REVIEWS

Henoch–Schönlein purpura nephritis

in children
Jean-Claude Davin and Rosanna Coppo
Abstract | Henoch–Schönlein purpura (HSP) is the most common vasculitis in children, in whom prognosis is
mostly dependent upon the severity of renal involvement. Nephritis is observed in about 30% of children with
HSP. Renal damage eventually leads to chronic kidney disease in up to 20% of children with HSP nephritis
in tertiary care centres, but in less than 5% of unselected patients with HSP, by 20 years after diagnosis.
HSP nephritis and IgA nephropathy are related diseases resulting from glomerular deposition of aberrantly
glycosylated IgA1. Although both nephritides present with similar histological findings and IgA abnormalities,
they display pathophysiological differences with important therapeutic implications. HSP nephritis is mainly
characterized by acute episodes of glomerular inflammation with endocapillary and mesangial proliferation,
fibrin deposits and epithelial crescents that can heal spontaneously or lead to chronic lesions. By contrast,
IgA nephropathy normally presents with slowly progressive mesangial lesions resulting from continuous low-
grade deposition of macromolecular IgA1. This Review highlights the variable evolution of similar clinical and
histological presentations among paediatric patients with HSP nephritis, which constitutes a challenge for
their management, and discusses the treatment of these patients in light of current guidelines based on
clinical evidence from adults with IgA nephropathy.

Davin, J.‑C. & Coppo, R. Nat. Rev. Nephrol. 10, 563–573 (2014); published online 29 July 2014; doi:10.1038/nrneph.2014.126

Introduction
Henoch–Schönlein purpura (HSP) is the most fre-
quently detected form of vasculitis in children. The
incidence of HSP decreases with age, 1 but the preva-
lence of the disease is not well established.2 In a Dutch
study,3 the yearly incidence of HSP was 6.1 per 100,000
children in a cohort of children aged 0–18 years versus
14.9 per 100,000 in children aged between 3–6 years. The
reported annual incidence of HSP varies from 6.1 cases
per 100,000 children in the Netherlands to 20.4 cases per
100,000 children in the UK.3–6 However, these figures
might be overestimated, as they were based on the
contro­versial 1990 American College of Rheumatology
criteria for classification of vasculitis.7 Diagnosis of
HSP required the presence of two or more of the fol-
lowing criteria: age ≤20 years at disease onset; palpa-
Emma Children’s ble purpura; acute abdominal pain; and biopsy samples
Hospital and Academic
Medical Centre, showing granulo­c ytes in the walls of small arterioles
University of or venules.7 These criteria risk confusion with various
Amsterdam,
Meibergdreef 9,
forms of hypersensitivity vasculitis and may result in
1105 AZ, Amsterdam, ­overdiagnosis of HSP in children.8
Netherlands (J.‑C.D.). In 1994, HSP was defined by the Chapel Hill
Nephrology, Dialysis
and Transplantation Consensus Conference on the Nomenclature of Systemic
Unit, Città della Salute Vasculitis as “a vasculitis with IgA-dominant immune
e della Scienza di
Torino, Regina
deposits, affecting small vessels (that is, capillaries,
Margherita University venules or arterioles) typically involving skin, gut and
Hospital, Piazza glomeruli, and associated with arthralgias or arthri-
Polonia 94, Turin 1016,
Italy (R.C.). tis.”8 The European League Against Rheumatism and
Correspondence to:
J.‑C.D. Competing interests
j.c.davin@amc.nl The authors declare no competing interests.
the Paediatric Rheumatology European Society (PRES)
published modified criteria in 2006—­palpable purpura
(mandatory criterion) in the presence of at least one of
the following: diffuse abdominal pain; any biopsy sample
showing predominant IgA deposition; arthritis or acute
arthralgia in any joint; and renal involvement (any
haema­turia and/or proteinuria).9 After a formal statistical
validation of these criteria, performed by the Paediatric
Rheumatology International Trials Organization and
PRES, the 2008 Ankara Consensus Conference updated
the 2006 PRES classification by adding the criterion that
IgA deposition in a biopsy sample is required in patients
who have an ­atypical distribution of purpura.10
Although complications resulting from lesions in
organs other than the kidney (lung, intestine, brain and
testis) are sometimes severe, the prognosis of HSP is
mainly dependent on the renal component of the disease.
Moreover, for decades HSP nephritis and primary IgA
nephropathy have been considered related diseases
because of their similar renal histologic features and IgA
abnormalities.11,12 This relationship is exemplified by
the occurrence of HSP nephritis and IgA nephropathy
in identical twins.13 The proportion of children with HSP
who present with renal involvement varies from 20% to
100% according to the level of nephrology care.14 In a
20 year follow-up series from a tertiary care centre, HSP
nephritis led to chronic kidney disease (CKD) in up to
20% of children.15 By contrast, in unselected patients,
the corresponding percentage is less than 5%.16,17 HSP
nephritis is an infrequent cause of end-stage renal disease

NATURE REVIEWS | NEPHROLOGY VOLUME 10  |  OCTOBER 2014  |  563

© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS

Key points Finnish series of 223 paediatric patients with HSP, the risk
factors for development of nephritis were age >8 years
■■ Henoch–Schönlein purpura (HSP)—the most common vasculitis in children—
at onset, abdominal pain and recurrence of vasculitis.19
is complicated by nephritis in about 30% of patients
■■ Long-term prognosis of HSP nephritis depends mainly on the development of The natural history of HSP in children has changed over
chronic kidney disease (CKD); often CKD risk cannot be predicted from the initial time; late diagnosis and late referral of patients with renal
clinical and histological presentation involvement were common in early reports, but diagnosis
■■ CKD can be observed at long-term follow-up even after apparent complete is now easily performed and the risk of progression is the
recovery from HSP nephritis most important factor to be considered.
■■ HSP nephritis and IgA nephropathy both result from glomerular deposition The frequency of renal involvement in HSP is variable.
of aberrantly glycosylated IgA1 but have different histological features and
In a population-based survey in the Netherlands, which
clinical courses
■■ Typically IgA nephropathy presents as slowly progressive mesangial lesions,
included 232 patients aged 0–18 years (average 6 years)
whereas HSP nephritis presents as acute episodes characterized by inflammatory with HSP, 60% of participants were male (a higher per-
glomerular lesions that require prompt resolution to avoid chronic progression centage of males than females also have IgA nephritis)
■■ Use of guidelines based on evidence obtained in adults with IgA nephropathy to and 29% presented with renal symptoms within 1 month
select treatment for children with HSP nephritis risks delaying the provision of of onset (although the majority of these patients had only
adequate therapies mild renal signs, such as haematuria and/or minimal
proteinuria).3 Proteinuria >1 g/l was detected in 3% of
patients, nephrotic syndrome in 2%, hypertension in
Table 1 | Predictors of long-term renal outcome in Henoch–Schönlein purpura 3% and renal insufficiency in 1%.9 These figures are
Factor Symptoms Outcome consistent with the contemporary experience of neph-
Initial renal Nephrotic and nephritic syndrome CKD >50%15 rologists and large published series, rather than with the
symptoms15,32 Nephrotic syndrome CKD 40%15 extremely broad range of 20–100% of patients present-
Nephritic syndrome CKD 15%15 ing with renal involvement described in the early litera-
Heavy non-nephrotic proteinuria CKD 15%15
Haematuria and/or minimal proteinuria CKD <5%15 ture.14 In some reviews, up to 97% of patients with HSP
Nephrotic syndrome persisting <3 months ESRD 0%32 nephritis developed renal symptoms within 6 months
Nephrotic syndrome persisting >3 months ESRD 41%32 after presentation,20 and 35% developed clinical signs
Renal symptoms GFR <70 ml/min/1.73 m2 at 3 years* ESRD 100%26 of renal impairment after 1 year, with a continuous
during follow-up increase in frequency thereafter.21 By contrast, a review
Initial symptoms Mean follow-up proteinuria (g per day) ESRD RR 1.78, published in 2009 concluded that 1–2% of all unselected
versus increasing P <0.0131 patients with HSP nephritis will develop CKD in the
proteinuria during Severely impaired versus normal GFR at onset ESRD RR 3.83, long term.17 This proportion is much higher in patients
follow-up P = 0.2031
Nephrotic versus minimal proteinuria at onset ESRD RR 4.74, from tertiary medical centres; up to 20–30% of such
P = 0.1731 patients have CKD or ESRD by 20 years after diagno-
*Univariate analysis of predictors related to renal survival that used dialysis therapy as an end point; GFR sis of HSP.15,22–24 As expected, patients with severe initial
was calculated using the Schwartz formula. Abbreviations: CKD, chronic kidney disease; ESRD, end-stage symptoms are at higher risk of CKD than are those with
renal disease; GFR, glomerular filtration rate; RR, relative risk. Reproduced and modified with permission
from the American Society of Nephrology © Davin, J. C. Clin. J. Am. Soc. Nephrol. 6, 687–689 (2011). http:// mild symptoms (Table 1).15,23
cjasn.asnjournals.org/.

(ESRD) in adults,1 but in historical series of children with
HSP nephritis, the incidence of CKD can reach 5%.18
Thanks to improvements in the care and short-term
outcomes of these children, only a few require dialysis
nowadays, although whether the long-term outcomes
of patients with HSP have also substantially improved
remains unclear.
This Review summarizes advances in our understand-
ing of the clinical symptoms, prognosis, pathology,
pathophysiology and treatment of HSP nephritis in chil-
dren. We also present contemporary perspectives on the
typical clinical presentation and course of HSP nephri-
tis. Similarities and differences with IgA nephropathy,
as well as the consequences of the latter for treatment of
HSP nephritis, are also discussed.
Clinical symptoms
A recent or simultaneous infection is reported in 33–66%
of patients with HSP.1 Although any of the four major
components of the syndrome (rash, joint pain, abdominal
symptoms and renal disease) might precede the others,
the renal symptoms are rarely first to develop. 1 In a
Prognosis
A major difficulty in the identification of prognostic
factors in HSP nephritis resides in the fact that CKD can
develop up to 20 years after disease initiation, even after
apparently complete resolution of symptoms.15,23 The
possibility of clinically silent progression towards CKD
was initially raised in 1992. 15 Progression towards
CKD is particularly likely to occur during pregnancy.15,23
Approximately 20 years after initial presentation, CKD
develops in 15% of patients who present with heavy
proteinuria without nephrotic syndrome, in 40% of
those who present with nephrotic syndrome, and in
>50% of those who present with both nephritic and
nephrotic syndromes.15 However some (rare) patients
who present with mild initial symptoms also eventually
develop CKD,15,23,25 sometimes after repeated episodes
of macroscopic haematuria.26,27 Even patients without
urinary abnormalities can later present with hyper­
tension28 and ESRD can occasionally result from rapidly
progressive glomerulonephritis.29,30
These reports show that any initial clinical renal
presentation can result either in CKD or in complete
healing.15,23,25–27 The unpredictability of the prognosis of

564  |  OCTOBER 2014  |  VOLUME 10  www.nature.com/nrneph

© 2014 Macmillan Publishers Limited. All rights reserved

HSP nephritis probably results from the varied patho­
genetic pathways involved and the response to treat-
ment. HSP nephritis is typically an acute disease, but
even severe inflammatory glomerular lesions can resolve
completely in patients who receive early adequate treat-
ment, or when the causal event has been of short dura-
tion. On the contrary, when the cause of active lesions
persists or when adequate treatment is delayed, a fibrotic
process leads to chronic damage. The reduction in func-
tional glomeruli can be insufficient to induce substan-
tial proteinuria or impair filtration function in the short
term, a clinical picture that can initially be confused
with apparent healing but could result in hyperfiltration
and relentless progression to CKD. These clinical fea-
tures imply that two important conditions would render
further studies more informative: they should have long
follow-up periods, and instead of only looking for cor-
relations between initial clinical and histological param-
eters and outcome at last follow-up, patients’ data should
be analysed at intervals throughout the study to investi-
gate the possible role of relapses or of persistently active
disease processes.
Difficulty in the interpretation of initial symptoms
for prognostic purposes has resulted in the considera-
tion of other clinical parameters related to the status
of patients at subsequent stages of the illness (Table 1).
Unfortunately, assessment of those parameters might not
improve the outcomes of patients with HSP nephritis
because all imply a period of observation, which could
potentially delay the initiation of treatment to prevent
the development of chronic lesions (Table 1).26,31,32 Pooled
data from three studies that investigated the outcomes
of children with HSP nephritis, classified according to
International Study for Kidney Diseases in Children
(ISKDC) criteria, showed that about 25% of patients
progressed to various stages of CKD during a median
follow-up of 6 years.24,25,30,33 Although more common
in patients with class IV and class V features of HSP
nephritis, such as moderate to severe extracapillary
proliferation and crescent formation (37% and 70% of
such patients progressed to CKD, respectively), than
in those with milder disease, progression to CKD was
eventually also observed in 15% of patients with class II
or III features.
The predictive value of crescents has been addressed
in several retrospective studies. This feature was often
associ­ated with outcome in univariate but not in multi­
variate analyses. 15,24,25,31,33–42 The latest research sup-
ports the idea that even extensive crescentic lesions can
undergo regression and healing, and that progression
to CKD can occur in individuals with relatively mild
disease.22,23,26,40 Several factors might influence the effect
of crescents on final outcome, including the possibility
that renal biopsy sampling might miss areas with focal
crescentic glomeruli, the healing of limited segmental
crescents (which can appear and disappear within a
few weeks) and changes in the percentage of glomeruli
involved.42 Moreover, the ISKDC classification does not
consider the concomitant presence of tubulointerstitial
fibrosis, endocapillary hypercellularity, arteriolar damage
NATURE REVIEWS | NEPHROLOGY VOLUME 10  |  OCTOBER 2014  |  565
© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
or segmental sclerosis, which might also predict CKD
­according to their extension.43
Another factor that is likely to modify the outcome
of patients with crescents is the prompt institution of
aggressive immunosuppressive treatment, which can
blunt the risk of progression. This observation might
explain the apparent paradox that patients with ISKDC
grades II–III disease tend to have worse outcomes than
those with grades IV–V disease in published series. 23
Notably, an adverse effect of crescentic HSP nephritis
on outcome was reported in early published series,15
although this association has not been detected in series
published after use of prednisone and immunosuppres-
sive drugs had become the rule.40 This observation sug-
gests that prompt immunosuppressive treatment might
prevent cellular crescents from progressing to fibrosis
and thereby contributing to CKD. Conversely, delaying
kidney biopsy and the initiation of immunosuppressive
treatment might enable crescentic lesions to progress to
complete glomerulosclerosis.42,44,45
Because considering only the approximate extent of
mes­angial proliferation (focal or diffuse) and the per-
centage of crescentic glomeruli (ISKDC classification)
has limited prognostic utility, it is necessary to improve
prognostic predictions by introducing other param-
eters. These could potentially include a precise evalu-
ation of mesangial hypercellularity (by calculating the
mesangial cell score), endocapillary hypercellularity,
inflammatory cell infiltration, Bowman capsule integ-
rity, extent and type of crescents (cellular versus fibrotic),
glomerulo­sclero­sis and interstitial fibrosis or tubular
atrophy, following the example of the new histological
classification of IgA nephropathy.43
Recurrence of glomerular IgA deposits after kidney
transplantation is very frequent in patients with HSP
nephritis, but does not significantly affect the rate of graft
loss during the first 5 years of follow-up.46 By 10 years
post-transplantation, however, the risk of graft loss owing
to recurrence of IgA deposits was 7.5% in one study.47
The same pattern is observed for IgA nephropathy.48,49
IgA nephropathy versus HSP nephritis
Presentation and clinical course
HSP nephritis frequently presents as acute nephritic and/
or nephrotic syndrome, reflecting a rapidly developing
pathophysiological mechanism.12 In patients with HSP
nephritis, repeated or prolonged episodes of acute glo-
merular inflammation lead to fibrous scars and hyper-
filtration in the remaining areas, which finally result in
CKD and ESRD. Hence, the number and severity of acute
episodes of HSP nephritis have a crucial role in the sub-
sequent progression and loss of renal function (Box 1).
However, HSP nephritis is more benign in children than
in adults, as children mostly enter stable remission,
which enables healing, whereas HSP nephritis in adults
tends to show a chronic and relentless course similar
to that of primary IgA nephropathy.16,17,34 By contrast,
IgA nephropathy is often discovered as an incidental
finding of asymptomatic haematuria, or as short-­lasting
episodes of gross haematuria that generally do not
REVIEWS
Box 1 | Factors linked to adverse renal outcome in HSP
Several observations support the role of the initial acute
episode in developing irreversible scarring and loss of
renal function at long term follow-up:
■■ Relationship between the severity of initial
clinical and histological signs and the long-term
prognosis15,24,25,30,33
■■ Correlation between the chronicity score and the time
elapsed between clinical onset of renal signs and the
kidney biopsy 45
■■ Rapid evolution of florid crescents to global glomerular
sclerosis44
■■ Relationship between delayed treatment and worse
evolution23,115,116
■■ Development of impaired renal function in the long
term after apparently complete resolution15
Abbreviation: HSP, Henoch–Schönlein purpura.
affect chronic progression. As a consequence, in IgA
nephropathy, initi­ation and progression of CKD occurs
slowly and often goes unnoticed, eventually leading to
­glomerulosclerosis and tubulointerstitial fibrosis.
Histological findings
The histologic lesions of HSP nephritis are categorized
as ISKDC grades I, II, III, IV and V according to the
presence and percentage of glomeruli with crescents.
Grade VI is used for lesions with a membranoprolifer­
ative aspect.30 Extent of mesangial proliferation (focal
or diffuse) further characterizes the different classes of
ISKDC (Box 2).30
The glomerular lesions found in patients with HSP
nephritis, particularly those in children, are charac­
terized in early stages after clinical onset by acute
inflammation with leucocyte influx and endocapillary
hypercellularity.29,30,42 Tuft necrosis is frequently found
in crescentic glomeruli, and this feature is, therefore,
thought to favour crescent formation.29,30,50 However,
these acute-phase inflammatory changes can also be
found in patients with IgA nephropathy, albeit mostly
in those with macroscopic haematuria or in the rare
patients with rapidly progressive disease. By contrast,
acute-phase inflammatory changes are uncommon in
patients with stable clinical conditions.29 Endocapillary
proliferation and inflammatory cell infiltration by poly-
nuclear neutrophils is another histological feature that
is more frequently observed in HSP nephritis than in
IgA nephropathy.29
The lesions observed in kidney biopsy samples from
patients with HSP nephritis are strongly dependent on
the timing of renal biopsy in relation to the onset of clini-
cal symptoms and detection of urinary abnormalities.
Early acute clinical phases of HSP nephritis are charac-
terized by mesangial and endocapillary hyper­cellularity,
often with fibrinoid necrosis and small cellular crescents,
whereas advanced stages present with segmental dis­
organized sclerosis, mostly coincident with fibrous cres-
cents. The presence of crescents is a prominent histologic
feature of HSP nephritis, and has also been considered
a pivotal feature for the ISKDC pathology classification,
566  |  OCTOBER 2014  |  VOLUME 10  www.nature.com/nrneph
© 2014 Macmillan Publishers Limited. All rights reserved
Box 2 | ISKDC histologic classification of HSP nephritis30
■■ I: Minimal histologic alterations
■■ II: Pure mesangial proliferation
■■ III: Focal (IIIa) or diffuse (IIIb) mesangial proliferation
with <50% crescentic glomeruli
■■ IV: Focal (IVa) or diffuse (IVb) mesangial proliferation
with 50–75% crescentic glomeruli
■■ V: Focal (Va) or diffuse (Vb) mesangial proliferation
with >75% crescentic glomeruli
■■ VI: Membranoproliferative-like glomerulonephritis
Abbreviations: HSP, Henoch–Schönlein purpura; ISKDC, International
Study of Kidney Disease in Children.
owing to its prognostic relevance (Box 2).30 Crescents
in patients with HSP nephritis rarely involve more than
50% of glomeruli.29 A pathophysiological role of inflam-
matory cells and fibrin in the formation of crescents is
suggested by the predominance of glomerular leucocyte
infiltration, glomerular necrosis and fibrin deposition in
glomeruli with crescents.42,50–53 The role of fibrin is also
supported by the observation that crescent formation can
be prevented by the anti­coagulant warfarin in experi-
mental models of crescentic nephritis.54 Degenerative
tubular changes and lesions of the interstitium (oedema,
infiltration of lymphocytes, macrophages and plasma
cells) are consistent with the severity and duration of
glomerular injury.29 Unlike in HSP nephritis, crescents
are uncommon in IgA nephropathy (detected only in
5–10% of renal biopsy samples) and generally occupy
only part of the capsule parietal wall.29
The most characteristic immunohistochemical finding
consists of predominant glomerular deposits of IgA,
mostly represented by the subclass IgA1.29 The typical
immunofluorescence pattern is diffuse, granular mes­
angial staining with associated capillary wall staining
in patients with endocapillary proliferation. The associ­
ation between an increased severity of diffuse endocapil-
lary proliferation (and/or extensive crescent formation)
and extensive capillary deposits of IgA and complement
strongly suggests the crucial pathophysiological role of
the latter.29 Differences in the extent and distribution
of crescent formation between the two disorders might
reflect the increased intensity of subendothelial depo-
sition of IgA circulating immune complexes (CIC) in
HSP nephritis.29,30 IgG and/or IgM and fibrin-related
antigen are also detected in 65–75% of biopsy samples.29
Complement factors B, C3 and C5b–9 complex are
frequently found in biopsy samples from patients with
HSP nephritis.29,30,55 Components of the lectin pathway
(mannose-binding protein C, ficolin‑2, mannan-binding
lectin serum protease [MASP-1] and C4d) are present in
50% of patients,55 whereas C1q is rarely found.29
Pathophysiology
Both diseases result from the glomerular deposition
of IgA-CIC; however, the pathophysiological mecha-
nisms activated in HSP nephritis (Figure 1) and IgA
nephro­pathy are complex and divergently modulated,
leading to relevant histological and clinical differences
that might result in disparate responses to the same
 REVIEWS

Intestinal
1 microvillus
Mucosal Mucosal
antigen villous

2
Increased antigen MALT
penetration due to
defective mucosal
immunity DC

Genetic factors
Epigenetic factors CD4+

3 4
Synthesis of galactose-deficient Formation of immune complexes
IgA1 exposing GalNAc residues and autoaggregates
and IgG anti-GalNAc IgA1
GalNAc

IgA IgG
GalNAc IgA1
–mucosal antigen GalNAc IgA1–IgG GalNAc IgA1
immune complexes immune complexes autoaggregates

9 Crescent formation

Neutrophil 5
Activation of complement
Lymphocyte in the circulation and at the
mesangial and endothelial level
Macrophage 6 7
Activation of circulating Activation of resident
Monocyte leucocytes endothelial cells
and mesangial cells

Immune complexes
and autoaggregates
Endothelial Glomerular
Dead cell basement
Damaged endothelial cell Podocyte membrane
podocyte
Cytokine
IgA
C3

Lumen
Cytokines

Mesangial cell
8
Leaky Attraction of Podocyte
membrane inflammatory cells foot process
and activation
of podocytes
Inflamed glomerular capillary Glomerular capillary
Figure 1 | Pathogenesis of Henoch–Schönlein purpura nephritis. Genetic and epigenetic factors contribute to activation
of the mucosal immune system after antigen exposure (1), favoured by increased antigen penetration owing to defective
mucosal immunity (2). Antigens reach the MALT and activate DCs and CD4+ lymphocytes, leading to synthesis of GalNAc-
IgA1 (3) and anti-GalNAc-IgA1 IgG autoantibodies. Several circulating macromolecular forms of IgA1: GalNAc-IgA1, IgA-CIC
(formed by binding of GalNAc-IgA1 to mucosal antigens or anti-GalNAc-IgA1 IgG) and galactose-deficient IgA1 autoaggregates,
(4) activate the complement system, mostly through the alternative and lectin pathways (5). Collectively, these factors
favour the activation and influx of PBMCs and neutrophils (6) and the activation of resident endothelial and mesangial cells
(7). Podocytes are activated via chemokines produced by mesangial cells (8). Subendothelial and mesangial deposition of
IgA-CIC leads to fibrin deposition, destruction of the glomerular basement membrane, attraction of macrophages and
inflammatory cells, and cytokine-induced proliferation of epithelial cells in the Bowman space. Eventually, capsular
destruction and fibrous crescent formation result from infiltration of the Bowman space by interstitial fibroblasts (9).
Abbreviations: C3, activated complement component 3 protein; CIC, circulating immune complex; DC, dendritic cell;
GalNAc-IgA1, galactose-deficient IgA1; MALT, mucosa-associated lymphoid tissue; PBMC, peripheral blood mononuclear cell.

NATURE REVIEWS | NEPHROLOGY VOLUME 10  |  OCTOBER 2014  |  567

© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 2 | Comparison of primary IgA nephropathy and HSP nephritis treatment (Table 2). The glomerular deposition of
IgA-CIC initiates a cascade of cross-talk between resi-
Clinical and Detection method IgA nephropathy HSP nephritis
histological features dent cells resulting in their mutual stimulation and the
attraction of inflammatory cells. These processes lead
Renal histology
to mes­angial expansion and both endocapillary and
Crescents Light microscopy +/– ++
extracapillary proliferation. The variability of patients’
Glomerular tuft necrosis +/– ++
Neutrophil infiltration +/– ++ symptoms might be explained by differences in the
duration of production, amount, composition and local-
Mesangial IgA deposits Immunofluorescence +++ +++
IgA deposits along – ++ isation of IgA-CIC, as well as by the intensity of reaction
capillary walls of the different cells involved.
Subendothelial deposits Electron microscopy +/– ++ HSP nephritis and IgA nephropathy are characterized
Clinical presentation
by an abnormal IgA1 glycosylation pattern with reduced
galactosylation (Figure 2). The lack of terminal β1,3-
Extrarenal symptoms NA +/– +++
galactosyl residues in the hinge region of IgA1 might be
Age at onset NA Mostly >15 years Mostly <15 years due to reduced activity of β1,3-galactosyltransferase in
Nephritic syndrome NA +/– ++ IgA1-producing peripheral B cells,56,57 which might be
Nephrotic syndrome NA +/– ++ regulated by miRNAs58 or the result of type 2 T-helper
Disease course NA Continuous Repeated acute
cell prevalence.59–62 This reduced galactosylation results
moderate activity episodes in exposure of N‑acetylgalactosamine (GalNAc) residues
with exacerbations at the IgA1 surface, forming a novel antigen and inducing
Renal outcomes a humoral IgG autoimmune response.63 Circulating com-
Clinical remission NA 30–50% 98%
plexes of mixed IgG and galactose-deficient IgA1 have
been detected in both HSP nephritis and IgA nephro­
CKD long after apparent NA – +
complete remission
pathy.64 The characteristic feature of IgA nephropathy is
an increased proportion of circulating mucosal-type IgA1
ESRD NA 20–40% after 1–3% in children,
20 years 30% in adults
antibodies.65 However, this ‘hypogalactosylated’ form
of IgA1 is not restricted to patients with IgA nephro­
Transplantation outcomes
pathy but is also observed in the serum of individuals in
IgA deposit recurrence Immunofluorescence Frequent Frequent response to mucosal infection.65 The pathophysiological
Graft loss at 5 years NA Rare Rare role of galactose-deficient IgA1 molecules in renal lesions
post-transplantation is suggested by the observation that they are only found
Graft loss at 10 years NA 9.7% 7.5% in the serum of patients with HSP when they are having
post-transplantation an episode of nephritis.57
–, +/–, +, ++ and +++ indicate the relative likelihood of each clinical or histological feature occurring in Since HSP nephritis and IgA nephropathy are both
a direct comparison of IgA nephropathy and HSP nephritis. Abbreviations: CKD, chronic kidney disease;
ESRD, end-stage renal disease; HSP, Henoch–Schönlein purpura; NA, not applicable. characterized by predominant glomerular IgA depos-
its and mesangial proliferative changes, they have been
suggested to be determined by the same pathophysio­
CH1 logical mechanisms. A relationship between these two
Pro Gal diseases was further confirmed by the detection of
CH1
Ser β1-3
Thr* familial clustering of individuals with HSP nephritis
GalNAc GalNAc α2,6 GalNAc
Pro
α1-0 α1-0
Neu5Ac
α1-0
and IgA nephropathy, who have similar circulating IgA
Pro abnormalities.12,13 However, the two diseases also have
Thr*
Pro important differences that should be considered in
Hinge I II III
region
Ser* therapeutic decision-making.
Pro
Ser* Neu5Ac Neu5Ac The increased intestinal permeability reported in
CH2 Pro α2-3 α2-3 patients with HSP nephritis,66 and the reduced mucosal
Thr Gal Gal Gal immune reaction to novel antigens in those with IgA
Pro β1-3 β1-3 β1-3
Thr* nephropathy, 67 suggest that resistance to mucosal
GalNAc GalNAc α2,6 GalNAc α2,6
CH3 Pro Neu5Ac Neu5Ac antigen penetration is impaired in both diseases. A
Ser α1-0 α1-0 α1-0
high antigen load stimulates the immune system to
Pro
Ser produce large amounts of galactose-deficient IgA 1,
IV V VI
CH2 forming circulating IgA-CIC that cannot be cleared by
Figure 2 | Human IgA1 O‑glycosylation sites and galactosylation patterns. The hinge asialoglycoprotein receptor 1 on hepatocytes. The size of
region of IgA1 contains up to six major glycosylation sites at serine and threonine macro­molecular IgA1 further increases after binding to
residues. The O‑glycans include a core GalNAc, usually extended with Gal to form IgA Fc receptors (FcαRI, also known as CD89) on leuco-
Galβ1,3GalNAc, which can bind to Neu5Ac. Thus, each IgA1 O‑glycan can have one
cytes, and IgA1–FcαRI complexes are released after cleav-
of four short carbohydrate structures (types III, IV, V and VI), leading to a mixture of
IgA1 forms with varying degrees of galactosylation. Patients with Henoch–Schönlein
age of the FcαRI extracellular domain.68,69 Why IgA-CICs
purpura nephritis or primary IgA nephropathy have a high prevalence of galactose- provoke multi-organ inflammation only in patients with
deficient (types I and II) IgA1. Abbreviations: Gal, galactosamine; NAc, N‑acetyl; HSP nephritis is not known. However, some data suggest
Neu5Ac, α2,6 sialic acid and/or α2,3 sialic acid. that IgE-associated mechanisms (such as an increased

568  |  OCTOBER 2014  |  VOLUME 10  www.nature.com/nrneph

© 2014 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Table 3 | Immune system reactants involved in the pathogenesis of HSP nephritis Mesangial cells
After penetration into the mesangium, which is driven by
Reactant* Details Reference(s)
the intraglomerular filtration pressure, macromolecules
Circulating IgA Galactose-deficient IgA1 34,35 containing galactose-deficient IgA1 can bind to resident
molecules Autoaggregated galactose-deficient IgA1 77
IgG antibodies to galactose-deficient IgA1 63 mesangial cells. Mesangial proliferation is an invari-
IgG–IgA1 circulating immune complexes 76 able histological finding in patients with HSP nephritis.29
IgA1–soluble CD89 complexes 69 Binding of IgA-CICs to mesangial cells, and subsequent
Receptors for IgA1 Myeloid FcαRI (also known as CD89) 68 activation of these cells, depends on the constituents of
Transferrin receptor (also known as CD71) 129 IgA-CICs, and particularly their galactose-deficient IgA1
on mesangial cells content.68,69,73–76 As circulating galactose-deficient IgA1 is
Cytokines‡ IL‑17 (increased ratio of IL‑17:TREG cells), TNF, 130 also found in healthy relatives of patients with HSP
IL‑1β, IL‑2, IL‑6, IL‑8, TGF‑β, VEGF, TWEAK, low nephritis and IgA nephropathy, the involvement of some
IFN‑γ and IL‑12, increased IL‑4 (imbalance of
TH1:TH2) additional factors is likely to be necessary to render
galactose-­deficient IgA1 pathogenic.76 This ‘second hit’
Mesangial cell C3, FcγRI, TNF, TGF‑β, PDGF-RB, IL‑1, IL‑6, IFN‑γ, 78,79
receptors§ fibronectin receptor, integrins, angiotensin II might be related to the degree of self-aggregation of
receptor, CD71, EGF, TLR‑3, TLR‑4, chemokines galactose-­deficient IgA 1,77 the presence and amount
Products of Cytokines: TNF, IL‑1β, IL‑6, TGF‑β 78 of molecules other than IgA1 in IgA-CIC, for which mes­
mesangial cells Chemokines: IL‑8, RANTES, MCP‑1 angial cells also display membrane receptors (including
Prostanoids, angiotensin II, nitric oxide, reactive IgG autoanti­bodies and complement breakdown ­products;
oxygen species Table 3)73,75–79 or the presence of oxidative stress.80
*This list is not exhaustive; the most relevant reactants are shown. ‡Produced by endothelial cells and/or Stimulation of mesangial cells by aberrantly glyco-
infiltrating blood mononuclear cells. §Activated by specific ligands released in the mesangium.
Abbreviations: C3, complement protein C3; EGF, epidermal growth factor; FcαRI, IgA fragment crystallizable sylated macromolecules triggers their proliferation and
receptor; FcγRI, IgG fragment crystallizable receptor; HSP, Henoch–Schönlein purpura; IFN, interferon;
MCP‑1, monocyte chemotactic protein 1; PDGF-RB, platelet-derived growth factor-receptor β polypeptide;
matrix production, along with the release of many factors
RANTES, regulated upon activation normal T cell expressed and secreted; TGF-β, transforming growth (Table 3): prostaglandins, angiotensin II, nitric oxide
factor-β; TH, T‑helper; TLR, Toll-like receptor; TNF, tumour necrosis factor; TREG, T regulatory; TWEAK, TNF-like
weak inducer of apoptosis; VEGF, vascular endothelial growth factor. synthase,56 chemokines that attract polymorphonuclear
leuco­cytes and monocytes,56,73,74,78 cytokines and medi­
ators that affect key functions of podocytes (including
Table 4 | Treatment of HSP nephritis local migration, adherence to the basement membrane
Type of treatment Evidence and control of proteinuria).81,82 Cytokines involved in
acute and chronic inflammatory responses stimulate mes­
Randomized clinical trials Nonrandomized studies
angial cells,78 and can be produced by the mesangial cells
Oral steroids Oral prednisone does not prevent No effect of prednisone themselves (in an autocrine process) or by inflammatory
the development of nephritis17 alone on existing
nephritis (retrospective cells that have migrated into glomeruli (Table 3).74,78,83–90
studies)25,36,122,123 Tumour necrosis factor seems to have a key role as a
Other In children, no advantage of Advantage of direct or indirect effector of podocyte damage and altered
immunosuppressive cyclophosphamide (used without methylprednisolone pulses nephrin expression.82,89,90
drugs or prednisone) over placebo109 followed by prednisone In summary, the stimulation of mesangial cells by glo-
combination In adults, no advantage of versus prednisone alone merular deposition of IgA-CIC can result in mesangial
cyclophosphamide plus (prospective study with
historical controls)116 hypercellularity, extracellular matrix expansion, oxi­
methylprednisolone pulses
followed by oral prednisone versus Favourable effect of dative stress, attraction of inflammatory cells and altered
the steroid schema alone111 immunosuppressive drugs podocyte function. These factors lead to glomerulo­
In patients with nephrotic-range in patients receiving several sclerosis and tubulointerstitial damage, owing to feed-
proteinuria, 1 year of treatment agents (retrospective studies back mechanisms involving the glomeruli and tubules,
with ciclosporin is not inferior to and one prospective
study)45,118–121 which have the clinical consequences of proteinuria,
methylprednisolone pulses
followed by 4 months prednisone110
hypertension and renal insufficiency.
ACE inhibitors NA Efficacy of ACE inhibitors in
moderately severe proteinuria
Complement activation
(retrospective study)131 In patients with HSP nephritis, extensive crescents are
Plasma exchange NA Favourable effects of plasma
often associated with capillary wall destruction and
exchange in patients with very endocapillary hypercellularity,30 along with the pres-
severe clinical and ence of subendothelial immune deposits of IgA and
histological features complement.29 These observations, combined with data
(retrospective study)114,115
on experimental crescentic glomerulonephritis,50,91–102
Abbreviations: ACE, angiotensin-converting enzyme HSP, Henoch–Schönlein purpura; NA, not available.
suggest that the mechanism of crescent formation
involves local complement activation.
abundance of mast cells in skin, gut and joints, peri­ Glomerular C3 deposits are seen in the vast majority
vascular IgE dermal deposits, high IgE plasma levels and of patients with HSP nephritis.30 Although no relation-
high concentration of eosinophil cationic protein) have ship between serum levels of C3 and C4 and the presence
a role in increasing vascular permeability and favour of nephritis has been demonstrated,103 in situ activation of
IgA-CIC deposition in various organs.70–72 the complement system is probably an important event in

NATURE REVIEWS | NEPHROLOGY VOLUME 10  |  OCTOBER 2014  |  569

© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
Box 3 | Treatment of HSP nephritis in children
Principles of treatment based on pathophysiology
■■ Reduce endocapillary hypercellularity owing to inflammatory cell infiltration
■■ Reduce endothelial cell activation
■■ Reduce mesangial cell activation and proliferation
■■ Reverse podocyte activation and crescent formation
■■ Reduce the production of aberrantly glycosylated IgA1 and anti-GalNAC IgG
■■ Enhance the removal of circulating IgA1-containing immune complexes
Important clinical observations and resulting principles of treatment
■■ Initial clinical and histological symptoms do not predict outcome accurately
except when very mild or very severe
■■ Renal lesions can rapidly change owing to dynamic pathogenetic events and
histology can deteriorate rapidly
■■ Renal biopsy samples might not be representative of the whole kidney
■■ Worsening of clinical features can follow detection of moderate lesions (in renal
biopsy samples) by only a few days
■■ Delaying adequate treatment can be harmful
■■ Owing to the unpredictability of disease evolution, prolonged follow-up is
mandatory
■■ In case of no response or deterioration under treatment, a repeated renal biopsy
is useful to adapt therapy
Abbreviation: HSP, Henoch–Schönlein purpura.
Box 4 | 2012 KDIGO guidelines for treatment of HSP nephritis112
■■ Only when proteinuria is >0.5g/1.73 m2 daily, treat children (including those
with nephrotic syndrome) with angiotensin-converting-enzyme inhibitors or
angiotensin II-receptor antagonists for 3–6 months
■■ In patients who fail to respond to the above treatment, administer corticosteroid
therapy
■■ More aggressive treatment, including methylprednisolone pulses plus
cyclophosphamide, is reserved for patients with >50% crescentic glomeruli and
rapidly progressive renal deterioration with or without nephrotic syndrome. Plasma
exchange sessions can be added when plasma creatinine is >500 μmol/l
Abbreviations: HSP, Henoch–Schönlein purpura; KDIGO, Kidney Disease: Improving Global
Outcomes.
the pathophysiology of glomerular lesions. Complement
activation might be initiated by subendothelial and
mesangial deposition of IgA-CIC, leading to fibrin depo-
sition, destruction of the glomerular basement membrane
and attraction of inflammatory cells and macrophages
as well as cytokine-induced proliferation of epithelial
cells in the Bowman capsule space. Ultimately, IgA-CIC
­deposition results in destruction of capsular integrity and
formation of fibrous crescents due to infiltration of the
Bowman space by interstitial fibroblasts (Figure 1).
Immunofluorescence studies, as mentioned above,
suggest a role of both the alternative and lectin path-
ways in activation of the complement system. However,
mannose-binding protein C, MASP‑1 and C4d depos-
its are associated with severe glomerular inflammation
and clinical signs in both patients with HSP nephri-
tis55 and IgA nephropathy,104,105 which suggests that the
­pathogenetic role of the lectin pathway is predominant.
Treatment
HSP nephritis in children was initially considered to be
a rather benign disease for which only supportive treat-
ment was necessary, as affected children mostly undergo
spontaneous recovery.106 However, after the publica-
tion of long-term follow-up studies showing delayed
570  |  OCTOBER 2014  |  VOLUME 10  www.nature.com/nrneph
© 2014 Macmillan Publishers Limited. All rights reserved
development of CKD in this population,33,36,42 the treat-
ment policy changed and the use of steroids and immuno­
suppressive treatments was recommended even in the
absence of rapidly progressive glomerulonephritis.107,108
Randomized controlled trials in HSP nephritis are
scarce and have provided little clinically useful informa-
tion109–111 (Table 4). In the absence of strong therapeutic
evidence, principles of treatment should take into account
existing knowledge of pathophysiological mechanisms
and important clinical observations (Box 3).
Updated Kidney Disease: Improving Global Outcomes
(KDIGO) guidelines for the treatment of HSP nephri-
tis were published in 2012.112 In light of the similarities
between HSP nephritis and primary IgA nephritis, the
KDIGO guidelines include similar indications for the two
diseases when their clinical features are similar (Box 4).112
However, these guidelines conflict with generally used
clinical protocols, which include the use of methyl­
predniso­lone pulses, other immunosuppressive drugs and
plasma exchange in patients at a lower threshold of long-
term CKD risk (Box 4).113–121 Moreover, although these
two diseases can present with clinical and histological
similarities, HSP nephritis differs from IgA nephropathy
in pathophysiology, the predictability of its prognosis and
the effects that delayed treatment can have on patients’
long-term outcomes. Following the KDIGO guidelines
might, therefore, delay the initiation of effective treatment
and increase the risk of CKD in the long-term.23,115,116
For example, in patients with nephrotic syndrome and
ISKDC grade III disease who have extensive inflamma-
tory lesions, the KDIGO guidelines recommend initial
treatment with angiotensin-converting-­enzyme (ACE)
inhibitors, and only suggest adding oral prednisone if
improvement does not occur (Box 4). However, treat-
ment with ACE inhibitors is unlikely to improve the
acute glomerular inflammatory changes in these patients,
and a delay of 3–6 months before initi­ation of a possibly
effective treatment might not prevent the development
of sclerotic changes. Although some retrospective case
series reported a lack of effect of oral prednisone,25,36,122,123
a possible beneficial effect of methyl­prednisolone pulses
followed by prednisone was suggested in a prospective
case series (in which parti­cipants were compared to a
cohort of historical controls treated at the same centre),116
as well as in the control arm of a randomized controlled
trial.111 The beneficial effect of methylprednisolone pulses
is also suggested in patients receiving combinations of
multiple immuno­suppressive drugs (Table 4),119–121 and
is supported by experimental models showing maximal
therapeutic benefits of steroids with 30 mg/kg intra-
venous methylprednisolone pulses in experimental
crescentic glomerulonephritis.124
Children with similar presentations of HSP nephritis
can experience either complete disappearance of urinary
signs or an unexpected late progression to CKD—­
sometimes in spite of an apparent long-term remission of
symptoms—hence, prolonged observation is necessary.
Repetition of renal biopsy is mandatory for individuals
with symptom aggravation or lack of improvement with
therapy, as useful information for choosing a therapeutic
 REVIEWS

strategy can result from the analysis of crescent patho-
physiology.51,92,93 The timing and frequency of biopsy is
dependent upon the severity of the symptoms and his-
tology, on the evolution of the disease and on response
to treatment.45
Conclusions
It seems obvious that the clinical, histological and patho-
physiological differences between IgA nephropathy and
HSP nephritis need different therapeutic approaches. In
our opinion, this renders the 2012 KDIGO guidelines,
which are mostly based on evidence obtained in adults
with IgA nephropathy, at risk of being inappropriate for
children with HSP nephritis.
A 2013 study has shown that even small acute changes
in kidney function, particularly in pre-existing kidney
disease, can be complicated later on by CKD.125 Improved
treatment strategies are necessary to detect patients at risk
of CKD at an early stage. This goal should be achieved
by advances in understanding of HSP nephritis patho-
physiology, the availability of new biomarkers of disease
activity, and improved knowledge of the pattern of pro-
gression to CKD (either scarring resulting from acute but
self-limited episodic disease, or a slowly progressive
but chronically active process, as occurs in IgA nephro­
pathy). A serious need remains for multicentre, inter­
national studies that prospectively enrol patients at their
initial presentation with the disease (including individu-
als with mild initial symptoms), and include continued
monitoring for long periods of time, even after apparent
complete healing. To modulate each patient’s treatment
according to the specific pathophysiological pathways
activated, the researchers will need to evaluate at regular
intervals the relationships between clinical signs, histo-
logic findings, treatments and levels of immune ­reactants
in plasma and/or urine that reflect an active process
(Table 3). Further research into the pathogenetic impor-
tance of circulating galactose-deficient IgA1, CIC consist-
ing of galactose-deficient IgA1 bound to anti-GalNac-IgA1
IgG, cleavage products of C3 and C5b‑9 resulting from
activation of the complement system, urinary podo-
cyte markers and expression of CD71 (the latter result-
ing from the high proliferation activity of mesangial
cells11,56,57,75,126–130), is highly recommended. This research
should be supported by online registry data coupled to all
prospective clinical trials in patients with HSP nephritis.
In the meantime, panels of paediatric experts should be
created by international societies to amend the KDIGO
guidelines in accordance with clinical practice, to propose
prospective studies on new biomarkers of disease activity,
and to design randomized controlled trials.
Review criteria
The PubMed online database was searched for full-text,
English language articles published in major medical
journals during 2010–2014. The search terms used were
“IgA nephropathy” and “Henoch–Schönlein purpura”. The
authors also utilized their personal libraries of original
papers and review articles on the same topic. Reference
lists of identified papers were reviewed for further leads.

1. Haycock, G. B. In Oxford Textbook of Clinical
Nephrology 1st edn, Ch. 4 (eds Cameron, S.
et al.) 595–612 (Oxford University Press, 1992).
2. Eleftheriou, D. & Brogan, P. A. Vasculitis in
children. Best Pract. Res. Clin. Rheumatol. 23,
309–323 (2009).
3. Aalberse, J., Dolman, K., Ramnath, G.,
Pereira, R. R. & Davin, J. C. Henoch Schönlein
purpura in children: an epidemiological study
among Dutch paediatricians on incidence and
diagnostic criteria. Ann. Rheum. Dis. 66,
1648–1650 (2007).
4. Dolezalova, P., Telekesova, P., Nemcova, D. &
Hoza, J. Incidence of vasculitis in children in the
Czech Republic: 2‑year prospective epidemiology
survey. J. Rheumatol. 31, 2295–2299 (2004).
5. Gardner-Medwin, J. M., Dolezalova, P.,
Cummins, C. & Southwood, T. R. Incidence of
Henoch–Schönlein purpura, Kawasaki disease,
and rare vasculitides in children of different
ethnic origins. Lancet 360, 1197–1202 (2002).
6. Yang, Y. H. et al. A nationwide survey on
epidemiological characteristics of childhood
Henoch–Schönlein purpura in Taiwan.
Rheumatology 44, 618–622 (2005).
7. Mills, J. A. et al. The American College
of Rheumatology 1990 criteria for the
classification of Henoch–Schönlein purpura.
Arthritis Rheum. 33, 1114–1121 (1990).
8. Jennette, J. C. et al. Nomenclature of systemic
vasculitides. Proposal of an international
consensus conference. Arthritis Rheum. 37,
187–192 (1994).
9. Ozen, S. et al. EULAR/PRES endorsed
consensus criteria for the classification of
childhood vasculitides. Ann. Rheum. Dis. 65,
936–941 (2006).
10. Ozen, S. et al. EULAR/PRINTO/PRES criteria
for Henoch–Schönlein purpura, childhood
polyarteritis nodosa, childhood Wegener
granulomatosis and childhood Takayasu arteritis:
Ankara 2008. Part II: Final classification criteria.
Ann. Rheum. Dis. 69, 798–806 (2010).
11. Lau, K. K. et al. Serum levels of galactose-
deficient IgA in children with IgA nephropathy
and Henoch–Schönlein purpura. Pediatr.
Nephrol. 22, 2067–2072 (2007).
12. Davin, J. C., Ten Berge, I. J. & Weening, J. J. What
is the difference between IgA nephropathy and
Henoch–Schönlein purpura nephritis? Kidney Int.
59, 823–834 (2001).
13. Meadow, S. R. & Scott, D. G. Berger disease:
Henoch–Schönlein without the rash. J. Pediatr.
106, 27–32 (1985).
14. White, R. H. R., Yoshikawa, N. & Feehally, J.
In Pediatric Nephrology 4th edn, Ch. 41 (eds
Barratt, T. M., Avner, E. D. & Harmon, W. E.)
691–706 (Williams & Wilkins, 1999).
15. Goldstein, A. R., White, R. H. R., Akuse, R. &
Chantler, C. Long-term follow-up of childhood
Henoch–Schönlein nephritis. Lancet 339,
280–282 (1992).
16. Koskimies, O., Mir, S., Rapola, J. & Vilska, J.
Henoch–Schönlein nephritis: long-term
prognosis of unselected patients. Arch. Intern.
Med. 56, 482–484 (1981).
17. Bogdanovic, R. Henoch–Schönlein purpura
nephritis in children: risk factors, prevention and
treatment. Acta Paediatr. 98, 1882–1889 (2009).
18. Broyer, M. In Néphrologie Pédiatrique [French]
3rd edn, Ch. 2 (eds Royer, P. et al.) 75–98
(Flammarion Médecine-Sciences, 1983).
19. Jauhola, O. et al. Renal manifestations of
Henoch–Schönlein purpura in a 6‑month
prospective study of 223 children. Arch. Dis.
Child. 95, 877–882 (2010).
20. Narchi, H. Risk of long term renal impairment
and duration of follow up recommended for
Henoch–Schönlein purpura with normal or
minimal urinary findings: a systematic review.
Arch. Dis. Child. 90, 916–920 (2005).
21. Kaku, Y., Nohara, K. & Honda, S. Renal
involvement in Henoch–Schönlein purpura:
a multivariate analysis of prognostic factors.
Kidney Int. 53, 1755–1759 (1998).
22. Coppo, R., Mazzucco, G., Cagnoli, L., Lupo, A.
& Schena, F. P. Long-term prognosis of Henoch–
Schönlein nephritis in adults and children.
Italian Group of Renal Immunopathology
Collaborative Study on Henoch–Schönlein
purpura. Nephrol. Dial. Transplant. 12,
2277–2283 (1997).
23. Ronkainen, J., Nuutinen, M. & Koskimies, O. The
adult kidney 24 years after childhood Henoch–
Schönlein purpura: a retrospective cohort study.
Lancet 360, 666–670 (2002).
24. Scharer, K. et al. Clinical outcome of Schönlein–
Henoch purpura nephritis in children. Pediatr.
Nephrol. 13, 816–823 (1999).
25. Counahan, R. et al. Prognosis of Henoch–
Schönlein nephritis in children. Br. Med. J. 2,
11–14 (1977).
26. Bunchman, T. E., Mauer, S. M., Sibley, R. K.
& Vernier, R. L. Anaphylactoid purpura:
Characteristics of 16 patients who progressed
to renal failure. Pediatr. Nephrol. 2, 393–397
(1988).
27. Niaudet, P., Murcia, I., Beaufils, H., Broyer, M. &
Habib, R. Primary IgA nephropathies in children:
prognosis and Treatment. Adv. Nephrol. Necker
Hospital 22, 121–140 (1993).

NATURE REVIEWS | NEPHROLOGY VOLUME 10  |  OCTOBER 2014  |  571

© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
28. Nussinovitch, N., Elishkevitz, K., Volovitz, B. &
Nussinovitch, M. Hypertension as a late sequela
of Henoch–Schönlein purpura. Clin. Pediatr. 44,
543–547 (2005).
29. Emancipator, S. N. In Pathology of the Kidney
4th edn, Ch. 6 (ed. Heptinstall, R. H.) 389–476
(Little Brown, 1993).
30. Haas, M. In Pathology of the Kidney 6th edn,
Ch. 10 (eds Jennette, J. C. et al.) 423–486
(Lippincott, Williams & Wilkins, 2007).
31. Coppo, R. et al. Predictors of outcome in
Henoch–Schönlein nephritis in children and
adults. Am. J. Kidney Dis. 47, 993–1003 (2006).
32. Wakaki, H. et al. Henoch–Schönlein purpura
nephritis with nephrotic state in children:
predictors of poor outcomes. Pediatr. Nephrol.
26, 921–925 (2011).
33. Yoshikawa, N., White, R. H. & Cameron, A. H.
Prognostic significance of the glomerular
changes in Henoch–Schönlein nephritis. Clin.
Nephrol. 16, 223–229 (1981).
34. Pillebout, E. et al. Henoch–Schönlein purpura
in adults: outcome and prognostic factors.
J. Am. Soc. Nephrol. 13, 1271–1278 (2002).
35. Fogazzi, G. B. et al. Long-term outcome of
Schönlein–Henoch nephritis in the adult.
Clin. Nephrol. 31, 60–66 (1989).
36. Meadow, S. R. et al. Schönlein–Henoch
nephritis. Q. J. Med. 41, 241–258 (1972).
37. Heaton, J. M., Turner, D. R. & Cameron, J. S.
Localization of glomerular “deposits” in
Henoch–Schönlein nephritis. Histopathology 1,
93–104 (1977).
38. Schillinger, F. et al. Severe Schönlein–Henoch
nephritis in adults: a report of twenty cases
[French]. Nephrologie 21, 247–252 (2000).
39. Shenoy, M., Bradbury, M. G., Lewis, M. A. &
Webb, N. J. Outcome of Henoch–Schönlein
purpura nephritis treated with long-term
immunosuppression. Pediatr. Nephrol. 22,
1717–1722 (2007).
40. Soylemezoglu, O. et al. Henoch–Schönlein
nephritis: a nationwide study. Nephron Clin.
Pract. 112, 199–204 (2009).
41. Shrestha, S. et al. Henoch Schönlein purpura
with nephritis in adults: adverse prognostic
indicators in a UK population. Q. J. Med. 99,
253–265 (2006).
42. Habib, R. In Néphrologie Pédiatrique 3rd edn,
Ch. 11 (eds. Royer, P. et al.) 342–350
(Flammarion Médecine-Sciences, 1983).
43. Working Group of the International IgA
Nephropathy Network and the Renal Pathology
Society et al. The Oxford classification of IgA
nephropathy: rationale, clinicopathological
correlations, and classification. Kidney Int. 76,
534–545 (2009).
44. Bennett, W. M. & Kincaid-Smith, P. Macroscopic
hematuria in mesangial nephropathy: correlation
with glomerular crescents and renal dysfunction.
Kidney Int. 23, 392–400 (1983).
45. Foster, B. J., Bernard, C., Drummond, K. N. &
Sharma, A. K. Effective therapy for severe
Henoch–Schönlein purpura nephritis with
prednisone and azathioprine: a clinical and
histopathologic study. J. Pediatr. 136, 370–375
(2000).
46. Kanaan, N. et al. Recurrence and graft loss after
kidney transplantation for Henoch–Schönlein
purpura nephritis: a multicenter analysis. Clin. J.
Am. Soc. Nephrol. 6, 1768–1772 (2011).
47. Thervet, E. et al. Histologic recurrence of
Henoch–Schönlein purpura nephropathy after
renal transplantation on routine allograft biopsy.
Transplantation 92, 907–912 (2011).
48. Ponticelli, C. et al. Kidney transplantation in
patients with IgA mesangial glomerulonephritis.
Kidney Int. 60, 1948–1954 (2001).
572  |  OCTOBER 2014  |  VOLUME 10  www.nature.com/nrneph
49. Briganti, E. M., Russ, G. R., McNeil, J. J.,
Atkins, R. C. & Chadban, S. J. Risk of renal
allograft loss from recurrent glomerulonephritis.
N. Engl. J. Med. 347, 103–109 (2002).
50. Kincaid-Smith, P., Nicholls, K. & Birchall, I.
Polymorphs infiltrate glomeruli in mesangial IgA
glomerulonephritis. Kidney Int. 36, 1108–1111
(1989).
51. Cunningham, M. A., Kitching, A. R., Tipping, P. G.
& Holdsworth, S. R. Fibrin independent
proinflammatory effects of tissue factor in
experimental crescentic glomerulonephritis.
Kidney Int. 66, 647–654 (2004).
52. Kamitsuji, H. et al. Urinary cross-linked fibrin
degradation products in glomerular disease
with crescents. Clin. Nephrol. 29, 124–128
(1988).
53. Szeto, C. C. et al. Grading of acute and chronic
renal lesions in Henoch–Schönlein purpura.
Mol. Pathol. 14, 635–640 (2001).
54. Vassalli, P. & McCluskey, R. T. The pathogenetic
role of the coagulation process in rabbit Masugi
nephritis. Am. J. Pathol. 45, 653–677 (1964).
55. Hisano, S., Matsushita, M., Fujita, T. &
Iwasaki, H. Activation of the lectin complement
pathway in Henoch–Schönlein purpura nephritis.
Am. J. Kidney. Dis. 45, 295–302 (2005).
56. Novak, J., Julian, B. A., Mestecky, J. &
Renfrow, M. B. Glycosylation of IgA1 and
pathogenesis of IgA nephropathy. Semin.
Immunopathol. 34, 365–382 (2012).
57. Allen, A. C., Willis, F. R., Beattie, T. J. &
Feehally, J. Abnormal IgA glycosylation in
Henoch–Schönlein purpura restricted to patients
with clinical nephritis. Nephrol. Dial. Transplant.
13, 930–934 (1998).
58. Serino, G., Sallustio, F., Cox, S. N., Pesce, F. &
Schena, F. P. Abnormal miR‑148b expression
promotes aberrant glycosylation of IgA1 in IgA
nephropathy. J. Am. Soc. Nephrol. 23, 814–824
(2012).
59. Chintalacharuvu, S. R. et al. T cell cytokines
determine the severity of experimental IgA
nephropathy by regulating IgA glycosylation.
Clin. Exp. Immunol. 126, 326–333 (2001).
60. Chintalacharuvu, S. R. et al. T cell cytokine
polarity as a determinant of immunoglobulin A
(IgA) glycosylation and the severity of
experimental IgA nephropathy. Clin. Exp.
Immunol. 153, 456–462 (2008).
61. Inoshita, H. et al. Disruption of SMAD4
expression in T cells leads to IgA nephropathy-
like manifestations. PLoS ONE 8, e78736
(2013).
62. Yamada, K. et al. Down-regulation of core 1 β1,
3-galactosyltransferase and Cosmc by TH2
cytokine alters O‑glycosylation of IgA1. Nephrol.
Dial. Transplant. 25, 3890–3897 (2010).
63. Suzuki, H. et al. Aberrantly glycosylated IgA1 in
IgA nephropathy patients is recognized by IgG
antibodies with restricted heterogeneity. J. Clin.
Invest. 119, 1668–1677 (2009).
64. Wyatt, R. J. & Julian, B. A. IgA nephropathy.
N. Engl. J. Med. 368, 2402–2414 (2013).
65. Smith, A. C., Molyneux, K., Feehally, J. &
Barratt, J. O‑glycosylation of serum IgA1
antibodies against mucosal and systemic
antigens in IgA nephropathy. J. Am. Soc. Nephrol.
17, 3520–3528 (2006).
66. Davin, J. C., Forget, P. & Mahieu, P. R. Increased
intestinal permeability to 51Cr EDTA is correlated
with IgA immune complex-plasma levels in
children with IgA-associated nephropathies.
Acta Paediatr. Scand. 77, 118–124 (1988).
67. de Fijter, J. W. et al. Deficient IgA1 immune
response to nasal cholera toxin subunit B
in primary IgA nephropathy. Kidney Int. 50,
952–961 (1996).
© 2014 Macmillan Publishers Limited. All rights reserved
68. Monteiro, R. C. The role of IgA and IgA Fc
receptors in inflammation. J. Clin. Immunol. 30,
1–9 (2010).
69. Vuong, M. T. et al. Association of soluble CD89
levels with disease progression but not
susceptibility in IgA nephropathy. Kidney Int.78,
1281–1287 (2010).
70. Davin, J. C. et al. Possible pathogenic role of IgE
in Henoch–Schönlein purpura. Pediatr. Nephrol.
8, 169–171 (1994).
71. Kawasaki, Y., Hosoya, M. & Suzuki, H. Possible
pathologenic role of interleukin‑5 and eosino
cationic protein in Henoch–Schönlein purpura
nephritis. Pediatr. Int. 47, 512–517 (2005).
72. Namgoong, M. K., Lim, B. K. & Kim, J. S.
Eosinophil cationic protein in Henoch–Schönlein
purpura and in IgA nephropathy. Pediatr. Nephrol.
11, 703–706 (1997).
73. Chen, A., Chen, W. P., Sheu, L. F. & Lin, C. Y.
Pathogenesis of IgA nephropathy: in vitro
activation of human mesangial cells by IgA.
J. Pathol. 173, 119–126 (1994).
74. Gomez-Guerrero, C., Lopez-Armada, M. J.,
Gonzalez, E. & Egido, J. Soluble IgA and IgG
aggregates are catabolized by cultured rat
mesangial cells and induce production of TNF‑α
and IL‑6, and proliferation. J. Immunol. 153,
5247–5255 (1994).
75. Oortwijn, B. D. et al. Differential glycosylation
of polymeric and monomeric IgA: a possible role
in glomerular inflammation in IgA nephropathy.
J. Am. Soc. Nephrol. 17, 3529–3539 (2006).
76. Suzuki, H. et al. The pathophysiology of IgA
nephropathy. J. Am. Soc. Nephrol. 22,
1795–1803 (2011).
77. Yan, Y., Xu, L. X., Zhang, J. J., Zhang, Y. &
Zhao, M. H. Self-aggregated deglycosylated IgA1
with or without IgG were associated with the
development of IgA nephropathy. Clin. Exp.
Immunol. 144, 17–24 (2006).
78. Schlöndorff, D. & Banas, B. The mesangial cell
revisited: no cell is an island. J. Am. Soc. Nephrol.
20, 1179–1187 (2009).
79. Wan, J. X. et al. Complement 3 is involved in
changing the phenotype of human glomerular
mesangial cells. J. Cell Physiol. 213, 495–501
(2007).
80. Camilla, R. et al. Oxidative stress and galactose-
deficient IgA1 as markers of progression in IgA
nephropathy. Clin. J. Am. Soc. Nephrol. 6,
1903–1911 (2011).
81. Wang, C. et al. Mesangial cells stimulated
by immunoglobin A1 from IgA nephropathy
upregulate transforming growth factor β1
synthesis in podocytes via renin–angiotensin
system activation. Arch. Med. Res. 41, 255–260
(2010).
82. Coppo, R. et al. Aberrantly glycosylated IgA1
induces mesangial cells to produce platelet-
activating factor that mediates nephrin loss in
cultured podocytes. Kidney Int. 77, 417–427
(2010).
83. Horii, Y. et al. Involvement of IL‑6 in mesangial
proliferative glomerulonephritis. J. Immunol. 143,
3949–3955 (1989).
84. Kashem, A. et al. Glomerular FcαR expression
and disease activity in IgA nephropathy. Am. J.
Kidney Dis. 30, 389–396 (1997).
85. Niemir, Z. I. et al. PDGF and TGF‑β contribute
to the natural course of human IgA
glomerulonephritis. Kidney Int. 48, 1530–1541
(1995).
86. Lopez-Armada, M. J., Gomez-Guerrero, C. &
Egido, J. Receptors for immune complexes
activate gene expression and synthesis of matrix
proteins in cultured rat and human mesangial
cells: role of TGF‑β. J. Immunol. 157,
2136–2142 (1996).

87. Terada, Y. et al. Expression of PDGF and PDGF
receptor mRNA in IgA nephropathy. J. Am. Soc.
Nephrol. 8, 817–819 (1997).
88. Yokoyama, H. et al. Urinary levels of chemokines
(MCAF/MCP‑1, IL‑8) reflect distinct disease
activities and phases of human IgA
nephropathy. J. Leukocyte Biol. 63, 493–499
(1998).
89. Camussi, G. et al. Tumor necrosis factor induces
contraction of mesangial cells and alters their
cytoskeletons. Kidney Int. 38, 795–802 (1990).
90. Ha, T. S. The role of tumor necrosis factor‑α in
Henoch–Schönlein purpura. Pediatr. Nephrol. 20,
149–153 (2005).
91. Couser, W. G. Glomerulonephritis. Lancet 353,
1509–1515 (1999).
92. Jennette, J. C. Rapidly progressive crescentic
glomerulonephritis. Kidney Int. 63, 1164–1177
(2003).
93. Bonsib, S. Glomerular basement membrane
discontinuities: scanning electron microscopic
study of acellular glomeruli. Am. J. Pathol. 119,
357–360 (1985).
94. Atkins, R. C., Nikolic-Paterson, D. J., Song, Q.
& Lan, H. Y. Modulators of crescentic
glomerulonephritis. J. Am. Soc. Nephrol. 7,
2271–2278 (1996).
95. Lan, H. Y., Nikolic-Paterson, D. J. & Atkins, R. C.
Involvement of activated periglomerular
leukocytes in the rupture of Bowman’s capsule
and crescent progression in experimental
glomerulonephritis. Lab. Invest. 67, 743–751
(1992).
96. Lan, H. Y., Nikolic-Paterson, D. J., Mu, W.,
Vannice, J. L. & Atkins, R. C. Interleukin‑1
receptor antagonist halts the progression of
established crescentic glomerulonephritis in
the rat. Kidney Int. 47, 1303–1309 (1995).
97. Isaka, Y. et al. Gene therapy by transforming
growth factor‑β receptor–IgG Fc chimera
suppressed extracellular matrix accumulation in
experimental glomerulonephritis. Kidney Int. 55,
465–475 (1999).
98. Adler, S. & Brady, H. R. Cell adhesion molecules
and the glomerulopathies. Am. J. Med. 107,
371–386 (1999).
99. Morel-Maroger Striker, L., Killen, P. D., Chi, E.
& Striker, G. E. The composition of
glomerulosclerosis. I. Studies in focal sclerosis,
crescentic glomerulonephritis, and
membranoproliferative glomerulonephritis.
Lab. Invest. 51, 181–192 (1984).
100. Mathieson, P. W. The ins and outs of glomerular
crescent formation. Clin. Exp. Immunol. 110,
155–157 (1997).
101. Emancipator, S. N. IgA nephropathy: morphologic
expression and pathogenesis. Am. J. Kidney. Dis.
23, 451–462 (1994).
102. Border, W. A. & Noble, N. A. Transforming growth
factor in tissue fibrosis. N. Engl. J. Med. 331,
1286–1292 (1994).
NATURE REVIEWS | NEPHROLOGY VOLUME 10  |  OCTOBER 2014  |  573
103. Lin, Q. et al. Henoch–Schönlein purpura with
hypocomplementemia. Pediatr. Nephrol. 27,
801–806 (2012).
104. Roos, A. et al. Glomerular activation of the lectin
pathway of complement in IgA nephropathy is
associated with more severe renal disease.
J. Am. Soc. Nephrol. 17, 1724–1734 (2006).
105. Espinosa, M. et al. Mesangial C4d deposition:
a new prognostic factor in IgA nephropathy.
Nephrol. Dial. Transplant. 24, 886–891 (2009).
106. McLean, R. H., Michael, A. F., Fish, A. J. &
Vernier, R. L. In Pediatric Nephrology 1st edn,
Ch. 25 (eds Rubin, M. I. & Barratt, T. M.)
584–588 (Williams & Wilkins, 1975).
107. Coppo, R. & Amore, A. In Pediatric Nephrology
5th edn, Ch. 25 (eds Avner, E. D. et al.) 851–864
(Lippincott Williams & Wilkins, 2004).
108. Rees, L., Webb, N. J. A. & Brogan, P. A. In
Paediatric Nephrology 2nd edn, Ch. 20 (eds
Rees, L. et al.) 312–313 (Oxford University
Press, 2007).
109. Tarshish, P., Bernstein, J. & Edelmann, C. M. Jr.
Henoch–Schönlein purpura nephritis: course of
disease and efficacy of cyclophosphamide.
Pediatr. Nephrol. 19, 51–56 (2004).
110. Jauhola, O. et al. Cyclosporine A vs.
methylprednisolone for Henoch–Schönlein
nephritis: a randomized trial. Pediatr. Nephrol.
26, 2159–2166 (2011).
111. Pillebout, E. et al. Addition of cyclophosphamide
to steroids provides no benefit compared with
steroids alone in treating adult patients with
severe Henoch Schönlein purpura. Kidney Int.
78, 495–502 (2010).
112. Kidney Disease: Improving Global Outcomes.
Chapter 11: Henoch–Schönlein purpura
nephritis. Kidney Int. Suppl. 2, 218–220 (2012).
113. Davin, J. C. Henoch–Schönlein purpura nephritis:
pathophysiology, treatment, and future strategy.
Clin. J. Am. Soc. Nephrol. 6, 679–689 (2011).
114. Hattori, M. et al. Plasmapheresis as the sole
therapy for rapidly progressive Henoch–
Schönlein purpura nephritis in children. Am. J.
Kidney Dis. 33, 427–433 (1999).
115. Shenoy, M., Ognjanovic, M. V. & Coulthard, M. G.
Treating severe Henoch–Schönlein and IgA
nephritis with plasmapheresis alone. Pediatr.
Nephrol. 22, 1167–1171 (2007).
116. Niaudet, P. & Habib, R. Methylprednisolone pulse
therapy in the treatment of severe forms of
Schönlein–Henoch purpura nephritis. Pediatr.
Nephrol. 12, 238–243 (1998).
117. Andersen, R. F., Rubak, S., Jespersen, B. &
Rittig, S. Early high dose immunosuppression
in Henoch–Schönlein nephrotic syndrome may
improve outcome. Scand. J. Urol. Nephrol. 43,
409–415 (2009).
118. Flynn, J. T., Smoyer, W. E., Bunchman, T. E.,
Kershaw, D. B. & Sedman, A. B. Treatment of
Henoch–Schönlein purpura glomerulonephritis
in children with high-dose corticosteroids plus
© 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS
oral cyclophosphamide. Am. J. Nephrol. 21,
128–133 (2001).
119. Kawasaki, Y., Suyama, K., Hashimoto, K. &
Hosoya, M. Methylprednisolone pulse plus
mizoribine in children with Henoch–Schönlein
purpura nephritis. Clin. Rheumatol. 30, 529–535
(2011).
120. Kawasaki, Y., Suzuki, J. & Suzuki, H. Efficacy of
methylprednisolone and urokinase pulse therapy
combined with or without cyclophosphamide in
severe Henoch–Schönlein nephritis: a clinical
and histopathological study. Nephrol. Dial.
Transplant. 19, 858–864 (2004).
121. Kawasaki, Y., Suzuki, J., Nozawa, R., Suzuki, S.
& Suzuki, H. Efficacy of methylprednisolone and
urokinase pulse therapy for severe Henoch–
Schönlein nephritis. Pediatrics 111, 785–789
(2003).
122. Ashton, H., Frenk, E. & Stevenson, C. J.
Therapeutics. XV. The management of Henoch–
Schönlein purpura. Br. J. Dermatol. 85, 199–203
(1971).
123. Borges, W. H. Anaphylactoid purpura. Med. Clin.
North Am. 56, 201–206 (1972).
124. Ou, Z. L. et al. Effective methylprednisolone dose
in experimental crescentic glomerulonephritis.
Am. J. Kidney Dis. 37, 411–417 (2001).
125. Lameire, N. H. et al. Acute kidney injury: an
increasing global concern. Lancet 382, 170–179
(2013).
126. Branten, A. J., Kock-Jansen, M., Klasen, I. S. &
Wetzels, J. F. Urinary excretion of complement
C3d in patients with renal diseases. Eur. J. Clin.
Invest. 33, 449–456 (2003).
127. Wickman, L. et al. Urine podocyte mRNAs,
proteinuria, and progression in human
glomerular diseases. J. Am. Soc. Nephrol. 24,
2081–2095 (2013).
128. Delanghe, S. E. et al. Soluble transferrin
receptor in urine, a new biomarker for IgA
nephropathy and Henoch–Schönlein purpura
nephritis. Clin. Biochem. 46, 591–597 (2013).
129. Moura, I. C. et al. Identification of the transferrin
receptor as a novel immunoglobulin (Ig)A1
receptor and its enhanced expression on
mesangial cells in IgA nephropathy. J. Exp. Med.
194, 417–425 (2001).
130. Park, S. J. et al. Advances in our understanding of
the pathogenesis of Henoch–Schönlein purpura
and the implications for improving its diagnosis.
Expert Rev. Clin. Immunol. 9, 1223–1238 (2013).
131. Ninchoji, T. et al. Treatment strategies for
Henoch–Schönlein purpura nephritis by
histological and clinical severity. Pediatr.
Nephrol. 26, 563–569 (2011).
Author contributions
The authors contributed equally to researching the
data for the article, discussions of its content, writing
the article and review and/or editing of the
manuscript before submission.