the pathology of lupus nephritis -...

KmwttnmipedccrHiccg

DA

0©

2

The Pathology of Lupus Nephritis

Melvin M. Schwartz

Summary: An international working group of clinicians and pathologists met in 2003 underthe auspices of the International Society of Nephrology (ISN) and the Renal Pathology Society(RPS) to revise and update the 1982 and 1995 World Health Organization classification oflupus glomerulonephritis. This article compares and contrasts the ISN/RPS classification andthe antecedent World Health Organization classifications. Although systemic lupus erythem-atosus is the prototypical systemic immune-complex disease, several non–immune-complexmechanisms of glomerular injury and dysfunction have been proposed, and this articlesummarizes the evidence supporting the pathogenic mechanisms of lupus vasculitis, glomer-ular capillary thrombosis, and lupus podocytopathy. The most significant and controversialfeature of the ISN/RPS classification is the separation of diffuse glomerulonephritis intoseparate classes with either segmental (class IV-S) or global (class IV-G) lesions. Several groupshave tested the prognostic significance of this separation, and this article discusses theimplications of these studies for the ISN/RPS classification.Semin Nephrol 27:22-34 © 2007 Elsevier Inc. All rights reserved.Keywords: Systemic lupus erythematosus, pathology, ISN/RPS classification, immune com-plex disease, in situ immune complex disease, podocytopathy, collapsing FSGS, thromboticmicroangiopathy, vasculitis, TTP-like

jTpcpigcutRtpo

TC

EspmimW

lemperer et al1 were the first to describethe light microscopic renal pathology ofsystemic lupus erythematosus (SLE) glo-

erulonephritis (GN) with cellular proliferation,ire-loops, hematoxylin bodies, and fibrin

hrombi in autopsied patients dying with the un-reated disease. However, informed by pathoge-etic insights and the newer techniques of im-unopathology and electron microscopy, the

nterpretation and the prognostic and thera-eutic implications of the pathology havevolved. Early in the renal biopsy era, theemonstration of immune deposits by fluores-ence and electron microscopy led to the con-lusion that all of the pathology observed in SLEenal disease was immune-complex mediated.owever, despite the strong evidence support-

ng the existence of DNA–anti-DNA immuneomplexes in the circulation and aggregatesontaining DNA and anti-DNA antibodies in thelomeruli, other mechanisms of glomerular in-

epartment of Pathology, Rush University Medical Center, Chicago, IL.ddress requests for reprints to Melvin M. Schwartz, MD, Department ofPathology, Rush University Medical Center, 1753 W. Congress Pkwy,Chicago, IL 60612. E-mail: [email protected]

270-9295/07/$ - see front matter
t2007 Elsevier Inc. All rights reserved. doi:10.1016/j.semnephrol.2006.09.005
Seminar2

ury have been described in patients with SLE.his article focuses on 2 aspects of the renalathology of SLE that are of concern both to thelinician, who uses the renal biopsy to assessrognosis and to guide therapy, and to the

nvestigator, who uses the biopsy findings toain pathogenetic insight. The first is the re-ently proposed classification of lupus glomer-lar disease developed under the auspices ofhe International Society of Nephrology and theenal Pathology Society (ISN/RPS classifica-

ion).2,3 The second is the non–immune-com-lex pathogenetic mechanisms proposed to beperative in SLE renal disease.

HE ISN/RPSLASSIFICATION OF GLOMERULAR DISEASE

arly renal biopsy studies in SLE patientshowed a spectrum of glomerular disease4 andrognosis was a function of the underlying glo-erular lesion,5 and the glomerular pathology

ncluded segmental (focal), diffuse (global), andembranous forms of GN.6,7 According toeening et al,2,3 Robert McCluskey codified

he first histologic classification of lupus ne-

s in Nephrology, Vol 27, No 1, January 2007, pp 22-34

poKi(fimacfiats

tddTfpglcdtIw

Pathology of lupus nephritis 23

hritis, and, after modification by the Pathol-gy Committee of the International Study ofidney Disease in Children, it was published

n 1982 by the World Health OrganizationWHO) (Table 1).8 Although the WHO classi-cation provided the pathologic context forany clinical studies, it was not universally

ccepted. The impetus to revisit the classifi-ation was the concern that the 1982 classi-cation and the 19959 version with 6 classesnd many subclasses were too cumbersomeo apply, in certain aspects difficult to under-tand, and key terms were undefined.

Table 1. Comparison Between the Modified WClass Modified WHO (1982) Critiqu

I Normal glomeruli No normclassifica) all techniques

b) EM/FM depositsII Pure mesangial

alterationsNo show

differenclassesa) Mild hypercellularity

b) Moderatehypercellularity

III Focal segmental GN* 1) Problegloba

2) No qudefini

IV Diffuse GN* 1) Categwidelsegmlesion

2) No qudefini

V Diffuse membranous GN 1) No declinicabetweand b

2) MixedcombmembactiveIII and

a) Pureb) Plus class IIc) Plus class IIId) Plus class IV

VI Advanced sclerosing GN No quandefiniti

*Subcategories: active necrotizing, active and sclerosing, and†Subcategories: active, active and chronic, and chronic lesio

Differences that are more fundamental were c

he interpretation and clinical implications of theistinction between focal segmental (class III) andiffuse (class IV) GN in the WHO classification.o some, segmental GN represented a different

orm of glomerular pathology and a differentathogenetic mechanism from the diffuselobal lesion seen in many cases of class IVupus glomerulonephritis. Others believed thatlass III and IV lesions represented quantitativeifferences in the same pathologic process. Al-hough distinguishing between classes III andV by the proportion of glomerular involvementithout considering sample size leads to mis-

1982) and the ISN/RPS (2004) ClassificationsHO 1982 ISN/RPS Classification (2004)

pathologic Minimal mesangial lupusnephritis. Normal by LM.EM/FM deposits

icaltween

d b)

Mesangial proliferative lupusnephritis with mesangialproliferation of any degree

focal

ativef focal

Focal lupus nephritis†�50% glomerular involvement

ompriseributedand global

ativef diffuse

Diffuse lupus nephritis† �50%glomerular involvement

a) IV-S �50% of glomerulihave segmental lesions

b) IV-G �50% of glomerulihave global lesions3

stratederenceasses a)

es

us withns in classonfusing

Membranous lupus nephritis

a) With or without mesangialproliferation or deposits

b) In biopsies with mixedmembranous and focal(class III) or diffuse (IV) GN,the glomerular lesions arediagnosed separately

eadvanced

Advanced sclerosing lupusnephritis �90% globalsclerosis without residualactivity

sing lesions.

HO (e of W

al ination

n clince bea) an

m ofl GNantit

tion oory cy distentalsantit

tion omonl diffen cl)class

iningranolesioIV c

titativon of

sclerons.

lassification of a significant number of biopsy

saucst

aIscaTiipccpnnitnmnlsafi

IM

TbbdtpWci

IP

Tgmmc

ditlchso

I

Cutditadrmsfwid

Famtei

24 M.M. Schwartz

pecimens,10 the latter interpretation prevailed,nd the ISN/RPS classification uses 50% glomer-lar involvement as the cut-off level betweenlasses III and IV. The implications of this deci-ion are discussed later (see section “Applyinghe ISN/RPS Classification”).

An international working group of cliniciansnd pathologists met under the auspices of theSN and the RPS with “the major objective oftandardizing definitions, emphasizing clini-ally relevant lesions, and encouraging uniformnd reproducible reporting between centers.”he classification of lupus GN that was proposed

s based on the 1982 WHO classification. Table 1,n which the 1982 WHO classification is com-ared with the ISN/RPS classification, lists spe-ific criticisms and problems that were ofoncern to the working group. This is a mor-hologic classification of lupus GN that doesot include coagulopathies, podocytopathy, oron-immunoglobulin-mediated glomerular cap-

llary necrosis (vasculitis) (see later). If present,he working group recommends separate diag-oses for these glomerular lesions and nonglo-erular vascular lesions, and tubulointerstitial

ephritis. The salient features of the 6 classes ofupus glomerular pathology in the ISN/RPS clas-ification are presented below and in Table 1,long with significant changes from the modi-ed 1982 and 1985 WHO classifications.

SN/RPS Class I: Minimalesangial Lupus Nephritis

he glomeruli are normal by light microscopy,ut they must have mesangial immune depositsy fluorescence microscopy or electron-denseeposits by electron microscopy. This defini-ion eliminates the normal glomeruli by all mor-hologic modalities in category (IA) in theHO classification, which is objectionable be-

ause a normal category should not be presentn a pathologic classification.

SN/RPS Class II: Mesangialroliferative Lupus Nephritis

he glomerular pathology is limited to mesan-ial hypercellularity of any degree or mesangialatrix expansion by light microscopy withesangial immune deposits seen by fluores-

ence microscopy or mesangial electron-dense I

eposits seen by electron microscopy. This def-nition eliminates the distinction drawn be-ween mild and moderate mesangial hypercel-ularity (categories IIA and IIB) in the WHOlassification because clinicopathologic studiesave produced no outcome data to support theeparation of mesangial lupus nephritis basedn the degree of mesangial cellularity.

SN/RPS Class III: Focal Lupus Nephritis

lass III is defined by the proportion of glomer-li with any active lesion or scar involving lesshan 50% of the glomeruli (Fig 1). It is notefined by the character of the pathology or its

ntraglomerular distribution. Although most ofhe active lesions seen in focal lupus nephritisre, in fact, segmental in their intraglomerularistribution, this strictly quantitative definitioneplaces the descriptive category III, “Focal seg-ental GN,” in the WHO classification. It

hould be noted that morphologically identicalocal segmental GN may be distributed moreidely, but when 50% or more of glomeruli are

nvolved, the biopsy specimen is classified asiffuse lupus nephritis (see section “Class IV:

igure 1. Focal segmental glomerulonephritis. There issegmental proliferative lesion that obliterates the glo-erular architecture and involves at least 50% of the

uft. The uninvolved glomerular lobules show mesangialxpansion and normal cellularity. Only 2 of 32 glomerulin the biopsy specimen were involved (ISN/RPS class
IIA). Hematoxylin and eosin, 66�.

Dv(

ID

Ddamw(mwlialIdspgtgm

Ftpg5tsdo


iffuse Lupus Nephritis”). Class III has 3 subdi-isions for active (III [A]), active and chronicIII [A/C]), and chronic (III [C]) lesions.

Table 2. Active Glomerular Lesions in SLE

Endocapillary hypercellularity with or withoutleukocyte infiltration and with substantialluminal reduction

KaryorrhexisFibrinoid necrosisRupture of glomerular basement membrane

(breaks)Crescents, cellular or fibrocellularSubendothelial deposits identifiable by light

microscopy (wire-loops)Intraluminal immune aggregates (hyaline

thrombi)Hematoxylin bodies2

NOTE. Although hematoxylin bodies are a classic featureof active lupus glomerulonephritis, the ISN/RPS classifi-cation does not include them among the active lesions.

Modified from Weening et al.2,3

igure 2. Focal segmental glomerulonephritis. (A) Theuft (arrow) and a focal fibrin exudate (arrowhead). Theatent capillaries and a mild increase in mesangial malomeruli (59%) in the biopsy specimen. This biopsy spe0% of the glomeruli are involved. Note the similarity ofhan 80% of the tuft and similar lesions involved 30 oegmental nature of the process indicated by the uninviffuse global (ISN/RPS IV-G [A]) because of the involvem
f the tuft. (A and B) Hematoxylin and eosin, 66�.
SN/RPS Class IV:iffuse Lupus Nephritis

iffuse lupus nephritis (ISN/RPS class IV) isefined by the proportion of glomeruli with anyctive lesion or scar (Table 2) involving 50% orore of the glomeruli. This class contains casesith global endocapillary and extracapillary

crescentic) proliferative GN involving 50% orore of the glomeruli and the rare cases withidespread subendothelial deposits (wire-

oops), unaccompanied by significant endocap-llary proliferation, seen by light microscopy. Itlso includes biopsy specimens with segmentalesions in 50% or more of the glomeruli. TheSN/RPS distinguishes between these 2 types ofiffuse lupus nephritis and categorizes themeparately as diffuse, segmental proliferative lu-us nephritis (segmental lesions in �50% oflomeruli) (Fig 2), and diffuse, global prolifera-ive lupus nephritis (global lesions in �50% oflomeruli) (Figs 2 and 3). The segmental glo-erular lesions involve variable proportions of

segmental proliferative lesion that involves 20% of thelved glomerular lobules show normal architecture withd cellularity. Similar lesions were present in 13 of 22is classified as ISN/RPS class IV-S (A) because more thansion to Figure 1. (B) A proliferative lesion involves moreglomeruli (85%) in the biopsy specimen. Despite thecapillaries (arrow), this biopsy specimen is classified asf more than 50% of the glomeruli and more than 80%

re is auninvotrix ancimenthe lef 35

olvedent o

tdocinCasaa

IM

GpssSwIclt�

mr2tml

IA

Ts

A

TWSsocctcpnIparpcsii

ncvpbib2ildlsv

Fmsfbse

26 M.M. Schwartz

he tuft, but by definition the ISN/RPS limits andefines segmental involvement as less than 50%f the glomerulus. The distinction betweenlasses IV-S and IV-G is based purely on thentraglomerular distribution of the lesion andot on the character of the inflammatory lesion.lasses IV-S and IV-G are each divided further,s are the focal lesions of category III, into 3ubdivisions for active (IV-S [A] and IV-G [A]),ctive and chronic (IV-S [A/C] and IV-G [A/C]),nd chronic (IV-S [C] and IV-G [C]) lesions.

SN/RPS Class V:embranous Lupus Nephritis

lobal or segmental subepithelial immune de-osits or light fluorescence or electron micro-copic findings that are the result of antecedentubepithelial deposits define ISN/RPS class V.mall subendothelial deposits are allowed ifire-loops are not visible by light microscopy.

n contrast to the modified WHO classification,lass V may have any degree of mesangial pro-iferation, but when class III or IV lesions occurogether with diffuse class V lesions (involving

igure 3. Diffuse global glomerulonephritis. The glo-erulus shows a lobular pattern with every capillary

howing endocapillary proliferation. This is a classic dif-use global glomerulonephritis (ISN/RPS class IV-G [A]),ut contrast the pathologic process to the similarly clas-ified lesion illustrated in Figure 2B. Hematoxylin andosin, 66�.

50% of the glomeruli and �50% of the glo- t

eruli, respectively), they are diagnosed sepa-ately. For example, membranous GN in which5% of the glomeruli have segmental prolifera-ive lesions or scars would be diagnosed asembranous lupus nephritis (class V) and focal

upus nephritis (class III).

SN/RPS Class VI:dvanced Sclerosing Lupus Nephritis

his class has 90% or more of glomeruli globallyclerosed without residual histologic activity.

PPLYING THE ISN/RPS CLASSIFICATION

he ISN/RPS classification is not simpler than theHO classification. However, the pathology of

LE GN is heterogeneous, and 6 categories withubcategories are necessary to catalogue all casesf lupus GN. The authors of the ISN/RPS classifi-ation modified the 1982 and 1995 WHO classifi-ations to remove ambiguity, to provide defini-ions of pathologic terms, and to eliminateategories that have not shown diagnostic orrognostic utility. For example, the category oformal by all modalities was removed from class, the subclasses based on the degree of mesangialroliferation were removed from classes II and V,nd segmental and diffuse proliferative GN wereemoved from class V. Although a predeterminedroportion of glomerular involvement (50%) washosen as an easy and reproducible criterion foreparating class III from class IV, the statisticalmplications of this division require careful exam-nation.

In SLE, glomerular inflammation (glomerulo-ephritis) may involve only some glomeruli (fo-al), leaving the remaining glomeruli unin-olved. The validity of the inference that theroportion of glomeruli with lesions seen in aiopsy specimen represent the proportion of

nvolved glomeruli in the kidneys can be testedy representing the probability of finding 0, 1,, . . . , n abnormal glomeruli in a biopsy spec-

men as a binomial distribution (Fig 4). It fol-ows from these statistical considerations thatichotomizing focal SLE GN with one class with

ess than 50% glomerular involvement and aecond class with 50% or more glomerular in-olvement will lead to misclassification of 10%
o 15% of the class III biopsy specimens as class

IsItwbcIgtkTeeccricmb1gotIofii

i

ismthiofIeibprugtci

tptdrgIavc

P

IairosAtpthant

C

I

Fm(R


V, and the same proportion of class IV biopsypecimens will be incorrectly included in classII. These statistically based errors of classifica-ion will be exaggerated in biopsy specimensith small numbers of glomeruli. Furthermore,y including the most severe class III lesions inlass IV and downgrading the least severe classV lesions to class III, the prognosis of bothroups will improve without changing the ac-ual distribution of involved glomeruli in theidneys or the overall outcome of the patients.his overlap in the distributions of biopsy cat-gories is analogous to the Will Rogers phenom-non, which refers to misleading statistical out-omes resulting from �stage migration” in thelassification of cancer patients.11 To avoid er-oneous classification, biopsy results must takento account these sampling problems. Practi-ally speaking, a biopsy specimen with 10 glo-eruli must have 8 involved glomeruli and a

iopsy specimen with 20 glomeruli must have4 involved glomeruli to ensure that 50% of thelomeruli in the kidneys will be involved. Usersf the ISN/RPS classification should be aware ofhe problems associated with separating classII from class IV on the basis of the proportionf glomerular involvement and of the potentialor misclassification of some biopsy specimensn this situation in which sample size is anmportant consideration.

The ISN/RPS diagnostic categories allow the

igure 4. Binomial distribution of percentage of abnor-al glomeruli in biopsy samples from kidneys with mild

10%), moderate (35%), and severe (60%) involvement.eprinted with permission from S. Karger AG, Basel.10

nclusion of active lesions, representing current f

nflammation, and scars, related to past epi-odes of disease activity. Classes III, IV, and Vay contain active or chronic lesions or a mix-

ure of both, and the active and chronic lesionsave very different prognostic and therapeutic

mplications. Therefore, the clinician shouldbserve the caveat that the abbreviated symbolsor the ISN/RPS categories (eg, III, IV-S, andV-G) are no substitute for an in-depth knowl-dge of the extent and nature of the glomerularnflammation and scars in the individual renaliopsy specimen. Therapeutic decisions andrognostication should be based on a pathologyeport that clearly states the number of glomer-li, the proportions of active and sclerosinglomerular lesions, the nature of the inflamma-ory lesion (proliferation, necrosis, thrombosis,rescents), and the distribution and pattern ofmmune deposits.

It was the consensus of the Working Grouphat the ISN/RPS classification is a work inrogress, and the participants anticipated thatesting of the classification would lead to vali-ation of some and elimination of other catego-ies and definitions. After discussing the patho-enetic implications of the renal biopsy findings,present the studies that have appeared since theppearance of the ISN/RPS classification that in-estigate the prognostic and pathogenetic signifi-ance of classes IV-S and IV-G.

ATHOGENESIS OF SLE GN

t is my thesis that multiple pathogenetic mech-nisms are capable of causing glomerular injuryn SLE, and this pathogenetic heterogeneity iseflected in the variable patterns of injury seenn renal biopsy and the different clinical pre-entations of patients with lupus renal disease.bundant data have accumulated concerning

he pathogenesis and etiology of lupus in ex-erimental models, but I wish to consider onlyhe implications of pathologic observations inuman beings. I begin by presenting 2 generallyccepted pathogenetic mechanisms in lupusephritis and follow with evidence for 3 addi-ional forms of glomerular injury.

irculating Immune-Complex Disease

mmune complexes deposited in the kidney
rom the circulation were the first pathogenic

mScsc(pMmmawlmciApmdlmiwwtpflR

ilItmitaSSGhdwIbIlIR

Hpmgdpomam

I

SapasptethwteslpcdtcirbmwtM

nplmTcp

28 M.M. Schwartz

echanism described in glomerular injury inLE.12 Serologic studies showing high levels ofirculating immune complexes, double-tranded DNA, and other nuclear antigens andonsumption of classic complement componentsC1, C4, and C3)13,14 support an immune-com-lex mechanism in patients with active lupus GN.orphologic demonstration of glomerular im-une reactants (immunoglobulins, comple-ent components, and presumably antigens)

nd electron-dense deposits also are consistentith an immune-complex mechanism. Sero-

ogic findings associated with circulating im-une complexes and activation of complement

orrelate best with large subendothelial depos-ts seen by light microscopy as wire-loops.15-17

lthough injection of preformed immune com-lexes into experimental animals gives, at best,esangial immune deposits and never repro-

uces the massive subendothelial deposits ofupus glomerulonephritis,18 this remains the

ost widely acknowledged form of glomerularnjury in SLE. The histologic findings associated

ith immune-complex deposition includingire loops, hyaline (nonfibrin, immune reac-

ant) capillary thrombi, and glomerular macro-hages19,20 are signs of active glomerular in-ammation (Table 2) and are seen only in ISN/PS classes III and IV.Furthermore, morphologic evidence of an

mmune-complex mechanism is more compel-ing in ISN/RPS class IV-G (diffuse global) thanV-S (diffuse segmental). Najafi et al21 comparedhe glomerular pathology of 35 biopsy speci-ens with diffuse SLE GN with 24 biopsy spec-

mens with segmental GN in more than 50% ofhe glomeruli (this classification only roughlypproximates ISN/RPS classes IV-G and IV-S; seeignificance of the Distinction Between Diffuseegmental [ISN/RPS Class IV-S] and Diffuselobal [ISN/RPS Class IV-G] GN). Wire-loops,yaline thrombi, and massive subendothelialeposits seen by electron microscopy occurredith significantly greater frequency in class

V-G biopsy specimens. Mittal et al22 comparediopsy specimens with ISN/RPS class IV-S withSN/RPS class IV-G. Wire-loops and subendothe-ial deposits were observed more frequently inSN/RPS class IV-G, but differences from ISN/
PS class IV-S were not statistically different. m
ill et al23 observed that biopsy specimens fromatients with ISN/RPS class IV-G (n � 31) wereore likely to have glomerular hyaline thrombi,

lomerular macrophage/monocytes, and suben-othelial deposits than biopsy specimens fromatients with ISN/RPS class IV-S (n � 15). Thesebservations imply that subendothelial im-une-complex deposits activate complement

nd mediate SLE GN, and this type of injury isost common in ISN/RPS class IV-G.

n Situ Immune-Complex Disease

LE membranous GN (MGN) (ISN/RPS class V)lso is immune-complex mediated, but the de-osits seen in SLE MGN are not usually associ-ted with glomerular inflammation. In fact, theubepithelial and intramembranous immune de-osits in SLE MGN are in the same location ashe deposits seen in idiopathic MGN and thexperimental model of MGN, Heymann nephri-is, and in both of these conditions there is aeavy load of immune deposits in the capillaryall without significant glomerular inflamma-

ion. The mechanisms responsible for the sub-pithelial localization of immune complexes inome patients with SLE and the subendothelialocalization in others are not understood com-letely, but the discrete patterns of immune-omplex deposition have been attributed toifferences in antibody and antigen characteris-ics and antigen location. However, there isonsiderable experimental evidence that themmune deposits in Heymann nephritis are theesult of the in situ reaction of unbound anti-ody with antigen, and although the in situechanism is unproven in human beings, theeight of experimental evidence supports

his mechanism in the pathogenesis of SLEGN.24,25

Thus, there appears to be 2 discrete pathoge-etic mechanisms of glomerular immune-com-lex localization in SLE associated with the pro-

iferative (principally ISN/RPS class IV-G) andembranous (ISN/RPS class V) forms of SLE GN.his dichotomy also is important clinically be-ause the therapy and prognosis for glomerularathology associated with these 2 forms of im-
une-complex disease are very different.

V

Gttiisctc(mllSntlsptitog

apapptaboicapactpGtf

T

Gos

FuTmpt


asculitic Lupus GN

lomerular involvement in the pauci-immune sys-emic vasculitides has a focal and segmental dis-ribution, and these forms of GN are destructivenflammatory lesions that are frequently necrotiz-ng and crescent-forming.26 Likewise, a focal andegmental distribution of glomerular pathology isharacteristic of many biopsy specimens from pa-ients with lupus GN including most ISN/RPSlass 3 (focal GN) and many in ISN/RPS class 4Fig 5).27 Although the pathogenesis of focal seg-ental lupus GN has been attributed to glomeru-

ar mesangial immune deposits, glomerular capil-ary thrombosis also has been invoked. Churg andobin8 noted that immune “deposits are found inearly every case (of lupus glomerulonephritis) inhe mesangium and less consistently in the capil-ary wall,” but the absence of deposits in the focalegmental glomerular lesions of focal lupus ne-hritis implies that they may not be produced byhe same mechanism as the diffuse form. A non–mmune-complex mechanism was suggested byhe report of necrotizing and crescentic GN thatccurred without significant glomerular immuno-lobulin deposits.28 In addition to the character

igure 5. Lupus vasculitis in a patient who had a positivendergone fibrinoid necrosis (arrow) without associated enhe uninvolved capillaries have normal architecture with thesangial matrix and cellularity. Although there were m

eriphery of the glomerular capillaries. (B) Fibrin in the sam
he segmental necrotizing nature of the lesion. (A) Hematoxylin
nd distribution of the glomerular lesions, theresence of a positive antineutrophil cytoplasmicutoantibody (ANCA) in more than one third ofatients with lupus GN, usually a P-ANCA sup-orts a vasculitic mechanism in lupus GN. Inhese patients a P-ANCA correlates with histologicctivity. The significance of this observation haseen questioned because the target antigens areften different from myeloperoxidase that usually

s seen in the systemic vasculitides and includeathepsin G and lactoferrin.29-31 However, Hill etl23 summarized the evidence concerning theathogenesis of segmental proliferative SLE GNnd concluded that the findings cast doubt on alassic immune-complex mechanism. Therefore,here is morphologic and serologic support for aauci-immune, pathogenetic mechanism in lupusN, which is separate and distinct from lesions

hat may result from antigen–antibody complexormation in the glomerular capillary wall.

hrombotic Microangiopathy

lomerular fibrin thrombi are a histologic featuref proliferative lupus GN,32 and in the presence ofevere glomerular inflammation, thrombus forma-

A (myeloperoxidase positive). (A) Several capillaries haveillary proliferation, and they contain karyorrhectic debris.

cate basement membranes, patent capillaries, and normalial immune deposits, they were not present around thetribution as the lesions seen by light microscopy supports

P-ANCdocap

in deliesange dis

and eosin, 163�. (B) Fibrin, 66�.

tasotiacantcerinWdtacttpto

diit

P

TvglHnimdptgriptpbtSctrhpiij(dntgcc

mhTietp

FtigTe

30 M.M. Schwartz

ion can result from endothelial activation or dam-ge with activation of the intrinsic coagulationystem. Thrombi also occur without other signsf glomerular inflammation in the thrombotichrombocytopenic purpura (TTP)-like lesion thats a rare complication of systemic lupus erythem-tosus (Fig. 6).33 Glomerular thrombosis also oc-urs in SLE in the absence of severe inflammation,nd this suggests several additional pathoge-etic mechanisms. A lupus anticoagulant or an-iphospholipid antibody is present in a signifi-ant proportion of patients with systemic lupusrythematosus, and these autoantibodies, di-ected against the phospholipid elements in thentrinsic clotting cascade, may arise by mecha-isms analogous to antinuclear antibodies.34

ith the onset of the antiphospholipid syn-rome, the patients may develop thromboses inhe arteries, veins, and glomerular capillaries. Inddition, patients with SLE may acquire a defi-iency of von Willebrand’s factor–cleaving pro-ease (ADAMTS-13), and the resulting large mul-imers of von Willebrand’s factor lead tolatelet aggregation and thrombosis.35 Auto-an-ibody against ADAMTS-13 is a common causef acquired von Willebrand’s factor-protease

igure 6. Thrombotic microangiopathy. There is ahrombus in the arteriole (arrow) that appears to extendnto the glomerulus (between the arrowheads), and thelomerular thrombus contains fragmented red blood cells.he remaining capillaries are congested. Hematoxylin and

fosin, 66�.

eficiency in patients with TTP, and althoughmmunoglobulin G anti–ADAMTS-13 antibodys unusual in patients with SLE,36,37 the associa-ion remains a potential cause of thrombosis.

odocytopathy

he nephrotic syndrome is a sign of renal in-olvement in SLE, and it usually is attributed tolomerular inflammation or disruption of capil-ary wall function caused by immune deposits.owever, there are cases of SLE in which theephrotic syndrome occurs in the absence of

nflammatory cell infiltrates, capillary wall im-une deposits, and no or minimal mesangial

eposits.38,39 The glomerular pathology in theseatients is usually minimal change disease, buthere also are cases of collapsing focal segmentallomerulosclerosis (Fig 7).40,41 The co-occur-ence of 2 independent diseases, lupus and min-mal glomerulopathy, could explain the ne-hrotic syndrome, but Hertig et al39 showedhat the nephrotic syndrome in this group ofatients was much more frequent than coulde explained by the chance association be-ween minimal change glomerulopathy andLE. Kraft et al40 showed that diffuse foot pro-ess effacement is the only morphologic featurehat distinguishes nephrotic and nonnephroticange proteinuric SLE patients who did notave glomerular inflammation or necrosis oreripheral capillary-wall immune deposits. The

nvestigators used the term podocytopathy tomply that non–immune-complex–mediated in-ury to the glomerular visceral epithelial cellpodocyte), rather than immune-complex–me-iated glomerular injury, was responsible forephrotic range proteinuria. These observa-ions raise the intriguing possibility that in thisroup of SLE patients a primary pathologic pro-ess involving the glomerular visceral epithelialell causes the nephrotic syndrome.

The cause of idiopathic minimal change glo-erulopathy is unknown, but the Shalhoub42

ypothesis holds that a product of aberrant-cell function is responsible for glomerular ep-

thelial cell injury resulting in diffuse podocyteffacement. There is growing evidence that ac-ivated T cells,43,44 abnormal T-cell cytokine ex-ression,45,46 and an unidentified circulating

actor of presumed T-cell origin47-49 are in-

vittbtSttoifmpmaodp

aismc

mawmwwmlmmca

SDSD

Tdpsfp

Fidgfd d-Schi


olved in the pathogenesis of podocyte diseasesn human beings that result in proteinuria andhe nephrotic syndrome. These findings raisehe question of whether activated T cells coulde implicated similarly in the pathogenesis ofhe podocyte lesion seen in some patients withLE. T-cell activation seems to play an impor-ant role in the pathogenesis of SLE by alteringhe production of cytokines and the expressionf cellular adhesion molecules, as well as induc-

ng B cells to produce autoantibodies.50 There-ore, one might speculate that the underlyingechanism of SLE-related podocytopathy, asroposed in idiopathic minimal-change glo-erulopathy, could similarly be mediated by

ctivated T cells. The concurrent presentationf clinical renal abnormalities (nephrotic syn-rome) and evidence of active SLE in someatients supports this hypothesis.These different pathogenetic mechanisms

re recognized most easily when they occur insolation. However, the glomerular pathologyeen on renal biopsy often contains evidence ofultiple types of injury, and it is important to

igure 7. Lupus podocytopathy. (A) There is diffuse cozation, and periodic acid-Schiff (PAS)-positive absorptioeposits of immunoglobulin G, this pattern of glomelomerulopathy rather than diffuse proliferative glomerulorom the biopsy specimen illustrated in A. There is diffueposits in the peripheral capillary wall. (A) Periodic aci

onsider the possibility that more than one i

echanism may be active at one time. For ex-mple, membranous lupus nephritis associatedith subepithelial and intramembranous im-une deposits (ISN/RPS class V) often co-existsith proliferative GN (ISN/RPS class IV-G [A]),hich is associated with subendothelial im-une deposits. In a similar fashion, the vascu-

itic lesions, podocytopathy, and thromboticicroangiopathy may be superimposed on im-une-complex–mediated lupus GN and may

ontribute to the complexity of the pathologynd the clinical manifestations.

IGNIFICANCE OF THEISTINCTION BETWEEN DIFFUSEEGMENTAL (ISN/RPS CLASS IV-S) ANDIFFUSE GLOBAL (ISN/RPS CLASS IV-G) GN

he rationale for dividing ISN/RPS class IV intoiffuse segmental and diffuse global lupus ne-hritis was based largely on the study pre-ented by Najafi et al.21 This was a 10-yearollow-up evaluation of patients with severe lu-us GN who participated in a prospective clin-

of the glomerular capillaries with hypertrophy, vacuol-plets in the podocytes. Although there were mesangialnjury is called the cellular lesion of FSGS or collapsingitis. (B) Electron micrograph of 2 glomerular capillariest process effacement, and there are no electron-denseff, 66�. (B) Uranyl acetate and lead citrate, 7,000�.

llapsen drorular inephrse foo

cal trial conducted by the Collaborative Study

Gsaieyspc[it(

cstdw(npmdatda

owNoulgwosiptrIgmtItc

lirmipfmImlumca

C

TiWepwBdoeloogrTm

R

32 M.M. Schwartz

roup. There were 24 SLE patients with activeegmental GN in 50% or more of the glomerulind 35 SLE patients with diffuse GN. The clin-cal parameters at study entry were not differ-nt, and therapy was identical for the first 4ears.51 After 10 years, the incidence of end-tage renal disease was significantly greater inatients with segmental GN of 50% or moreompared with those with diffuse GN (9/2438%] versus 5/35 [14%], P � .05), and thencidence of remission with stable renal func-ion was greater in patients with diffuse GN22/35 [63%] versus 9/24 [38%], P � .05).

Three studies, focusing on the prognosis oflasses ISN/RPS IV-S and IV-G, have appearedince the publication of the ISN/RPS classifica-ion,52 and none was able to show a significantifference in any outcome between patientsith diffuse segmental (IV-S) and diffuse global

IV-G) nephritis. Two of these studies had smallumbers of patients, a relatively short follow-uperiod, and uncontrolled and variable treat-ent. However, it is possible that observed

ifferences in outcome between Najafi et al21

nd these 3 studies result from differences inhe way in which segmental and diffuse GN areefined and marked differences in treatmentmong these cohorts.

Segmental GN is characteristic of the pathol-gy in some biopsy specimens from patientsith lupus GN, and in the definition used byajafi et al,21,53 segmental GN could involvenly one glomerulus or virtually all the glomer-li in a biopsy. Although there was no upper

imit to the extent of involvement within thelomerulus, the segmental nature of the lesionas discerned readily, prospectively, by a panelf 4 experienced renal pathologists, in the pre-erved architecture and patent capillaries seenn a portion of the glomerulus.54 Although theatients with segmental GN in 50% or more ofhe glomeruli and diffuse GN of Najafi et al21

oughly correspond to ISN/RPS classes IV-S andV-G, respectively, the categories are not con-ruent, chiefly because segmental GN involvingore than 50% of the tuft and more than 50% of

he glomeruli are included in ISN/RPS classV-G, by definition, despite the clearly segmen-al nature of the pathology. As a result, ISN/RPS
lass IV-G contains all of Najafi et al’s21 diffuse
esions and some of their segmental lesions thatnvolve more than 50% of the glomeruli. Theelegation of the most widely distributed seg-ental lesions to ISN/RPS class IV-G could eas-

ly conceal differences in outcomes betweenatients with class IV-S and IV-G lesions. There-

ore, the arbitrary limits of glomerular involve-ent that define classes IV-S and IV-G in the

SN/RPS classification and that ignore the seg-ental nature of widely distributed segmental

esions that involve 50% or more of the glomer-li and more than 50% of the glomerular tuftay be incapable of detecting important clini-

al and biological differences between globalnd segmental forms of diffuse lupus GN.

ONCLUSIONS

he ISN/RPS classification of lupus nephritis elim-nated some ambiguities from the antecedent

HO classification. However, it adds the math-matic confusion of the meaning of an absoluteercentage of abnormal glomeruli in a situation inhich sample size is an important consideration.ecause it is mainly a histologic and immune-eposit–based classification of glomerular pathol-gy, the ISN/RPS classification does not includeither non–immune-complex–mediated glomeru-ar pathology or extraglomerular renal pathologyf lupus. The appropriateness and prognostic usef separating diffuse segmental (IV-S) from diffuselobal (IV-G) GN using the ISN/RPS definitionsemain to be proven.he author is grateful for the critical review of theanuscript by Edmund J. Lewis, MD.

EFERENCES1. Klemperer P, Pollack AD, Baehr G. Pathology of dis-

seminated lupus erythematosus. Arch Pathol. 1941;32:569-631.

2. Weening JJ, D’Agati VD, Schwartz MM, Seshan SV,Alpers CE, Appel GB, et al. The classification of glo-merulonephritis in systemic lupus erythematosus re-visited. Kidney Int. 2004;65:521-30.

3. Weening JJ, D’Agati VD, Schwartz MM, Seshan SV,Alpers CE, Appel GB, et al. The classification of glo-merulonephritis in systemic lupus erythematosus re-visited. J Am Soc Nephrol. 2004;15:241-50.

4. Muehrcke RC, Kark RM, Pirani CL, Pollak VE. Lupusnephritis: a clinical and pathologic study based onrenal biopsies. Medicine. 1957;36:1-146.

5. Pollak VE, Pirani CL, Schwartz FD. The natural history

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3


of the renal manifestations of systemic lupuserythematosus. J Lab Clin Med. 1964;63:537-50.

6. Baldwin DS, Lowenstein J, Rothfield NF, Gallo G,McCluskey RT. The clinical course of the proliferativeand membranous forms of lupus nephritis. Ann InternMed. 1970;73:929-42.

7. Baldwin DS, Gluck MC, Lowenstein J, Gallo GR. Lu-pus nephritis. Clinical course as related to morpho-logic forms and their transitions. Am J Med. 1977;62:12-30.

8. Churg J, Sobin LH. Lupus Nephritis. In: Renal disease,classification and atlas of glomerular diseases. NewYork: Igaku-Shoin; 1982. p. 127-49.

9. Churg J, Bernstein J, Glassock RJ. Lupus Nephritis. In:Renal disease. Classification and atlas of glomerulardiseases. New York: Igaku-Shoin; 1995. p. 151-80.

0. Corwin HL, Schwartz MM, Lewis EJ. The importanceof sample size in the interpretation of renal biopsy.Am J Nephrol. 1988;8:85-9.

1. Feinstein AR, Sosin DM, Wells CK. The Will Rogersphenomenon. Stage migration and new diagnostictechniques as a source of misleading statistics forsurvival in cancer. N Engl J Med. 1985;312:1604-8.

2. Koffler D, Agnello V, Thoburn R, Kunkel HG. Systemiclupus erythematosus: prototype of immune complexnephritis in man. J Exp Med. 1971;134:169s-79s.

3. Hill GS, Hinglais N, Tron F, Bach JF. Systemic lupuserythematosus. Morphologic correlations with immu-nologic and clinical data at the time of biopsy. Am JMed. 1978;64:61-79.

4. Cameron JS, Turner DR, Ogg CS, Williams DG, LessofMH, Chantler C, et al. Systemic lupus with nephritis:a long-term study. QJM. 1979;48:1-24.

5. Dujovne I, Pollak VE, Pirani CL, Dillard MG. The distri-bution and character of glomerular deposits in systemiclupus erythematosus. Kidney Int. 1972;2:33-50.

6. Appel GB, Silva FG, Pirani CL, Meltzer JI, Estes D.Renal involvement in systemic lupus erythematosus(SLE): a study of fifty-six patients emphasizing histo-logic classification. Medicine. 1978;57:371-410.

7. Sinniah R, Feng PH. Lupus nephritis: correlation be-tween light, electron microscopic and immunofluo-rescent findings and renal function. Clin Nephrol.1976;6:340-51.

8. Wilson CB. Animal models of immunologic renal dis-eases. In: Neilson EG, Couser WG, editors. Immuno-logic renal diseases. Philadelphia: Lippincott-Raven;1997. p. 729-73.

9. Hill GS, Delahousse M, Nochy D, Tomkiewicz E,Rémy P, Mignon F, et al. A new morphologic indexfor evaluation of renal biopsies in lupus nephritis.Kidney Int. 2000;58:1160-73.

0. Ichiryu MS, Magil AB. Intraglomerular monocyte infil-tration and immune deposits in diffuse lupus glomer-ulonephritis. Am J Kidney Dis. 1999;33:866-71.

1. Najafi CC, Korbet SM, Lewis EJ, Schwartz MM, ReichlinM, Evans J. Significance of histologic patterns of glomer-ular injury upon long-term prognosis in severe lupus
glomerulonephritis. Kidney Int. 2001;59:2156-63.
2. Mittal B, Hurwitz S, Rennke H, Singh AK. New sub-categories of class IV lupus nephritis: are there clini-cal, histologic, and outcome differences? Am J KidneyDis. 2004;44:1050-9.

3. Hill GS, Delahousse M, Nochy D, Bariety J. Class IV-Sversus class IV-G lupus nephritis: clinical and mor-phologic differences suggesting different pathogene-sis. Kidney Int. 2005;68:2288-97.

4. Couser WG. Mediation of immune glomerular injury.J Am Soc Nephrol. 1990;1:13-29.

5. Kerjaschki D. Pathogenetic concepts of membranousglomerulopathy (MGN). J Nephrol. 2000;13 Suppl3:S96-100.

6. Jennette JC. Antineutrophil cytoplasmic autoanti-body-associated diseases: a pathologist’s perspective.Am J Kidney Dis. 1991;18:164-70.

7. Schwartz MM, Lan SP, Bonsib SM, Gephardt GN,Sharma HM, and the Lupus Nephritis CollaborativeStudy Group. Clinical outcome of three discrete his-tologic patterns of injury in severe lupus glomerulo-nephritis. Am J Kidney Dis. 1989;13:273-83.

8. Schwartz MM, Roberts JL, Lewis EJ. Necrotizing glo-merulitis of systemic lupus erythematosus. HumPathol. 1983;14:158-67.

9. Chin HJ, Ahn C, Lim CS, Chung HK, Lee JG, Song YW,et al. Clinical implications of antineutrophil cytoplas-mic antibody test in lupus nephritis. Am J Nephrol.2000;20:57-63.

0. Zhao MH, Liu N, Zhang YK, Wang HY. Antineutrophilcytoplasmic autoantibodies (ANCA) and their targetantigens in Chinese patients with lupus nephritis.Nephrol Dial Transplant. 1998;13:2821-4.

1. Pradhan VD, Badakere SS, Bichile LS, Almeida AF.Anti-neutrophil cytoplasmic antibodies (ANCA) insystemic lupus erythematosus: prevalence, clinicalassociations and correlation with other autoantibod-ies. J Assoc Physicians India. 2004;52:533-7.

2. Kant KS, Pollak VE, Weiss MA, Glueck HI, Miller MA,Hess EV. Glomerular thrombosis in systemic lupuserythematosus: prevalence and significance. Medi-cine. 1981;60:71-86.

3. Hamasaki K, Mimura T, Kanda H, Kubo K, SetoguchiK, Satoh T, et al. Systemic lupus erythematosus andthrombotic thrombocytopenic purpura: a case reportand literature review. Clin Rheumatol. 2003;22:355-8.

4. Nzerue CM, Hewan-Lowe K, Pierangeli S, Harris EN.“Black swan in the kidney”: renal involvement in theantiphospholipid antibody syndrome. Kidney Int.2002;62:733-44.

5. Gungor T, Furlan M, Lammle B, Kuhn F, Seger RA.Acquired deficiency of von Willebrand factor-cleavingprotease in a patient suffering from acute systemiclupus erythematosus. Rheumatology (Oxford). 2001;40:940-2.

6. Mannucci PM, Vanoli M, Forza I, Canciani MT, ScorzaR. Von Willebrand factor cleaving protease (ADAMTS-13) in 123 patients with connective tissue diseases(systemic lupus erythematosus and systemic sclero-
sis). Haematologica. 2003;88:914-8.

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

34 M.M. Schwartz

7. Rieger M, Mannucci PM, Kremer Hovinga JA, HerzogA, Gerstenbauer G, Konetschny C, et al. ADAMTS13autoantibodies in patients with thrombotic microan-giopathies and other immunomediated diseases.Blood. 2005;106:1262-7.

8. Dube GK, Markowitz GS, Radhakrishnan J, Appel GB,D’Agati VD. Minimal change disease in systemic lupuserythematosus. Clin Nephrol. 2002;57:120-6.

9. Hertig A, Droz D, Lesavre P, Grunfeld JP, Rieu P. SLEand idiopathic nephrotic syndrome: coincidence ornot? Am J Kidney Dis. 2002;40:1179-84.

0. Kraft SW, Schwartz MM, Korbet SM, Lewis EJ. Glo-merular podocytopathy in patients with systemic lu-pus erythematosus. J Am Soc Nephrol. 2005;16:175-9.

1. Laurinavicius A, Hurwitz S, Rennke HG. Collapsingglomerulopathy in HIV and non-HIV patients: a clini-copathological and follow-up study. Kidney Int. 1999;56:2203-13.

2. Shalhoub RJ. Pathogenesis of lipoid nephrosis: a dis-order of T-cell function. Lancet. 1974;2:556-60.

3. Koyama A, Fujisaki M, Kobayashi M, Igarashi M, NaritaM. A glomerular permeability factor produced by hu-man T cell hybridomas. Kidney Int. 1991;40:453-60.

4. Cunard R, Kelly CJ. T cells and minimal change dis-ease. J Am Soc Nephrol. 2002;13:1409-11.

5. Yap HK, Cheung W, Murugasu B, Sim SK, Seah CC,Jordan SC. Th1 and Th2 cytokine mRNA profiles inchildhood nephrotic syndrome: evidence for in-creased IL-13 mRNA expression in relapse. J Am SocNephrol. 1999;10:529-37.

6. van den Berg JG, Aten J, Annink C, Ravesloot JH,Weber E, Weening JJ. Interleukin-4 and -13 promotebasolateral secretion of H(�) and cathepsin L byglomerular epithelial cells. Am J Physiol Renal
Physiol. 2002;282:F26-F33.
7. Korbet SM, Schwartz MM, Lewis EJ. Recurrent ne-phrotic syndrome in renal allografts. Am J Kidney Dis.1988;11:270-6.

8. Trachtman H, Greenbaum LA, McCarthy ET, SharmaM, Gauthier BG, Frank R, et al. Glomerular permeabil-ity activity: prevalence and prognostic value in pedi-atric patients with idiopathic nephrotic syndrome.Am J Kidney Dis. 2004;44:604-10.

9. McCarthy ET, Sharma M, Sharma R, Falk RJ, JennetteJC. Sera from patients with collapsing focal segmentalglomerulosclerosis increase albumin permeability ofisolated glomeruli. J Lab Clin Med. 2004;143:225-9.

0. Wüthrich RP, Jevnikar AM, Kelley VR. Understandinglupus nephritis: unraveling the pathogenesis throughmolecular approaches in murine models. In: Lewis EJ,Schwartz MM, Korbet SM, editors. Lupus nephritis.New York: Oxford University Press; 1999. p. 103-25.

1. Lewis EJ, Hunsicker LG, Rohde RD, Lachin JM, for theLupus Nephritis Collaborative Study Group. A con-trolled trial of plasmapheresis therapy in severe lupusnephritis. The Lupus Nephritis Collaborative StudyGroup. N Engl J Med. 1992;326:1373-9.

2. Yokoyama H, Wada T, Hara A, Yamahana J, Nakaya I,Kobayashi M, et al. The outcome and a new ISN/RPS2003 classification of lupus nephritis in Japanese.Kidney Int. 2004;66:2382-8.

3. Schwartz MM, Kawala KS, Corwin HL, Lewis EJ. Theprognosis of segmental glomerulonephritis in systemiclupus erythematosus. Kidney Int. 1987;32:274-9.

4. Schwartz MM, Bernstein J, Hill GS, Holley K, PhillipsEA, The Lupus Nephritis Collaborative Study Group.Predictive value of renal pathology in diffuse prolif-erative lupus glomerulonephritis. Kidney Int. 1989;
36:891-6.

the pathology of lupus nephritis -...

Documents