carriage of Staphylococcus aureus in patients with GPA compared with non-GPA patients ; reduced risk of relapse in GPA patients who are on antibacterial maintenance treatment against S. aureus ; high frequency of respiratory tract infections preceding or accompanying onset or flares of GPA ; and increased antibody levels against several other infectious agents in patients with GPA have been reported. Several mechanisms linking infection to autoimmunity in AAV have been proposed.

The presence of ANCAs implicates neutrophils as key effector cells in GPA. The predominant antigenic target of ANCA in GPA is PR3. Patients with GPA have an increased percentage of neutrophils expressing PR3 on their membranes as compared to healthy individuals, and among patients with GPA increased membrane expression of PR3 is associated with severity and rate of relapse. Success in studies of B-cell depletion therapy with rituximab argues for a pathogenic role of autoantibodies, as well as the observation that activated B lymphocytes are increased in patients with GPA compared with healthy controls and are higher in patients with active disease than in those individuals in remission. Although autoantibodies probably play some role in GPA pathogenesis, the characteristic presence of granulomas suggest a more complex process with a pre­dominance of Th-1 cells in the cell-mediated hypersensitivity model of disease. Arguably, inflammation of the upper airways by S. aureus , other pathogens, or environmental exposure induces cytokine production. Elucidation of the intricate immune pathways and processes continues, and recently a role for Th-17 cells in the pathogenesis of GPA and other autoimmune disease has also been proposed.

Clinical Manifestations

The triad of upper and lower respiratory tract inflammation and renal disease is characteristic of childhood GPA as described in the five largest pediatric series. The largest cohort describes 130 patients collected within A Registry for Children with Vasculitis: e-entry (ARChiVe) and includes patients accumulated since the earlier report of 65 patients. The median age at diagnosis was 14.9 years (range 4 to 19) and the median interval from symptom onset to diagnosis was 4.8 months (range 0 to 67). At disease onset, the most common features were constitutional symptoms including fatigue, weight loss, and fever, then followed by pulmonary (81%); renal (79%); and ear, nose, and throat (ENT) (75%) manifestations. Less frequently involved were the musculoskeletal system (64%), skin (55%), gastrointestinal tract (39%), nervous system (27%), and eyes (30%). The frequencies of other clinical features according to organ system as described in the two most recent large pediatric series are listed in Table 36-2 . Features of MPA patients in the largest cohort are listed for comparison. In contrast to the adult experience and two smaller pediatric series, so-called limited or localized GPA defined by the absence of kidney disease occurred in a minority of children at the time of diagnosis and was even less frequent in follow-up, as the frequency of renal disease may accumulate with time. The spectrum of pulmonary manifestations in the ARChiVe cohort included chronic cough (61%), shortness of breath (49%), and hemoptysis or alveolar hemorrhage (42%). Pulmonary function tests were abnormal in 61% of the 67 patients tested. Upper respiratory tract signs and symptoms at presentation were as common as lower respiratory tract features and in follow-up were reported in 91% and 96% of patients, respectively. One study reported that subglottic stenosis was five times as common in pediatric-onset GPA compared with adults, and this feature was subsequently included in the EULAR/PRES classification criteria. In the ARChiVe cohort, subglottic stenosis was no more frequent than otitis/mastoiditis or hearing loss, and less frequent than nasal and sinus involvement. Its inclusion as a classification criterion reflected its specificity. In a cohort of 28 patients from the Cleveland Clinic, a quaternary referral center for vasculitis, airway stenosis with careful screening at diagnosis occurred in 36% of children, increasing to 50% at follow-up; among the 14 patients with airway stenosis, the tracheobronchial tree was involved in 5. Damage to the nasal cartilage, characteristic of long-standing disease, may result in a saddle-nose deformity, similar to that seen in relapsing polychondritis ( Fig. 36-2 ). Renal involvement as manifested by significantly elevated serum creatinine was found in 36% of the patients in the ARChiVe cohort and 28% of patients in the Toronto cohort ; renal dialysis was necessary in 13% of the ARChiVe cohort and 20% in the Toronto cohort.

Saddle-nose deformity in a girl with erosive sinusitis and granulomatosis with polyangiitis.

Diagnosis

The diagnosis of GPA is based on a combination of clinical features (e.g., pulmonary-renal vasculitis syndrome), the presence of serological markers (specifically ANCA, and most commonly PR3-ANCA or cANCA), and characteristic histopathological findings (pauci-immune granulomatous inflammation of predominantly small to medium arteries, capillaries or small veins, or pauci-immune glomerulonephritis). If GPA is suspected, it is crucial to take a careful and specific history of upper respiratory tract involvement and consider formal otolaryngological assessment. Because one third of adult patients initially have asymptomatic renal and pulmonary involvement, it is crucial to examine fresh-spun urinary sediment, and look for pulmonary changes both radiographically and with pulmonary function testing. A decrease in the diffusion capacity of carbon monoxide (DL _CO ) may be the earliest sign of pulmonary hemorrhage. If the disease is limited to a single organ system , tissue diagnosis is desirable to confirm the diagnosis and exclude other diseases. The differential diagnoses include infections (notably mycobacteria, fungi, or helminths, which may also be associated with granulomatous vasculitis), neoplastic disease, sarcoidosis and in young children, chronic granulomatous disease. Pulmonary manifestations in ANCA-positive ulcerative colitis patients have also mimicked GPA. Other forms of vasculitis that can manifest as pulmonary-renal syndromes, such as antiglomerular basement membrane (anti-GBM) disease, systemic lupus erythematosus, mixed connective tissue disease, or PAN, should also be considered. Guided kidney biopsy is now a relatively safe procedure with a high yield, and the finding of pauci-immune glomerulonephritis together with characteristic serology provides a relatively secure diagnosis. GPA can be more readily distinguished from MPA by the presence of upper airway disease including saddle-nose deformity, nasal septal perforation, subglottic stenosis, dacryocystitis, proptosis associated with periorbital tumor, characteristic lung nodules (distinguished from hemorrhage), or the presence of granuloma in biopsy specimens. Characteristic histological features may be patchy in the lungs, and the yield from biopsies may be low especially with fine needle aspirations, but are better with transbronchial or open lung procedures. Frequent involvement of the upper respiratory tract in patients with GPA (e.g., trachea, ears, nose, sinuses, and eyes) may offer sites for relatively noninvasive biopsy procedures; however, the yield from such sites is disappointingly low. When the disease is isolated to these sites, a definitive diagnosis may be difficult, and the diagnostician may need to rely on nonspecific or incomplete histopathological confirmation, serological findings, and the combined expertise of multiple disciplines including the otorhinolaryngologist, ophthalmologist, pathologist, radiologist, and rheumatologist. The need for tissue diagnosis in such cases may be influenced by disease severity, the need to embark upon toxic chemotherapy, and the need to exclude other diseases.

Laboratory Examination

White blood cell counts may be elevated particularly in generalized disease, and associated with normochromic normocytic anemia, thrombocytosis, and markedly elevated erythrocyte sedimentation rates (ESRs) or C-reactive protein levels. In early disease, or in disease limited to a single or few systems, these tests may be normal or only slightly abnormal. Proteinuria, microscopic hematuria, and red blood cell casts in a fresh urine sample indicate glomerular disease. Gross hematuria is uncommon. Elevation of blood urea nitrogen and serum creatinine levels indicates the presence of significant renal disease.

Antinuclear antibodies of unknown specificity are present in 20% to 36% of children tested. Rheumatoid factors are present in approximately 50% of adult and pediatric patients. Children with GPA may be at risk for thrombosis because of antiphospholipid antibodies and factor V Leiden mutations. Antiphospholipid antibodies were evident at presentation in 6 (18%) of 34 children tested in one series but were not associated with venous thrombosis. The antibodies were present in two of nine children in another series who had venous thrombosis.

ANCAs of any kind (cANCA, pANCA, PR3-ANCA, or MPO-ANCA) were present in 98% of 124 children with GPA who were tested. Either cANCA or PR3-ANCA was present in 73%, and either pANCA or MPO-ANCA was present in 29%. Similar frequencies of ANCA positivity were found in other pediatric series and relatively recent adult series, although rates of cANCA positivity were lower and pANCA positivity were higher in the ARChiVe cohort. While cANCA and anti-PR3 are highly sensitive and specific for GPA, pANCA and anti-MPO antibodies may occur especially in so-called renal limited GPA and in Chinese patients with multisystem GPA.

von Willebrand factor antigen (vWFAg), a nonspecific marker of endothelial injury, has been shown to be elevated in other forms of vasculitis during active disease and may be elevated in AAV. Other potential biomarkers of vascular inflammation and injury include matrix metalloproteinase-3 (MMP-3), tissue inhibitor of metalloproteinase-1, pentraxin3 (PTX3), and CXL13 (B-cell attracting chemokine 1).

Pathology

The full histological spectrum of GPA includes necrotizing granulomas of the upper and/or lower respiratory tract, with necrotizing or granulomatous vasculitis involving predominantly small arteries and veins (commonly in the lungs but also in some other systemic organs), and focal segmental necrotizing glomerulonephritis. Granulomata characterize the disease and show acute and chronic inflammation with central necrosis and histiocytes, lymphocytes, and giant cells; eosinophils may be present in small numbers that should not be confused with the large numbers found in specimens from EGPA patients ( Fig. 36-3 ). The granulomas may be discrete, confluent, or poorly formed with scattered giant cells; at times, lung specimens demonstrate nonspecific inflammation. Renal glomeruli are infiltrated with lymphocytes and histiocytes ( Fig. 36-4 ). The earliest renal change may be glomerular thrombosis but the most commonly reported renal lesions are extracapillary proliferation (with or without fibrinoid necrosis) and crescent formation found in a focal and segmental pattern, followed by necrotizing glomerulonephritis. Renal granulomata are rare. Immunofluorescence microscopy is characteristic of a pauci-immune pattern with scanty deposition of immunoglobulins and complement. Dense subendothelial deposits are visible on electron microscopy.

Lung biopsy specimen from a patient with granulomatosis with polyangiitis. Necrotizing granulomata and fibrous tissue have obliterated the normal alveolar architecture. Hematoxylin-eosin stain, magnification ×480.

A, Renal biopsy specimen from a child with granulomatosis with polyangiitis. The glomerulus on the right shows areas of hypercellularity and fibrinoid necrosis with interstitial inflammation. Hematoxylin-eosin stain, magnification ×480. B, Positive immunofluorescent stain for fibrin in a renal biopsy specimen from a patient with granulomatosis with polyangiitis. Magnification ×480.

Imaging

Approximately 78% of children with GPA have abnormalities on chest radiographs, with nodules being more common than fixed infiltrates ( Fig. 36-5 ). Cavitations, pleural effusions, and pneumothoraces may also occur. Although these gross abnormalities are readily detectable on conventional radiography, high-resolution computed tomography (CT) is more effective at detecting other characteristic changes such as small nodules; linear opacities; focal low attenuation infiltrates ; and fluffy centrilobular, perivascular densities ( Fig 36-6 ). The CT halo sign (a rim of ground-glass opacity surrounding the pulmonary lesion) is seen in up to 15% of adult cases and is usually the result of hemorrhage. Sinus radiographs or CT may demonstrate thickening of the sinus lining, opacification of the frontal or maxillary sinuses, or bony thickening, cavitation, and destruction ( Fig 36-7 ). However, magnetic resonance imaging (MRI) is superior in defining highly active mucosal disease. MRI is also helpful in visualizing soft tissue changes that involve the nose, orbits, mastoids and upper airways (i.e., subglottic stenosis), and the characteristic patterns that have been described. Several case reports and a recent retrospective study also suggest a role for 18-F-fluorodeoxyglucose positron emission tomography (PET)/CT scans as an imaging modality in GPA, especially for delineating disease distribution and guiding biopsies of active lesions.

Patchy, lower lobe infiltrates and several well-defined granulomata ( left , lower lobe; right , upper and middle lobes) are evident in the lung of an adolescent with granulomatosus with polyangiitis.

Cavitating lung disease and main bronchus stenosis in a child with granulomatosis with polyangiitis.

Erosive sinusitis in a 12-year-old girl with granulomatosis with polyangiitis. Computed tomography of the sinuses in (A) September 2008 and (B) October 2009 showing almost complete destruction of the medial walls of the maxillary sinuses bilaterally.

Treatment

Prior to the aggressive use of glucocorticoids and cyclophosphamide (CYC) for the treatment of GPA, the disease was fatal in the majority of pediatric and adult patients. Oral CYC originally prescribed together with glucocorticoids for more than 2 years, often regardless of disease severity, induced remission in more than 90% of adult and pediatric patients. Although this strategy was lifesaving, disease relapses were frequent, and CYC-related toxicity and morbidity (infections, bone marrow failure, infertility, hemorrhagic cystitis, and bladder cancer) was high. The clinical challenge, and aim of ongoing clinical trials in adults described below, is to balance the risks associated with current or emerging therapies, against the damage associated with either undertreating aggressive disease or overtreating less aggressive disease. Thus in considering strategies to reduce CYC burden, criteria have been developed for subclassifying or “staging” patients according to disease severity ( Table 36-3 ). It has been recommended that treatment be based on disease severity and be divided into two phases described as remission induction and remission maintenance . The following sections and Tables 36-4 and 36-5 summarize the conclusions from several clinical trials in adults. It should be noted that the large majority of trials have evaluated cohorts under the rubric of AAV. These trials have included both GPA and MPA patients, but not patients with EGPA. There have been no trials in children.

Remission induction: Administering CYC for 3 to 6 months only, until remission is achieved and then switching to less aggressive therapy for remission maintenance, has been the predominant strategy used to reduce the CYC cumulative burden. The European Vasculitis Study Group (EUVAS) designed and tested a regimen to administer the CYC by intermittent intravenous pulses every 2 to 3 weeks (see Table 36-6 ) as an alternative to oral induction therapy that uses 50% less CYC. In the CYCLOPS (Daily Oral Versus Pulsed Cyclophosphamide) trial, the intravenous EUVAS protocol was not inferior to daily oral therapy at inducing remission and was associated with lower rates of leukopenia and infection. However, during a median of 4.3 years of follow-up, the rate of relapse was higher in the pulsed group compared to the oral daily group. Of note, for patients who fail to achieve remission on an intravenous CYC regimen, subsequently switching to oral CYC can be an effective rescue treatment (WEGENT trial).

Glucocorticoids remain a cornerstone of therapy for remission induction and maintenance of remission, but they also carry a toxicity burden. There are no clinical trials examining their role, but their use is integral in all studies evaluating other immunosuppressive treatments for GPA. In critically ill patients, high-dose intravenous methylprednisolone can be used initially. Otherwise, for remission induction prednisone should be started at a dose of 1 to 2 mg/kg/day to a maximum of 60 mg in two or three divided doses to be given for a minimum of 2 weeks ( Table 36-6 ). Reduction in therapy to twice daily and/or to a single daily dose should take place over 4 to 6 weeks. Dose reductions of about 10% to 15% should be on a regular schedule every 2 to 4 weeks as long as the patient remains well with no evidence of worsening disease activity, ultimately aiming to establish the patient on a prolonged low “maintenance” dose, for at least 12 months. In one study where glucocorticoids were withdrawn rapidly and another study where maintenance therapy of any kind was discontinued at 12 months, relapse rates were high.

Several other agents are being used or are being evaluated for use as an alternative to CYC for remission induction. For those patients with less severe GPA (non–life-threatening or non–organ-threatening), methotrexate effectively induces remission, and in the randomized NORAM (Nonrenal Wegener’s Granulomatosis Treated Alternatively with Methotrexate) trial, it was not inferior to CYC at inducing remission in patients without significant renal disease. For both CYC and methotrexate, remission rates were 93.5% and 89.8%, respectively; comparable but high relapse rates of 46.5% and 69.5% suggest the need for more prolonged maintenance therapy.

Rituximab, a monoclonal antibody that binds to the CD20 antigen on the surface of activated B cells, is being increasingly used and evaluated for the treatment of AAV. In the RAVE (Rituximab in ANCA-Associated Vasculitis) trial, 197 patients with newly diagnosed or relapsing GPA or MPA were randomized to receive either rituximab or oral CYC for remission induction. Rituximab was not inferior to CYC at inducing remission by 6 months (64% vs. 53%, respectively), and for the 100 patients with relapsing disease, rituximab was superior to CYC (67% vs. 42%). In the RITUXVAS trial (Rituximab versus Cyclophosphamide in ANCA-Associated Renal Vasculitis), newly diagnosed patients with ANCA-associated renal vasculitis were assigned to receive rituximab plus one or two doses of intravenous CYC or intravenous CYC for 3 to 6 months followed by azathioprine. By 12 months both groups had achieved high remission rates (the rituximab group, 76%; the CYC-only group, 82%), but rituximab was not found to be superior to intravenous CYC for remission induction in newly diagnosed ANCA-associated renal vasculitis. Adverse event rates in both groups were similar. Longer-term studies evaluating the safety and efficacy of rituximab are still needed.

Preliminary results from a trial comparing mycophenolate mofetil with pulse CYC (MYCYC trial) for inducing remission by 6 months in AAV suggest that mycophenolate mofetil is not inferior to CYC; however, subsequent risk of relapse appears higher with mycophenolate mofetil.

For patients with pulmonary hemorrhage, intensive care unit management with ventilatory support and even extracorporeal membrane oxygenation (ECMO) may be required for initial life support. For such patients there may be a role for plasmapheresis in addition to other aggressive immunosuppressive therapy. The routine role for plasma exchange in remission induction in GPA remains unclear. When used in conjunction with standard CYC and glucocorticoid therapy, renal survival improved in a small randomized trial of 32 patients, and when compared to methylprednisolone pulse therapy in the MEPEX (Methylprednisolone versus Plasma Exchange) trial, there was reduced dialysis dependency and end-stage renal disease (ESRD). However, there did not seem to be improved overall survival in either study, and in a median 4-year follow-up of patients from the latter study there were no differences in ESRD. The PEXIVAS (Plasma Exchange and Glucocorticoid Dosing in the Treatment of ANCA-Associated Vasculitis) trial is an international randomized controlled trial that is currently enrolling patients and will further examine the role of plasma exchange in the treatment of GPA.

Remission Maintenance

Relapse rates following remission induction are high, but in view of toxicity, long-term remission maintenance therapy with CYC is no longer an acceptable option. Azathioprine used for at least 18 months following remission induction was shown to be not inferior to CYC for remission maintenance in a large randomized trial and has been adopted as a standard of care for adult patients with AAV. Several other drug regimens have been compared to azathioprine, but none have proved to be significantly better. A summary of the major trials and their outcomes are shown in Table 36-4 . Methotrexate may be considered as an alternative; however, it should be used with caution in patients with impaired renal function. Leflunomide has also been shown to be effective in remission maintenance in GPA but may be associated with more adverse effects than methotrexate. Mycophenolate mofetil appears to be less effective than azathioprine for remission maintenance. Trials evaluating optimal duration of maintenance therapy as well as the use of rituximab for remission maintenance are currently underway (European Vasculitis Study Group. Clinical Trial Protocol: REMAIN, http://www.vasculitis.nl/media/documents/remain.pdf (2006); US National Library of Medicine. Clinicaltrials.gov, http://clinicaltrials.gov/ct2/show/NCT00748644 (2013); US National Library of Medicine. Clinicaltrials.gov, http://clinicaltrials.gov/ct2/show/NCT01697267 ).

Localized Disease

Trimethoprim sulfamethoxazole alone or in combination with glucocorticoids may be considered for remission induction for select cases of localized GPA, especially when limited to the upper respiratory tract ; however, rates of relapse may be high for patients who receive trimethoprim sulfamethoxazole alone. Trimethoprim sulfamethoxazole may have a role as an adjunctive remission-maintenance treatment for patients in limiting the rate of relapse of disease of the ear, nose, and throat. For patients with nasal disease and chronic nasal carriage of Staphylococcus aureus , treatment with topical antibiotics might help limit relapse at that site. Certain disease manifestations such as tracheal and subglottic stenosis may develop or persist despite optimal systemic immunosuppression and may require other treatments such as intralesional glucocorticoid injections, tracheostomy, stent placement, endoscopic dilatation, and laser surgery followed by mitomycin-C application.

Refractory Vasculitis

A small proportion of patients will have disease that persists or progresses despite standard therapy. Several studies support the role of rituximab for refractory disease. Other treatments that may be beneficial but which require further study include antithymocyte globulin; alemtuzumab (monoclonal antibody against CD52); stem-cell transplantation; 15-deoxyspergualin; and tumor necrosis factor antagonists. Of note, the use of etanercept is contraindicated in the treatment of GPA following the results of the Wegener’s Granulomatosis Etanercept Trial (WGET), where 6 out of 91 (7%) patients who received etanercept plus standard therapy developed solid tumors; those who received standard therapy alone developed no tumors.

Current Recommendations

In 2008 EULAR provided recommendations for the management of primary small- and medium-vessel vasculitis in adults based on the results of clinical trials at that time. Intravenous CYC (15 mg/kg [max 1.2 g] every 2 weeks for three pulses followed by three to six pulses every 3 weeks), combined with high-dose glucocorticoids, is recommended for remission induction in life-threatening or organ-threatening generalized primary small- and medium-vessel vasculitis. Plasma exchange is recommended for select patients with rapidly progressive severe renal disease. Methotrexate combined with glucocorticoids is recommended for remission induction in non–organ-threatening or non–life-threatening AAV. For remission maintenance therapy, a combination of low-dose glucocorticoids and either aza­thioprine, leflunomide, or methotrexate is recommended. Prophylactic treatment against Pneumocystis jiroveci with trimethoprim sulfamethoxazole on alternative days was recommended for patients being treated with CYC.

The EULAR recommendations were intended for use in adults, and currently there are no pediatric-specific recommendations. In the absence of any pediatric clinical trials, current treatment in children is largely based on adult recommendations and summarized in Table 36-6 . In ARChiVe, the majority of pediatric patients in the United States and Canada are given CYC and glucocorticoids for remission induction, but maintenance therapy is not reported. The most common treatment alternative to CYC in this series and other single-center case series was methotrexate. Plasma exchange and rituximab in refractory disease have also been reported in case reports or case series of children with GPA, and rituximab for refractory disease. The case series of Fowler et al. focuses on local and systemic treatment of airway disease.

Disease Measurement Tools

Staging disease severity and measuring disease activity and damage are an important part of clinical practice and an essential part of conducting clinical trials. Various tools have been developed and used in adult patients with AAV, including the disease severity scales described above, the Birmingham Vasculitis Activity Score (BVAS) and the Vasculitis Damage Index (VDI). Such tools have become widely established in adult AAV and are accepted by the Outcome Measures in Rheumatology Clinical Trials (OMERACT) initiative. When the performance of BVAS was examined in pediatric AAV patients in ARChiVe, there was a weak correlation with disease activity by physician’s global assessment (PGA), and only a moderate correlation with ESR and treatment decision. In the same cohort of patients, the EUVAS and WGET adult severity subclassification systems were adapted to pediatrics and applied retrospectively. They were found to have a strong correlation with physician choice of treatment; however, they have not been prospectively used to guide therapy. Recently, a pediatric tool based on modifications to the BVAS—the Pediatric Vasculitis Activity Score (PVAS)—was developed and preliminarily validated. Further validation of this tool on an independent dataset and the development of a pediatric adapted tool for assessing damage are planned.

Clinical Course and Outcome/Prognosis

Prognosis arguably depends in part on the clinical state or stage of disease at diagnosis, which in turn may be influenced by the interval from symptom onset to diagnosis. The 1-year mortality of adult patients with untreated GPA was approximately 80%, whereas more recent studies report 5-year mortality rates of treated patients on the order of 10% to 25%. The cause of death of 28 adult patients with GPA in a long-term follow-up of subjects used in development of the ACR classification criteria were as follows: infection (29%), cardiac disease (18%), renal failure (18%), and malignancy (14%). Twenty-two children younger than 15 years of age died of the disease in the United States in a 10-year period ending in 1988. In adults treated with prednisone and CYC over 90% of patients responded completely or partially; however, more than 50% relapsed within 5 years. Similar rates of remission and relapse are reported in two pediatric series. The pediatric series reported 23 and 25 patients who were followed for a mean of 8.7 years and a mean of 2.7 years, respectively. In the former series reported by Rottem and associates, one patient died of severe lung disease and cor pulmonale, and one died of sepsis. No patients died in the series reported by Akikusa and associates.

Reports of long-term morbidity from adult cohorts describe an increased risk of cardiac disease and cardiovascular related events (cardiovascular death, stroke, coronary artery disease, and myocardial infarction). The identified increased risk of malignancy in patients with AAV is lower in studies of more contemporary cohorts, and this might be attributable to the reduced cumulative exposure to CYC. No patient developed malignancy in either pediatric series. Other important morbidities in both adult and pediatric patients include osteoporosis, chronic kidney disease, ESRD, infertility, cystitis, diabetes, and avascular necrosis.

For pediatric GPA there is an increasingly high rate of ENT involvement accumulating with time described in general follow up series. In two of the series, approximately half of patients developed ENT manifestations including hearing impairment, nasal septal, or upper airway deformities over the course of their disease. Two additional single-center case series from 2011 and 2013 each with 28 children having GPA, focused primarily on ENT disease and airway stenoses. In the 2011 study, during followup seven children (25%) developed airway stenosis, predominantly subglottic, but also tracheal, bronchial, with multilevel involvement in four children. In the 2013 study, 14 children (50%) reported the same conditions, with three children developing multilevel airway stenosis. Both studies recommended a multidisciplinary treatment approach involving ENT, pulmonology, and rheumatology, and described the need for several surgical interventions and biological therapy.