pathology of thin basement membrane...

PK

Thufa

hggsscessddTsm

aad

D

A

0d

athology of Thin Basement Membrane Nephropathyirk Foster, Glen S. Markowitz, and Vivette D. D’Agati

Thin basement membrane nephropathy (TBMN) is a glomerular disorder characterizedclinically by isolated hematuria and pathologically by diffuse thinning of the glomerularbasement membrane (GBM) on ultrastructural examination. The pathologic diagnosis ofTBMN is problematic, in part because of the wide range of GBM thicknesses in the normalpopulation. GBM thickness varies with age, sex, and the different methods of tissuepreparation and measurement. In addition, there are no standardized diagnostic criteriadefining the degree or extent of GBM thinning required for the diagnosis of TBMN. GBMthinning is often seen in other glomerulopathies, where it may represent an overlap withTBMN or may be secondary to GBM damage and remodeling. Importantly, TBMN must bedifferentiated from the GBM thinning seen in some renal biopsy specimens from boys andfemale heterozygotes with X-linked Alport syndrome because of the very different prog-noses of these two conditions.Semin Nephrol 25:149-158 © 2005 Elsevier Inc. All rights reserved.

olimtmsdcmbics(auti

PItmcfcTt

hin basement membrane nephropathy (TBMN) is theglomerular disorder characterized clinically by isolated

ematuria and pathologically by a diffusely thinned glomer-lar basement membrane (GBM). This condition often is re-erred to as benign familial hematuria because of its familialggregation and lack of progression.

The genetic basis of TBMN is heterogeneous and includeseterozygous mutations in the COL4A3 or the COL4A4enes, which encode the �3 and �4 chains of type IV colla-en, respectively.1-3 Mutations usually result in missense,plice site, or frame shift changes. Most cases show an auto-omal-dominant pattern of inheritance, although sporadicases also occur. Because homozygous or compound het-rozygous mutations in COL4A3 or COL4A4 result in auto-omal-recessive Alport syndrome, some investigators con-ider TBMN to be the carrier state of this form of Alportisease,4 and the distinction between TBMN and Alport syn-rome becomes less absolute. However, some kindreds withBMN do not show linkage to the COL4A3/COL4A4 locus,uggesting the existence of further, still to be identified, chro-osomal loci.5

TBMN is aptly named for its defining morphologic char-cteristic, diffuse thinning of the GBM, which is evident onlyt the ultrastructural level. The pathology of TBMN appearseceptively simple, but in practice the diagnosis of TBMN is

epartment of Pathology, Columbia University, College of Physicians andSurgeons, New York, NY.

ddress reprint requests to Vivette D. D’Agati, MD, Department of Pathology,Columbia University, College of Physicians and Surgeons, 630 West 168th

dSt, Room VC14-224, New York, NY 10032. E-mail: [email protected]

270-9295/05/$-see front matter © 2005 Elsevier Inc. All rights reserved.oi:10.1016/j.semnephrol.2005.01.006

ften difficult. The major problems are the absence of guide-ines on how to measure the GBM thickness, and of standard-zed criteria to define GBM thinning. GBM thickness nor-

ally varies with age and can be influenced by the type ofissue fixation, embedding medium, and method of measure-ent. TBMN must be differentiated from the thinned GBM

een in some patients with Alport syndrome. It also must beistinguished from the nonspecific localized thinning thatan occur in the course of GBM remodeling in diverse glo-erular diseases. Interpretation of the literature is hampered

y the lack of genotype-phenotype correlations in early stud-es and the difficulty in distinguishing the overlapping clini-al and pathologic features of TBMN from those of Alportyndrome. In addition, in part because TBMN is so commonit affects up to 5% of the general population6), it may occurs an incidental finding in renal biopsies performed for othernrelated renal diseases. This article outlines the approach tohe renal biopsy diagnosis of TBMN in the context of thesessues.

athologic of TBMNn 1973, Rogers et al7 were the first to link familial hematuriao thinning of the GBM in a family with isolated persistenticroscopic hematuria over 4 generations. By light micros-

opy, the glomeruli and tubulointerstitium of an affectedamily member appeared normal. Direct immunofluores-ence examination showed no immunoglobulin deposition.he GBM was reported to show “marked thinning” by elec-

ron microscopy with “irregular narrowing of the lamina
ensa” that varied from segmental to circumferential. Com-
149

pdwwtlit

WaGonlgiomGeumflmcw

ctskmptmbap

t

omm3ewg

GgwHtl

vs(wmetBmotntGm

aosem

mspmwtcimbmtimfnn

amodttcyaa1ttgwd

150 K. Foster, G.S. Markowitz, and V.D. D’Agati

arison with the size of latex beads measuring 312 nm intro-uced onto the tissue sections indicated the GBM thicknessas 150 nm. The rare breaks in the GBM that were observedere proposed as the source of entry of red blood cells into

he urine. Since this historic report, numerous clinicopatho-ogic studies have confirmed the association of benign famil-al hematuria with the renal biopsy findings of diffuse GBMhinning.8-19

hat is Normal GBM Thicknessnd How Should it be Measured?BM thickness varies with age and is influenced by methodsf tissue preparation and measurement. Because the trilami-ar GBM is composed of a central lamina densa, an inner

amina rara interna, and an outer lamina rara externa, there iseneral agreement that the most accurate means of determin-ng GBM thickness is to measure the distance between theuter limits of the endothelial cell and the podocyte cellembranes. Measuring the electron-dense portion of theBM accounts only for the lamina densa and not the morelectron-lucent laminae rarae. The GBM thickness is mostniform over the peripheral capillary filtering surface, butay appear thickened, folded, and redundant where it re-ects over the mesangium. Thus, GBM thickness should beeasured only in the peripheral capillaries. Because the GBM is

urved, measurements are most accurate when they are madehere the GBM is cut in cross-section, rather than tangentially.The methods of GBM measurement are variable and in-

lude the introduction of latex beads of known diameter ontohe tissue sections, the orthogonal intercept method whichuperimposes a grid of perpendicularly oriented lines ofnown dimension onto the electron micrograph, and simpleeasurement with a magnification graticule or metric rulerlaced over the electron microscope prints. Once the GBMhickness has been measured in millimeters from the electronicroscopic print, its value in nanometers can be calculated

y correcting for the magnification of the photographic neg-tive and the factor of enlargement from the negative to therint according to the following equation:

Thickness (nm) � thickness on print (mm) � (magnifica-ion � enlargement factor)

For example, if the GBM thickness (T) measures 1.5 mmn an electron microscopic print photographed at 2,000�agnification and if the print represents a 2.5-fold enlarge-ent from the negative, then T � 1.5 mm � (2,000 � 2.5) or

00 nm. To simplify matters, the digital cameras on manylectron microscopes are now equipped with computer soft-are that calculates the distance between 2 points for anyiven magnification.

There is no consensus on how many measurements ofBM width should be made per biopsy specimen, how manylomerular capillaries and glomeruli should be sampled, andhether the arithmetic or harmonic mean should be used.owever, there is general agreement that the mean from mul-

iple representative measurements made over multiple capil-
aries of multiple glomeruli is most accurate. f
Normal GBM thickness varies with age and sex, and thealues reported in the literature differ widely. Osawa et al20

tudied a series of normal adult biopsy (n � 7) and autopsyn � 14) specimens. An average of 587 GBM measurementsere made for each biopsy and 254 for each autopsy speci-en. In each case, the thickness was measured from the

ndothelial to the epithelial cytoplasmic borders, avoidingangentially cut areas, and only in the peripheral capillaries.y using these criteria, the mean GBM thickness (arithmeticean) for all cases was calculated to be 315 nm with a range

f means from 239 to 453 nm. However, the mean GBMhickness in individual glomeruli ranged from 225 to 500m, indicating even wider interglomerular variability. Fur-hermore, these investigators showed that the distribution ofBM thickness is relatively Gaussian in some individual glo-eruli, whereas in others it is much more irregular.In 1981, Haynes21 outlined his methods for ultrastructural

nalysis of the normal human glomerulus. By overlaying a gridf lines of known dimensions, measurements were taken at po-itions where the GBM clearly was bordered by epithelial andndothelial cytoplasm. For normals aged 11 to 26 years, theean GBM thickness was 394 nm (range, 372-632 nm).21

The orthogonal intercept technique is a modification of theethod originally described by Weibel and Knight22 for

tudying the thickness of alveolar membranes. A grid of per-endicularly oriented lines is superimposed on the electronicrograph, and the GBM thickness is measured at any pointhere a line intersects the interface between endothelial cy-

oplasm and membrane. The inverse length is then used toalculate the harmonic mean. The advantage of this methods that tangential sections may be included in the measure-

ents and are accounted for by the formula. Multiplicationy a correction factor yields the mean GBM thickness. Theathematic basis of this method is quite complicated, and

he reader is referred to the detailed descriptions.23,24 By us-ng this method with an average intercept number of 164, the

ean GBM thickness was 326 nm in a group of 59 normalemale kidney donors and 373 nm in 59 normal male do-ors.25 These investigators also showed that the GBM thick-ess increases with age until age 40 and then plateaus.The definition of normal GBM thickness is more problem-

tic in the pediatric population. Vogler et al26 calculated nor-al GBM thickness in an autopsy study of 37 children with-

ut known renal disease by using latex microbeads 500 nm iniameter as comparative markers. GBM measurements wereaken at the thinnest points to minimize error introduced byangential sectioning. The GBM thickness was found to in-rease progressively with age until preadolescence (�11ears of age), at which time it reached the adult width ofbout 300 nm. The growth rate was maximal at 2 years of agend thereafter decreased. The GBM thickness ranged from32 to 208 nm at ages 2 days to 1 year, and increased to 208o 245 nm at ages 1 to 6 years, and 244 to 307 nm at ages 6o 11 years.26 In addition to being thin, the GBM of immaturelomeruli may show apparent splitting of the lamina densa,hich represents the individual basement membranes pro-uced by the endothelial cell and the podocyte before their
usion into a single membrane.

cOpaswsficgetas

eoBsmp2d2

tctscmt5a1

HGptpt2nmg

h28�Bb2Ig

oGcsa

LBhadmacacucgmm

u(rrupggofiahgtttayragtohsci

msdicc

Pathology of Thin BM Nephropathy 151

Measurements of GBM thickness also are influenced by thehoice of embedding compound for electron microscopy.sawa et al20 found thicker measurements using Vestopal, aolyester-polystyrene copolymer compared with Ethacrylatend Araldite epoxy resins. Haynes21 described thicker mea-urements using the low-viscosity Spurr’s resin comparedith Vestopal and Methacrylate. We consistently have ob-

erved that GBMs appear thinner in reprocessed formalin-xed tissue that has been deparaffinized for electron micros-opy (as a salvage technique for inadequate tissue) than inlutaraldehyde-fixed, Epox-embedded tissue. This differ-nce probably relates to increased tissue dehydration withhe former method. For this reason, it is not possible to di-gnose TBMN accurately in deparaffinized, reprocessed tis-ue.

Optimally, individual renal pathology laboratories shouldstablish their own criteria for normal GBM thickness basedn their fixation, embedding, and measurement techniques.ecause biopsy specimens of normal kidneys are rare, sometudies have used cases with either minimal change disease oresangial proliferative glomerulonephritis as the controlopulations, with normal mean GBM thickness ranging from87 to 399 nm.8,10,17,27,28 As a result, the threshold used toefine thin in various studies is highly variable, ranging from00 to 330 nm.12,13,17,28-31

In summary, it is best to measure the GBM thickness fromhe endothelial to the podocyte cell membranes in peripheralapillary walls (not overlying the mesangial reflection). Mul-iple representative measurements from many capillarieshould be recorded, and the arithmetic or harmonic meansalculated. The results must be interpreted based on the nor-al values for the patient’s age and sex. Mean adult GBM

hickness is approximately 370 � 50 nm in men and 320 �0 nm in women.25 GBM thickness is approximately 150 nmt birth, 200 nm at 1 year, and approaches adult thickness by1 years of age.26

ow Thin is Thin?iven the wide variation of GBM thickness in the normalopulation, it has been difficult to define what constituteshinning of the GBM in a particular individual. For practicalurposes, the World Health Organization has promulgated ahreshold of 250 nm for adults and 180 nm for children agedto 11 years.32 Some investigators prefer a threshold of 200m in adults, especially in women.33 There is general agree-ent that the thinning should affect at least 50% of the total

lomerular surface area.The Columbia University Nephropathology Laboratory

as set its own standards for normal based on tissue fixed in.5% glutaraldehyde and embedded in Epoxy Resin (Epox12; Fullam Inc., Latham, NY). Mean GBM thickness is 35050 nm for adult males and 300 � 40 nm for adult females.

y using a definition of thin that requires greater than 2 SDselow the mean value, we thus define thin GBM as less than50 nm in adult males and less than 220 nm in adult females.n addition, we require greater than 50% involvement of
lomerular capillaries for the diagnosis of TBMN. In the case d
f biopsies performed for hematuria and showing isolatedBM thinning that involves less than 50% of the glomerularapillaries, this ultrastructural abnormality is duly noted asegmental GBM thinning in the electron microscopic report,lthough its significance remains unclear at this time.

ight Microscopic Findings in TBMNy light microscopy, the majority of renal biopsy specimensave little or no pathologic alterations. Usually, the glomerulippear unremarkable. Light microscopy is too insensitive toetect thinning of the basement membrane accurately, and inost cases the GBM appears normal in thickness, texture,

nd contour. In some cases, delicate attenuation of the GBMan be appreciated by Jones methenamine silver or periodiccid-Schiff stain, suggesting GBM thinning, but this must beonfirmed by electron microscopy (Fig 1A). The mesangiumsually is normal in size and cellularity. In a minority ofases, however, the mesangium may show mild segmental orlobal expansion by either increased cell number, increasedatrix material, or both (Fig 1B).10,11,13,29,30 Erythrocytesay be identified free within the urinary space.Focal glomerulosclerosis occurs in 5% to 25% of glomer-

li, especially when aging or hypertension are presentFig 1C).28,34 In the older adult, coexistent arterionephroscle-osis of aging or hypertension, with secondary patchy arte-iosclerosis, tubular atrophy, and interstitial fibrosis are notncommon. It is controversial whether TBMN everrogresses to global glomerulosclerosis. In a Dutch study,lobal glomerulosclerosis in TBMN involved 13% � 17% oflomeruli and correlated with the subsequent developmentf hypertension, proteinuria, and progression to renal insuf-ciency, suggesting a nonbenign course.35 Because the meange of participants in the study was 36 years and none hadypertension at the time of biopsy examination, the globallomerulosclerosis could not be attributed to aging or hyper-ension. In these patients, Alport syndrome was excluded byhe absence of hearing loss and ophthalmologic disease, byhe normal GBM staining for the �3 subunit of collagen IV,nd by repeat biopsy specimens taken after a median of 10ears that confirmed persistent GBM thinning with minimaleduplication. Unlike in Alport syndrome, repeat biopsy ex-minations in these patients all showed increased focal globallomerulosclerosis with patchy tubular atrophy and intersti-ial fibrosis, but without focal segmental glomerulosclerosisr interstitial foam cells.35 In rare instances in which patientsave both TBMN and the nephrotic syndrome, segmentalclerosis including a tip lesion have been reported.36 Suchases probably represent the coincidental superimposition ofdiopathic nephrotic syndrome on TBMN.

In children, the tubulointerstitial compartment is unre-arkable or shows only minimal tubular atrophy and inter-

titial fibrosis.18 In adults, the amount of tubulointerstitialisease is expectedly more variable, with tubular atrophy and

nterstitial fibrosis involving from 0% to 60% of the sampledortical tissue.11,12,15 Aggregates of interstitial foam cells, aommon finding in the more advanced stages of Alport syn-
rome, are not observed in TBMN. Depending on the severity


Figure 1 Light microscopic and immunofluorescence findings in TBMN. (A) A representative glomerulus is normocel-lular with fully patent capillaries. The GBMs appear attenuated, even by light microscopy, in this section stained withJones methenamine silver (JMS) stain (magnification, 250�). (B) An example of a case with mild mesangial prominenceby light microscopy. GBMs appear thin and delicate, with normal texture and contour. Erythrocytes are noted in theurinary space, consistent with a glomerular source of hematuria (hematoxylin-eosin stain, magnification, 400�). (C)A low-power view shows 2 globally sclerotic glomeruli. The other 2 glomeruli and the tubulointerstitial compartmentappear unremarkable (JMS, magnification, 100�). (D) Immunostaining for the �1 subunit of collagen IV shows thenormal distribution of positivity in all renal basement membranes and the mesangial matrix (magnification, 400�). (E)Immunostaining for the �3 subunit of collagen IV shows the normal distribution of global positivity in the GBM, as wellas distal tubular basement membranes. Bowman’s capsule stains negative over most of its circumference (in contrast tothe positive staining of Bowman’s capsule for �1 and �5) (magnification, 400�). (F) Staining for the �5 subunit ofcollagen IV shows retention of the normal global positivity in the GBM, with positivity in Bowman’s capsule and distal
tubular basement membranes (magnification, 400�).

oeptbihalpht

IItcmodcipw

ETttaTlw2nMcsm

coeeacsvicaTf

Tfm

caceBuGppw

Gfleerfmpgss

Tem(

SIpfHbsrdptn

sia(Ac(smptbm�


f the hematuria, tubules may contain free intraluminalrythrocytes or red blood cell casts. The presence of dysmor-hic red blood cells or casts helps differentiate in vivo hema-uria from traumatic blood introduced into the tubules by theiopsy procedure. Red blood cell casts have been described

n patients with TBMN who have loin pain and macroscopicematuria.34 Rarely, gross hematuria results in tubular injurynd acute renal failure, analogous to the cases of acute tubu-ar necrosis from hematuria in immunoglobulin (Ig)A ne-hropathy.37,38 In these patients, the release of iron fromemoglobin probably produces toxic acute tubular injuryhrough the generation of reactive oxygen species.39

mmunofluorescence Findings in TBMNn most cases, direct immunofluorescence staining is nega-ive for immunoglobulins and complement. In a minority ofases, there may be a weak (trace to 1�) and segmentalesangial positivity, usually for IgM or C3, and rarely for IgG

r IgA.8,13,15,30,36 In the absence of corresponding electronense deposits at the ultrastructural level, this staining isonsidered nonspecific. Strong or diffuse staining for anymmune reactants suggests a superimposed immune com-lex–mediated glomerular disease, the most common ofhich is IgA nephropathy.

lectron Microscopic Findings in TBMNhe renal biopsy diagnosis of TBMN can be made only by elec-

ron microscopy. The cardinal feature is thinning of the GBM inhe absence of other significant glomerular alterations (Figs 2And 2B). The reported mean GBM thickness in adults withBMN varies widely (Figs 2C-2E).10,12,14,28-31,36,40,41 In a compi-

ation of series analyzing renal biopsy findings in 104 adultsith TBMN, the mean GBM thickness was 250 to 300 nm in2% of biopsy specimens, 200 to 250 nm in 34%, 150 to 200m in 38%, and less than 150 nm in only 6%.10,12,28,30,31,36,40

ost studies report global glomerular involvement, but thehanges appear segmental in some individuals.10,13,15 Sometudies have used a cut-off level of 50% or greater GBM involve-ent.13,18

Some investigators have described the attenuation asaused primarily by thinning of the lamina densa, whereasthers have noted that the lamina densa appears accentuat-d.15,29 We have observed that typically there is relative pres-rvation of the lamina rara interna and externa (these are lessffected), which makes the thinned lamina densa appear ac-entuated. The contour of the lamina densa generally ismooth and uniform. In a minority of cases, there may beery mild and focal splitting or lamellation, as well as foci ofntramembranous rarefaction producing electron lucent la-unae.10,15,18 These textural changes are extremely localizednd far less developed than those seen in Alport syndrome.he lamina rara interna and externa also generally are uni-

orm but can have a scalloped appearance (Fig 2F).8

Despite the characteristic clinical feature of hematuria inBMN, ruptures in the GBM are rarely seen. However, a

ortuitous plane of section may show an erythrocyte in
idtransit across the GBM.42 The difficulty in detecting gaps a
ould be explained by the relatively small glomerular surfacerea sampled by 2-dimensional transmission electron mi-roscopy. There are no reports of studies using scanninglectron microscopy to quantitate the GBM gaps in TBMN.ased on the infrequency of detected gaps in this conditionsing transmission electron microscopy, it is likely that theBM in TBMN is sufficiently thin for erythrocytes to be pro-elled through the gel-like structure at the heights of systolicressure in susceptible areas, and that these then close overithout the need for any fixed gaps.The mesangium usually is unremarkable. The attenuated

BM maintains its normally folded contours where it is re-ected over the mesangium (Fig 2E). There may be mildxpansion of the mesangial matrix, and small scant mesangiallectron densities are rare.10,13,28 The podocytes usually haveelatively intact foot processes. However, mild segmental ef-acement of foot processes may be observed overlying theore attenuated segments (Fig 2B).11,12,28 If the patient hasronounced proteinuria and the foot process effacement iseneralized or complete, minimal change disease or focalegmental glomerulosclerosis superimposed on TBMNhould be considered.36

Little is known of the genotype-phenotype correlations inBMN. As yet, no study has reported morphologic differ-nces by either light or electron microscopy based on theutant gene (COL4A3 and COL4A4) or type of mutation

missense, splice site, frame shift, and so forth).43-46

pecial Studies for Collagen IV Subtypest has been known for some time that the glomeruli fromatients with Alport syndrome usually fail to stain with serarom patients with Goodpasture (or antiGBM) disease.47

owever, it was 25 years after this observation first was madeefore its molecular basis was elucidated. This absence oftaining with Goodpasture antisera by indirect immunofluo-escence often is used as an ancillary test to confirm theiagnosis of Alport syndrome. By contrast, the glomeruli ofatients with TBMN retain positive staining with Goodpas-ure antisera, although with decreased intensity compared toormals.48

In recent years, the use of Goodpasture antisera has beenuperceded by commercially available monospecific antibod-es to the � subunits of collagen IV.49 These antibodies can bepplied to frozen sections of renal or skin biopsy specimensTable 1). The commercially available Wieslab kit (WieslabB, Lund, Sweden) includes antibodies to the �1, �3, and �5hains of collagen IV. Antibodies to the subunits �1 and �2which comprise the major collagen IV network) normallytain all renal basement membranes including the mesangialatrix. Antibodies to subunits �3, �4, and �5 (which com-rise the minor collagen IV network) normally stain the en-ire thickness of the GBM as well as distal tubular mem-ranes. Normal Bowman’s capsule and normal epidermalembrane have immunoreactivity for the �1, �2, �5, and6, but not the �3 or �4 chains.50,51

In TBMN, the glomeruli retain their normal immunore-
ctivity for the �1, 2, 3, 4, and 5 subunits of collagen IV, as

wuti

hrei�locmsddtnchXmGm

(tfcd

uptstt�wiwLsTid

DTstoHlonlecswh

T

NTXXA

A

*


ell as for Goodpasture antisera (directed to the �3 sub-nit) (Figs 1D-1F).48,52,53 However, because of the GBMhinning, the staining often appears attenuated, giving anmpression of weaker positivity than normal.48

By contrast, males with X-linked Alport syndrome whoave mutations in the COL4A5 gene typically lack immuno-eactivity for the �5 subunit of collagen IV in their GBM andpidermal membranes. Because mutations in the gene encod-ng the �5 subunit cause defective incorporation of �3 and4 subunits into the supramolecular collagen IV network,

ack of reactivity for �3 and �4 usually accompanies the lossf �5 from the GBM.54 However, because the �3 and �4hains are not expressed normally in the epidermal basementembrane, immunostaining for �3 and �4 in skin biopsy

pecimens has no diagnostic utility. In males with X-linkedisease, Bowman’s capsule and the distal tubular membraneso not stain for the �5 and �6 chains. Immunoreactivity forhe �1 chain in the glomerular and epidermal membranes isormal or even increased because this major collagen IVhain is part of a collagen IV network that does not formeterotrimers with the �5 subunit.55,56 Females with-linked Alport syndrome usually show mosaicism with seg-ental or patchy loss of the �3, �4 and �5 chains from theBM and segmental loss of the �5 chain from their epidermalembrane.In patients with autosomal-recessive Alport syndrome

caused by homozygous or compound heterozygous muta-ions in COL4A3 or COL4A4), there is absent GBM stainingor the �3, �4, and �5 chains, but in contrast to X-linkedases, normal staining of �5 and �6 in Bowman’s capsule,istal tubular basement membranes, and skin.57

Unfortunately, immunostaining for the collagen IV sub-

Figure 2 Ultrastructural findings in TBMN. (A) On low-ation of the GBM compared with (B) a normal age-matcthe foot processes are well preserved (magnification, 2,0lamina densa. There is relative accentuation of the lamin(magnification, 6,000�). (D) A more extreme examplesegmental foot process effacement (magnification, 4,0peripheral capillary walls. Where the GBM reflects overless attenuated. For this reason, measurement of GBM thsegments (magnification, 5,000�). (F) On high-powerlarities including slight lamellation of the lamina den

able 1 Expression Patterns of Type IV Collagen Chains in K

Alpha-1in Kidney

Ain

ormal All renal BMs GBMBMN All renal BM GBM-linked Alport syndrome (male) All renal BM Neg-linked Alport syndrome (female) All renal BM Moutosomal-recessive Alport syndrome All renal BM Neg

bbreviations: BM, basement membrane; GBM, glomerular basemenbasement membrane; BC, Bowman’s capsule

Mosaic pattern, segmental loss of positivity.

8,000�).

nits has only a 60% to 70% sensitivity for detection of Al-ort syndrome, and positive staining for �5 does not excludehe disease, especially in patients with single nucleotide sub-titutions.49,58 To this point, Lajoie52 tested a group of hema-uric women with mean GBM thicknesses ranging from 161o 239 nm and found that most had diffuse staining for the3, �4, and �5 chains consistent with TBMN, including 3omen with segmental GBM lamellation who had male fam-

ly members with end-stage renal failure, suggesting theomen were heterozygotes for X-linked Alport syndrome.iapis et al53 also found that preserved �3 and �5 immuno-taining did not always differentiate thinning caused byBMN from that caused by X-linked Alport syndrome. Thus,

mmunostaining is a helpful, but not an absolute means toistinguish TBMN from Alport syndrome.

ifferential Diagnosishe major disease to be differentiated from TBMN is Alportyndrome. The characteristic and predominant ultrastruc-ural feature of Alport syndrome is lamellation of the GBM,ften alternating with segments of thinning (Fig 3A).59,60

owever, in some kindreds with Alport syndrome, particu-arly with autosomal-recessive inheritance, thinning is thenly detectable ultrastructural abnormality.43 Diffuse thin-ing also is observed commonly in young children with X-

inked Alport syndrome and in some females who are het-rozygotes for this condition (Figs 3B and 3C) indicatingonsiderable phenotypic overlap between TBMN and Alportyndrome. Such cases can be classified as “Alport syndromeith thin basement membrane phenotype”.53,61 A full clinicalistory including the mode of inheritance, a family history of

view, a case of TBMN displays uniform diffuse attenu-ntrol photographed at the same magnification. Overall,(C) A case of TBMN with characteristic thinning of theae. Mild segmental effacement of foot processes is seenM thinning (mean, 100 nm) is shown. There is mild(E) There is uniform thinning of the GBM over thesangium, it retains its usual folded pattern and appearss should be performed only on peripheral capillary wallation, the thinned GBM exhibits mild textural irregu-scalloping of the lamina rara interna (magnification,

and Skin

-3ey

Alpha-5in Kidney

Alpha-1in Skin

Alpha-3in Skin

Alpha-5in Skin

BM GBM, D-TBM, BC EBM Negative EBMBM GBM, D-TBM, BC EBM Negative EBM

Negative EBM Negative NegativeMosaic* EBM Negative Mosaic*D-TBM, BC EBM Negative EBM

brane; D-TBM, distal tubular basement membrane; EBM, epidermal

powerhed co00�).ae rarof GB

00�).the meicknesexaminsa and

idney

lphaKidn

, D-T, D-T

ativesaic*ative

t mem

rihAtmp

ojvglacmiltGds

si

ODBsocsfiTossalcrdg


enal failure and deafness, as well as the immunohistochem-cal examination of the GBM for collagen type IV chains allelp differentiate between TBMN and thinning in X-linkedlport syndrome. Furthermore, in Alport syndrome, in con-

rast to TBMN, progression to secondary focal segmental glo-erulosclerosis with prominent interstitial foam cells androgressive chronic tubulointerstitial disease is often found.GBM thinning also occurs as an acquired non-specific sec-

ndary lesion in diverse glomerular diseases because of in-ury and remodeling. Hill et al59 reported focal thinning in aariety of glomerulopathies. Thinning may occur because oflomerular capillary distention by thrombosis or endocapil-ary proliferation. Thus, focal GBM thinning may occur incute postinfectious glomerulonephritis, lupus nephritis, ne-rotizing pauci-immune glomerulonephritis, and thromboticicroangiopathies. Localized GBM thinning may be detected

n immune complex diseases such as membranous glomeru-opathy or resolving postinfectious glomerulonephritis, par-icularly in foci of resorbed deposits where there has beenBM remodeling.59 In all of these conditions, an accurateiagnosis can be reached by careful consideration of the ob-
ervations on electron microscopy together with the clinical, o
erologic, light microscopic, and immunofluorescence find-ngs.

verlap Between TBMN and Other Renaliseases

ecause TBMN occurs in up to 5% of the population,6 it is noturprising that TBMN often is recognized coexistent withther renal diseases at the time of renal biopsy. The mostommon of these is IgA nephropathy (Fig 4).33,62 In thisetting, it is not clear whether TBMN occurs as a coincidentalnding caused by the overlap of 2 unrelated entities, whetherBMN predisposes to IgA disease, or whether the thinningccurs because of basement membrane remodeling. In onetudy, 31% of biopsy specimens with IgA nephropathy hadegmental GBM thinning defined as less than 250 nm, oftenccompanied by small foci of lamellation or splitting of theamina densa.63 Yoshikawa et al64 hypothesized that thesehanges resulted from lysis and attenuation of GBM duringesorption of immune deposits. Others have suggested thatiffuse GBM thinning in IgA nephropathy represents duallomerulopathies that are diagnosed coincidentally because

igure 3 Ultrastructural findings in Alport syndrome. (A) Low-ower view shows the classic lamellation and irregular thickening ofhe GBM, alternating with thinned segments. The lamellation evennvolves the GBM reflections over the mesangium. There is exten-ive effacement of foot processes (magnification, 4,000�). (B) Aemale heterozygote with X-linked Alport syndrome has a predom-nantly thin basement membrane with only rare textural irregulari-ies. A few minute electron lucent lacunae are seen within the laminaensa (magnification, 5,000�). (C) Another female heterozygoteor X-linked Alport syndrome has diffuse GBM thinning and mini-al segmental lamellation that mimics TBMN (magnification,

,000�).

Fptisfitdfm5

f their similar presentations with hematuria.65

iaapmgtltbdincataitbhnstdg

corutt

otfs

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

FTmi2


There are other reports of TBMN occurring as an underly-ng condition in renal biopsy examinations performed fornother glomerular disease. TBMN has been described inssociation with membranous glomerulopathy.66 Whereasatients with idiopathic membranous glomerulopathy had aean GBM thickness of 876 nm, patients with membranous

lomerulopathy and thin membranes had an average GBMhickness of 163 nm.66 Of interest, the patients with under-ying TBMN all exhibited stage 1 membranous glomerulopa-hy, whereas all 4 stages were observed in the group of mem-ranous glomerulopathy controls without TBMN. Thus, theefective GBM synthesis in patients with TBMN appears to

nhibit the formation of spikes and overlying neomembranesormally seen in response to the subepithelial deposits in theourse of membranous glomerulopathy. Thin membraneslso have been reported in patients with membranoprolifera-ive glomerulonephritis and lupus nephritis.67,68 Some met-bolic nonimmune renal diseases show thinned GBM too,ncluding diabetic nephropathy.69 Interestingly, in a study ofype 1 and type 2 diabetic patients, the patients with overlapetween diabetic nephropathy and TBMN were reported toave a significantly lower incidence of Kimmelstiel-Wilsonodules, suggesting dysregulated matrix synthesis.69 Biopsypecimens showing early diabetic nephropathy had a GBMhickness of less than 200 nm, whereas those with advancedisease had progressively more thickened GBM segmentsreater than 1,000 nm.69

In conclusion, the diagnosis of TBMN by renal biopsyhallenges the skills of the renal pathologist. The clinicopath-logic diagnosis of TBMN is problematic because of the wideange of GBM thickness in the normal population, the lack ofniform criteria for the diagnosis of TBMN, and the observa-ion that GBM thinning also occurs in other glomerulopa-

igure 4 An example of overlap between IgA nephropathy andBMN. The mesangium is expanded by increased cell number andatrix, which contains abundant mesangial electron dense depos-

ts. The GBM is diffusely and uniformly thinned (magnification,,000�).

hies. In the future, pathologic examination of the renal bi- 2

psy specimen may be supplemented by molecular or geneticesting that further refines our ability to differentiate TBMNrom Alport syndrome, a disease with very different progno-is.

eferences1. Badenas C, Praga M, Tazon B, et al: Mutations in the COL4A4 and

COL4A3 genes cause familial benign hematuria. J Am Soc Nephrol13:1248-1254, 2002

2. Longo I, Porcedda P, Mari F, et al: COL4A3/COL4A4 mutations: Fromfamilial hematuria to autosomal-dominant or recessive Alport syn-drome. Kidney Int 61:1947-1956, 2002

3. Ozen S, Ertoy D, Heidet L, et al: Benign familial hematuria associatedwith a novel COL4A4 mutation. Pediatr Nephrol 16:874-877, 2001

4. Hudson BG, Tryggvason K, Sundaramoorthy M, et al: Alport’s syn-drome, Goodpasture’s syndrome, and type IV collagen. N Engl J Med348:2543-2556, 2003

5. Piccini M, Casari G, Zhou J, et al: Evidence for genetic heterogeneity inbenign familial hematuria. Am J Nephrol 19:464-467, 1999

6. Dische FE, Anderson VER, Keane SJ, et al: Incidence of thin membranenephropathy: Morphometric investigation of a population sample.J Clin Pathol 43:457-460, 1990

7. Rogers PW, Kurtzman NA, Bunn SM, et al: Familial benign essentialhematuria. Arch Intern Med 131:257-262, 1973

8. Yoshikawa N, Hashimoto H, Katayama Y, et al: The thin glomerularbasement membrane in children with haematuria. J Pathol 142:253-257, 1984

9. Milanesi C, Rizzoni G, Braggion F, et al: Electron microscopy for mea-surement of glomerular basement membrane width in children withbenign familial hematuria. Appl Pathol 2:199-204, 1984

0. Dische FE, Weston MJ, Parsons V: Abnormally thin glomerular base-ment membranes associated with hematuria, proteinuria or renal fail-ure in adults. Am J Nephrol 5:103-109, 1985

1. Coleman M, Haynes WDG, Dimopoulos P, et al: Glomerular basementmembrane abnormalities associated with apparently idiopathic hema-turia: Ultrastructural morphometric analysis. Hum Pathol 17:1022-1030, 1986

2. Abe S, Amagasaki Y, Iyori S, et al: Thin basement membrane syndromein adults. J Clin Pathol 40:318-322, 1987

3. Yoshikawa N, Matsuyama S, Iijima K, et al: Benign familial hematuria.Arch Pathol Lab Med 112:794-797, 1988

4. Shindo S, Yoshimoto M, Kuriya N, et al: Glomerular basement membranethickness in recurrent and persistent hematuria and nephrotic syndrome:Correlation with sex and age. Pediatr Nephrol 2:196-199, 1988

5. Aarons I, Smith PS, Davies RA, et al: Thin membrane nephropathy: Aclinico-pathological study. Clin Nephrol 32:151-158, 1989

6. Gauthier B, Trachtman H, Frank R, et al: Familial thin basement mem-brane nephropathy in children with asymptomatic microhematuria.Nephron 51:502-508, 1989

7. Tiebosch ATMG, Frederik PM, van Breda Vriesman PJC, et al: Thin-basement-membrane nephropathy in adults with persistent hematuria.N Engl J Med 320:14-18, 1989

8. Gubler MC, Beaufils H, Noel LH, et al: Significance of thin glomerularbasement membranes in hematuric children. Contrib Nephrol 80:147-156, 1990

9. Lang S, Stevenson B, Risdon RA: Thin basement membrane nephrop-athy as a cause of recurrent haematuria in childhood. Histopathology16:331-337, 1990

0. Osawa G, Kimmelstiel P, Seiling V: Thickness of glomerular basementmembranes. Am J Clin Pathol 45:7-20, 1966

1. Haynes WDG: The normal human renal glomerulus. Virchows Arch35:133-158, 1981

2. Weibel ER, Knight BW: A morphometric study on the thickness of thepulmonary air-blood barrier. J Cell Biol 21:367-396, 1964

3. Gundersen HJG, Jensen TB, Osterby R: Distribution of membranethickness determined by lineal analysis. J Microsc 113:27-43, 1978

4. Jensen EB, Gundersen HJG, Osterby R: Determination of membrane

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

5

5

6

6

6

6

6

6

6

6

6

6


thickness distribution from orthogonal intercepts. J Microsc115:19-33, 1979

5. Steffes MW, Barbosa J, Basgen JM, et al: Quantitative glomerular mor-phology of the normal human kidney. Lab Invest 49:82-86, 1983

6. Vogler C, McAdams AJ, Homan SM: Glomerular basement membraneand lamina densa in infants and children: An ultrastructural evaluation.Pediatr Pathol 7:527-534, 1987

7. Basta-Jovanovic G, Venkataseshan VS, Gil J, et al: Morphometric anal-ysis of glomerular basement membranes (GBM) in thin basement mem-brane disease (TBMD). Clin Nephrol 33:110-114, 1990

8. Suh KS, Kim JO, Kang GH: Thin glomerular basement membrane dis-ease: Light microscopic and electron microscopic studies. J Korean MedSci 12:234-239, 1997

9. Saxena S, Davies DJ, Kirsner RLG: Thin basement membranes in min-imally abnormal glomeruli. J Clin Pathol 43:32-38, 1990

0. McLay ALC, Jackson R, Meyboom R, et al: Glomerular basement mem-brane thinning in adults: Clinicopathological correlations of a newdiagnostic approach. Nephrol Dial Transplant 7:191-199, 1992

1. Marquez B, Stavrou F, Zouvani I, et al: Thin glomerular basementmembranes in patients with hematuria and minimal change disease.Ultrastruct Pathol 23:149-156, 1999

2. Churg J, Bernstein J, Glassock RJ (eds): Renal Disease: Classificationand Atlas of Glomerular Diseases (ed 2). New York, Igaku-Shoin, 1995

3. Cosio FG, Falkenhain ME, Sedmak DD: Association of thin glomerularbasement membrane with other glomerulopathies. Kidney Int 46:471-474, 1994

4. Hebert LA, Betts JA, Sedmak DD, et al: Loin pain-hematuria syndromeassociated with thin glomerular basement membrane disease and hem-orrhage into renal tubules. Kidney Int 49:168-173, 1996

5. Nieuwhof CMG, de Heer F, de Leeuw P, et al: Thin GBM nephropathy:Premature glomerular obsolescence is associated with hypertensionand late onset renal failure. Kidney Int 51:1596-1601, 1997

6. Nogueira M, Cartwright J, Horn K, et al: Thin basement membranedisease with heavy proteinuria or nephrotic syndrome at presentation.Am J Kidney Dis 35:E15, 2000

7. Abt AB, Carroll LE, Mohler JH: Thin basement membrane disease andacute renal failure secondary to gross hematuria and tubular necrosis.Am J Kidney Dis 35:533-536, 2000

8. Lee HS, Pyo HJ, Koh HI: Acute renal failure associated with hematuriain IgA nephropathy. Am J Kidney Dis 12:236-239, 1988

9. Paller MS: Hemoglobin- and myoglobin-induced acute renal failure inrats: Role of iron in toxicity. Am J Physiol 255:F539-F544, 1988

0. Das AK, Pickett TM, Tungekar MF: Glomerular basement membranethickness—a comparison of two methods of measurement in patients withunexplained haematuria. Nephrol Dial Transplant 11:1256-1260, 1996

1. Dische FE: Measurement of glomerular basement membrane thicknessand its application to the diagnosis of thin-membrane nephropathy.Arch Pathol Lab Med 116:43-49, 1992

2. Collar JE, Ladva S, Cairns TDH, et al: Red cell traverse through thinglomerular basement membranes. Kidney Int 59:2069-2072, 2001

3. Buzza M, Wang YY, Dagher H, et al: COL4A4 mutation in thin base-ment membrane disease previously described in Alport syndrome. Kid-ney Int 60:480-483, 2001

4. Lemmink HH, Nillesen WN, Mochizuki T, et al: Benign familial hema-turia due to mutation of the type IV collagen alpha4 gene. J Clin Invest98:1114-1118, 1996

5. Buzza M, Wilson D, Savige J: Segregation of hematuria in thin basementmembrane disease with haplotypes at the loci for Alport syndrome.Kidney Int 59:1670-1676, 2001

6. Buzza N, Dagher H, Wang YY, et al: Mutations in the COL4A4 gene inthin basement membrane disease. Kidney Int 63:447-453, 2003

7. McCoy RC, Johnson HK, Stone WJ, et al: Absence of nephritogenicGBM antigen(s) in some patients with hereditary nephritis. Kidney Int21:642-652, 1982

8. Pettersson E, Tornroth T, Wieslander J: Abnormally thin glomerular

basement membrane and the Goodpasture epitope. Clin Nephrol33:105-109, 1990

9. Nakanishi K, Yoshikawa N, Iijima K, et al: Immunohistochemical studyof alpha 1-5 chains of type IV collagen in hereditary nephritis. KidneyInt 46:1413-1421, 1994

0. Heidet L, Cai Y, Guicharnaud L, et al: Glomerular expression of type IVcollagen chains in normal and x-linked Alport syndrome kidneys. Am JPathol 156:1901-1910, 2000

1. Pirson Y: Making the diagnosis of Alport’s syndrome. Kidney Int 56:760-775, 1999

2. Lajoie G: Approach to the diagnosis of thin basement membrane ne-phropathy in females with the use of antibodies to type IV collagen.Arch Pathol Lab Med 125:631-636, 2001

3. Liapis H, Gokden N, Hmiel P, et al: Histopathology, ultrastructure, andclinical phenotypes in thin glomerular basement membrane diseasevariants. Hum Pathol 33:836-845, 2002

4. Gunwar S, Ballester F, Noelken ME, et al: Glomerular basement mem-brane: Identification of a novel disulfide-cross-linked network of �3,�4, and �5 chains of type IV collagen and its implications for thepathogenesis of Alport syndrome. J Biol Chem 273:8767-8775, 1998

5. Kleppel MM, Fan WW, Cheong HI, et al: Evidence for separate net-works of classical and novel basement membrane collagen: Character-ization of �3(IV)-Alport antigen heterodimers. J Biol Chem 267:4137-4142, 1992

6. Zhou J, Reeders ST: The � chains of type IV collagen. Contrib Nephrol117:80-104, 1996

7. Gubler MC, Knebelmann B, Beziau A, et al: Autosomal recessive Alportsyndrome: Immunohistochemical study of type IV collagen chain dis-tribution. Kidney Int 47:1142-1147, 1995

8. Barsotti P, Muda AO, Mazzucco G, et al: Distribution of �-chains oftype IV collagen in glomerular basement membranes with ultrastruc-tural alterations suggestive of Alport syndrome. Nephrol Dial Trans-plant 16:945-952, 2001

9. Hill GS, Jenis EH, Goodloe S: The nonspecificity of the ultrastructuralalterations in hereditary nephritis with additional observations on be-nign familial hematuria. Lab Invest 31:516-532, 1974

0. Rumpelt HJ, Langer KH, Scharer K, et al: Split and extremely thinglomerular basement membranes in hereditary nephropathy (Alport’ssyndrome). Virchows Arch 364:225-233, 1974

1. Meleg-Smith S, Magliato S, Cheles M, et al: X-linked Alport syndromein females. Hum Pathol 29:404-408, 1998

2. Berthoux FC, Laurent B, Koller JM, et al: Primary IgA glomerulonephri-tis with thin glomerular basement membrane: A peculiar pathologicalmarker versus thin membrane nephropathy association. Contrib Neph-rol 111:1-7, 1995

3. Shigematsu H, Kobayashi Y, Tateno S, et al: Ultrastructural glomerularloop abnormalities in IgA nephritis. Nephron 30:1-7, 1982

4. Yoshikawa N, Yoshiara S, Yoshiya K, et al: Lysis of the glomerularbasement membrane in children with IgA nephropathy and Henoch-Shonlein nephritis. J Pathol 150:119-126, 1986

5. Monga G, Mazzucco G, Roccatello D: The association of IgA glomeru-lonephritis and thin glomerular basement membrane disease in a he-maturic patient: Light and electron microscopic and immunofluores-cence investigation. Am J Kidney Dis 18:409-412, 1991

6. Toth T, Naito I, Takebayashi S: Diffuse thin glomerular basement mem-brane in association with idiopathic membranous glomerulonephritis.Clin Nephrol 50:137-143, 1998

7. Kawasaki Y, Suzuki J, Nozawa R, et al: Thin basement membranedisease with membranoproliferative glomerulonephritis in a 13-yearold girl. Pediatr Int 44:321-323, 2002

8. Baranowska-Daca E, Choi YJ, Barrios R, et al: Nonlupus nephritides inpatients with systemic lupus erythematosus: A comprehensive clinico-pathologic study and review of the literature. Hum Pathol 32:1125-1135, 2001

9. Matsumae T, Fukusaki M, Sakata N, et al: Thin glomerular basementmembrane in diabetic patients with urinary abnormalities. Clin Neph-
rol 42:221-226, 1994

pathology of thin basement membrane...

Documents