380 Arch Dis Child 1999;80:380–383

RHEUMATOLOGY Arch Dis Child: first published as 10.1136/adc.80.4.380 on 1 April 1999. Downloaded from

Henoch-Schönlein

E J Tizard

Henoch-Schönlein purpura (HSP) is the most syndrome may be associated with renal insuY- common vasculitic disease of childhood. It is a ciency, , or both. Hyper- multisystem disease most commonly aVecting tension may be associated with acute skin, joints, gastrointestinal tract, and kidneys, but has also been reported in the absence of but other organs may be aVected. Epidemio- urinary abnormality. logical studies have shown HSP to have an Florid cerebral manifestations including sei- annual incidence of approximately 13.5–18/ zures, paresis or coma are uncommon but have 100 000 children.12Although this is a condition been reported.7 It has been suggested that mild that can occur from age 6 months to adulthood, cerebral involvement presenting as headaches 50% of cases occur in children under 5 years of may occur in as many as a third of cases, and age and 75% are under 10 years. In most this may coincide with electroencephalogram reports HSP is more common in boys.3 abnormalities.8 Scrotal involvement is not The cause remains unknown although there uncommon and occasionally resembles tes- is often an antecedent respiratory infection. ticular torsion, which must be excluded. Other Streptococcal infections have been implicated rare manifestations are cholecystitis and myo- but other organisms including adenovirus, par- cardial infarction. Interstitial lung disease with vovirus, and mycoplasma have also been impairment of lung diVusion capacity has been reported to precede HSP. The increased reported, although this was clinically incidence in winter and spring supports an insignificant.9 infectious trigger in a susceptible individual. The American College of Immunopathology 1990 criteria for the classification of HSP The pathogenetic mechanisms underlying http://adc.bmj.com/ aimed to identify diagnostic criteria to diVeren- HSP are poorly understood. Widespread ab- tiate HSP from other vasculitic diseases.4 The normalities in IgA have been described includ- four criteria identified, of which two are neces- ing raised serum IgA concentrations, IgA sary to make the diagnosis, are: immune complexes, IgA class such + age < 20 years at onset as IgA rheumatoid factor (RF),10 IgA ANCA + palpable purpura (antineutrophil cytoplasmic ),11 12 and + “bowel angina” (diVuse abdominal pain or IgA AECA (antiendothelial cell antibody).13

bowel ischaemia usually with bloody diar- IgA deposits are also found in skin biopsies and on September 25, 2021 by guest. Protected copyright. rhoea) deposited within glomeruli. There are reports + biopsy evidence of granulocytes in the walls of HSP developing in patients with IgA of arterioles or venules. nephropathy14 as well as simultaneous develop- The diagnosis is usually made after the ment of IgA nephropathy and HSP within appearance of the classic rash that primarily sibships.15 The strikingly similar histological aVects the extensor surfaces and may be abnormalities of mesangial IgA deposition urticarial or purpuric. Joint involvement occurs suggest a common immunopathogenic proc- in 60–84% of cases and generally aVects the ess. IgA and IgG ANCA mirrored disease ankles and knees. It is often the most incapaci- activity in a girl with frequently relapsing tating part of the initial illness. Gastrointestinal disease.12 AECA has been detected in patients disease occurs in up to 76% of patients varying with vasculitic disorders such as Kawasaki dis- from colicky abdominal pain, nausea, and ease and reported to be associated with renal vomiting to intestinal haemorrhage, intussus- involvement in HSP, although complement ception, pancreatitis, and hydrops of the gall dependent cytotoxicity to glomerular endothe- bladder. There is an increased risk of renal dis- lial cells has not been shown.13 Consultant Paediatric ease in those with bloody stools. Complement abnormalities have been de- Nephrologist, Department of The reported incidence of renal disease scribed in association with HSP: C2 deficiency, 5 Paediatrics, ranges from 20–100%. In 80% of those with homozygous null C4 phenotypes, and C4B Southmead Hospital, renal involvement it becomes apparent within deficiency. Other abnormalities including Westbury-on-Trym, the first four weeks of the illness. The remain- glomerular C3 and deposition, low Bristol BS10 5NB, UK der predominantly occurs over the next two CH50 and properdin, and raised C3d concen- E J Tizard months although a few are further delayed.6 trations in the acute phase of the disease have Correspondence to: Haematuria with or without is the suggested complement activation. However, a Dr Tizard. most common renal feature. Acute nephritic study of three multimolecular complement Henoch-Schönlein purpura 381

activation protein complexes has failed to sup- Treatment port a role for complement activation in HSP.16 The natural history of HSP is a self limiting ill- Arch Dis Child: first published as 10.1136/adc.80.4.380 on 1 April 1999. Downloaded from ness in most cases. Acutely, abdominal pain Laboratory markers and joint pain may be debilitating, and There are no specific diagnostic markers of have been used with some success in alleviating HSP. If present, anaemia is usually a result of the abdominal symptoms.21 Gut may gastrointestinal or associated with result in significant blood loss and occasionally acute nephritis. Thrombocytosis is a feature ischaemic bowel may require resection. There that may be related to disease severity. Renal is no clear evidence that treatment alters the involvement results in haematuria with or natural history of the disease. It is, however, the without proteinuria, which may be in the significant long term morbidity associated with nephrotic range. A rapidly progressive nephri- renal disease that has led to trials of steroids tis may result in renal insuYciency and and other immunosuppressive drugs. requires histological assessment. In 1992 Mollica et al reported a prospective, Von Willebrand factor is synthesised randomised, controlled study of steroids in the by and is present in endothelial cells. High prevention of HSP nephritis.22 Of the 221 chil- concentrations have been detected in patients dren with HSP, 168 without nephritis at with HSP as with other vasculitic diseases presentation were randomised to receive ster- probably because of vascular endothelial oids (1 mg/kg/day) for two weeks or no damage.17 18 A relation with disease activity has steroids. None of those treated with steroids also been noted and this may be a useful but 10 of the controls developed nephropathy marker of disease severity. Soluble thrombo- within six weeks and two other controls devel- modulin is derived from damaged endothelium oped nephropathy at 24 and 72 weeks, respec- and may reflect endothelial injury, and raised tively. The diVerence in the groups was highly serum thrombomodulin concentrations have significant suggesting steroids have a role in been demonstrated in patients with HSP preventing HSP nephritis, although even the nephritis.13 Routine coagulation tests are nor- untreated group had relatively mild involve- mal while factor XIII activity has been reported ment. In contrast, a retrospective study of 69 to be low, especially in patients with severe children with HSP of whom 50 did not have gastrointestinal disease.18 nephritis at diagnosis demonstrated a similar incidence of nephritis developing in those Histology treated or not treated with steroids.23 Steroids Skin biopsy specimens show a leukocytoclastic had been prescribed to manage abdominal or vasculitis with perivascular infiltration of poly- joint pain. More recently, in a multivariate morphs and mononuclear cells. Necrosis of analysis of prognostic factors for developing small blood vessels and platelet thrombi may renal involvement in HSP, corticosteroids were also be seen. Immunostaining reveals IgA in found to decrease the risk, although univariate 24

most purpuric skin lesions and often also in analysis showed no benefit. This was a retro- http://adc.bmj.com/ non-aVected areas. Renal biopsy findings may spective study with treatment more be graded according to the classification of the common in patients with severe abdominal International Study of Disease in Chil- pain, which itself was found to be a risk factor dren (ISKDC) from grades I–VI.19 The pri- for renal disease, perhaps explaining the mary lesion is an endocapillary proliferative discrepancy in these findings. In addition, per- involving both endothelial sistent purpura and decreased factor XIII were and mesangial cells, but proliferation of significant risk factors for developing renal dis-

extracapillary cells may result in crescent ease, and the authors recommended treatment on September 25, 2021 by guest. Protected copyright. formation. Immunofluorescence usually re- with steroids for these patients. veals mesangial IgA with IgG, C3, and fibrin. More intensive treatment has been investi- gated in patients with severe nephritis. Niaudet DiVerential diagnosis and Habib reported a prospective study of In the presence of an atypical rash other vascu- intravenous pulsed methylprednisolone fol- litic conditions should be considered. Micro- lowed by oral prednisolone in patients present- scopic polyarteritis, Wegener’s granulomatosis, ing with nephrotic syndrome, > 50% crescen- and systemic erythematosus (SLE) may tic nephritis, or both.25 At 1–16 years of follow all be accompanied by a crescentic nephritis. up 27 patients had clinically recovered, three The associated clinical features with the had minor abnormalities, four had persistent presence of ANCA or ANA (antinuclear nephropathy, and four had end stage renal fail- antibody) can help to diVerentiate these condi- ure. Comparison with historical data showed tions. Cytoplasmic ANCA (C-ANCA) is more an improvement in outcome with this treat- commonly associated with Wegener’s granulo- ment. Oner et al described 12 patients present- matosis while perinuclear ANCA (P-ANCA) is ing with rapidly progressive glomerulonephritis more often associated with microscopic polyar- of whom nine had renal biopsies demonstrating teritis. Serology is also a distinguishing feature crescentic glomerulonephritis (60–90% in SLE, although ANA positive HSP has been crescents).26 These patients were treated with described.20 methylprednisolone, oral , Sepsis may cause a purpuric rash as may dipyridamole, and prednisolone. There was a clotting disorders or thrombocytopenia. The good response in 11 patients—complete re- clinical picture, particularly the distribution of sponse in seven and partial response in four— the rash, with haematological investigations while one had a persistently decreased should identify these patients. glomerular filtration rate. The follow up of 382 Tizard

these patients was relatively short, from 9–39 Predicting the patients that are at risk of long months. A similar study of patients with term complications remains diYcult. In an Arch Dis Child: first published as 10.1136/adc.80.4.380 on 1 April 1999. Downloaded from ISKDC grade IV or V HSP nephritis with a attempt to define risk factors Bunchmann et al longer follow up of 7.5 (0.9) years also reviewed 16 patients that developed end stage suggested benefit from intensive treatment.27 renal failure and matched them with a group of Plasma exchange is used in the management children whose clearance had re- of some adults with vasculitis and idiopathic turned to normal at 10 year follow up.32 A cre- rapidly progressive nephritis. There are few atinine clearance of > 125 ml/min/1.73m2 at data in children, and controlled, prospective three years predicted recovery while a creati- trials would be diYcult to do because of the nine clearance < 70 ml/min/1.73m2 predicted small numbers and disease variability. Plasma progression to end stage renal failure. Gross exchange has been reported in children with haematuria was more frequent in the group rapidly progressive nephritis or cerebral vascu- that progressed. 28 litis associated with HSP. In this retrospec- There is debate as to the risks of recurrence tive study 17 patients, of whom 14 had severe of HSP in transplanted kidneys. The rate of renal involvement (30–100% crescents or isolated histological recurrence—that is, with- dialysis dependent) associated with HSP and out a change in urinalysis or renal function, three had cerebral vasculitis, were treated with varies from 0–89%. In a review of 78 renal plasma exchange with combinations of ster- transplants in both adults and children there oids, cyclophosphamide, and . All was histological recurrence in about 50%. three patients with cerebral disease recovered. Clinical recurrence occurred in 20%, graft All nine patients with renal disease treated failure in 12%, and graft loss in 9%.33 Delaying within a month of presentation responded and transplantation by six months to a year does have sustained improvement whereas five of not necessarily prevent recurrence. There is six treated later in the course of the disease some suggestion that live related transplants developed end stage renal failure. It has previ- have a higher risk of recurrence, possibly ously been reported that the presence of owing to a familial predisposition to the fibrous crescents as opposed to cellular condition. crescents is a poor prognostic factor in the Although the incidence of significant long outcome of crescentic nephritis, implying that term morbidity in unselected series is only a longer duration of disease has a worse 134 prognosis.29 1–5%, HSP has been reported to account for between 5% and 15% of patients entering end stage renal failure.532 Treatment with Prognosis intensive regimens should be reserved for The long term morbidity of HSP is predomi- patients with significant renal disease associ- nantly associated with renal involvement. In ated with a crescentic nephritis or children with severe neurological involvement. However, a

1988 Stewart et al described an unselected http://adc.bmj.com/ group of 270 patients with HSP who were fol- further prospective study of the use of an early lowedupover13years.1 The overall prognosis short course of steroids in all patients with HSP was good, with < 1% mortality and 1% renal would help to confirm or refute the view that morbidity. In an earlier unselected series of 141 steroids can ameliorate the long term outlook. patients, 39 (28%) of whom had abnormal uri- It has been demonstrated that those with the nary sediment for a month, one developed end most severe acute clinical and histological stage renal failure and two chronic renal failure findings have the highest risk of long term renal

giving an incidence of 1.5% for chronic renal impairment. However, in view of the unex- on September 25, 2021 by guest. Protected copyright. impairment.30 An extensive long term follow up pected late deterioration in some patients, long of HSP nephritis included 78 patients from an term follow up must be recommended for all initial cohort of 88 with a mean follow up of who develop nephritis. 23.4 years.31 Of 39 who had had a nephritic or nephrotic presentation, 17 had or 1 Stewart M, Savage JM, Bell B, McCord B. Long term renal impaired renal function. Eighty two per cent of prognosis of Henoch-Schönlein purpura in an unselected those presenting with haematuria with or with- childhood population. Eur J Pediatr 1988;147:113–15. 2 Neilsen HE. Epidemiology of Schönlein-Henoch purpura. out proteinuria were normal but two had renal Acta Paediatr Scand 1988;77:125–31. insuYciency and one died. Of concern is that 3 Robson WLM, Leung AKC. Henoch-Schönlein purpura. of the 17 patients that deteriorated overall, Advances in Pediatrics 1994;41:163–94. 4 Mills JA, Michel BA, Bloch DA, et al. The American College seven had recovered completely after a follow of Rheumatology 1990 criteria for the classification of up of 10 years. Of the 15 patients in the worst Henoch-Schönlein purpura. Arth Rheum 1990;33:1114– 21. clinical outcome group, five died and six had 5 Meadow SR. The prognosis of Henoch Schönlein nephritis. 11 transplants. Overall those with a severe Clin Nephrol 1978;9:87–90. 6 Meadow SR, Glasgow EF, White RHR, MoncrieV MW, clinical course and the most abnormal renal Cameron JS, Ogg CS. Schönlein-Henoch nephritis. QJ biopsy had the poorest outcome, but significant Med 1972;41:241–58. 7 Ha T-S, Cha S-H. Cerebral vasculitis in Henoch-Schönlein deterioration did occur in some without these purpura: a case report with sequential magnetic resonance poor prognostic indicators. This review also imaging. Pediatr Nephrol 1996;10:634–6. 8 Ostergaard JR, Storm K, Neurologic manifestations of identified 24 women who had 44 liveborn Schönlein-Henoch purpura. Acta Paediatr Scand 1991;80: infants at term. Significantly, 16 of these preg- 339–42. 9 Chaussain M, de Boissieu D, Kalifa G, et al. Impairment of nancies were complicated by proteinuria or lung diVusion capacity in Schönlein-Henoch purpura. J hypertension indicating the particularly close Pediatr 1992;121:12–16 10 Saulsbury FT. The role of IgA1 rheumatoid factor in the for- surveillance that is required in pregnant mation of IgA-containing immune complexes in Henoch- women. Schönlein purpura. J Clin Lab Immunol 1987;23:123–7. Henoch-Schönlein purpura 383

11 Ronda N, Esnault VLM, Layward L, et al. Antineutrophil 23 Saulsbury FT. Corticosteroid therapy does not prevent

cytoplasm antibodies (ANCA) of IgA isotype in adult nephritis in Henoch-Schönlein purpura. Pediatr Nephrol Arch Dis Child: first published as 10.1136/adc.80.4.380 on 1 April 1999. Downloaded from Henoch-Schönlein purpura. Clin Exp Immunol 1994;95: 1993;7:69–71. 49–55. 24 Kaku Y, Nohara K, Honda S. Renal involvement in 12 Lin J-J, Stewart CL, Kaskel FJ, Fine RN. IgG and IgA Henoch-Schönlein purpura: a multivariate analysis of classes of anti-neutrophil cytoplasmic antibodies in a 13 prognostic factors. Kidney Int 1998;53:1755–9. year old girl with recurrent Henoch-Schönlein purpura. 25 Niaudet P, Habib R. Methylprednisolone pulse therapy in Pediatr Nephrol 1993;7:143–6. the treatment of severe forms of Schonlein-Henoch 13 Fujieda M, Oishi N, Naruse K, et al. Soluble thrombomodu- purpura nephritis. Pediatr Nephrol 1998;12:238–43. lin and antibodies to bovine glomerular endothelial cells in 26 Oner A, Tinaztepe K, Erdogan O. The eVect of triple patients with Henoch-Schönlein purpura. Arch Dis Child therapy on rapidly progressive type of Henoch-Schönlein 1998;78:240–4. nephritis. Pediatr Nephrol 1995;9:6–10. 14 Hughes FJ, Wolfish NM, McLaine PN. Henoch-Schönlein 27 Iijima K, Ito-Kariya S, Nakamura H, Yoshikawa N. Multiple syndrome and IgA nephropathy: a case report suggesting a combined therapy for severe Henoch-Schönlein nephritis common pathogenesis. Pediatr Nephrol 1988;2:389–92. in children. 1998; :244–8. 15 Meadow SR, Scott DG. Berger disease: Henoch Schönlein Pediatr Nephrol 12 28 Gianviti A, Trompeter RS, Barratt TM, Lythgoe MF, Dillon syndrome without the rash. J Pediatr 1985;106:27–32. 16 Smith GC, Davidson JE, Hughes DA, Holme E, Beattie TJ. MJ. Retrospective study of plasma exchange in patients Complement activation in Henoch-Schönlein purpura. with idiopathic rapidly progressive glomerulonephritis and vasculitis. 1996; :186–90. Pediatr Nephrol 1997;11:477–80. Arch Dis Child 75 17 Ates E, Bakkaloglu A, Saatci U, Soylemezoglu O. Von Will- 29 Jardim HM, Leake J, Risdon RA, Barratt TM, Dillon MJ. ebrand factor antigen compared with other factors in Crescentic glomerulonephritis in children. Pediatr Nephrol vasculitic syndromes. Arch Dis Child 1994;70:40–3. 1992;6:231–5. 18 De Mattia D, Penza R, Giordano P, et al. Von Willebrand 30 Koskimies O, Mir S, Rapola J, Vilska J. Henoch-Schönlein factor and factor XIII in children with Henoch-Schönlein nephritis: long term prognosis of unselected patients. Arch purpura. Pediatr Nephrol 1995;9:603–5. Dis Child 1981;56:482–4. 19 Counahan R, Winterborn MH, White RHR, et al. Prognosis 31 Goldstein AR, White RHR, Akuse R, Chantler C. Long of Henoch-Schönlein nephritis in children. BMJ 1977;ii: term follow-up of childhood Henoch-Schönlein nephritis. 11–14. Lancet 1992;339:280–2. 20 Saulsbury FT. Case report: antinuclear antiboby in 32 Bunchman TE, Mauer SM, Sibley RK, Vernier RL. Henoch-Schönlein purpura. Am J Med Sci 1986;291:180– Anaphylactoid purpura:characteristics of 16 patients who 2. progressed to renal failure. Pediatr Nephrol 1988;2:393–7. 21 Rosenblum ND, Winter HS. Steroid eVects on the course of 33 Meulders Q, Pirson Y, Cosyns J-P, SquiZet J-P, van abdominal pain in Henoch-Schönlein purpura. Pediatrics Ypersele de Strihou C. Course of Henoch-Schönlein 1987;79:1081–21. nephritis after renal transplantation. Transplantation 1994; 22 Mollica F, LiVolti S, Garozzo R, Russo G. EVectiveness of 58:1179–86. early prednisone treatment in preventing the development 34 White RHR. Henoch-Schönlein purpura. In: Churg A, of nephropathy in anaphylactoid purpura. Eur J Pediatr Churg J, eds. Systemic vasculitides. New York: Igaku- Shoin 1992;151:140–4. 1991:203–17. http://adc.bmj.com/ on September 25, 2021 by guest. Protected copyright.