Pathologic Classiﬁcation of Focal Segmental...

Pathologic Classification of Focal Segmental Glomerulosclerosis

By Vivette D’Agati

Focal segmental glomerulosclerosis (FSGS) is defined as a clinical-pathologic syndrome manifesting proteinuriaand focal and segmental glomerular sclerosis with foot process effacement. The pathologic approach to theclassification of FSGS is complicated by the existence of primary (idiopathic) forms and multiple subcategorieswith etiologic associations, including human immunodeficiency virus (HIV)-associated nephropathy, heroin ne-phropathy, familial forms, drug toxicities, and a large group of secondary FSGS mediated by structural-functionaladaptations to glomerular hyperfiltration. A number of morphologic variants of primary and secondary focalsclerosis are now recognized, including FSGS not otherwise specified (NOS), perihilar, cellular, tip, and collapsingvariants. The defining features of these morphologic variants and of the major subcategories of FSGS arediscussed with emphasis on distinguishing light microscopic patterns and clinical-pathologic correlations.© 2003 Elsevier Inc. All rights reserved.

THE FIRST IMAGE of focal segmental glo-merulosclerosis (FSGS) was published in

1925 in a remarkably accurate drawing by Fahr1

from an atlas of human pathology. Even at thatearly time, Fahr1 recognized the relatedness of thisnovel lesion to minimal change disease when heaptly entitled it lipoid nephrosis with degeneration.It would be another 32 years before Rich2 provideda more detailed pathologic description of focalsclerosis in autopsy specimens of children dyingwith nephrotic syndrome caused by apparent lipoidnephrosis. Rich2 was the first to observe the pref-erential distribution of the segmental sclerosinglesions in juxtamedullary glomeruli early in thedisease, indicating the focality of the sclerosingprocess. He postulated that the development ofsclerosis probably accounted for the progression torenal failure seen in a subset of children withidiopathic nephrotic syndrome. However, it wasnot until the 1970s, in a report by the InternationalStudy of Kidney Diseases in Children, that focalsegmental glomerulosclerosis emerged as a sepa-rate clinical-pathologic entity that was distin-guished from minimal change disease by its greatersteroid resistance and progression to renal failure.3

Over the past 30 years, concepts of FSGS havebeen refined in more detailed clinical-pathologicstudies from many centers. FSGS is defined as aclinical-pathologic syndrome manifesting protein-uria, usually of nephrotic range, associated withlesions of focal and segmental glomerular sclerosisand foot process effacement.4,5 Although hyalineinsudation is common, the condition lacks glomer-ular immune complex deposits. Early in the diseaseprocess, the pattern of glomerular sclerosis is focal,involving a subset of glomeruli, and segmental,involving a portion of the glomerular tuft. As thedisease progresses, a more diffuse and global pat-

tern of sclerosis evolves. Alterations of the podo-cyte cytoarchitecture constitute the major ultra-structural findings.

The approach to a diagnosis of FSGS is prob-lematic because the morphologic features are non-specific and can occur in a variety of other condi-tions or superimposed on other glomerularprocesses.6 In addition, because the defining glo-merular lesion is focal, it may not be adequatelysampled in small needle biopsies.

The diagnosis of FSGS is further complicated bythe existence of a primary (or idiopathic) form andmany secondary forms (Table 1).6,7 Before a diag-nosis of primary FSGS can be reached, secondaryforms must be carefully excluded. Idiopathic FSGSmust be distinguished from human immunodefi-ciency virus (HIV)-associated nephropathy andheroin nephropathy, as well as the large group ofsecondary FSGS caused by structural-functionaladaptations mediated by intrarenal vasodilatationand by increased glomerular capillary pressuresand plasma flow rates.7 Such maladaptive glomer-ular hemodynamic alterations can arise through:(1) a reduction in the number of functioningnephrons (such as after unilateral renal agenesis,surgical ablation, oligomeganephronia, or any ad-vanced primary renal disease), or (2) mechanismsthat place hemodynamic stress on an initially nor-mal nephron population (as in morbid obesity,cyanotic congenital heart disease, and sickle cell

From the Department of Pathology, Columbia University,College of Physicians and Surgeons, New York, NY.

Address reprint requests to Vivette D’Agati, MD, Departmentof Pathology, Columbia University, College of Physicians andSurgeons, 630 West 168th St, New York, NY 10032.E-mail: [email protected]

© 2003 Elsevier Inc. All rights reserved.0270-9295/03/2302-0002$30.00/0doi:10.1053/snep.2003.50012

117Seminars in Nephrology, Vol 23, No 2 (March), 2003: pp 117-134

anemia). Finally, primary and secondary FSGSalso must be differentiated from the nonspecificpattern of focal and segmental glomerular scarringthat can follow a variety of inflammatory, prolif-erative, thrombotic, and hereditary conditions.

FSGS comprises a number of morphologic sub-types that may have different prognostic and ther-apeutic implications. These morphologic variantswere defined at a recent consensus conference ofrenal pathologists in New York City (Table 2). Thecategorization outlined in Table 2 encompasses thespectrum of primary FSGS, as well as some sec-ondary forms. This schema presumes prior exclu-sion of secondary FSGS caused by glomerular

scarring in the course of other primary glomerulardiseases (such as chronic glomerulonephritis, dia-betic glomerulosclerosis, membranous glomeru-lopathy, hereditary nephritis, and so forth). Fivemain light microscopic patterns of FSGS havebeen defined, including FSGS (not otherwise spec-ified [NOS]), perihilar variant, cellular variant, tipvariant, and collapsing variant. Although the ap-pearance of the glomerular tuft differs in theseforms, all share the common feature of podocytealterations at the ultrastructural level. At thepresent time, it is unclear if these morphologicvariants reflect pathogenetic differences or whetherthey are the consequence of different severities ofpodocyte injury or tempos of histopathologic evo-lution. Future studies are needed to address thesequestions. The pathologic features and major clin-ical correlates of each of these morphologic vari-ants are discussed.

FSGS (NOS)

FSGS (NOS) constitutes the generic lesion ofFSGS. The synonyms classic FSGS or FSGS of theusual type often are applied. This category requiresthat other morphologic categories (perihilar, cellu-lar, tip, and collapsing) be excluded. FSGS (NOS)is the most common morphologic pattern of FSGS.Evidence from repeat biopsy examinations sug-gests that other variants may evolve into this pat-tern in the course of disease progression and in-creasing chronicity.

Pathologic Features

FSGS (NOS) is defined as a discrete segmentalconsolidation of the glomerular tuft by increasedextracellular matrix, causing obliteration of theglomerular capillary lumen/lumina (Figs 1A–1B).Early in the disease, the segmental lesions have apredilection for the juxtamedullary glomeruli. Le-sions of sclerosis can affect the perihilar (ie, vas-cular pole) region or the periphery of the tuft. Insome glomeruli, segmental lesions may affectmore than one lobule, involving both the perihilarand peripheral regions. According to one studyusing serial sections, peripheral lesions tend to bemore common in childhood FSGS than the adultdisease.8 Any number of glomeruli can be affectedby segmental sclerosis, with or without associatedglobal sclerosis.

Glomerular capillaries are occluded segmentallyby relatively acellular matrix material, often asso-

Table 1. Etiologic Classification of FSGS

Primary (idiopathic) FSGSC1q nephropathyHIV-associated nephropathyHeroin nephropathyFamilial FSGS

Mutations in �-actinin 4 (autosomal dominant)Mutations in podocin (autosomal recessive)Mitochondrial cytopathies

Drug toxicityPamidronateLithiumInterferon-�

Secondary FSGS (adaptive structural-functionalresponse likely mediated by glomerularhypertrophy/hyperfiltration)

Reduced renal massOligomeganephroniaUnilateral renal agenesisRenal dysplasiaReflux nephropathySequela to cortical necrosisSurgical renal ablationAny advanced renal disease with reduction in

functioning nephronsChronic allograft nephropathy

Initially normal renal massDiabetes mellitusHypertensionObesityCyanotic congenital heart diseaseSickle cell anemia

Nonspecific pattern of FSGS caused by renal scarringFocal proliferative glomerulonephritis (IgA

nephropathy, lupus nephritis, pauci-immune focalnecrotizing and crescentic glomerulonephritis)

Hereditary nephritisDiabetic nephropathyHypertensive arterionephrosclerosisMembranous glomerulopathyThrombotic microangiopathies

VIVETTE D’AGATI118

ciated with inframembranous hyalinosis, endocap-illary foam cells, and wrinkling of glomerularbasement membrane (GBM). Hyalinosis, alsoknown as plasmatic insudation, consists of theaccumulation beneath the GBM of amorphousglassy material that is eosinophilic, nonargyro-philic, and trichrome-red (Fig 1B). Clear lipidvacuoles may be entrapped in the hyaline material.There often is continuity between the glomerularhyalinosis and hyalinosis involving the contiguousafferent arteriole. There may be segmental GBMcollapse without podocyte hypertrophy or hyper-plasia. Mesangial hypercellularity and glomerulo-megaly may be present. Mesangial hypercellularityis more prevalent in pediatric FSGS, especiallyearly in the disease course.9-11

Adhesions or synechiae to Bowman’s capsuleare common. The overlying visceral epithelial cellsoften appear swollen and can form a cellular capover the sclerosing segment (Fig 1B). Detachmentof podocytes from the sclerosing segment withintervening accumulation of newly formed matrixmaterial may produce an apparent halo of weaklyperiodic-acid Schiff (PAS)-positive or weaklytrichrome blue-positive matrix between the scle-rosed segment and the detached podocytes (Fig1B). There is no restriction on the degree of podo-cyte hypertrophy and hyperplasia seen in FSGS(NOS) provided that these podocyte alterations donot affect collapsed capillaries.

The number of glomeruli affected by segmentallesions depends on the severity of the disease pro-cess and the number of serial sections examined.As the lesions evolve globally, there is progressionto complete glomerular obsolescence. Lobules un-affected by the segmental sclerosis usually appearnormal by light microscopy but for mild swellingof the podocytes.

There is typically patchy tubular atrophy andinterstitial fibrosis commensurate with the severityand distribution of the glomerular sclerosis. Prox-imal tubules frequently contain intracellular lipidand protein resorption droplets (Fig 1C). Interstitialfoam cells may be identified, either as isolated cellsor aggregates.

In some cases, there is disproportionately severetubulointerstitial damage relative to the degree ofglomerular sclerosis. In such cases, the tubulesmay display degenerative and regenerativechanges including epithelial simplification, en-larged hyperchromatic nuclei, and nucleoli. Suchdamage tends to be more common in cases withsevere unremitting nephrotic syndrome in whichlong-standing heavy proteinuria itself is likely tomediate the progressive tubulointerstitial damage.Evidence from human and animal studies indicatesthat increased protein trafficking through tubularreabsorption of filtered protein promotes progres-sive tubulointerstitial damage.12,13 Protein over-load of tubular cells leads to increased tubular

Table 2. Morphologic Variants of FSGS

Variant Positive Criteria Negative Criteria

FSGS (NOS) At least one glomerulus with segmental increase in matrixobliterating the capillary lumina

There may be segmental glomerular basement membranecollapse without podocyte hyperplasia

Exclude perihilar, cellular, tip, andcollapsing variants

Perihilar variant Perihilar sclerosis and hyalinosis involving �50% ofsegmentally sclerotic glomeruli

Exclude cellular, tip, and collapsingvariants

Cellular variant At least one glomerulus with segmental endocapillaryhypercellularity occluding lumina, with or without foamcells and karyorrhexis

Exclude tip and collapsing variants

Tip variant At least one segmental lesion involving the tip domain(outer 25% of tuft next to origin of proximal tubule)

The tubular pole must be identified in the defining lesionThe lesion must have either an adhesion or confluence of

podocytes with parietal or tubular cells at the tubularlumen or neck

The tip lesion may be sclerosing or cellular

Exclude collapsing variantExclude any perihilar sclerosis

Collapsing variant At least one glomerulus with segmental or global collapseand podocyte hypertrophy/hyperplasia

None

PATHOLOGIC CLASSIFICATION OF FSGS 119

Fig 1. Light and fluorescence microscopic findings in FSGS (NOS), FSGS, perihilar variant and FSGS, cellularvariant. (A) FSGS (NOS): Low power view showing segmental sclerosis involving many glomeruli. The lesions ofsegmental sclerosis are discrete and readily distinguished from the adjacent preserved portions of the tuft. In thisexample, there is no evidence of tubular atrophy. (Jones methenamine silver; JMS). (B) FSGS (NOS): Trichromestain helps to distinguish the wrinkled and retracted glomerular basement membranes, shown in blue, from thered-staining hyaline insudation. There is detachment and capping of the overlying podocytes, with a halo-like zoneof pale neomembrane overlying the site of detachment. (C) FSGS (NOS): The tubules contain abundant intracyto-plasmic protein resorption droplets that stain red and blue with the trichrome stain. (D) FSGS (NOS): By immuno-fluorescence, there is segmental staining for IgM involving the sclerotic portion of the tuft, with weaker mesangialpositivity involving the adjacent nonsclerotic segments. (E) FSGS, perihilar variant: The lesion of segmentalsclerosis and hyalinosis is located at the glomerular vascular pole, or hilus, which is identified by the presence ofthe juxtaglomerular apparatus with macula densa. The hyalinosis appears glassy and eosinophilic. There is nopodocyte reactivity. (Hematoxylin and eosin; H&E). (F) FSGS, cellular variant: The glomerular capillary lumina aresegmentally obliterated by endocapillary hypercellularity including numerous infiltrating leukocytes, resembling aproliferative glomerulonephritis. There is hypertrophy of the overlying podocytes. (H&E).


expression of chemokines such as monocyte che-moattractant protein-1 and osteopontin that in turnattract inflammatory cells to the interstitium andlead to enhanced expression of fibrogenic cyto-kines that promote interstitial fibrosis. Tubular syn-thesis of endothelin-1 also may contribute to inter-stitial fibrosis by local vasoconstrictive effects andresultant ischemic injury.13

By immunofluorescence, there is typically focaland segmental granular deposition of immunoglob-ulin (Ig)M, C3, and, more variably, C1 in thedistribution of the segmental glomerular sclerosisand hyalinosis (Fig 1D). More generalized weak(�2�) mesangial deposition of IgM also may bepresent (Fig 1D). Staining for albumin and someimmunoglobulins (particularly IgA as well as IgG)may be found within the podocytes, correspondingto intracytoplasmic protein resorption droplets.One should be careful to distinguish this intracel-lular staining from immune deposits within theglomerular tuft itself. Similarly, intracytoplasmicdeposits of albumin, immunoglobulins, and some-times C3 may be found involving proximal tubulesthat are engaged in active protein resorption.

By electron microscopy, the lesions of segmen-tal sclerosis display wrinkling and retraction ofGBM and accumulation of inframembranous hya-line, with resulting narrowing or occlusion of theglomerular capillary lumina (Fig 4A). The electrondense hyaline material is usually more waxy inappearance than true immune complex depositsand tends to pool beneath the GBM, conforming tothe contours of the delimiting membrane. Hyalinedeposits frequently contain curvilinear membra-nous particles or entrapped electron lucent lipidglobules. Endocapillary foam cells appear as largeintracapillary cells containing abundant electronlucent vacuoles (Fig 4A).

Directly overlying the lesions of segmental scle-rosis there usually is complete effacement of footprocesses, accompanied by podocyte alterationsthat include hypertrophy, increased organellar con-tent, and focal microvillous transformation. Thismicrovillous appearance is caused by the forma-tion of slender cellular projections resembling villialong the surface of the podocytes facing the uri-nary space. The hypertrophied podocytes displayrounded cell bodies that adhere smoothly to theGBM, with frequent loss of primary processes.

There may be detachment of podocytes from thesclerosing segment (Fig 4B), as well as from non-

sclerotic segments (Fig 4C). In these areas, inter-vening accumulation of lamellated neomembranematerial commonly is observed between the nakedGBM and the retracted podocyte cell body (Fig4C). These foci of detachment correspond to thehalos observed by light microscopy surroundingthe sclerosing segments.

The major ultrastructural finding involving non-sclerotic glomerular capillaries is foot process ef-facement. The degree of foot process fusion ob-served overlying these open capillaries varies frommild to severe, but usually involves greater than50% of the total glomerular capillary surface area.In general, the degree of fusion correlates roughlywith the severity of the proteinuria, such that pa-tients with subnephrotic proteinuria tend to haveless foot process fusion than those who are fullynephrotic. In the areas of foot process effacement,there usually is loss of recognizable slit dia-phragms and mat-like condensations of cytoskel-etal filaments oriented parallel to the direction ofthe GBM itself. Thus, although the lesions ofFSGS are focal at the light microscopic level, thepodocyte alterations are relatively diffuse at theelectron microscopic level.

Clinical-Pathologic Correlations

The most prognostically significant clinical fea-tures in FSGS are serum creatinine and the severityof proteinuria at presentation.14 The presence ofnephrotic proteinuria (�3.0–3.5 g/d) has been as-sociated with a worse outcome in primary FSGS,with mean time course to end-stage renal diseaseof 6 to 8 years.14 This is compared with over 80%10-year renal survival in patients with nonne-phrotic proteinuria. Not surprisingly, severe pro-teinuria of over 10 g/d is associated with an evenmore rapid course to renal failure (�3 y).15,16

Prognosis is better in those who undergo a remis-sion of nephrotic syndrome than those with persis-tent nephrotic syndrome. Most studies agree thatthe best pathologic predictor of poor outcome isthe degree of interstitial fibrosis.17-20 Interestingly,the percentage of glomeruli with segmental scarsor global sclerosis has not been found to be inde-pendently predictive of outcome.

FSGS (NOS) accompanied by mesangial hyper-cellularity is most common in the pediatric agegroup. Studies by Yoshikawa et al9 and by theSouthwest Pediatric Nephrology Study Group10 in57 and 75 patients, respectively, did not show


prognostic import of diffuse mesangial hypercellu-larity occurring in idiopathic FSGS. The only cor-relation reported by the Southwest Pediatric Ne-phrology Study Group study was a shorter timecourse from clinical presentation to biopsy exam-ination in the patients with FSGS and diffuse mes-angial hypercellularity compared with FSGS with-out mesangial hypercellularity.10 Thus, similar tocellular FSGS, the presence of diffuse mesangialhypercellularity in FSGS (NOS) appears to repre-sent an early stage in the development of FSGS butdoes not correlate with outcome. Therefore, theconsensus conference concluded that the presenceof diffuse mesangial hypercellularity in FSGS didnot warrant a separate morphologic classification.

FSGS, PERIHILAR VARIANT

This category requires that the cellular variant,tip variant, and collapsing variant be excluded. It isdefined by the presence of perihilar sclerosis andhyalinosis involving greater than 50% of segmen-tally sclerotic glomeruli (Fig 1E). Glomerulo-megaly and adhesions are common. There often isarteriolar hyalinosis, sometimes in continuity withhyalinosis in the perihilar segment. Foam cells maybe entrapped in the sclerotic lesions. Podocytehypertrophy and hyperplasia may be present buttypically are less frequent than in the other vari-ants. Other glomeruli may show lesions of segmen-tal and/or global glomerulosclerosis, as describedfor FSGS (NOS) earlier. Immunofluorescence andultrastructural findings are similar to those de-scribed in the section on FSGS (NOS).


This variant of FSGS may occur in primaryFSGS. However, when accompanied by glomeru-lomegaly, it is particularly common in patientswith secondary forms of FSGS mediated by anadaptive response to increased glomerular capil-lary pressures and flow rates (as in association withobesity, cyanotic congenital heart disease, refluxnephropathy, renal agenesis, dysplasia, oligomega-nephronia, or any advanced renal disease with areduced number of functioning nephrons; see dis-cussion of secondary forms of FSGS later).

FSGS, CELLULAR VARIANT

The cellular variant of FSGS was first describedby Schwartz and Lewis21 in 1985. A diagnosis of

cellular FSGS requires that tip and collapsing vari-ant be excluded.

Pathologic Features

The cellular variant is defined by the presence ofat least one glomerulus with segmental endocapil-lary hypercellularity involving at least 25% of thetuft and causing occlusion of the capillary lumen/lumina (Fig 1F). Any segment (perihilar or periph-eral) may be affected. When numerous glomeruliare affected, the process may mimic focal prolif-erative glomerulonephritis.6 The lesions of endo-capillary hypercellularity typically are expansile,causing engorgement of the glomerular capillaries.The endocapillary cells may include endothelialcells, foam cells, and infiltrating leukocytes, in-cluding monocyte/macrophages (Fig 1F). Otherleukocytes, including lymphocytes and neutro-phils, also may be present. Some of these lesionsare accompanied by foamy hyaline material, fibrin,and karyorrhexis, resembling segmental necrotiz-ing lesions, but without rupture of GBM.

Although the cellular variant is defined by thepresence of segmental endocapillary hypercellular-ity, it also frequently displays extracapillary hyper-cellularity owing to hyperplasia of the podocytes.Podocytes may appear swollen and crowded, some-times forming pseudocrescents. These pseudo-crescents usually can be distinguished from truecrescents by their lack of attachment to Bowman’scapsule or continuity with the parietal epithelialcells. Moreover, the extracapillary cells tend to beplump, rounded, and poorly cohesive, with fre-quent intracellular protein resorption droplets.These extracapillary cells lack the spindled mor-phology or pericellular matrix typically observedin true crescents. Another distinguishing feature isthat Bowman’s capsule itself is intact, without theruptures typical of cellular crescents of the inflam-matory type.

In some cases, cellular lesions are identified inall affected glomeruli, whereas in others they affecta minority of glomeruli, possibly reflecting differ-ent stages in the evolution of sclerosis. Other glo-meruli may contain lesions of segmental or globalglomerulosclerosis of the usual type, as describedfor FSGS (NOS).

By immunofluorescence there is focal and seg-mental glomerular positivity for IgM and C3. Atthe ultrastructural level, the cellular variant usuallydisplays severe foot process effacement, correlat-


ing with the generally high levels of proteinuria(Fig 4D). Cellular lesions consist of segmentalocclusion of glomerular capillaries by endocapil-lary hypercellularity including foam cells andmonocytes (Figs 4E and 4F). The GBM is intact,without evidence of rupture.


Compared with FSGS (NOS), the cellular vari-ant is characterized by more severe proteinuria anda shorter time course from clinical onset of renaldisease to biopsy examination, suggesting an earlyphase in the evolution of the segmental sclerosis.21

Schwartz and Lewis21 found a shorter interval be-tween onset of proteinuria (3.4 versus 71.9 mo) inpatients with cellular versus classic FSGS. More-over, 90% of patients in the cellular group hadurine protein levels greater than 3 g/d, comparedwith 49% of those without cellular lesions. Simi-larly, the incidence of full nephrotic syndrome atpresentation was significantly higher (70% versus23%). Indeed, patients with cellular FSGS oftenhave a very abrupt onset of severe nephrotic syn-drome resembling the presentation of minimalchange disease (personal observations). Repeat bi-opsy examinations in some cases have shown evo-lution to more typical segmental scars, supportingthe concept that the hypercellularity is an earlystage in the development of the segmental sclero-sis. A similar evolution from cellular to more scle-rosing lesions has been documented by repeat biopsyexamination in recurrent FSGS in the transplant.22

The cellular variant may be responsive to im-munosuppressive therapy.23 This favorable treat-ment response probably relates to the early andrelatively active stage of glomerular injury in thecellular variant.

FSGS, TIP VARIANT

The tip variant of FSGS is defined by the pres-ence of at least one glomerulus with a segmentallesion involving the tip domain (ie, the peripheral25% of the glomerular tuft next to the origin of theproximal tubule). There must be either adhesionbetween the tuft and Bowman’s capsule at thetubular lumen or neck, or confluence of podocyteswith parietal epithelial or tubular epithelial cells atthe tubular pole or neck (Figs 2A–2B). The prox-imal tubular pole must be identified in the definingglomerulus. The designation of tip lesion requiresthat the collapsing variant be excluded.

Pathologic Features

As defined originally by Howie and Brewer,24

the early lesion is characterized by confluence ofswollen, hypertrophied visceral epithelial cellswith parietal or tubular epithelial cells at the tubu-lar pole. The affected lobule may display endocap-illary hypercellularity with endocapillary foamcells and hyalinosis. In some cases, the affectedsegment appears to herniate into the tubular lumen(Fig 2B). As the lesion evolves, there is adhesionof the glomerular tuft to Bowman’s capsule at thepoint of transition to the proximal tubular base-ment membrane (Fig 2A).24 Later lesions mayform segmental scars.

Other glomeruli may show segmental sclerosisor endocapillary hypercellularity in the peripherybut not involving the tip, or in a portion of the tuftthat cannot be identified as tip or perihilar. How-ever, the presence of segmental sclerosis or endo-capillary hypercellularity in a perihilar locationrules out the tip variant of FSGS. Global sclerosismay be present.

The segmental lesions usually stain for IgM andC3 by immunofluorescence. But for their locationat the tubular pole, the lesions often resemble thoseof cellular FSGS at the ultrastructural level.

Differential Diagnosis

Tip lesions are not specific, but may occur in thesetting of a variety of glomerular diseases includ-ing membranous glomerulopathy, IgA nephropa-thy, diabetic glomerulosclerosis, and others.25

Thus, the designation of glomerular tip lesionshould be applied only to those cases with glomer-uli that look otherwise like minimal change diseaseor focal segmental glomerulosclerosis, and inwhich other glomerular conditions have beenexcluded.


It is uncertain how the pathogenesis of tip lesionmay differ, if at all, from that of the more classiclesions of FSGS. A study of autopsy kidneys fromchildren dying with minimal change disease in thepresteroid era revealed focal tip lesions in a smallpercentage of glomeruli.26 This study concludedthat tip lesions may arise as a nonspecific responseof the peritubular segment of the glomerular tuft tofluxes of protein-rich filtrate in the setting of ne-phrotic syndrome.


Fig 2. Light microscopic findings in FSGS, tip variant and FSGS, collapsing variant. (A) FSGS, tip variant: Thereis a segmental lesion with endocapillary foam cells that forms an adhesion to Bowman’s capsule at the origin ofthe tubular pole. The podocytes are capped over this segment and merge with the tubular epithelial cells.(Periodic-acid Schiff, PAS). (B) FSGS, tip variant: A segmental lesion herniates into the tubular pole. The tip zoneis engorged with endocapillary foam cells and forms an adhesion to Bowman’s capsule at the tubular origin. Thereis confluence of the podocytes and the tubular epithelial cells surrounding the tip lesion. (Jones methenaminesilver, JMS). (C) FSGS, collapsing variant: The glomerular collapse is diffuse and global in distribution, withassociated severe tubular degenerative changes, (JMS). (D) FSGS, collapsing variant: In the collapsed tuft, theglomerular basement membranes are imploded, without appreciable increase in matrix material. The podocytesoverlying the collapsed tuft are marked hyperplastic with enlarged vesicular nuclei, focal binucleated forms, andcrowding of the urinary space. At the periphery, some of the podocytes appear to be falling off into the urinaryspace (JMS). (E) FSGS, collapsing variant: Immunostaining for proliferation marker Ki-67 (MIB1) shows manycycling cells in the urinary space, including visceral and parietal cells, as well as adjacent tubular epithelial cells,(Immunoperoxidase). (F) FSGS, collapsing variant: Trichrome stain delineates the collapsed blue-staining tuft andthe abundant hyperplastic podocytes forming a pseudocrescent containing many intracytoplasmic fuchsinophilicprotein resorption droplets. The podocyte hyperplasia lacks the spindled cellular morphology and pericellularmatrix seen in true crescents of parietal cell origin.


The relationship of tip lesion to minimal changedisease and FSGS has been hotly debated. Whereassome groups have reported a greater likelihood ofsteroid responsivity and excellent long-term prog-nosis resembling that of minimal change disease,others have described an evolution toward moretypical FSGS.27 Indeed, repeat biopsy examina-tions in some of these patients have shown pro-gression to focal segmental and global glomerulo-sclerosis with development of renal failure.

Recently, Hogan-Moulton et al28 described 80%steroid responsivity in glomerular tip lesion com-pared with 33% in FSGS, but with similar long-term renal survival (87% at 4 years in both groups).Clearly, evaluation of larger series of patients willbe required to determine the significance of thislesion.

FSGS, COLLAPSING VARIANT

The designation of collapsing variant (alsoknown as collapsing glomerulopathy) is applied tocases of FSGS in which at least one glomerulusdisplays segmental or global obliteration of theglomerular capillary lumina by wrinkling and col-lapse of GBMs associated with podocyte hypertro-phy and hyperplasia. Collapse involving a singleglomerulus is considered significant, such that thepresence of any glomerular collapse pre-empts theother morphologic categories of FSGS.

The term glomerular collapse was first intro-duced by Weiss et al29 in 1986 to describe anunusual clinicopathologic complex of severe ne-phrotic syndrome, rapidly progressive renal fail-ure, and glomerular collapse occurring in 6 blackpatients. Two patients required dialysis within10 weeks of clinical presentation and 5 had anill-defined febrile illness. Although the clinicaland pathologic findings suggested possible HIV-associated nephropathy, only one of these pa-tients subsequently developed acquired immunedeficiency syndrome. Two subsequent serieshave reported a similar malignant course to renalfailure in patients with collapsing FSGS wholack HIV infection.30,31 The incidence of thecollapsing variant is increasing. CollapsingFSGS comprised 11% of all primary FSGS atColumbia Presbyterian Medical Center from1979 to 1985, 20% from 1986 to 1989, and 24%from 1990 to 1993.31,32

Pathologic Features

Collapsing FSGS presents a dramatic pattern ofinjury. Glomerular capillary lumina are occludedby an implosive wrinkling and collapse of theGBMs that is more often global than segmental,without predilection for the perihilar segments(Figs 2C–2D). This GBM collapse is best delin-eated with the use of the PAS or the Jones methe-namine silver stains (Figs 2C and 2D). The acutenature of the glomerular injury is evidenced by thelack of appreciable increase in intracapillary ormesangial matrix.

The glomerular collapse must be accompaniedby striking hypertrophy and hyperplasia of theoverlying podocytes, which have enlarged, openvesicular nuclei with frequent nucleoli, occasionalbinucleated forms, and rare mitotic figures (Fig2D). Proliferation marker Ki-67 (MIB1) is fre-quently positive in the distribution of the podo-cytes, indicating that they are cell-cycle engaged(Fig 2E).33 Podocytes may be so crowded as to fillthe urinary space, forming pseudocrescents, andoften contain prominent intracytoplasmic proteinresorption droplets (Fig 2F).

In collapsing FSGS, the podocyte displays adysregulated phenotype with increased rates ofproliferation and apoptosis.33 Podocytes normallyare endowed with high levels of constitutive ex-pression of cyclin kinase inhibitors to safeguardagainst easy entry into the cell cycle that mightjeopardize the cell’s ability to maintain its highlydifferentiated cytoarchitecture. However, in col-lapsing FSGS, this system is perturbed, leading tostructural deterioration. The injured podocytes ex-hibit reduced expression of cyclin kinase inhibitorsp27 and p57, promoting permissive cellular prolif-eration.34,35 As the cells enter the cell cycle theylose their mature podocyte markers (such as syn-aptopodin, WT-1, GLEPP-1, C3b receptor, andCALLA).33 Transdifferentiation of podocytes tomacrophage/monocytes also has been de-scribed.36,37 The podocytes become less cohe-sive, round up, and may actually detach and shedinto the urinary space.37 Thus, in collapsingFSGS podocytes appear to lose their highly dif-ferentiated cytoarchitecture through activationof a genetic program that involves re-entry intothe cell cycle, disruption of the cytoskeleton, andcellular dedifferentiation.


Collapsing FSGS is distinguished from the cel-lular form by the absence of endocapillary hyper-cellularity. In fact, there is often an apparent re-duction in the number of glomerular endothelialcells in collapsed lobules. Unlike FSGS (NOS),glomeruli with collapsing sclerosis usually lackhyalinosis, endocapillary foam cells, and adhesionsto Bowman’s capsule. Mesangial hypercellularity,glomerulomegaly, and arteriolar hyalinosis are un-common.

Tubulointerstitial disease is an important com-ponent of this condition and often appears out ofproportion to the degree of glomerular sclerosis. Inaddition to tubular atrophy, interstitial fibrosis,edema, and inflammation, there are widespreadtubular degenerative and regenerative changes.31

These include tubular epithelial simplification withenlarged hyperchromatic nuclei, nucleoli, mitoticfigures, and focal apoptosis (Fig 2C). About 40%of cases may have tubular microcysts that containloose proteinaceous casts.31

By immunofluorescence there are segmental toglobal deposits of IgM, C3, and, less commonly,C1 in collapsing segments. Visceral epithelial pro-tein resorption droplets often stain for IgG, IgA,and albumin, with similar staining in the tubularepithelial protein droplets.

At the ultrastructural level, the collapsed lobulesdisplay wrinkling and little or no thickening ofGBMs (Fig 5A). The overlying podocytes aremarkedly hypertrophied with severe foot processeffacement, focal detachment, and increased num-bers of organelles including electron dense proteinresorption droplets, electron lucent transport vesi-cles, and rough endoplasmic reticulum. The actincytoskeleton usually appears disrupted, giving thecells a relatively open appearing cytoplasm (Fig5A). Noncollapsed capillaries also display severefoot process effacement. No electron dense depos-its are observed, with the exception of rare smallparamesangial electron densities corresponding tothe mesangial deposits of IgM. In contrast to HIV-associated nephropathy, no tubuloreticular inclu-sions are identified in idiopathic collapsing FSGS.

Collapsing glomerulopathy may be confusedwith forms of crescentic glomerulonephritis. Theproliferating podocytes in collapsing FSGS lackthe spindled morphology and pericellular matrixseen surrounding the proliferating parietal cells oftrue crescents. Crescentic glomerulonephritis alsois distinguished by the usual presence of necrotiz-

ing lesions in the underlying tuft and breaks in theGBM.


Once a diagnosis of collapsing FSGS has beenreached, the possibility of HIV-associated ne-phropathy must be ruled out. Exclusion of HIV-associated nephropathy is based on the demonstra-tion of negative HIV serologies, and is supportedby the absence of endothelial tubuloreticular inclu-sions.

When compared with patients with FSGS (NOS)or classic FSGS, patients with idiopathic collaps-ing FSGS are more likely to be black and topresent with more severe markers of nephroticsyndrome, including more severe proteinuria, hy-poalbuminemia, and hypercholesterolemia. More-over, these patients have a higher presenting serumcreatinine level (3.5 versus 1.3 mg/dL and 4.2versus 2.0 mg/dL) despite a shorter time coursefrom clinical onset to biopsy examination.30,31

Primary FSGS with collapsing features typicallyhas a rapid course to renal failure and is oftenunresponsive to steroid therapy.30-31,38 One groupfound a median renal survival of 13.0 monthscompared with 62.5 months for controls with clas-sic FSGS.31 Thus, this variant has been consideredthe morphologic counterpart of the malignantFSGS proposed years earlier by Brown et al16 andCameron et al.39

Recently, some cases of FSGS with collapsingfeatures have been identified as a form of drugnephrotoxicity in older patients treated with pam-idronate, an osteoclast inhibitor that reduces boneresorption, for myeloma, or carcinoma metastaticto bone.40 Collapsing FSGS has been reported tooccur in some patients with viral infections owingto parvovirus B1941 or SV40,42 suggesting a rolefor direct viral infection of the kidney.

A collapsing pattern of FSGS may also occur denovo or as a recurrent disease in the allograft.43-45

In this setting, the de novo disease often is unac-companied by nephrotic syndrome and appears tobe a less specific morphologic lesion that does notnecessarily represent primary FSGS. Nadasdy etal45 has reported a zonal distribution of collapsingFSGS in some allografts with transplant arteriopa-thy, suggesting a role for ischemic injury. Thesefindings are reminiscent of the collapsing or cellu-lar pattern of focal sclerosis observed in somenative kidney specimens with ischemia and pro-


teinuria on the basis of renovascular hypertensionor cholesterol embolization.46,47

OTHER FORMS OF FSGS

Among the entities listed in Table 1, only C1qnephropathy, HIV-associated nephropathy, andsecondary FSGS mediated by structural-functionaladaptations will be discussed in any depth owing tospace constraints.

C1q NEPHROPATHY

This controversial entity was first described byJennette and Hipp48 in 1985. It is defined as a formof idiopathic nephrotic syndrome caused by a glo-merulopathy with dominant paramesangial depos-its of C1q (of at least 2� intensity on a scale of0–4�).

Pathologic Features

The pattern by light microscopy is usually oneof FSGS with variable mesangial hypercellularity(Fig 3A). Although early reports emphasized themesangial proliferative features,48 subsequent re-ports have stressed the resemblance to FSGS.49

By immunofluorescence, in addition to stainingfor C1q (of at least 2� intensity), most cases havecodeposits of IgG (90%) of mean 1.6 intensity, orIgM (94%) of mean 1.1 intensity, and C3 (90%) ofmean 1.1 intensity on a scale of 0 to 4�.50 Depositsof IgA are less common and were observed in 56%of biopsy examinations with mean intensity 0.7.The mesangial deposits are often comma shapedbecause of their paramesangial location and con-formation to the overlying GBM reflection (Fig 3B).

Electron dense deposits are primarily or exclu-sively located in the mesangium. In over 90% ofcases, electron microscopy reveals prominentparamesangial electron dense deposits located sub-jacent to the GBM reflection (Fig 5B). In a minor-ity of cases, rare subendothelial and subepithelialdeposits also may be seen. There is prominent butvariable foot process effacement.

Clinical-Pathologic Features

Many investigators believe that C1q nephropa-thy represents a morphologic variant of primaryFSGS with similar pathogenesis and outcome.48-50

Most patients present with idiopathic nephroticsyndrome and renal insufficiency (with mean urineprotein levels of 5.9 g and mean serum creatininelevels of 1.5 mg/dL).50 Edema and hypertension

are observed in almost half of the cases.50 Serolo-gies for lupus and HIV are negative, serum com-plement levels are normal, and there is no clinicalevidence of systemic disease. Most cases are ste-roid resistant, although some have achieved fullremissions.50 Three-year renal survival was re-ported at 84% in a cohort of 79 patients.50

HIV-ASSOCIATED NEPHROPATHY

Pathologic Features

HIV-associated nephropathy (HIV-AN) is de-fined as a form of focal segmental glomeruloscle-rosis occurring in HIV-infected patients. The lightmicroscopic findings in HIV-associated nephropa-thy are qualitatively similar to those described ear-lier in collapsing glomerulopathy.51,52 The charac-teristic lesion is a collapsing sclerosis withprominent podocyte alterations. As the lesionsprogress, the glomerular tuft may be reduced to anacellular sclerotic ball with crowning of the over-lying podocytes and relative dilatation of the uri-nary space (Fig 3C). A proteinaceous filtrate fre-quently is identified within the enlarged urinaryspace (Fig 3C).

Tubular microcysts are common, affecting ap-proximately 30% to 40% of cases51 (Fig 3C). Somepatients with advanced disease at autopsy havealmost complete replacement of the cortical paren-chyma by massive microcyst formation. This mi-crocystic transformation likely contributes to theenlarged kidneys and increased echogenicity byultrasound observed even in patients with end-stage renal failure.

The immunofluorescence findings are similar tothose of primary FSGS with collapsing features.Differences between HIV-associated nephropathyand idiopathic collapsing glomerulopathy are seenonly at the ultrastructural level (Figs 5C-5E). Themajor distinguishing feature is the abundance oftubuloreticular inclusions in the glomerular endo-thelial cells of HIV-associated nephropathy (Fig5D).51 Tubuloreticular inclusions, also known asinterferon footprints consist of 24-nm interanasta-mosing tubular structures located within dilatedcisternae of endoplasmic reticulum. They may belarge and multiple per cell. Although they areidentified most readily in the glomerular endothe-lium, they also occur in arterial or interstitial cap-illary endothelial cells as well as infiltrating leuko-cytes. Tubuloreticular inclusions were noted with


Fig 3. Light and fluorescence microscopic findings in C1q nephropathy, HIV-AN, and secondary FSGS. (A) C1qnephropathy. Low-power view showing a characteristic field of focal segmental and global glomerulosclerosiswith patchy tubular atrophy and interstitial fibrosis (PAS). (B) C1q nephropathy. There are global comma-shapeddeposits of C1q outlining the paramesangial regions (immunofluorescence). (C) HIV-associated nephropathy.There is collapse of the glomerular tuft with dilatation of the urinary space. The tubules are markedly dilatedforming microcysts with voluminous proteinaceous casts (PAS). (D) Secondary FSGS caused by obesity. Theglomerulus on the left from an obese patient shows marked glomerulomegaly compared with that of an age- andsex-matched nonobese control on the right. (E) Secondary FSGS caused by obesity. There are 2 discrete lesions ofsegmental sclerosis, one arising from the vascular pole, and the other involving the periphery of the tuft. Note theabsence of podocyte reactivity (PAS). (F) Secondary FSGS caused by hypertensive nephrosclerosis. In the subcapsulararea there are numerous atubular glomeruli forming glomerular microcysts with shrunken and simplified tuft, dilatationof the urinary space, and proteinaceous filtrate. The adjacent tubules are severely atrophied (PAS).


Fig 4. Electron microscopic findings in FSGS (NOS) and FSGS, cellular variant. (A) FSGS (NOS). A lesion ofsegmental sclerosis with inframembranous hyalinosis, endocapillary foam cells, and podocyte detachment withcomplete effacement of foot processes. (B) FSGS (NOS). Overlying the segmental lesion there is podocytedetachment with large intracytoplasmic electron lucent transport vesicles. The GBM are wrinkled and the glomer-ular capillaries are obliterated by matrix material. (C) FSGS (NOS). A nonsclerotic segment showing detachment ofpodocytes with lamellation of neomembrane material. The podocytes are hypertrophied with microvillous trans-formation of their cytoplasm. (D) FSGS, cellular variant. Nonsclerotic capillaries display complete foot processeffacement with extensive microvillous transformation. (E) FSGS, cellular variant. The glomerular capillary lumenis obliterated by endocapillary hypercellularity including many foamy cells with clear lipid vacuoles. The overlyingpodocytes are detached with neomembrane formation. (F) FSGS, cellular variant. The glomerular capillary lumenis engorged with cells, including infiltrating monocytes and foam cells, admixed with inframembranous hyaline.There is marked hypertrophy and hyperplasia of the overlying podocytes.


Fig 5. Electron microscopic findings in FSGS, collapsing variant, C1q nephropathy, HIV-AN, and secondaryFSGS caused by obesity. (A) FSGS, collapsing variant. There is tight collapse of the GBMs causing luminalobliteration. Note the marked hypertrophy and hyperplasia of the overlying podocytes with loss of primaryprocesses, complete effacement of foot processes, and podocyte detachment. (B) C1q nephropathy. Electrondense deposits are present in the mesangium. The deposits tend to pool in the paramesangial regions, subjacentto the GBM reflection over the mesangium. There is mild focal foot process effacement. (C) HIV-AN. The GBMs arecollapsed with secondary luminal narrowing. The overlying podocytes form a cellular cap with marked hypertrophyand numerous intracytoplasmic protein resorption droplets. (D) HIV-AN. A glomerular endothelial cell contains alarge tubuloreticular inclusion located within a cistern of endoplasmic reticulum. (E) HIV-AN. A tubular epithelialcell displays granular degeneration of its nuclear chromatin. (F) Secondary FSGS caused by obesity. The footprocesses show minimal effacement involving less than 20% of the glomerular capillary surface area, despite thepresence of nephrotic-range proteinuria.


frequency in renal biopsy examinations of HIV-infected patients in the 1980s. However, they maybe far less frequent in the modern era of highlyactive antiretroviral therapy, probably owing toreduced viral burden (personal observations).

Other characteristic but relatively less frequentultrastructural findings include nuclear bodieswithin tubular and interstitial cells, granular-fibril-lar transformation of the tubular nuclei, and con-fronting cylindrical cisternae (Fig 5E).51

Differential Diagnosis

The major differential diagnosis is from primaryFSGS with collapsing features. HIV serologies arerequired for definitive distinction between theseentities.

Included within the spectrum of HIV-associatednephropathies are diffuse mesangial hypercellular-ity and minimal change disease, which are morecommon in HIV-infected children than adults.These milder variants usually present with ne-phrotic syndrome and normal renal function, with-out the rapid course to renal failure that character-izes the collapsing form of HIV-associatednephropathy.

HIV-associated nephropathy must be differenti-ated from a variety of other glomerular and tubu-lointerstitial diseases that occur in HIV-infectedpatients.52 One of the most common immune com-plex–mediated glomerular lesions is membrano-proliferative glomerulonephritis, particularly inHIV-infected intravenous drug abusers with hepa-titis C virus coinfection. Another common glomer-ular disease in the HIV population is IgA nephrop-athy, which may be associated with IgA-containingcryoglobulins. IgA nephropathy has been reportedin both HIV-infected blacks and whites. Glomeru-lar immune deposits eluted from the glomeruli ofsome of these patients have shown specificity forHIV envelope or core proteins. Other glomerularlesions occurring in HIV-infected patients includemembranous glomerulopathy, lupus-like glomeru-lonephritis, acute postinfectious glomerulonephri-tis, and thrombotic microangiopathies.


The incidence of heroin nephropathy has de-creased reciprocally with the advent of HIV-AN inthe early 1980s.32 In 1984, a distinctive form ofFSGS occurring in HIV-infected patients was firstdescribed at Downstate Medical Center in New

York City.53 HIV-AN is predominantly a diseaseof blacks (90%). Although it can occur in bothsexes and with any HIV risk factor, it is mostcommon in male intravenous drug abusers.

Presenting features include proteinuria, usuallyin the nephrotic range, and renal insufficiency.Despite the high frequency of nephrotic-range pro-teinuria and hypoalbuminemia, hypercholesterol-emia and edema are relatively uncommon. Theabsence of hypercholesterolemia likely reflects thereduced hepatic synthesis of lipoproteins in ac-quired immune deficiency syndrome. Hyperten-sion is relatively uncommon.54 By ultrasound, thekidneys usually are large and echogenic.

Early reports of HIV-AN described a rapidcourse to renal failure, with mean time to dialysisof less than 2 months.53 However, in the modernera of highly active antiretroviral therapy, reducedviral load and improved renal survival have beenachieved. After highly active antiretroviral therapythere may be dramatic improvement in the renalbiopsy findings, with reversal of tubular micro-cysts.55 A more detailed description of this entitycan be found in the article by Herman and Klotmanin this issue.

SECONDARY FSGS

Secondary FSGS denotes the pattern of focaland segmental glomerulosclerosis that develops inthe course of a number of renal diseases in whichthere is reduced number of functioning nephrons orhemodynamic stress on an initially normal nephronpopulation (see Table 1).6,7 Secondary FSGS ismost common in the setting of obesity, hyperten-sive nephrosclerosis, sickle cell anemia, and anyadvanced renal process with significant loss offunctioning nephrons.

Pathologic Features

Glomerular hypertrophy is a relatively constantfinding (Fig 3D). Although actual measurements ofglomerular diameter are not performed routinely inclinical practice, an experienced renal pathologistcan recognize hypertrophied glomeruli readily byroutine light microscopy. As a simple rule ofthumb, hypertrophied glomeruli usually fill a 40�high-dry microscopic field. Because the glomerulartuft is a sphere, glomerular hypertrophy is bestassessed in a plane of section that transects thehilus of the glomerulus, at the epicenter of theglomerular tuft. The largest glomerular volumes


recorded are seen in oligomeganephronia, a con-genital disorder with markedly reduced nephronnumber from birth.

In forms of secondary FSGS resulting from lossof renal mass, FSGS usually is seen in a back-ground of extensive global glomerulosclerosis withcorresponding tubular atrophy and interstitialfibrosis.

Most forms of secondary FSGS have discretesegmental scars, often involving the perihilar re-gions of hypertrophied glomeruli (Fig 3E). Podo-cyte hypertrophy and hyperplasia are less frequentthan in primary FSGS (Fig 3E). In secondaryFSGS, the degree of foot process fusion is gener-ally mild, affecting less than 50% of the totalglomerular capillary surface area (Fig 5F). Becauseof its variability, however, the percentage of footprocess fusion cannot be used as an absolute orspecific criterion by which to distinguish primaryfrom secondary FSGS.

In secondary FSGS caused by hypertensive ne-phrosclerosis, there is prominent arteriolosclerosis,arteriolar hyalinosis, and arteriosclerosis. Predilec-tion for glomerular sclerosis to occur in the outercortex with formation of subcapsular scars is ap-preciated readily in biopsy specimens in which therenal capsule has been sampled. There is invariablyprominent chronic tubulointerstitial disease andglomerular hypertrophy. Lesions of segmentalsclerosis and hyalinosis are often perihilar anddevelop in glomeruli that also are hypertrophied.Most lesions of segmental sclerosis are of the clas-sic type, with solidification of the tuft by increasedmatrix. However, in some cases there also arecellular or collapsing features that may mimicidiopathic FSGS. This is especially the case inpatients with secondary FSGS in the setting ofcholesterol embolization or renovascular hyperten-sion.46,47 A common feature of secondary FSGSowing to hypertensive nephrosclerosis is the devel-opment of atubular glomeruli with cystic dilatationof Bowman’s space (Fig 3F).56 The tuft isshrunken and partially resorbed into Bowman’scapsule. Atubular glomeruli are most abundant inareas of subcapsular scarring with severe tubularatrophy (Fig 3F).

In obesity-associated FSGS, there is prominentglomerular hypertrophy and the lesions of focalsclerosis typically affect a minority of hypertro-phied glomeruli.57,58 Lesions are characteristicallyperihilar with associated hyalinosis. There is little,

if any, podocyte reactivity (Fig 3E). FSGS usuallydevelops in the absence of significant backgroundtubulointerstitial disease. Mild mesangial sclerosismay be seen, resembling the changes of early di-abetic glomerulosclerosis.57

Pathogenesis

Secondary FSGS is often mediated by increasedglomerular capillary pressures and flow rates thatoccur as an adaptive response to a reduced numberof functioning nephrons.7 Increased wall tensioncauses mechanical stress on the podocyte foot pro-cess-GBM connection. This in turn leads to localdilatation of capillaries and strain on the podo-cytes.59-61 If the tension on the podocyte is severeand prolonged, there is progressive cell body at-tenuation, pseudocyst formation, and, ultimately,detachment from the GBM.61 Overload of the ly-sosomal system with protein resorption dropletsmay promote cell autoinjury.

Podocyte detachment is the first committed le-sion to segmental sclerosis, leaving bare patches ofGBM. The concept of relative podocyte insuffi-ciency proposed by Kriz explains how these de-nuded segments of GBM come into contact withparietal epithelial cells, promoting synechiae toBowman’s capsule.61 If patent capillaries remaincaught in the adhesion, a route of filtration into theperiglomerular interstitium toward the tubular polecan lead to obliteration of the tubular pole andformation of atubular glomeruli with microcysticdilatation of Bowman’s capsule.56


Knowledge of the presenting clinical featuresand associated medical conditions is essential todifferentiate primary FSGS from secondary FSGScaused by structural-functional adaptations.7 Typ-ically, patients with secondary FSGS manifest ne-phrotic range or subnephrotic proteinuria withoutfull nephrotic syndrome. Thus, they may have pro-teinuria levels greater than 3.0 g/d, but they usuallylack hypoalbuminemia, hypercholesterolemia, andedema. Because the development of FSGS is oftena response to a loss of functioning nephrons, mostpatients have a history of renal insufficiency formonths or years preceding the development ofnephrotic proteinuria. An exception is FSGS sec-ondary to morbid obesity, in which patients oftenhave an initially supernormal glomerular filtrationrate, reflecting a hyperfiltration state imposed by an


increased ratio of body mass to renal mass.57 Theresulting glomerular overwork is associated withincreased renal plasma flow rates, as well as anelevated glomerular filtration rate. Hypoxia causedby sleep apnea may play a role in this process bystimulating the renin-angiotensin system throughsympathetic activation.

The therapeutic approach to secondary FSGSdepends on the underlying condition. Correction ofthe underlying process, such as surgical repair ofreflux or congenital heart disease, should be soughtwhere appropriate. In patients with hyperfiltrationcaused by reduced renal mass (such as after vesi-coureteral reflux, hypertensive nephrosclerosis, re-nal agenesis, and so forth), maneuvers to reduceglomerular capillary pressures, such as angiotensinconverting enzyme inhibition or angiotensin II re-ceptor blockade are generally offered. Steroids areuniformly ineffective in secondary FSGS and mayeven promote weight gain and progressive sclero-sis in patients with obesity and latent diabetes. Thecourse of obesity-related glomerulopathy is gener-ally indolent and few patients progress to end-stagerenal disease. In this condition, reductions of pro-teinuria can be achieved with weight loss, sleepapnea therapy, and angiotensin receptor blockadeor angiotensin converting enzyme inhibitors.57

ACKNOWLEDGMENT

The author acknowledges the intellectual contributions of themembers of the consensus conference on the pathologic clas-sification of FSGS, including Drs. Jan Bruijn, Agnes Fogo, andJ. Charles Jennette.

REFERENCES

1. Fahr T: Pathologische anatomie des morbus brightii, inHenke F, Lubarsch O (eds): Handbuch der speziellen patholo-gischen anatomie und histologie (vol 6). Berlin, Springer, 1925,p 156

2. Rich AR: A hitherto undescribed vulnerability of thejuxta-medullary glomeruli in lipoid nephrosis. Bull Johns Hop-kins Hosp 100:173-186, 1957

3. Churg J, Habib R, White RH: Pathology of the nephroticsyndrome in children: A report for the International Study ofKidney Disease in Children. Lancet 760:1299-1302, 1970

4. Habib R: Focal glomerulosclerosis. Kidney Int 4:355-361,1973

5. Southwest Pediatric Nephrology Study Group: Focal seg-mental glomerulosclerosis in children with idiopathic nephroticsyndrome. Kidney Int 27:442-449, 1985

6. D’Agati V: The many masks of focal segmental glomer-ulosclerosis. Kidney Int 46:1223-1241, 1994

7. Rennke H, Klein PS: Pathogenesis and significance ofnon-primary focal and segmental glomerulosclerosis. Am JKidney Dis 13:443-455, 1989

8. Fogo A, Glick AD, Horn SL, et al: Is focal segmentalglomerulosclerosis really focal? Distribution of lesions in adultsand children. Kidney Int 47:1690-1696, 1995

9. Yoshikawa N, Ito H, Akamatsu R, et al: Focal segmentalglomerulosclerosis with and without nephrotic syndrome inchildren. J Pediatr 109:65-70, 1986

10. Southwest Pediatric Nephrology Study Group: Focalsegmental glomerulosclerosis in children with idiopathic ne-phrotic syndrome. Kidney Int 27:442-449, 1985

11. Schoeneman MJ, Bennett B, Greifer I: The natural historyof focal segmental glomerulosclerosis with and without mesangialhypercellularity in children. Clin Nephrol 9:45-54, 1978

12. Eddy AA, Giachelli CM: Renal expression of genes thatpromote renal interstitial inflammation and fibrosis in rats withprotein-overload proteinuria. Kidney Int 47:1546-1557, 1995

13. Zoja C, Benigni A, Remuzzi G: Protein overload acti-vates proximal tubular cells to release vasoactive and inflam-matory mediators. Exp Nephrol 7:420-428, 1999

14. Korbet SM: Primary focal segmental glomerulosclerosis.J Am Soc Nephrol 9:1333-1340, 1998

15. Velosa J, Holley K, Torres V, et al: Significance ofproteinuria on the outcome of renal function in patients withfocal segmental glomerulosclerosis. Mayo Clin Proc 58:568-577, 1983

16. Brown C, Cameron J, Turner D, et al: Focal segmentalglomerulosclerosis with rapid deterioration in renal function(“malignant FSGS”). Clin Nephrol 10:51-61, 1978

17. Wehrmann M, Bohle A, Held H, et al: Long-term prog-nosis of focal sclerosing glomerulonephritis: Analysis of 250cases with particular regard to tubulointerstitial changes. ClinNephrol 33:115-122, 1990

18. Schwartz MM, Korbet SM: Primary focal segmentalglomerulosclerosis: Pathology, histologic variants, and patho-genesis. Am J Kidney Dis 22:874-883, 1993

19. Schwartz MM, Korbet SM, Rydel JJ, et al. Primary focalsegmental glomerular sclerosis in adults: Prognostic value ofhistologic variants. Am J Kidney Dis 25:845-852, 1995

20. Mongeau JG, Robitaille PO, Clermont MJ, et al: Focalsegmental glomerulosclerosis (FSG) 20 years later: From tod-dler to grown up. Clin Nephrol 40:1-6, 1993

21. Schwartz MM, Lewis EJ: Focal segmental glomerularsclerosis: The cellular lesion. Kidney Int 28:968-974, 1985

22. Korbet SM, Schwartz MM, Lewis EJ: Recurrent ne-phrotic syndrome in renal allografts. Am J Kidney Dis 11:270-276, 1988

23. Schwartz MM, Evans J, Bain R, et al: Focal segmentalglomerulosclerosis: Prognostic implication of the cellular le-sion. J Am Soc Nephrol 10:1900-1907, 1999

24. Howie AJ, Brewer DB: The glomerular tip lesion: Apreviously undescribed type of segmental glomerular abnormal-ity. J Pathol 142:205-220, 1984

25. Huppes W, Hene RJ, Kooiker CJ: The glomerular tiplesion: A distinct entity or not? J Pathol 154:187-190, 1988

26. Haas M, Yousefzadeh N: Glomerular tip lesion in min-imal change nephropathy: A study of autopsies before 1950.Am J Kidney Dis 39:1168-1175, 2002

27. Beaman M, Howie AJ, Hardwicke J, et al: The glomer-ular tip lesion; a steroid responsive nephrotic syndrome. ClinNephrol 27:217-221, 1987

28. Hogan-Moulton A, Hogan S, Falk R, et al: Glomerular


tip lesion: Clinical features, response to corticosteroids andcomparison to FSGS. J Am Soc Nephrol 8:87A, 1997

29. Weiss MA, Daquioag E, Margolin EG, et al: Nephroticsyndrome, progressive irreversible renal failure, and glomerular“collapse.” A new clinico-pathologic entity? Am J Kidney Dis7:20-28, 1986

30. Detwiler RK, Falk RJ, Hogan SL, et al: Collapsingglomerulopathy: A clinically and pathologically distinct variantof focal segmental glomerulosclerosis. Kidney Int 45:1416-1424, 1994

31. Valeri A, Barisoni L, Appel GB, et al: Idiopathic col-lapsing focal segmental glomerulosclerosis: A clinico-patho-logic study. Kidney Int 50:1734-1746, 1996

32. Barisoni L, D’Agati V: The changing epidemiology offocal segmental glomerulosclerosis in New York City. ModPathol 7:156A, 1994

33. Barisoni L, Kriz W, Mundel P, et al: The dysregulatedpodocyte phenotype; a novel concept in the pathogenesis ofcollapsing idiopathic focal segmental glomerulosclerosis andHIV-associated nephropathy. J Am Soc Nephrol 10:51-61,1999

34. Shankland SJ, Eitner F, Hudkins KL, et al: Differentialexpression of cyclin-dependent kinase inhibitors in human glo-merular disease: Role in podocyte proliferation and maturation.Kidney Int 58:674-683, 2000

35. Barisoni L, Mokrzycki M, Sablay L, et al: Podocyte cellcycle regulation and proliferation in collapsing glomerulopa-thies. Kidney Int 58:137-143, 2000

36. Bariety J, Nochy D, Mandet C, et al: Podocytes undergophenotypic changes and express macrophagic-associated mark-ers in idiopathic collapsing glomerulopathy. Kidney Int 53:918-928, 1998

37. Bariety J, Bruneval P, Hill G, et al: Post-transplantationrelapse of FSGS is characterized by glomerular epithelial celltransdifferentiation. J Am Soc Nephrol 12:261-274, 2001

38. Laurinavicius A, Hurwitz S, Rennke HG: Collapsingglomerulopathy in HIV and non-HIV patients: A clinicopatho-logical and follow-up study. Kidney Int 56:2203-2213, 1999

39. Cameron JS, Turner DR, Ogg CS, et al: The long-termprognosis of patients with focal segmental glomerulosclerosis.Clin Nephrol 10:213-218, 1978

40. Markowitz GS, Appel GB, Fine PL, et al: Collapsingfocal segmental glomerulosclerosis following high dose pam-idronate. J Am Soc Nephrol 12:1164-1172, 2001

41. Moudgil A, Nast CC, Bagga A, et al: Association ofparvovirus B19 infection with idiopathic collapsing glomeru-lopathy. Kidney Int 59:2126-2133, 2001

42. Li R-M, Branton MH, Tanawattanacharoen S, et al:Molecular identification of SV40 infection in human subjectsand possible association with kidney disease. J Am Soc Nephrol13:2320-2330, 2002

43. Stokes MB, Davis CL, Alpers CE: Collapsing glomeru-lopathy in renal allografts: A morphological pattern with di-verse clinicopathologic associations. Am J Kidney Dis 33:658-666, 1999

44. Meehan SM, Pascual M, Williams WW, et al: De novo

collapsing glomerulopathy in renal allografts. Transplantation65:1192-1197, 1998

45. Nadasdy T, Allen C, Zand MS: Zonal distribution ofglomerular collapse in renal allografts: Possible role of vascularchanges. Hum Pathol 33:437-441, 2002

46. Greenberg A, Bastacky SL, Iqbal A, et al: Focal seg-mental glomerulosclerosis associated with nephrotic syndromein cholesterol embolization: Clinicopathologic correlations.Am J Kidney Dis 29:334-344, 1997

47. Thadhani R, Pascual M, Nickeleit V, et al: Preliminarydescription of focal segmental glomerulosclerosis in patientswith renovascular disease. Lancet 347:231-233, 1996

48. Jennette JC, Hipp CG: C1q nephropathy: A distinctpathologic entity usually causing nephrotic syndrome. Am JKidney Dis 6:103-110, 1985

49. Iskandar SS, Browning MC, Lorentz WB: C1q nephrop-athy: A pediatric clinicopathologic study. Am J Kidney Dis18:459-465, 1991

50. Jennette JC, Wilkman AS, Hogan SL, et al: Clinical andpathologic features of C1q nephropathy. J Am Soc Nephrol4:681, 1993

51. D’Agati V, Suh J-I, Carbone L, et al: Pathology ofHIV-associated nephropathy: A detailed morphologic and com-parative study. Kidney Int 35:1358-1370, 1989

52. D’Agati V, Appel GB: Renal pathology of human im-munodeficiency virus infection. Semin Nephrol 18:406-421,1998

53. Rao TK , Filippone EJ, Nicastri AD, et al: Associatedfocal and segmental glomerulosclerosis in the acquired immu-nodeficiency syndrome. N Engl J Med 310:669-673, 1984

54. Carbone L, D’Agati V, Cheng JT, et al: The course andprognosis of human immunodeficiency virus-associated ne-phropathy. Am J Med 87:389-395, 1989

55. Winston JA, Bruggeman LA, Ross MD, et al: Nephrop-athy and establishment of a renal reservoir of HIV type 1 duringprimary infection. N Engl J Med 344:1979-1984, 2001

56. Kriz W, Hosser H, Hahnel B, et al: From segmentalglomerulosclerosis to total nephron degeneration and interstitialfibrosis: A histopathologic study in rat models and humanglomerulopathies. Nephrol Dial Transplant 13:2781-2798, 1998

57. Kambham N, Markowitz GS, Valeri AM, et al: Obesity-related glomerulopathy: An emerging epidemic. Kidney Int59:1498-1509, 2001

58. Verani RR: Obesity-associated focal segmental glomer-ulosclerosis: Pathologic features of the lesion and relationshipwith cardiomegaly and hyperlipidemia. Am J Kidney Dis 20:629-634, 1992

59. Kriz W, Gretz N, Lemley KV: Progression of glomerulardiseases: Is the podocyte the culprit? 54:687-697, 1998

60. Kriz W, Elger M, Nagata M, et al: The role of podocytesin the development of glomerular sclerosis. Kidney Int 45:S64-S72, 1994 (suppl 45)

61. Nagata M, Kriz W: Glomerular damage after unine-phrectomy in young rats. II. Mechanical stress on podocytes asa pathway to sclerosis. Kidney Int 42:148-160, 1992


Pathologic Classiﬁcation of Focal Segmental...

Documents

Transcript of Pathologic Classiﬁcation of Focal Segmental...