Pediatric Nephrology

, Volume 20, Issue 9, pp 1214–1218 | Cite as

Problems in classifying vasculitis in children



Vasculitis is inflammation of blood vessels, which may also affect the arteries of the kidney. The definition and classification of vasculitides derive from the Chapel Hill nomenclature criteria (CHCC) and American College of Rheumatology (ACR) criteria, respectively. These criteria are not perfect, and adult literature refers to problems with misclassification and a low concordance between the two criteria. The problems associated with defining a child patient according to these criteria are even more apparent, since both sets of criteria have been based on adult data alone and have never been validated in children. When classifying their patients as having polyarteritis, paediatricians do not keep strictly to the CHCC or ACR criteria. A recent series has shown that the microscopic polyarteritis associated with MPO-ANCA and the HBs associated with classic polyarteritis nodosa are rarer in children than in adults. Paediatricians should revise the existing criteria based on registries and international consensus in the field.


Vasculitis Children Juvenile polyarteritis 


Vasculitis is inflammation of the blood vessel, which can lead to tissue damage. The kidney is one of the major target organs of vasculitis. Although vasculitic diseases are increasingly observed in everyday practice, the aetiology of vasculitic diseases has not been completely elucidated. This remains a major obstacle when defining diagnostic criteria for vasculitides. Without the pathognomonic features to distinguish these syndromes, classification criteria use an arbitrary combination of clinical, laboratory and histological features [1]. Since childhood vasculitides lack pathognomonic clinical symptoms or a specific diagnostic test, classification criteria are used to guide physicians in diagnosing these diseases. The purpose of this paper is to review existing classification and nomenclature criteria for vasculitides and to emphasise relevant current problems with them.

Classifications would help to distinguish one type of disease from another and are not diagnostic criteria [2]. They are useful for grouping patients for studies, comparisons and epidemiological studies. Patients in the early stages of the disease or with minimal manifestations may not fulfil the classification criteria [3]. Sensitivity and specificity are never 100%, and they would serve as diagnostic criteria if they were. The most widely applied classification criteria for vasculitides are the American College of Rheumatology (ACR) criteria. They have been developed for each type of vasculitis separately [4, 5, 6, 7]. ACR criteria rely on the clinical and laboratory features of the specific diseases.

To define a disease is to identify the essential elements of the disease that are present in all cases [8]. In Chapel Hill, an international group led by rheumatologists, nephrologists and pathologists gathered to reach consensus in constructing root definitions for vasculitides [2]. The Chapel Hill Nomenclature criteria (CHCC) clearly state that the proposed criteria are not diagnostic criteria. CHCC mainly requires a biopsy diagnosis for the patients [2] (Table 1).
Table 1

Chapel Hill classification of vasculitides

I. Predominantly large vessel

Giant cell (temporal) arteritis

Takayasu arteritis

II. Predominantly medium size vessel

Kawasaki disease

Classical polyarteritis nodosa (PAN)

III. Predominantly small vessel disease

Microscopic PAN

Wegener’s granulomatosis

Henoch-Schonlein purpura

Churg-Strauss syndrome

Cutaneous leukocytoclastic vasculitis

Essential cryoglobulinemic vasculitis

Table reused from Arthritis and Rheumatism (1994) 37:189 with permission from Wiley Publishers [2]

Neither the ACR nor the CHCC have been validated in children. They have been suggested by adult rheumatologists based on experience with adult diseases. This causes problems that will be discussed in the following sections of this paper.

Problems with the ACR and CHCC in adults

A relatively low concordance between the CHCC and ACR criteria is evident from studies comparing them [9, 10]. It has been stated by adult rheumatologists that ACR criteria perform reasonably well when comparing vasculitic diseases with each other. However, they do not perform well when the clinical syndrome is poorly defined and some patients without vasculitis may falsely fulfil the criteria [9]. Furthermore, microscopic polyarteritis/polyangiitis is not included in the ACR criteria and many adult series have shown that these patients end up being classified as having Wegener’s granulomatosis (WG) with the ACR criteria [9, 10]. Anti-neutrophil cytoplasmic antibody (ANCA) is not included in either CHCC or ACR.

On the other hand, the CHCC are based on histologic findings and almost always call for a biopsy. Another limitation of CHCC in adults is with the term polyarteritis nodosa (PAN). With polyarteritis the CHCC restricts the definition of PAN to a disease involving medium or small arteries and excludes smaller vessel involvement [2, 8]. Again, in a number of adult series it has been shown that this elimination poses problems since adult patients with classical PAN occasionally do have small arterial involvement as well. The same is true in children. PAN, as restricted according to CHCC, may therefore be exceedingly rare [8]. Clinical diagnosis of PAN depends on the findings at biopsy and/or arteriography, which in certain settings may not be possible. One may also question the inclusion of HBSAg-associated classical PAN in a group of primary vasculitic diseases.

In the case of Takayasu arteritis, patients with early symptoms may not meet the CHCC or the ACR criteria [8]. An elevated ESR is not included in either criteria although it is very useful in the diagnosis.

The absence of the use of ANCA is another important problem in the existing classifications. The main obstacle to a perfect classification is the lack of understanding of the aetiology of that particular disease. In fact, recent studies have provided quite convincing data that ANCA has a role in the aetiology of a group of vasculitides, microscopic polyarteritis and WG [11, 12]. A perinuclear staining by indirect immunofluorescence (IIF) technique confirmed to be against myeloperoxidase (MPO) is associated with microscopic polyarteritis, whereas a granular cytoplasmic staining technique confirmed to be against proteinase 3 (PR3) is associated with WG [11, 12]. These two specific ANCAs are highly specific to ANCA-associated vasculitis, although they are not of course present in all cases. Studies mainly from Dr Savage’s work [12] have shown that leukocytes primed with tumour necrosis factor (TNF)-alpha and infectious agents, an impaired apoptosis and the genetic make-up of the host also play a role in the outcome of the vasculitic disease (Fig. 1). The polymorphisms for the various relevant genes the individual is carrying predisposes them to vasculitic damage. A recently reported mouse model proved that MPO-ANCA played an important role in the etiopathogenesis of ANCA-associated vasculitis and necrotizing crescentic glomerulonephritis [13]. Many authors suggest that a new classification should include ANCA not as a prerequisite but as one of the classification criteria for WG and microscopic PAN [11, 14].
Fig. 1

A simplified schema for the pathogenesis of ANCA-associated vasculitis

Problems in children

The problems with the existing criteria are even more pronounced with children, since neither the ACR nor CHCC have been validated in children [15]. ACR has not even proposed criteria for Kawasaki disease, which is the most common paediatric vasculitis in some parts of the world. Our daily paediatric practice suggests that the ACR criteria probably have a low specificity and sensitivity due to overlapping patients, infectious complications in children and early cases of vasculitis.

Of the ten main headings of CHCC, two are not seen in children (giant cell arteritis and essential cryoglobulinemic vasculitis). Actually, giant cell arteritis is excluded in the definition since it requires an age of >50 [2]. Another heading, cutaneous leukocytoclastic angiitis, is again rarely diagnosed in children. Children with leukocytoclastic vasculitis would have HSP, cutaneous PAN or secondary cutaneous vasculitis. Actually there are problems with how we paediatricians classify a child as having Henoch-Schonlein purpura (HSP), where we use neither the CHCC nor the ACR criteria. There are also problems with what we classify as PAN, which we will refer to later as well.

On the other hand, some vasculitides we observe in children (and adults), such as cutaneous PAN and Behcet’s disease, have not been included.

Paediatricians have attempted to propose classifications for use in paediatric practice. Petty and Cassidy have suggested a classification with polyarteritis, leukocytoclastic vasculitis, granulomatoid vasculitis, giant cell arteritis and others constituting the main headings [16]. Dillon modified the classification proposed by Fink in 1997 [17, 18]. However, these classifications have not been approved by international and multi-professional groups.

Problems related to specific vasculitides in childhood

Since Kawasaki disease (KD) and Henoch-Schonlein purpura (HSP) are mainly childhood diseases, the problems apply specifically to the paediatric age group. Juvenile PAN will also be reviewed because of recent new suggestions in the approach to its classification.

Henoch-Schönlein purpura

In an adult series, the sensitivity of the ACR criteria for HSP was found to be only 64.8% and that of the CHCC definition was only 31% [19]. The concordance of these two systems for HSP was very low, with only 12 out of 71 patients fulfilling criteria for both classifications. Furthermore, according to the ACR criteria, HSP can easily be misclassified as other forms of vasculitis [8].

The problem is more pronounced in children. None of us apply the two out of four criteria suggested by the ACR: age <20, palpable purpura, bowel angina and biopsy showing granulocytes [6]. CHCC defines HSP according to IgA deposits at biopsy. Paediatricians would not routinely—if ever—biopsy a child to diagnose HSP.

Kawasaki disease

The diagnosis of Kawasaki disease (KD) is based on the presence of at least five days of fever plus four of the five principal clinical features suggested by Kawasaki [20]. KD is the leading cause of acquired heart disease in children in the US [21]. Unfortunately, a significant portion of the patients do not fulfil these criteria and are defined as “atypical” or “incomplete” KD. These “incomplete” patients are a challenge for the paediatrician since management and treatment in terms of cardiac disease is of vital importance. The American Academy of Pediatrics and the American Heart Association have proposed a new algorithm to aid clinicians in deciding which child deserves cardiac work-up and treatment [21]. However, further work is clearly needed to define KD, and the subsequently revised criteria needs to be validated in various groups.

Juvenile polyarteritis

The problems cited for adults apply to children as well [15]. As for CHCC, patients with overlapping features pose a problem. Children may have small artery involvement in the setting of polyarteritis nodosa defined by CHCC [22]. Overlap patients may occur, as in the case of a child reported by us who was initially classified as classic PAN, according to CHCC, with mid-size renal artery involvement, who after a couple of years of remission went on to develop microscopic disease with necrotizing glomerulonephritis [23]. In a series of 35 children with juvenile polyarteritis, only five had the typical necrotizing glomerulonephritis and high MPO-ANCA, whereas eight had renal artery involvement in their angiograms [22]. However, the classification of the remaining half was not as definite.

For the ACR criteria, the specificity is regarded to be low, and is especially hampered with HSP and other causes of rapidly progressive glomerulonephritis. Hepatitis B surface antigen (HBsAg) is among the criteria; however, recent vaccination policies have decreased this to nil in children in the last ten years. The same problem applies to the criteria suggested by us for paediatric PAN in 1992 [24]. We suggested two major and ten minor criteria for the childhood disease and concluded that the presence of five of these criteria, including at least one major criterion, was highly suggestive of childhood PAN [24]. One of the major criteria was renal involvement in any form. Brogan et al [25] suggested modified ACR criteria for use in childhood cases. None of these criteria have been formally tested for specificity nor sensitivity in multicentre studies.

We have recently collected data from 21 centres from Europe, and South and North America through a survey [26]. This survey aimed to collect retrospective data on paediatric patients diagnosed as having PAN. The registry form was distributed via electronic mail. Paediatricians caring for vasculitic disorders were asked to group their patients into one of the following groups that were separated by aetiology and/or clinical features [26]:
  1. 1.

    Microscopic polyarteritis of adults associated with MPO-ANCA: This was regarded as a separate group since the aetiology was associated with ANCA. The clinical presentation was determined by necrotizing glomerulonephritis with/without pulmonary disease.

  2. 2.

    Classic PAN associated with HBsAg: This was regarded as a separate group since it was the only subgroup of PAN that was determined by immunecomplexes against HBsAg. This calls for a different approach in management and therapy as well [27].

  3. 3.

    Cutaneous polyarteritis: Represented disease confined to the skin with no organ involvement except for musculoskeletal involvement (arthralgia, arthritis, myalgia). ANCA was negative.

  4. 4.

    Systemic PAN: Necrotizing arteries in any organ in mid- or smaller-sized arteries. Although ANCA was occasionally positive by IIF, it was not directed against MPO. Patients were diagnosed either by biopsy of the affected tissue and/or arteriographies. This constituted the major group of paediatric patients.


One feature that stood out in the latter two was the association with the upper respiratory tract, especially streptococcal infections that may have a role in the etiopathogenesis of these subgroups, especially for the cutaneous and systemic PANs [26]. The association of streptococci has been previously referred to in a number of paediatric papers.

The majority of the 110 patients were simply classified as systemic PAN by paediatricians. This group was not associated with ANCA. Mild kidney disease was present in about 10%; however, they did not present with rapidly progressive glomerulonephritis and had no pulmonary involvement. Some also had mid-size artery involvement elsewhere in the kidney even in the form of renal artery aneurysms [26]. HBsAg-associated classical PAN was very rare (<5%), which is probably due to successful vaccination programs. Microscopic polyarteritis, as defined for adults, was again rarer in children. One handicap of this study was that there was poor participation from nephrology centres, and this may be the explanation for the lower rate of this group when compared to a previous single-centre paediatric series.

Diagnostic approach to a patient with crescentic glomerulonephritis

The diagnosis of this condition requires the knowledgeable integration of clinical, histological (immunohistological) and laboratory data. A biopsy is required for the diagnosis of a patient with crescentic glomerulonephritis. A negative immunofluorescence study would suggest an ANCA-related vasculitis. ANCA should be tested by IIF and subsequently by ELISA against MPO and PR3 specificity [11, 12, 28]. ANCAs directed against other antigens do not have any clinical significance; thus an IIF result is not sufficient and antigen specificity should be tested for by ELISA. On the other hand, one must remember that renal limited vasculitis and Churg-Strauss syndrome may also present with high ELISA titres against these antigens, especially MPO.

Chest and computed tomography scans may give information regarding upper and lower respiratory tract disease [11]. An eosinophil count and a history of asthma would be needed to rule out Churg-Strauss syndrome, although it is very rare in children.

On the other hand, a positive immunofluorescence would suggest an immune complex pathology. Systemic lupus erythematosus would be the most common one in children. IgA deposits would suggest HSP.

A number of other rare diseases have also been associated with this presentation. Lymphomatoid granulomatosis would be suggested by the pathology [11]. Behcet’s disease has also been reported; however, this would have the typical clinical features [29].

One must also remember that not all PAN and WG patients present with crescentic glomerulonephritis [30]. We have reported patients with mild clinical manifestations of kidney disease [30]. ANCA was not studied, since these patients constituted a retrospective group. The histopathology of microscopic polyarteritis and WG differences were reviewed in an adult series of 173 patients [31]. These pointed out certain differences, such as increased glomerulosclerosis in the former group. The authors have concluded that ANCA may be helpful in the classification of ANCA-associated vasculitis and that the pathogenesis of renal disease could be different in these two vasculitides [31].

Future prospects

We need classification criteria to categorize paediatric vasculitides and to manage them appropriately. This would also enable us to form international study groups to compare treatment protocols in vasculitic diseases. We need international data and/or consensus to serve as a basis for satisfactory classification criteria. To increase the sensitivity and specificity of the existing classification criteria, we need to revise them with the CHCC and paediatric experience. Then we can validate these data to improve and use them as diagnostic criteria.


  1. 1.
    Sundel R, Szer I (2002) Vasculitis in childhood. Rheum Dis Clin N Am 28:625–54CrossRefGoogle Scholar
  2. 2.
    Jennette JC, Falk RJ, Andrassy K, Bacon PA, Churg J, Gross WL, Hagen EC, Hoffman GS, Hunder GG, Kallenberg CG (1994) Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum 37:187–92PubMedGoogle Scholar
  3. 3.
    Gonzalez-Gay MA, Garcia-Porrua C (2002) Systemic vasculitides. Best Pract Res Clin Rheumatol 16:833–45CrossRefPubMedGoogle Scholar
  4. 4.
    Lightfoot RW Jr, Michel BA, Bloch DA et al (1990) The ACR 1990 criteria for the classification of polyarteritis nodosa. Arthritis Rheum 33:1088PubMedGoogle Scholar
  5. 5.
    Leavitt RY, Fauci AS, Blach DA et al (1990) The ACR 1990 criteria for the classification of Wegener’s granulomatosis. Arthritis Rheum 33:1101PubMedGoogle Scholar
  6. 6.
    Mills JA, Michel BA, Bloch DA, Calabrese LH, Hunder GG, Arend WP, Edworthy SM, Fauci AS, Leavitt RY, Lie JT, et al. (1990) The American College of Rheumatology 1990 criteria for the classification of Henoch-Schonlein purpura. Arthritis Rheum 33:1114–1121PubMedGoogle Scholar
  7. 7.
    Arend WP, Michel BA, Bloch DA et al (1990) The ACR 1990 criteria for the classification of Takayasu arteritis. Arthritis Rheum 33:1129–34PubMedGoogle Scholar
  8. 8.
    Luqmani RA, Robinson H (2001) Introduction to and classification of the systemic vasculitides. Best Prac Res Clin Rheumatol 15:187–202CrossRefGoogle Scholar
  9. 9.
    Watts RA, Scott DG Watts RA (2000) Systemic vasculitis: epidemiology, classification and environmental factors. Ann Rheum Dis 59:161–3CrossRefPubMedGoogle Scholar
  10. 10.
    Bruce IN, Bell AL (1997) A comparison of two nomenclature systems for primary systemic vasculitis. Br J Rheumatol 36:453–8CrossRefPubMedGoogle Scholar
  11. 11.
    Gross WL, Trabandt A, Reinhold-Keller E (2000) Diagnosis and evaluation of vasculitis. Rheumatol 39:245–52CrossRefGoogle Scholar
  12. 12.
    Day CJ, Hewins P, Savage CO (2003) New developments in the pathogenesis of ANCA-associated vasculitis. Clin Exp Rheumatol 21(Suppl 32):S35–S48Google Scholar
  13. 13.
    Xiao H, Heeringa P, Hu P, Liu Z, Zhao M, Aratani Y, Maeda N, Falk RJ, Jennette JC (2002) Antineutrophil cytoplasmic autoantibodies specific for myeloperoxidase cause glomerulonephritis and vasculitis in mice. J Clin Invest 110:955–63CrossRefPubMedGoogle Scholar
  14. 14.
    Sorensen SF, Slot O, Tvede N, Petersen J (2000) A prospective study of vasculitis patients collected in a five year period: evaluation of the Chapel Hill nomenclature. Ann Rheum Dis 59:478–82CrossRefPubMedGoogle Scholar
  15. 15.
    Ozen S (2002) The spectrum of vasculitis in children. Best Pract Res Clin Rheumatol 16:411–425CrossRefPubMedGoogle Scholar
  16. 16.
    Petty RE, Cassidy JT (2001) Vasculitis and its classification. In: Textbook of pediatric rheumatology, 4th edn. WB Saunders Co, Philadelphia, PA, pp 564–579Google Scholar
  17. 17.
    Fink CW (1997) Polyarteritis and other diseases with necrotizing vasculitis in childhood. Arthritis Rheum 20:378–84Google Scholar
  18. 18.
    Dillon M (1998) Primary vasculitis in children. In: Maddison PJ, Isenberg DA, Woo P, Glass DN (eds) Oxford textbook of theumatology. Oxford University Press, Oxford, UK, pp 1402–12Google Scholar
  19. 19.
    Murali NS, George R, John GT, Chandi SM, Jacob M, Jeyaseelan L, Thomas PP, Jacob CK (2002) Problems of classification of Henoch Schonlein purpura: an Indian perspective. Clin Exp Dermatol 27:260–3CrossRefPubMedGoogle Scholar
  20. 20.
    Kawasaki T (1967) Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desquamation of the fingers and toes in children. Arerugi 16:178–222PubMedGoogle Scholar
  21. 21.
    Newburger JW, Takahashi M, Gerber MA, Gewitz MH, Tani LY, Burns JC, Shulman ST, Bolger AF, Ferrieri P, Baltimore RS, Wilson WR, Baddour LM, Levison ME, Pallasch TJ, Falace DA, Taubert KA;Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association (2004) Diagnosis, treatment and long-term management of Kawasaki disease: a statement for health professionals from the committee on rheumatic fever, endocarditis, and Kawasaki disease, Council on Cardiovascular Disease in the Young, American Heart Association. Pediatrics 114:1708–1733CrossRefPubMedGoogle Scholar
  22. 22.
    Bakkaloglu A, Ozen S, Baskin E, Besbas N, Gur-Guven A, Kasapcopur O, Tinaztepe K (2001) The significance of antineutrophil cytoplasmic antibody in microscopic polyangitis and classic polyarteritis nodosa. Arch Dis Child 85:427–30CrossRefPubMedGoogle Scholar
  23. 23.
    Ozen S, Ben-Chetrit E, Bakkaloglu A et al (2001) Polyarteritis nodosa in patients with Familial Mediterranean Fever (FMF): a concomitant disease or a feature of FMF? Semin Arthritis Rheum 30:281–7CrossRefPubMedGoogle Scholar
  24. 24.
    Ozen S, Besbas N, Saatci U, Bakkaloglu A , Gur H, Tinaztepe K, Calguneri M, Turgan C, Turkmen A, Akpolat I, Danaci M, Besbas N, Akpolat T (1992) Diagnostic criteria for polyarteritis nodosa in childhood. J Pediatr 120:206–9PubMedGoogle Scholar
  25. 25.
    Brogan PA, Davies R, Gordon I, Dillon MJ (2002) Renal angiography in children with polyarteritis nodosa. Pediatr Nephrol 17:277–283CrossRefPubMedGoogle Scholar
  26. 26.
    Ozen S, Anton J, Arisoy N, Bakkaloglu A, Besbas N, Brogan P, Garcia-Consuegra J, Dolezalova P, Dressler F, Duzova A, Ferriani VP, Hilario MO, Ibanez-Rubio M, Kasapcopur O, Kuis W, Lehmann TJ, Nemcova D, Nielsen S, Oliveira SK, Schikler K, Sztajnbok F, Terreri MT, Zulian F, Woo P (2004) Juvenile polyarteritis: results of a multicenter study of 110 children. J Pediatr 145:517–22CrossRefPubMedGoogle Scholar
  27. 27.
    Guillevin L, Mahr A, Cohen P, Larroche C, Queyrel V, Loustaud-Ratti V, Imbert B, Hausfater P, Roudier J, Bielefeld P, Petitjean P, Smadja D; French Vasculitis Study Group (2004) French Vasculitis Study Group. Short-term corticosteroids then lamivudine and plasma exchanges to treat hepatitis B virus-related PAN. Arthritis Rheum 15:482–7CrossRefGoogle Scholar
  28. 28.
    Weidner S, Geuss S, Hafezi-Rachti S, Wonka A, Rupprecht HD (2004) ANCA-associated vasculitis with renal involvement: an outcome analysis. Nephrol Dial Transplant 19:1403–11CrossRefPubMedGoogle Scholar
  29. 29.
    Akpolat T, Akkoyunlu M, Akpolat I, Dilek M, Odabas AR, Ozen S (2002) Renal Behcet’s disease: a cumulative analysis. Semin Arthritis Rheum 31:317–37CrossRefPubMedGoogle Scholar
  30. 30.
    Besbas N, Ozen S, Saatci U, Topaloglu R, Tinaztepe K, Bakkaloglu A (2000) Renal involvement in polyarteritis nodosa: evaluation of 26 Turkish children. Pediatr Nephrol 14:325–7CrossRefPubMedGoogle Scholar
  31. 31.
    Hauer HA, Bajema IM, van Houwelingen HC et al on behalf of the EUVAS (2002) Renal histology in ANCA-associated vasculitis. Kidney Int 61:80–89CrossRefPubMedGoogle Scholar

Copyright information

© IPNA 2005

Authors and Affiliations

  1. 1.Department of Pediatrics, Faculty of MedicineHacettepe UniversityAnkaraTurkey

Personalised recommendations