Definition and Classification

HSP is one of the most common vasculitides of childhood. It is characterized by nonthrombocytopenic purpura, arthritis and arthralgia, abdominal pain and gastrointestinal hemorrhage, and glomerulonephritis. A diagnostic triad of purpuric rash, arthritis, and abnormalities of the urinary sediment was proposed by Schönlein in 1837, and Henoch described the association of purpuric rash, abdominal pain with bloody diarrhea, and proteinuria in 1874. The term anaphylactoid purpura was applied by Gairdner in 1948. The American College of Rheumatology (ACR) criteria are now superseded by new pediatric criteria. The original ACR criteria were derived by comparing 85 patients with HSP to 722 controls with other vasculitides and had a sensitivity of 87.1% and specificity of 87.7% for the classification of HSP. In 2010 the vasculitis working group of the Pediatric Rheumatology European Society (PRES) proposed new classification criteria for pediatric vasculitides, endorsed by the European League Against Rheumatism (EULAR), which required palpable purpura with lower limb predominance (as mandatory criteria) plus at least one among the following four features: (1) diffuse abdominal pain, (2) biopsy showing typical leukocytoclastic vasculitis or proliferative glomerulonephritis with predominant immunoglobulin A (IgA) deposition, (3) arthritis or arthralgia, and (4) renal involvement (any hematuria and/or proteinuria). In the case of purpura with atypical distribution, a demonstration of IgA deposit in a biopsy is required. The PRES/EULAR criteria have been prospectively validated through international, Web-based prospective data that included 827 patients with HSP and 349 with other vasculitides. The sensitivity and specificity of the new classification criteria were 100% and 87%, respectively. Table 33-1 summarizes the new criteria, and the sensitivity and specificity of individual clinical features for the classification HSP.

In 2012, a new definition was proposed as part of an overall updating of nomenclature of systemic vasculitides undertaken at an international consensus conference that took place in Chapel Hill. It must be emphasized that definitions of disease are not the same as formal, validated classification criteria, however. Using that nomenclature, HSP is now referred to as IgA vasculitis (Henoch–Schönlein) and is defined as vasculitis, with IgA1-dominant immune deposits affecting small vessels (predominantly capillaries, venules, or arterioles); it often involves the skin and the gastrointestinal tract, and frequently causes arthritis. Glomerulonephritis indistinguishable from IgA nephropathy may occur.

Epidemiology

HSP is predominantly a disease of childhood, although a similar syndrome has been reported in adults. It occurs most frequently between the ages of 3 and 15 years and is more common in boys than in girls (1.5 : 1). The condition is rare in children younger than 2 years old. Prospective data from an international cohort of 827 patients (49% girls) with HSP showed a mean age at onset of 6.9 ± 3 (range, 1 to 16.2) years, and at diagnosis at 7 ± 3 years.

An incidence of 13.5 cases per 100,000 children per year was observed in an unselected childhood population in Belfast, Northern Ireland. An incidence of 9 to 86 per 100,000 children was estimated by Farley and colleagues. In the latter study, incidence was highest among Hispanic children (86 per 100,000) and children in lower socioeconomic groups (69 per 100,000), compared with children in higher socioeconomic or different racial groups (9 to 11 per 100,000). In a study from Britain, although the combined incidence was 20.4 per 100,000, it was 70.3 per 100,000 in the 4- to 6-year-olds age group. Striking seasonal variations have been observed, with most cases occurring in winter, often (30% to 50%) preceded by an upper respiratory tract infection.

Etiology and Pathogenesis

Many reports have implicated infection, particularly with β-hemolytic streptococci, as a potential trigger for this disease. Some investigators have, however, doubted this association. Other preceding infections, including vaccination, viral infection (e.g., varicella, rubella, rubeola, hepatitis A and B), Mycoplasma pneumoniae , Bartonella henselae, and Helicobacter pylori have been described.

As the term anaphylactoid purpura suggests, allergy has been regarded by some as the basis for development of this disease after insect bites and exposure to drug and dietary allergens. A possible pathogenic role of IgE in HSP has been suggested for some cases of HSP nephritis; it has been suggested that stimulation of IgE-sensitized mast cells by specific antigens in the presence of IgA circulating immune complexes leads to release of vasoactive substances, increased capillary permeability, and perivascular deposition of IgA immune complexes.

The characteristic vascular deposition of IgA strongly suggests, however, that HSP is a predominantly IgA-mediated dysregulated immune response to antigen and may operate through the alternative complement pathway. Although the pathogenic mechanisms of nephritis are still not delineated, studies suggest that galactose-deficient IgA1 is recognized by antiglycan antibodies, leading to the formation of circulating immune complexes and their mesangial deposition, which results in renal injury in HSP. An increase in the levels of poorly galactosylated IgA1 appears insufficient in itself to cause HSP nephritis or IgA nephropathy, because first-degree relatives may have high serum levels of poorly galactosylated IgA1 but no signs of nephritis. Thus it is likely that a “second hit” is required for high levels of poorly galactosylated IgA1 to form immune complexes that result in nephritis. It is suggested that this second hit is the formation of antiglycan IgG or IgA antibodies (perhaps triggered by infection) that then go on to form large circulating immune complexes with the poorly galactosylated IgA1 that are prone to deposition. Currently, the genes controlling IgA1 glycosylation are unknown. Other factors that may modulate IgA synthesis and galactosylation include B-cell programming at the time of antigen encounter, Toll-like receptor activation, and local cytokine production. It has been suggested that the genetic mechanisms controlling these processes should be further explored to try to better understand the genetic contribution to the pathogenesis of HSP nephritis and IgA nephropathy. Thus, although there remain many unanswered questions in relation to this pathogenetic model, increased serum levels of poorly galactosylated IgA1 remain the most consistent finding in patients with HSP nephritis and IgA nephropathy, and these almost certainly predispose to the formation of IgA immune complexes with resulting vasculitis.

Disorders of coagulation and its activation are also associated with the development of HSP or an HSP-like vasculitis. It is recognized that disease activity may be linked to a rapid decline of factor XIII, particularly in patients with severe abdominal involvement. This may be useful as a prognostic or diagnostic marker because the decline occurs before classic skin rash and thus could allow early diagnosis of HSP, when abdominal symptoms and signs predominate. There is no information regarding the diagnostic specificity of a fall in factor XIII, however. Anecdotal reports of factor XIII replacement to successfully treat severe abdominal symptoms in HSP are described ; no high-quality controlled data relating to this treatment are available, however. Factor XIII also declines prior to recurrence of HSP, but it is unknown whether this is a merely a secondary epiphenomenon or a true causative association.

One study concluded that oxidant stress, especially lipid peroxidation, was involved in the origin of renal injury. Vasculitis may develop after antirheumatic therapy, including administration of methotrexate and anti–tumor necrosis factor agents.

Genetic Background

Familial clusters of the disease may occur, with siblings affected simultaneously or sequentially. Investigations from Spain have presented preliminary data on genetic associations. The frequency of human leukocyte antigen (HLA)-B35 was increased in patients who developed nephritis. The incidence of HLA-DRB1*01 was also increased compared with matched controls, and HLA-DRB1*07 was decreased. A study of unselected children with HSP from Turkey showed that HLA-A2, -A11, and -B35 antigens were associated with a significantly increased risk of HSP, whereas HLA-A1, -B49, and -B50 antigens were associated with decreased risk for the disease. There was no association, however, between HLA class 1 alleles and renal involvement.

Although there were no general associations with the expression of intercellular adhesion molecule 1 in patients compared with controls, a K/E polymorphism at codon 469 was significantly decreased in those who did not develop severe gastrointestinal manifestations (and possibly in patients without renal sequelae). In studies from Israel and Turkey, mutations in the familial Mediterranean fever (MEFV) gene were frequent in patients with HSP. Other genetic polymorphisms have also been implicated and have been reviewed elsewhere.

Several polymorphisms relating to disease susceptibility, severity, and/or risk of renal involvement have been described. Studies of this nature have been hampered by relatively small patient numbers and thus lack the power to be definitive or necessarily applicable to all racial groups. It is, however, increasingly apparent that the genetic contribution is complex and probably polygenic in nature.

Clinical Manifestations

Clinical characteristics of HSP are presented in Table 33-2 . The onset is often acute, with the principal manifestations appearing sequentially over several days to weeks. Nonspecific constitutional signs, such as a low-grade fever or malaise, are often present.

Cutaneous Involvement

The presence of palpable purpura is characteristic. This rash is most prominent on dependent or pressure-bearing surfaces, especially the lower extremities and buttocks, but it may occur in other areas. The cutaneous lesions range from small petechiae to large ecchymoses to rare hemorrhagic bullae; they tend to occur in crops and progress in color from red to purple to brown ( Figs. 33-2 and 33-3 ). Ulceration may occasionally develop in large ecchymotic areas. The rash is often preceded by maculopapular or urticarial lesions. Subcutaneous edema over the dorsa of the hands and feet and around the eyes, forehead, scalp, and scrotum may occur early in the disease, particularly in the very young child.

Henoch–Schönlein purpura. A , These purpuric lesions appeared on the lower extremities of a 10-year-old boy who had an acute, self-limited illness characterized by fever, arthritis, melena, and transient hematuria. Notice the periarticular swelling around the ankles and knees. B, Purpura on dorsum of the hand of a 14-year-old boy with Henoch–Schönlein purpura.

Palpable purpuric lesions involve the heels and ankles of a patient with Henoch–Schönlein purpura.

Pathology

The pathological lesion of HSP is leukocytoclastic vasculitis ( Fig. 33-4 ), although in routine practice biopsies are not always performed. In the skin, this is demonstrated in the dermal capillaries and postcapillary venules. Deposition of IgA (principally IgA ₁ ) in these lesions is characteristic. It is possible to fail to detect IgA deposition in vascular tissue in some cases of HSP, especially if the biopsy was obtained from the middle of a lesion where the presence of proteolytic enzymes can result in negative staining for IgA.

Leukocytoclastic vasculitis in the skin of a patient with Henoch–Schönlein purpura. Hematoxylin-eosin stain; original magnification ×480.

In the kidneys, proliferative glomerulonephritis ranges from focal and segmental lesions to severe crescentic disease ( Fig. 33-5 ). Group A streptococcal antigen (i.e., nephritis-associated plasmin receptor) was identified by immunofluorescent microscopy in the mesangium of 10 of 33 children. Levy and co-workers provided a comprehensive review of the renal pathology. The principal lesion is an endocapillary proliferative glomerulonephritis with an increase in endothelial and mesangial cells. All gradations of severity may be present in the same biopsy. There may be marked interstitial inflammatory disease, but vasculitis per se is usually not present. Fluorescence microscopy confirms deposits of Ig, principally IgA, but it is often accompanied by IgG, fibrin, C3, and properdin in most involved glomeruli. These deposits are invariably in mesangial cells, but peripheral capillary loops also are involved in more severe cases. Dense deposits in the mesangium and occasionally in the subendothelial and paramesangial regions are present on electron microscopy. Thickening and splitting of the basement membrane, caused by the interposition of mesangial cell cytoplasmic material, are notable.

A, Diffuse, mesangial, proliferative nephritis in Henoch–Schönlein purpura. Hematoxylin-eosin stain. B, Mesangial and capillary wall deposition of immunoglobulin A (immunoperoxidase stain) in Henoch–Schönlein purpura.

Differential Diagnosis

HSP must be distinguished from immune thrombocytopenic purpura, acute poststreptococcal glomerulonephritis, SLE, septicemia, disseminated intravascular coagulation, hemolytic-uremic syndrome, the papular-purpuric gloves-and-socks syndrome, and other types of vasculitis. FMF can also mimic or occur in association with HSP in areas where FMF is endemic.

The more common causes of an acute surgical abdomen with abdominal pain and gastrointestinal tract bleeding must be considered. A tender abdominal mass may indicate intussusception, and abdominal tenderness with an elevated level of serum amylase and/or lipase suggests acute pancreatitis. Punch biopsy of a cutaneous lesion may assist in the diagnosis of difficult cases by demonstrating leukocytoclastic vasculitis characterized by deposition of IgA and C3. A renal biopsy is indicated only in children with persistent or significant renal manifestations. Indications for diagnostic renal biopsy in children with HSP are as follows :

• •
  

  Acute nephritic or nephrotic syndrome at presentation

  
  
• •
  

  Raised blood level of creatinine; hypertension or oliguria

  
  
• •
  

  Heavy proteinuria (early morning urine protein to creatinine ratio greater than 100 mg/mmol)

  
  
• •
  

  Persistent proteinuria (not declining) after 4 weeks

  
  
• •
  

  Impaired renal function (GFR less than 80 mL/min/1.73 m ² )

Laboratory Examination

There are no diagnostic laboratory abnormalities. The platelet count is normal or increased, differentiating this form of purpura from that caused by thrombocytopenia. A moderate leukocytosis up to 20,000 white cells/mm ³ (20 × 10 ⁹ /L) with a shift to left is identified in some children. Normochromic anemia is often related to gastrointestinal blood loss, confirmed by a positive stool guaiac examination in 80% of patients with abdominal complaints. Antinuclear antibody (ANA) and rheumatoid factor (RF) are usually absent.

Although renal disease may occur in the absence of overt urinary findings and minimal abnormalities, such as hematuria, are not necessarily associated with a severe glomerular lesion, these laboratory abnormalities usually demonstrate a direct correlation with the severity of the proliferative changes. Patients occasionally have decreased concentrating ability and creatinine clearance. Proteinuria, severe enough to result in hypoalbuminemia, may occur.

Although levels of C1q, C3, and C4 are usually normal, activation of the alternate complement pathway is demonstrated in half of patients during the acute illness by presence of C3d, low levels of total hemolytic complement, and decreased serum concentrations of properdin and factor B. Plasma levels of von Willebrand factor antigen are elevated, indicating endothelial cell damage.

Circulating IgA-containing immune complexes and cryoglobulins may be present. Serum IgA and IgM concentrations are increased in half of the patients during the acute phase of the disease. An increased number of circulating IgA-producing cells was found in one study in definite HSP cases but not in other forms of leukocytoclastic vasculitis. ANCAs are typically absent.

Radiological Examination

Plain radiographs may demonstrate decreased intestinal motility with dilated loops of bowel in children with abdominal involvement. Ultrasound studies can identify specific gastrointestinal abnormalities in children with abdominal complaints. Magnetic resonance imaging and magnetic resonance angiography of the brain can define the extent of cerebral vasculitis. Occasionally, intussusception is identified on a barium study and is relieved by it if performed early in the course. Epididymal enlargement, subcutaneous scrotal swelling, hydrocele, or, rarely, testicular torsion can be confirmed if necessary by scrotal ultrasonography.

Treatment

Treatment is supportive with maintenance of good hydration, nutrition, and electrolyte balance; control of pain is accomplished with simple analgesics such as acetaminophen. If necessary, control of hypertension is attempted. Although glucocorticoids dramatically decrease the severity of joint and cutaneous disease, they are not usually indicated for management of these manifestations. Short-term glucocorticoid therapy is effective in relieving the pain of severe orchitis. Prednisone has been advocated in children with severe gastrointestinal disease or hemorrhage. The severity of disease may occasionally prompt the use of intravenous corticosteroids. However, studies do not demonstrate a clear advantage of prednisone over supportive therapy (e.g., nasogastric suction, parenteral nutrition, antibiotics). Pulmonary hemorrhage is an extremely rare and sometimes fatal complication, which requires aggressive immunosuppressive treatment, combining intravenous (IV) methylprednisolone with another immunosuppressive agent such as cyclophosphamide or cyclosporine, and supportive care.

Course of the Disease and Prognosis

In two thirds of children, HSP runs its entire course within 4 weeks of onset. Younger children have a shorter course and fewer recurrences than do older patients. One third to half of the children have at least one recurrence that commonly consists of a rash and abdominal pain, with each episode usually being similar but briefer and milder than the preceding one. Most exacerbations take place within the initial 6-week period but may occur as late as 2 years after onset. They may be spontaneous or coincide with repeated respiratory tract infections. The severity of the cutaneous leukocytoclastic vasculitis does not correlate with visceral involvement.

Prognosis is excellent for most children. Significant morbidity or mortality is associated with gastrointestinal tract lesions in the short term and with nephritis in the long term. The development of major indications of renal disease, particularly those with a mixed nephritic-nephrotic syndrome within the first 6 months after onset or the occurrence of numerous exacerbations associated with nephropathy suggests a poor prognosis for renal function. Additional poor prognostic factors are decreased factor XIII activity; hypertension; renal failure at onset, and, if a renal biopsy had been performed, an increased number of glomeruli with crescents; macrophage infiltration; and tubulointerstitial disease.

The reported outcome of children with renal disease is highly variable. With minimal lesions, more than 75% recover within 2 years; in contrast, 66% of children with crescentic glomerulonephritis in more than 80% of glomeruli progress to renal failure within the first year. The worst outcome is associated with the presence of the nephrotic or nephritic syndrome at onset. Almost half of such children have active renal disease or renal insufficiency at follow-up periods of 6 or more years. Thus the extent of the renal disease is an important determinant of long-term outcome.

Overall, less than 5% of children progress to end-stage renal failure. HSP accounts for less than 1% of children with renal failure from all causes. In a follow-up evaluation of 64 children, the renal survival rate at 10 years was 73%, and initial renal insufficiency was the best predictor of the future course of nephritis. Similar results were found in a multicenter study on 443 patients with HS nephritis from Turkey, in which 87% patients had a favorable outcome and 13% an unfavorable one; 1.1% children showed end-stage renal disease at follow-up. All patients who showed end-stage renal disease had nephritic-nephrotic syndrome at presentation and greater than 50% crescents on renal biopsy.

Patients who have had clinical nephritis should be followed closely for at least 5 years. In one study, 8 of 12 patients had mesangial IgA deposition at follow-up of 2 to 9 years, despite an apparent clinical remission of their renal disease. Later, 16 of 44 pregnancies in adults were complicated by proteinuria or hypertension, even in the absence of active renal disease. Of 18 patients in the Belfast study who had a nephrotic or nephritic syndrome at onset and who were followed for a mean of 8.3 years, one died and three had persistent urinary abnormalities but no azotemia. The overall mortality rate was less than 1%, and the morbidity rate was 1.1%. This somewhat optimistic outcome is tempered by a study of 16 children from Minnesota, which indicated that the longer a child was followed, the more likely it was that renal disease would become clinically evident.