SEARCH

SEARCH BY CITATION

Keywords:

  • Wegener's granulomatosis;
  • Children;
  • Vasculitis

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Objective

Wegener's granulomatosis (WG) is a predominantly small-vessel vasculitis associated with antineutrophil cytoplasmic antibodies (ANCAs). There are few reports describing its clinical features and outcome in children. We report on the experience at a single tertiary referral center over 21 years.

Methods

We conducted a retrospective chart review of all patients diagnosed with WG at The Hospital for Sick Children between 1984 and 2005.

Results

Twenty-five patients were identified. Median age at diagnosis and median followup were 14.5 years and 32.7 months, respectively. Male-to-female ratio was 1:4. Median duration of symptoms before diagnosis was 2 months. Of 22 patients, 21 were ANCA positive during their disease course (classic ANCA 78.9%). Constitutional symptoms were the most common clinical feature at presentation (24 of 25). Glomerulonephritis was present in 22 patients at presentation. Only 1 of 11 patients who presented with or developed renal impairment had normalization of serum creatinine. Upper airway involvement occurred in 21 patients at presentation and 24 over followup; only 1 had subglottic stenosis. Twenty patients had initial pulmonary involvement, most commonly nodules (44%) and pulmonary hemorrhage (44%). Five patients required ventilation for pulmonary hemorrhage. Four patients (16%) had venous thrombotic events (VTEs). Treatment included prednisone (100%), cyclophosphamide (76%), azathioprine (40%), and methotrexate (32%).

Conclusion

Pediatric WG typically presents in adolescence and has a female predominance. Glomerulonephritis and pulmonary disease are common at diagnosis and frequently present as a pulmonary-renal syndrome. Loss of renal function is common and rarely completely reversible. As in adults, children with WG are at risk of VTEs.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Wegener's granulomatosis (WG) is a necrotizing predominantly small-vessel vasculitis associated with granulomatous inflammation, pauci-immune necrotizing crescentic glomerulonephritis, and, in most patients, the presence of antineutrophil cytoplasmic antibodies (ANCAs). The clinical and laboratory features of this disease and its outcome have been reported in several large series of predominantly adult patients (1–3). These reports have emphasized the frequent involvement of the upper and lower respiratory tract at presentation and the development of renal involvement over the long term, and have highlighted the frequency of disease relapse and disease- and treatment-related morbidity. The rarity of WG in children has resulted in descriptions in the pediatric population that have been based on smaller numbers of patients, most often in the form of case reports, case series, and literature reviews. The 2 largest single-center pediatric series to date have reported on a combined total of only 40 patients (4, 5). These studies have confirmed that WG in children has many features in common with the adult disease, particularly with respect to respiratory tract and renal involvement. However, they have also suggested that there may be differences between the 2, including a predominance of female patients in pediatric WG and a higher incidence of subglottic stenosis. In addition, there remain aspects of the disease in children that are incompletely defined, most importantly the prevalence of ANCA, which, despite its increasing use in diagnosis, has been infrequently reported in pediatric WG (5–7). The objective of the present study was to report the clinical and laboratory features, treatment, and outcome of a large cohort of children with WG followed at a single center.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

All children <18 years of age diagnosed with WG at The Hospital for Sick Children (SickKids) in Toronto, Canada were eligible for inclusion. To maximize patient ascertainment, searches of databases kept by the hospital records department and the Division of Rheumatology were conducted using diagnostic codes for all ANCA-associated vasculitides. In addition, databases in the Department of Pathology were reviewed for histopathologic or immunopathologic results suggestive of ANCA-associated vasculitis. The charts of all patients identified by these searches were reviewed. Patients were entered into the study if they had ≥2 of the 4 disease features according to the 1990 American College of Rheumatology (ACR) criteria for classification of WG (8) and were seen at SickKids at diagnosis. In the absence of formal documentation of the state of a patient's nasal or oral mucosa, computed tomographic evidence of turbinate mucosal thickening with associated sinusitis was accepted as evidence of nasal involvement, provided other features of WG were present.

Demographic details and predetermined clinical, radiographic, laboratory, and pathologic data were abstracted from the chart for the index admission and visits at 3, 6, 9, 12, 18, and 24 months and yearly thereafter until last review at SickKids. Where patient visits did not coincide with designated time points, the next closest assessment was used. Disease flare was defined as any of the following: 1) note of a new or recurrent clinical finding thought likely to relate to active WG, 2) any increase in prednisone dose (other than stress dosing), and 3) a sustained increase in serum creatinine ≥30% from preceding values in a non–dialysis-dependent patient. Remission was defined as the absence of 1) clinical signs considered indicative of active disease and 2) radiographic signs of reversible pulmonary disease (where radiography was performed), with 3) stable serum creatinine when compared with the immediately preceding visit. No duration requirement was specified for these criteria. Normal renal function was defined by reference to the age-adjusted normal ranges for serum creatinine used at the time patients were seen. Glomerulonephritis was defined as present if urinalysis demonstrated simultaneous proteinuria and hematuria or if glomerular inflammation was detected on renal biopsy when reviewed by a renal histopathologist.

Descriptive statistics were used to summarize patient demographics and clinical features at presentation, and to describe the accrual of clinical features during followup. A Kaplan-Meier plot was used to estimate survival with normal serum creatinine for patients with normal serum creatinine at presentation. This study was approved by the ethics board at SickKids.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Forty-eight patients with ANCA-associated vasculitis were identified. Nineteen were excluded based on the presence of the following diagnoses: microscopic polyangiitis (n = 12), idiopathic pauci-immune necrotizing crescentic glomerulonephritis (n = 5), and undifferentiated ANCA-associated vasculitis (n = 2). Of the 29 patients meeting study criteria for the diagnosis of WG, 4 were excluded because they had been diagnosed at an outside institution (n = 3) or had IgA present in their renal biopsy sample (n = 1). A total of 25 patients were entered into the study. Dates of diagnosis ranged from March 1984 to January 2005. Twenty of the 25 patients were female with a median age at presentation of 14.5 years (range 8.7–17.1 years). The majority were white (21 of 25) and had been symptomatic for a median of 2 months prior to diagnosis (range 0.3–12 months). Median followup was 32.7 months (0–81.4 months). Five patients had <6 months of followup.

Clinical features: presentation and course.

The clinical findings of patients at presentation and over time are outlined in Table 1 by descending order of frequency at presentation. Constitutional symptoms were the most common clinical feature at presentation, of which fever (median 38.7°C) was the leading symptom. Arthralgia, often not associated with signs of arthritis, was the next most frequent symptom and a common early sign of disease flare.

Table 1. Clinical features at presentation and as accrued over followup by descending order of frequency at presentation*
Clinical featureAt presentation (n = 25)At any time (n = 25)
  • *

    Values are the number (percentage).

Constitutional symptoms24 (96.0)24 (96.0)
 Fever18 (72.0)19 (76.0)
 Arthralgias16 (64.0)19 (76.0)
 Weight loss14 (56.0)15 (60.0)
Renal involvement22 (88.0)22 (88.0)
 Glomerulonephritis22 (88.0)22 (88.0)
 Elevated serum creatinine7 (28.0)11 (44.0)
 Requirement for dialysis5 (20.0)6 (24.0)
Ear, nose, throat involvement21 (84.0)24 (96.0)
 Sinusitis11 (44.0)14 (56.0)
 Epistaxis10 (40.0)15 (60.0)
 Oral ulcers7 (28.0)8 (32.0)
 Otitis media6 (24.0)6 (24.0)
 Nasal ulcers6 (24.0)11 (44.0)
 Conductive/sensorineural deafness4 (16.0)4 (16.0)
 Saddle nose2 (8.0)2 (8.0)
 Subglottic stenosis1 (4.0)1 (4.0)
 Nasal septal perforation0 (0.0)2 (8.0)
Pulmonary involvement20 (80.0)21 (84.0)
 Alveolar hemorrhage11 (44.0)12 (48.0)
 Nodules11 (44.0)13 (52.0)
 Airspace disease (nonhemorrhagic)4 (16.0)6 (24.0)
 Required ventilation4 (16.0)5 (20.0)
 Pleuritis2 (8.0)2 (8.0)
Eye involvement13 (52.0)15 (60.0)
 Conjunctivitis11 (44.0)14 (56.0)
 Scleritis/episcleritis3 (12.0)3 (12.0)
 Proptosis2 (8.0)2 (8.0)
Skin involvement8 (32.0)12 (48.0)
 Petechiae/palpable purpura8 (32.0)10 (40.0)
 Urticaria0 (0.0)2 (8.0)
 Panniculitis/erythematous nodules0 (0.0)2 (8.0)
Arthritis8 (32.0)11 (44.0)
Hypertension6 (24.0)13 (52.0)
Gastrointestinal involvement3 (12.0)4 (16.0)
Venous thrombotic event3 (12.0)4 (16.0)
 Deep vein thrombosis3 (12.0)4 (16.0)
 Pulmonary embolus2 (8.0)3 (12.0)
Nervous system involvement2 (8.0)3 (12.0)

Glomerulonephritis, diagnosed on urinalysis and renal biopsy in 16 patients and urinalysis alone in 6, was present in 22 patients (88%) at diagnosis. Seven (28%) had elevated serum creatinine and 5 required hemodialysis. The 2 patients presenting with elevated serum creatinine who did not require hemodialysis continued to have abnormal renal function throughout followup. Of the 5 patients who needed hemodialysis at presentation, 3 regained sufficient renal function to discontinue dialysis, 1 with normal serum creatinine and 2 with persistent elevation of serum creatinine; 1 of these 2 required dialysis permanently 8 months after presentation. Four of the 18 patients with normal renal function at presentation developed abnormal renal function in followup, 1 briefly requiring hemodialysis. All 4 patients had had biopsy-proven glomerulonephritis at presentation and all 4 continued to have abnormal renal function at their last visit. No patient underwent renal transplantation by last followup. Survival rates with normal serum creatinine during followup in all patients with normal serum creatinine at diagnosis are presented in Figure 1.

thumbnail image

Figure 1. Kaplan-Meier plot showing survival with normal serum creatinine over time of all patients with normal serum creatinine (SeCr) at presentation (n = 18). Failure was defined as the development of persistent elevation of serum creatinine above upper limit of normal range for age.

Download figure to PowerPoint

Upper airway involvement, predominantly sinusitis and recurrent epistaxis, was the third most common clinical feature at presentation. Chronic upper airway damage occurred in 4 patients by last followup and included, either singly or in combination, nasal septal perforation (2 patients), saddle nose deformity (2 patients), conductive deafness (1 patient), and chronic sinus disease (1 patient). Subglottic stenosis was rarely seen.

Pulmonary involvement was frequent at presentation and often the dominant clinical feature. Eleven of the 20 patients with pulmonary involvement at diagnosis had pulmonary hemorrhage, with 4 patients requiring ventilation. The median duration of intubation for patients requiring ventilation was 8 days. Pulmonary nodules were as frequent at presentation as pulmonary hemorrhage, and were seen concurrently in 5 patients. Pulmonary involvement was frequent during disease flares; 9 (45%) of the 20 patients followed for ≥6 months developed new or recurrent lung disease in followup, 5 with >1 episode. Pulmonary involvement during disease flares was seen with equal frequency in patients with and without pulmonary disease at presentation. Although all forms of pulmonary involvement were seen during flares, alveolar hemorrhage was the least common. The median time to recurrent or new pulmonary disease after presentation was 25.5 months. In all cases, patients had other clinical symptoms or laboratory findings suggestive of active disease.

The significance of ocular and skin involvement at presentation was frequently missed. Conjunctivitis was often treated as allergy and palpable purpura was diagnosed as Henoch-Schönlein purpura. No patient had uveitis.

Arthritis occurred in the context of active disease and resolved quickly with disease control. Chronic arthritis was not seen.

Three patients had venous thrombotic events (VTEs) at presentation and a fourth had VTE at 6-month followup. Two of the 3 patients with VTEs at presentation had events following femoral vessel punctures, leading to recurrent pulmonary emboli (PE) in one. The third patient had a deep vein thrombosis (DVT) in the left calf 3 months prior to diagnosis during active but unrecognized disease and presented with multiple PE. The patient who developed a VTE in followup was found to have a left-calf DVT during a routine clinic visit and PE on a subsequent ventilation-perfusion scan.

Neurologic involvement occurred in 2 patients at presentation, one with disorientation and hallucinations and the other with focal seizures. Both patients had significant concurrent hypertension, and brain magnetic resonance imaging in both revealed findings consistent with either hypertensive encephalopathy or vasculitis. Neither patient was well enough to undergo lumbar puncture. One further patient developed mononeuritis multiplex and a peripheral neuropathy in followup.

Laboratory features.

The most common laboratory features at presentation were an elevated erythrocyte sedimentation rate (ESR) and anemia, found in 95.8% and 88% of patients tested, respectively. Less common were thrombocytosis (52%) and leukocytosis (44%). The antibody results of the study population are outlined in Table 2. Overall, 95.5% of patients tested were ANCA positive (78.9% had classic ANCA [cANCA]). Specificity for proteinase 3 (PR3) on enzyme-linked immunosorbent assay (ELISA) was demonstrated in all cANCA specimens tested. Antiphospholipid antibodies (aPL) were found in 2 patients, both with VTEs; both had lupus anticoagulant and anticardiolipin antibodies.

Table 2. Autoantibody results at presentation and as determined over disease course*
 At presentationOver disease course
  • *

    Values are the number/total number (percentage). ANCA = antineutrophil cytoplasmic antibody; cANCA = classic ANCA; pANCA = perinuclear ANCA; anti-GBM = anti-glomerular basement membrane.

  • Within 4 weeks of diagnosis.

  • Proteinase 3 specificity by enzyme-linked immunosorbent assay in 10 of 10 tested.

  • §

    Myeloperoxidase specificity by enzyme-linked immunosorbent assay in 2 of 2 tested.

  • Positive if at least 1 result with titer >1/160.

Positive ANCA17/19 (89.5)21/22 (95.5)
 Pattern not stated2/17 (11.8)2/21 (9.5)
 cANCA13/15 (86.7)15/19 (78.9)
 pANCA2/15 (13.3)4/19 (21.1)§
Rheumatoid factor10/16 (62.5)12/23 (52.2)
Antinuclear antibodies2/24 (8.3)5/25 (20.0)
Anti-GBM0/11 (0)0/14 (0)
Anticardiolipin2/9 (22.2)2/11 (18.2)
Lupus anticoagulant1/4 (25.0)2/5 (40.0)

Histology.

Twenty-five renal biopsy specimens were obtained from 17 patients. The most common finding was pauci-immune necrotizing crescentic glomerulonephritis, present in 16 of 17 initial biopsy samples and 5 of 8 repeat biopsy samples. Granuloma (1 of 25) and/or vasculitis (2 of 25) were rarely found.

Ten biopsy specimens of the upper respiratory tract (9 nasal septum, 1 maxillary sinus) were obtained from 9 patients; all but 1 had nasal and/or sinus symptoms at the time of biopsy. Five biopsy specimens had features suggestive of WG (granulomata/giant cells or necrotizing vasculitis), including 1 from the patient without symptoms.

Five lung biopsies were performed. Three were of radiographically ill-defined airspace disease: 2 revealed vasculitis without granulomatous inflammation and 1, from a patient 8 months after a pulmonary hemorrhage, revealed metastatic calcification and hemosiderosis without inflammation. The fourth biopsy sample was obtained during a thoracotomy to drain a hemothorax complicating PE, and confirmed embolic disease without vasculitis or granuloma. The fifth biopsy sample, of a nodular lesion, revealed chronic necroinflammatory changes with occasional giant cells but no granulomata.

Twenty-one biopsy samples of the gastrointestinal (GI) tract were obtained from 4 patients with either abdominal pain, GI bleeding, or diarrhea. The most common finding was mucosal erosive change; none revealed vasculitis or granuloma.

Fifteen skin biopsies were performed in 9 patients. Leukocytoclastic vasculitis was the most common pathologic finding, present in 7 of 15 specimens. One biopsy sample, of a tender nodular lesion, revealed a predominantly septal panniculitis without evidence of vasculitis or granuloma.

Therapy and outcome.

The treatment and outcome of the 20 patients in this study with >6 months of followup, divided into groups according to therapy received during the first 12 months of disease, are outlined in Table 3. Three patients, all with glomerulonephritis, were treated initially with corticosteroids alone. In 2 patients, this was the only therapy administered during followup. The third patient had several disease flares during followup, which resulted in the addition of azathioprine (AZA). At last followup, this patient had persistently elevated serum creatinine.

Table 3. Therapy and selected clinical parameters of patients followed for >6 months (n = 20) divided into groups according to therapy received in the first 12 months of followup*
 No.Median F/U (mo.)Median no. flares in F/U (range)Median time to 1st flare (mo.)Median time to 1st additional agent (mo.)At presentation, no.At last followup, no.Median cum. pred. (gm)Median cum. CYC (gm)
Normal SeCrDialysisPHNormal SeCrDialysisPH
  • *

    F/U = followup; mo. = months; SeCr = serum creatinine; PH = pulmonary hemorrhage; cum. = cumulative; pred. = prednisone; CYC = cyclophosphamide; poCYC = oral cyclophosphamide; ivCYC = pulse intravenous cyclophosphamide; AZA = azathioprine; MTX = methotrexate.

  • The number of patients with pulmonary hemorrhage any time after presentation.

  • One patient in each group indicated also having a pulmonary embolus within 5 weeks of diagnosis.

Steroid only332.73.0 (0–4)6.625.630120017.0N/A
Steroid + poCYC444.32.0 (0–10)10.118.131121020.5101.0
Steroid + ivCYC229.12.5 (2–3)18.619.501210020.58.15
Steroid + ivCYC switched to AZA (n = 3) or MTX (n = 4)728.81.0 (0–3)7.912.643342013.97.04
Steroid + MTX (n = 2)/AZA (n = 2)448.62.5 (2–4)17.725.530110218.79.37

Six patients were initially treated with corticosteroids and either oral cyclophosphamide (poCYC) or monthly intravenous-pulse CYC (ivCYC). The 4 treated with poCYC all had glomerulonephritis (1 requiring dialysis) and 1 additionally had pulmonary hemorrhage. Two patients subsequently received ivCYC and intravenous immunoglobulin (IVIg) for disease flares, 1 of whom also underwent plasmapheresis. At last followup, 2 of the 4 patients had elevated serum creatinine. The median duration of poCYC therapy was 37.5 months. Both of the patients who were treated initially with corticosteroids and ivCYC had pulmonary-renal syndrome with renal impairment, 1 requiring dialysis. They received 7 and 8 monthly pulses of CYC, respectively, followed by corticosteroids alone. Both received additional therapy for disease flares: 1 patient, having been off all therapy for 5 months, received AZA at 27 months, and the other patient, while receiving corticosteroids alone, received mycophenolate mofetil at 12 months. The patient needing dialysis at presentation regained normal renal function 3 months into therapy and maintained this at 22 months of followup. The other patient continued to have stable renal impairment at 36 months of followup.

Seven patients were treated with a remission induction/maintenance regimen in which, after 6–7 months of therapy with corticosteroids and ivCYC, the latter was switched to either AZA or methotrexate (MTX). All 7 had glomerulonephritis at presentation, 3 required dialysis, and 3 had pulmonary hemorrhage (including 2 of the patients receiving dialysis). Three of the 7 patients required a change in therapy at a median of 12.6 months of followup as a result of disease flare. No patient developed new-onset renal impairment during 28.8 months of followup and 1 patient, initially dialysis dependent, regained independent renal function, although this patient continued to have elevated serum creatinine.

Four patients were treated with corticosteroids and either MTX or AZA at diagnosis. All 4 had glomerulonephritis, 1 with elevated serum creatinine and 1 with pulmonary hemorrhage. All 4 were switched to ivCYC at a median of 25.5 months of followup as a result of disease flares in which 2 patients developed elevated serum creatinine for the first time (1 with concurrent pulmonary hemorrhage), 1 patient developed pulmonary hemorrhage with increased urinary sediment activity, and 1 patient developed new-onset cavitating pulmonary nodules and an elevated serum creatinine. Both patients with pulmonary hemorrhage during their flares subsequently also received either poCYC and IVIg (1 patient) or AZA (1 patient). At last followup, 3 of the 4 patients had persistently elevated serum creatinine; in 2 patients this was the result of disease flares while receiving treatment.

Therapy received by the 5 patients with <6 months of followup was as follows: prednisone and ivCYC in 2 patients, prednisone and MTX in 1 patient, and prednisone only in 2 patients for whom the choice of a second agent had not been made at the time data collection ceased.

Irrespective of treatment group, all patients went into remission. The median time to first remission across the 5 groups was 5 months (range 3–6 months). Three of the 20 patients were weaned off all immunosuppressive therapy; all experienced disease flares within 10 months. Overall, 15 of the 20 patients with >6 months of followup experienced disease flares. The median time to first flare was 10.4 months (range 2.4–26.7 months). The median dose of prednisone at the time of first disease flare was 10 mg (range 0–45 mg). There were no patient deaths.

Complications of therapy.

During followup, 5 patients (20%) had 12 hospital admissions for infectious complications of therapy: febrile neutropenia (n = 3), herpes zoster (n = 3), central venous or peritoneal catheter sepsis (n = 4), Enterobacter urinary tract infection (n = 1), and Staphylococcus aureus sepsis (n = 1). Specific complications of corticosteroid therapy occurred in 4 patients: cataracts/glaucoma in 3 patients and steroid myopathy in 1. One patient with cataracts also had symptomatic vertebral compression fractures. Growth effects of corticosteroids were prominent; the median change in height and weight percentiles during followup of the 20 patients followed for >6 months were −17.0% and 10.5%, respectively, and the median peak change in age- and sex-corrected body mass index Z scores was 1.0. No patient developed a malignancy while followed at SickKids.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

This study confirms several observations from previous series characterizing pediatric WG. We found a significant female predominance, with a male-to-female ratio of 1:4. This is consistent with the findings of 2 previous, large, single-center pediatric series and stands in contrast to adult studies where sex incidence has been equal or slightly greater in men (1–5). We found that disease onset most commonly occurs in adolescence, again consistent with the findings of several previous series (4, 9–12), although at variance with that of a more recent study in which the median age at disease onset was just 6 years (5).

In contrast to previous studies, both in adults and children, the majority of patients in this series had generalized WG at diagnosis (i.e., renal involvement). In the reports by Hoffman et al (1) and Rottem et al (4), 18% of adults and 9% of children, respectively, had renal disease at presentation, with the majority having asymptomatic disease. By comparison, at presentation, 88% of our patients had renal involvement, almost 30% had elevated serum creatinine, and 20% needed dialysis. The reason for the high prevalence of generalized disease at diagnosis in our patients is unclear but similar findings have been reported previously. All 6 patients under 18 years of age in the study by Stegmayr et al (12) and all 4 in the report by Hall et al (10) had renal involvement at presentation. Loss of renal function in our patients was common: at last followup, 40% had sustained elevation of serum creatinine and 12% had end-stage renal failure, figures comparable with those in the study by Rottem et al (35% and 9%, respectively) (4). In our experience, this was rarely completely reversible; all but 1 patient who presented with or developed elevated serum creatinine continued to have elevated serum creatinine during followup. With therapy, however, significant improvement in renal function may be achieved; 3 of the 6 patients who required dialysis during their disease course regained sufficient renal function to allow discontinuation of dialysis. Interestingly, 2 of these 3 patients underwent plasmapheresis in addition to treatment with corticosteroids and CYC and were the only patients of the 5 needing dialysis at diagnosis who remained off dialysis in the long term. Plasmapheresis has been suggested to be potentially beneficial in patients with ANCA-associated vasculitides requiring dialysis acutely (13).

Pulmonary disease was common at presentation and pulmonary hemorrhage, often as part of a pulmonary-renal syndrome, was as common as nodular lesions at diagnosis; indeed, the 2 were frequently seen together. Pulmonary involvement was frequent during disease flares, although pulmonary hemorrhage was less frequent. Previous investigators have found that between 34% and 41% of radiographic abnormalities in adult and pediatric patients with WG, respectively, are asymptomatic (1, 4). In the present study, all new or recurrent pulmonary disease that was demonstrated radiographically was associated with other clinical or laboratory findings suggestive of active disease.

The high prevalence of upper airway involvement in our patients, both at diagnosis and during the disease course, is consistent with previous reports in children and is comparable with that in adults (1, 4). Unlike other studies, we did not find the incidence of subglottic stenosis to be higher than that reported in adult WG (4, 5, 9). The prevalence of other disease features at presentation including constitutional, eye, and skin symptoms was higher than has been previously reported and likely reflects the number of patients with generalized disease.

Four patients (16%) in this study had VTEs, which is increasingly recognized as an important complication of adult WG (14). Our data, and those of Von Scheven et al who reported VTEs in 5 pediatric patients with WG (15), suggest that this phenomenon is also a significant complication in children. The reasons for the increased risk of VTEs in patients with WG are unknown. Involvement of the venous circulation in small-vessel vasculitides, particularly ANCA-associated vasculitides, is well recognized (16) and has been theorized to be a contributor to the development of VTEs in WG (14). Another contributor may be the presence of aPL, found in up to 19% of adult patients (17) and reported in association with VTEs in both adults and children with WG (15, 18). Two patients with VTEs in our study were found to have aPL, and 2 (including 1 patient with aPL) had events in the context of vessel instrumentation. This is similar to the experience reported by Von Scheven et al (15) in which 3 of 5 patients with VTEs had aPL at the time of their events and 2 had events associated with central venous catheters. It is likely that VTEs in patients with WG are the result of several factors, singly or in combination, including active disease, aPL, and vessel trauma.

The nonspecific laboratory features of active WG are well described and include elevated ESR, anemia, and to a lesser extent thrombocytosis, all of which were common in our patients. Less well described in pediatric WG is the prevalence of ANCA. Of the 5 patients under 18 years of age in the study by Stegmayr et al in whom ANCA status was determined, 4 had a positive cANCA and 1 had a positive perinuclear ANCA (12). In the series by Belostotsky et al, ANCA was positive in only 59% of 17 patients (all cANCA), although if just those patients satisfying the ACR classification criteria are considered, this number increases to ∼82% (5). We found a positive ANCA in 95.5% of patients tested, 80% with a cANCA pattern, which, when tested, had PR3 specificity on ELISA, suggesting that pediatric WG has the same serologic profile as that described in adult disease (19). The association between ANCA titers and disease flares has been the subject of considerable interest in adult ANCA-associated vasculitides (20–22). Our data were insufficient to allow an examination of this issue.

Conclusions regarding outcome based on initial treatment are not possible because the distribution of patients between different therapy groups in this study was not random, therapy was not provided according to standard protocols, and followup duration between groups was unequal. However, some general observations can be made. Irrespective of the type of initial therapy, all patients' disease went into remission, usually while they were still receiving high-dose corticosteroids. Disease flares were common, seen in 75% of patients with >6 months of followup, and often occurred in the context of prednisone tapered to relatively low doses. Cumulative exposure to CYC was greatest in patients whose initial treatment included poCYC, and was more than 10 times the exposure of those initially treated with ivCYC. Modern treatment protocols in adult WG involve shorter courses of poCYC to induce remission before switching to a less toxic immunosuppressive agent (23, 24). Although it is not possible, from our data, to comment definitively on the relative efficacy of one form of CYC administration over the other, the use of ivCYC in combination with corticosteroids at presentation did not appear to confer an obviously worse outcome when compared with the more conventional regimen of poCYC and corticosteroids. Studies in adults have suggested that selected patients with WG can be treated from diagnosis with steroids and MTX (25–28). In our experience, all patients treated in this manner, or with AZA substituted for MTX, eventually required therapy with CYC for disease flares in which 2 patients developed pulmonary hemorrhage and 2 developed persistently elevated serum creatinine. Although the longer followup of this group relative to the others may have contributed to their apparent poor outcome, the results suggest that patients in whom this therapeutic approach is to be used should be selected and monitored carefully. Our experience over the last 22 years highlights the lack of uniform treatment guidelines for WG in children and the significant difficulties in assessing efficacy of treatment regimens in anything other than a prospective study.

Morbidity related to corticosteroids and cytotoxic therapy has been a major concern in adult WG series (1, 2). Serious infections occur in up to 46% of adult patients and the risk of malignancy increases by a factor of 2.4 (1). Similar rates of serious infections, but not of increased malignancy risk, have been reported in pediatric WG (4). Twenty percent of our patients required hospital admission for complications of therapy related to either presumed or proven infection, 16% experienced specific complications of steroid therapy, and almost all had significant corticosteroid-induced changes in growth. Although none of our patients developed malignancy in the short term, their long-term outcome is unknown.

Several limitations of this study must be considered when interpreting our data. The patients presented were derived from a tertiary referral hospital population and are likely to represent the sickest patients with this disease. It is possible that patients with milder or limited disease, especially without renal or pulmonary involvement, were not referred to SickKids for evaluation. Many large reported series of patients with WG, with both adults and children, have been based on tertiary referral populations and as such are likely to have had similar biases (1, 2, 4, 5, 12). Another limitation of this study relates to the criteria chosen to select patients. At the present time there are no validated criteria for the diagnosis of WG in children, although consensus criteria have recently been published (29). Although the ACR classification criteria for WG used in this study were not intended for the diagnosis of individual patients, they have been used for this purpose in other studies (26) and, in modified form, have been the basis for inclusion of patients in large multicenter therapy trials in adult WG (30). In this respect we believe the use of these criteria for diagnosis in this study is consistent with that in other studies in this field.

We and others have shown that pediatric WG shares many characteristics with adult disease. The consistent observation of a female predominance in pediatric series is intriguing and, as yet, is unexplained. Generalized disease at diagnosis may be more common than previously thought and for this reason early detection is essential. As in adults, WG in children has a significant tendency to relapse and erode renal function. The prospects for prolonged, treatment-free remission, at least in the short term, are poor. Therapy in pediatric WG has been derived from adult data; our study indicates the significant cumulative drug exposure this may entail and the short-term morbidity with which it is associated. Large, multicenter, prospective treatment studies in pediatric WG are needed. The long-term disease- and therapy-related morbidity and outcome for patients treated in childhood remains unknown.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES

Drs. Silverman and Schneider had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Akikusa, Schneider, Hebert, Thorner, Laxer, Silverman.

Acquisition of data. Akikusa, Hebert, Thorner, Laxer, Silverman.

Analysis and interpretation of data. Akikusa, Schneider, Harvey, Hebert, Thorner, Laxer, Silverman.

Manuscript preparation. Akikusa, Schneider, Harvey, Hebert, Thorner, Laxer, Silverman.

Statistical analysis. Akikusa.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. REFERENCES