Interlaboratory Comparison of Assessments of Alzheimer Disease-Related Lesions

Copyright @ 2006 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.

ORIGINAL ARTICLE

Interlaboratory Comparison of Assessments of AlzheimerDisease-Related Lesions: A Study of the

BrainNet Europe Consortium

Irina Alafuzoff, MD, PhD,* Maria Pikkarainen, PhD, Safa Al-Sarraj, FRCPath,* Thomas Arzberger, MD,

Jeanne Bell, MD,* Istvan Bodi, MD, Nenad Bogdanovic, MD, PhD,* Herbert Budka, MD,*

Orso Bugiani, MD, Isidro Ferrer, MD,* Ellen Gelpi, MD, Giorgio Giaccone, MD,*

Manuel B. Graeber, MD, PhD, Jean-Jacques Hauw, MD,* Wouter Kamphorst, MD, PhD,

Andrew King, MD, Nicolas Kopp, MD, Penelope Korkolopoulou, MD, PhD,

Gabor G. Kovacs, MD, PhD,* David Meyronet, MD,* Piero Parchi, MD, PhD,*

Efstratios Patsouris, MD, PhD,* Matthias Preusser, MD, Rivka Ravid, PhD,*

Wolfgang Roggendorf, MD, Danielle Seilhean, MD, PhD,* Nathalie Streichenberger, MD,

Dietmar R. Thal, MD, and Hans Kretzschmar, MD*

AbstractThis interlaboratory study evaluated the reproducibility of the

assessments of neuritic plaques and neurofibrillary tangles

(NFTs)Vthe hallmark lesions of Alzheimer diseaseVand compared

the staining between the BrainNet Europe centers. To reduce the

topography-related inconsistencies in assessments, we used a 2-mm

tissue microarray (TMA) technique. The TMA block included 42 core

samples taken from 21 paraffin blocks. The assessments were done on

Bielschowsky and Gallyas silver stains using an immunohistochem-

ical (IHC) method with antibodies directed to beta-amyloid (IHC/AA)

and hyperphosphorylated tau (IHC/HPtau). The staining quality and

the assessments differed between the participants, being most diverse

with Bielschowsky (good/acceptable stain in 53% of centers)

followed by Gallyas (good/acceptable stain in 57%) and IHC/AA

(good/acceptable stain in 71%). The most uniform staining quality

and assessment was obtained with the IHC/HPtau method (good/

acceptable stain in 94% of centers). The neuropathologic diagnostic

protocol (Consortium to Establish a Registry for Alzheimer Disease,

Braak and Braak, and the National Institute of Aging and Reagan

[NIA-Reagan] Institute) that was used significantly influenced the

agreement, being highest with NIA-Reagan (54%) recommendations.

This agreement was improved by visualization of NFTs using the

IHC/HPtau method. Therefore, the IHC/HPtau methodology to

visualize NFTs and neuropil threads should be considered as a method

of choice in a future diagnostic protocol for Alzheimer disease.

Key Words: Alzheimer disease, Neuropathologic diagnosis,

Immunohistochemistry, Hyperphosphorylated tau, Beta-amyloid,

Tissue microarray, Interlaboratory study.

INTRODUCTIONThe neuropathologic diagnosis of Alzheimer disease

(AD) is based on the detection and distribution of hallmark

J Neuropathol Exp Neurol � Volume 65, Number 8, August 2006740

J Neuropathol Exp NeurolCopyright � 2006 by the American Association of Neuropathologists, Inc.

Vol. 65, No. 8August 2006pp. 740Y757

From Departments of Neuroscience and Neurology (IA, MPi) and Pathology(IA), Kuopio University, Kuopio University Hospital, Kuopio, Finland;the Center for Neuropathology and Prion Research (TA, HK), MunchenLudwig-Maximilians-University, Munchen, Germany; the Department ofPathology (JB), Edinburgh University, Western General Hospital,Edinburgh, U.K; the Department of Geriatrics (NB), Karolinska Institutet,Huddinge, Sweden; the Department of Neuropathology (WB), Georg-August-Universitaet Goettingen, Goettingen, Germany; the Institute ofNeurology (HB, EG, MPr), Medical University of Vienna, Vienna,Austria; the Division of Neuropathology and Neurology (OB, GG),Istituto Nazionale Neurologico Carlo Besta, Milano, Italy; the Institut deNeuropathologia (IF), Universitat de Barcelona, Barcelona, Spain;Laboratoire de Neuropathologie Raymond Escourolle (J-JH, DS), Pitie-Salpetriere Hospital and Paris VI University, Paris, France; the Depart-ment of Clinical Neuropathology (SA-S, IB, AK), London Institute ofPsychiatry, London, U.K; The Netherlands Brain Bank (WK, RRa),Amsterdam, The Netherlands; the Department of Pathology (PK, EP),National and Capodistrian University of Athens, Athens, Greece; theDepartment of Neurological Sciences (PP), University of Bologna,Bologna, Italy; OPNI Neuropathology (GGK), National Institute ofPsychiatry and Neurology, Budapest, Hungary; the Institute for Neuro-pathology (DRT), University of Bonn, Bonn, Germany; the UniversityDepartment of Neuropathology (MBG), Imperial College London andHammersmith Hospitals Trust, London, U.K; the Center de PathologieEst-Neuropathologie (NK, DM, NS), Hospices civils de Lyon, Lyon,France; Pathologisches Institut (PR, WR), Abteilung Neuropathologie,der Universitat Wurzburg, Wurzburg, Germany; the Department of Psy-chiatry and Psychotherapy (PF), University of the Saarland, Homburg/Saar, Germany; Human Brain Tissue Bank Budapest, Department ofAnatomy, Histology and Embryology (MPa), Semmelweis University,Budapest, Hungary; the Department of Cell & Molecular Neuroscience(RRe), Imperial College London, London, U.K; and GABO:milliariummbH & Co. KG (AS), Munich, Germany.

Send correspondence and reprint requests to: Irina Alafuzoff, MD, PhD,Department of Neuroscience and Neurology, Section of Neuropa-thology, Kuopio University, P.O. Box 1627, Fin 70 211 Kuopio; E-mail:[email protected]

Team leaders of the BNE Consortium: Wolfgang Bruck, MD, Peter Falkai,MD, Miklos Palkovits, MD, Richard Reynolds, PhD, Peter Riederer,PhD, Ameli Schwalber, and all authors designated by an asterisk (*).

This work was supported by EU Grant FP6, BNEII No LSHM-CT-2004-503039.


lesions such as neuritic plaques (NPs) and neurofibrillarytangles (NFTs). In 1985, the Diagnosis of Alzheimer DiseaseResearch Workshop neuropathology panel stated that bothsenile plaques and NFTs were helpful diagnostic markers.This group recommended the use of Bielschowsky silverimpregnation as one possible staining technique (1). Variousstaining techniques used for senile plaques or NPs andNFTs were evaluated in 1980s, and the results indicatedthat the highest number of both NPs and NFTs weredetected using modified Bielschowsky silver impregnation(2, 3). Subsequently, in 1991, to produce more accurate andreliable neuropathologic criteria for AD, the Consortium toEstablish a Registry for Alzheimer Disease (CERAD)published more detailed instructions (4). This protocolrecommended the use of 6- to 8-Km-thick sections, silverstains such as the modified Bielschowsky, and a semi-quantitative assessment of neocortical NPs (4), and definedsenile plaques of neuritic type (NPs) as being those plaqueswith thickened silver-positive neurites. Also in 1991, Braakand Braak launched the staging of AD-related changes inwhich the emphasis was on the regional distribution ofGallyas silver-stained NFTs in 50-Km-thick sections (5). Sixyears later, in 1997, the consensus recommendations by the

National Institute of Aging and Reagan (NIA-Reagan)Institute Working Group recommended that assessment ofNPs and NFTs should be carried out according to both theCERAD protocol (4) and Braak and Braak staging (5) toestimate the likelihood that AD pathologic changes wouldunderlie the symptoms of dementia (6). It was also recom-mended by the NIA-Reagan protocol that, in addition tosilver-impregnation techniques (modified Bielschowsky orGallyas), specific immunohistochemical (IHC) stains shouldalso be used (6).

In summary, the likelihood that dementia is the resultof AD lesions is high, intermediate, or low based on thedetection and semiquantitative assessment of AD-relatedlesions in the postmortem brain. This emphasizes the im-portance of the staining used because it might be influencedby nonpathology-related conditions such as postmortem de-lay, fixation, embedding and technical practices, and thereliability in the quantification of lesions used by differentevaluators.

The BrainNet Europe (BNE) consortium includes 20centers. Brain banking for research purposes and neuro-pathologic diagnostic examination is carried out in 17 of thesecenters. It is imperative for a brain-banking consortium that

FIGURE 1. The flowchart delineates the logistics of this study. Work done by the 17 participating BrainNet Europe (BNE) centersand by the coordinating center is presented in their own panels. Arrows numbered 2, 4A, and 6 delineate shipment of the studymaterial between the centers. Results of this study with the reference to the corresponding tables and figures are given in thebottom panel. AD, Alzheimer disease; HPtau, hyperphosphorylated tau; IHC, immunohistochemical; TMA, tissue microarray.

J Neuropathol Exp Neurol � Volume 65, Number 8, August 2006 Interlaboratory Assessments of AD

� 2006 American Association of Neuropathologists, Inc. 741


the neuropathologic evaluations are comparable between thecenters. A previous pilot study done by the BNE consortiumassessed the comparability of the neuropathologic diagnosisof AD following the CERAD (4) and Braak and Braak (5)recommendations. The results of this pilot study indicatedpoor agreement in the evaluations among the participants(data not published). Therefore, this interlaboratory studywas designed to compare the reliability of the semiquanti-tative assessment of NPs and NFTs among the members ofthe BNE consortium using the detection methods recom-

mended by the NIA-Reagan (6), CERAD (4), and the Braakand Braak (5) protocols.

The emphasis of this study was on the quantification ofNPs and NFTs by different evaluators using a 2-mm tissuemicroarray (TMA) technique (7, 8). The TMA block includednumerous brain samples received from the 17 participatingBNE centers. This report summarizes the results obtainedwhen participating neuropathologists semiquantitativelyassessed AD-related lesions in 42 core samples, each measur-ing 2 mm in diameter, using 4 different staining techniques.

TABLE 1. Brain Samples in the Tissue Microarray Block

SampleBlock

CoreSample Gender

Age atDeath(years)

PostmortemDelay (hours) Fixative

FixationTime (days)

Embedding Medium/Highest Temperature

StorageTime(days)

1 1A, 1B Male 88 G12 Approximately 4%buffered formaldehyde

420 Histotec Merck/62-C 60

2 1C, 1D Female 91 62 Approximately 4%buffered formaldehyde

14 White paraffin/56-C 180

3 1E, 1F Female 86 24 Approximately 4%buffered formaldehyde

14 White paraffin/56-C 360

4 1G, 1H Male 71 30 Approximately 4%unbuffered formaldehyde

21 Tissue Tech3 paraffinwax/60-C

600

5 2C, 2D Female 74 not known 4.5% bufferedformaldehyde

30Y90 Histosec 1,11609Merck/61-C

1,170

6 2E, 2F Female 78 G12 4.5% bufferedformaldehyde

60 Histosec 1,11609Merck/61-C

240

7 2G, 2H Male 57 20 Approximately 4%buffered formaldehyde

30 Bioplast Plus/60-C 720

8 3A, 3B Male 81 3 4% bufferedparaformaldehyde

21 Panreac/57-C 7

9 3C, 3D Female 89 25 Approximately 4%unbuffered formaldehyde

21 Paraplast Plus/56-C 720

10 3E, 3F Female 71 15 Approximately 4%unbuffered formaldehyde

30 Paraplast Plus/56-C 1,440

11 3G, 3H Female 96 28 Approximately 4%buffered formaldehyde

240Y270 White/bees_/dentalwax mix/62-C

150

12 4A, 4B Female 88 5Y6 Approximately 4%buffered formaldehyde

30 Histowax/62-C 1,950

13 4C, 4D Male 75 24 Approximately 8%unbuffered formaldehyde

21 Paraplast/60-C 1

14 4E, 4F Male 69 18 Approximately 4%buffered formaldehyde

24 Paraplast Kendall/59-C

150

15 4G, 4H Male 83 7 Approximately 4%unbuffered formaldehyde

22 White paraffin mixedwith wax/56-C

210

16 5A, 5B Female 83 26 5% buffered formaldehyde 10 Paraplast Plus/60-C 210

17 5C, 5D Male 80 120 Approximately 4%buffered formaldehyde

7 Paraplast/60-C Few months

18 5E, 5F Male 68 15 4% bufferedparaformaldehyde

120 Lambwax/57Y58-C 720

19 6A, 6B Male 74 8 Approximately 4%unbufferedformaldehyde*

1 Bayers paraffin/54Y57-C

Few months

20 6C, 6D Female 87 9 Approximately 4%unbufferedformaldehyde andacetic acid*

2 Bayers paraffin/54Y57-C

1,440

21 6E, 6F Male 63 24 5% buffered formaldehyde 33 Histosec Merck/56Y58-C

1,440

*, Before fixation, formic acid treatment for 24 hours.

Alafuzoff et al J Neuropathol Exp Neurol � Volume 65, Number 8, August 2006

� 2006 American Association of Neuropathologists, Inc.742


MATERIALS AND METHODS

Tissue Microarray Block Constructionand Sectioning

Each participating BNE center (n = 17) was asked toprovide to the coordinating center a routinely processed,paraffin-embedded block of temporal cortex with various AD-related lesions as well as details of the postmortem delay, thefixatives used, and the storage time (Fig. 1; Table 1). Briefly,the male/female distribution was approximately equal. Theage at death ranged from 57 to 96 years. The postmortemdelay varied from 3 hours to 5 days and the fixation timefrom 1 day to 14 months. The maximum temperature of the

embedding medium ranged from 54-C to 62-C and thestorage duration for blocks ranged from 1 day to 5.5 years.

Paraffin blocks (at least 3-mm-thick) from a total of 21cases were obtained for the construction of the TMA block(Table 1), which was constructed as described elsewhere (9).Briefly, to give a wide range of detectable lesions, 2 coresamples were taken from each case, one from the sulcal andthe other from gyral grey matter (n = 42) (Fig. 2). Thecore samples were taken using a Manual Tissue Arrayer 1instrument (Beecher Instruments, Inc., Sun Prairie, WI).Representative samples were obtained using a 2.0-mm-diameter needle. The core samples were placed into therecipient TMA block. Serial 7-Km-thick sections were cut

FIGURE 2. The flowchart summa-rizes the construction of the tissuemicroarray (TMA)block, recommendedstainings, the lesions to be rated, andthe assessment. (A) Paraffin sectionsfrom tissue are first cut and stainedwith hematoxylin and eosin andregions of interest are marked onthese slides (circles). (B) Serial TMAsections are cut and identifications ofthose core samples included and ex-cluded into the assessments are illus-trated. (C) 1) The staining methodsused. 2) Optimal staining results: (A)Bielschowsky staining (original mag-nification: 400�) with neuritic pla-ques (arrow) where the dark brownto black thickened neurites (arrow-head) are seen; (B) Gallyas staining(original magnification: 400�) whereblack neurofibrillary tangles (arrow)and neuropil threads or neurites(arrowhead) are seen; (C) immuno-histochemical (IHC) staining of beta-amyloid (IHC/AA) -labeled proteinaggregates, i.e. plaques (arrow); and(D) hyperphosphorylated tau (IHC/HPtau) labeled neurofibrillary tangles(arrow) and neuropil threads/neurites(arrowhead). 3) Instructions for assess-ment of the lesions.




from the TMA block with a rotating microtome without theuse of supportive methods such as a tape-transfer system(8, 10). The serial sections were placed on commercial Super-Frost Plus microscope slides and dried overnight.

BNE Participant EffortsEach BNE participant received 5 consecutive slides to

be stained by modified Bielschowsky (2, 11), Gallyas silver-impregnation techniques (12), and IHC methodology using

TABLE 2. Immunohistochemical Stainings

Detectionof

Mode ofStaining Pretreatments

Antibodies

AssessmentCode (quality)Type Clone Code Company* Dilution

IncubationTime

AA Manual 98% FA 2 minutes Mono 6F/3D M0872 Dako 1:50 +4-C overnight

1 (G)

Manual 98% FA 5 minutes Mono 6F/3D M0872 Dako 1:100 +4-C overnight

2 (G)


3 (G)

Manual 96% FA 2 minutes Mono 6F/3D M0872 Dako 1:50 RT 30minutes

4 (A)


5 (G)


6 (P)


7 (A)

Manual 90% FA 40 minutes Mono 6F/3D M0872 Dako 1:100 RT 30minutes

8 (A)


9 (P)

Manual 80% FA 6 hours Mono 6F/3D M0872 Dako 1:100 +5-C overnight

10 (G), 11 (P),12 (G)

Manual Microwave in CB Mono 6F/3D M0872 Dako 1:25 +4-C overnight

13 (P)

Automatic Cell conditioning solution† Mono 6F/3D M0872 Dako 1:20 RT 30minutes

14 (P)

Manual 98% FA 3 minutes Mono 6F/3D NCL-B-Amyloid

Novocastra 1:100 +5-C overnight

15 (A)

Automatic 98% FA 2 minutes and autoclave100-C 10 minutes in CB

Mono 6F/3D NCL-B-Amyloid

Novocastra 1:200 RT 60minutes

16 (G)

Manual 80% FA 6 hours Poly V 44Y136 Biosource 1:1000 RT overnight

17 (G)

HPtau Manual None Mono AT8 BrY3‡ Innogenetics 1:30 RT 30minutes

14 (G)

Manual None Mono AT8 BrY3‡ Innogenetics 1:50 +5-C overnight

5 (G)

Manual None Mono AT8 BrY3‡ Innogenetics 1:200 RT 30minutes

8 (G), 4 (G)


2 (G)

Manual None Mono AT8 BrY3‡ Innogenetics 1:300 RT overnight

17 (G)


7 (A)


10 (G), 11 (A),12 (G)


9 (A)

Manual Microwave in CB Mono AT8 BrY3‡ Innogenetics 1:200 +4-C overnight

1 (G)

Manual Microwave in CB Mono AT8 BrY3‡ Innogenetics 1:800 +4-C overnight

13 (A)

Automatic None Mono AT8 BrY3‡ Innogenetics 1:500 +5-C overnight

15 (G)




antibodies directed to beta-amyloid (AA) and hyperphos-phorylated tau (HPtau) (Figs. 1, 2). BNE participants alsoreceived the protocol with recommended staining practicesand data sheets for recording assessments (Fig. 2). Accord-ing to the assessment instructions, Bielschowsky silver stainwas to be used only for assessment of NPs (i.e. plaques withdark brown to black silver-positive thickened neurites) (4).Gallyas silver stain was to be used for assessment of NFTsand NTs (neuropil threads/neurites), and IHC/AA stainingwas to be used for assessment of protein aggregates (i.e.plaques) and IHC/HPtau for NFTs and NTs. For theassessment, each detectable lesion (regardless of the size)within the core sample of 2 mm in diameter was to becounted. If the core sample was only partly grey matter, thishad to be taken into account; that is, if there were 4 lesionsin a core sample but only 50% of the core was grey matter,

the estimated count was twice as high, or 8 lesions. Thecounts were reported by the participating neuropathologistsas semiquantitative scores on a 5-step scale (Fig. 2).

Coordinating Center EffortsThe coordinating center assessed the comparability of

the sections, ensuring that the shipped sections were alike.Every 20th slide of the serial TMA sections (n = 6) was IHC/HPtau-stained, inspected, and assessed (Fig. 1).

The slides stained by the participating BNE centers, datasheets, and staining protocols were collected and analyzed(Fig. 1). The IHC methods are listed and the results of theassessments are identified based on the assessment code inTable 2. The missing core samples were calculated from thestained TMA sections by light microscopy at 25� magnifi-cation. In addition, the damage in the core samples was

FIGURE 3. Bielschowsky silver stain. Thesame core sample (2G) in (AYC) originalmagnification 50� (scale bar = 100 Km)and in (DYF) original magnification400� (scale bar = 10 Km), assessmentcodes 10, 17, and 12, respectively. Notethe difference in visualization of neuriticplaques. (GYI) demonstrate the stainingof 3 different core samples (3A, 1C, and6D, respectively) within the same tissuemicroarray section, assessment code 10(scale bar = 100 Km).

TABLE 2. (continued)

Automatic 99% FA 5 minutes Mono AT8 BrY3‡ Innogenetics 1:500 RT 30minutes

6 (P)

Automatic Autoclave 100- C 10 minutes inCB

Mono AT8 BrY3‡ Innogenetics 1:200 RT 60minutes

16 (G)

Manual 98% FA 10 minutes Poly V A0024 Dako 1:3000 4- C overnight

3 (A)

*, Innogenetics, Ghent, Belgium; DakoCytomation, Glostrup, Denmark; Biosource, Camarillo, CA; Novocastra Laboratories Ltd., Newcastle, U.K.†, Pretreatment buffer used on Ventana Benchmark staining (Ventana Medical Systems, Inc., Tucson, AZ).‡, New company and (code) for AT8 antibody: Endogen (90206) or Pierce Biotechnology, Inc., Rockford, IL (MN1020).A, acceptable; AA, beta-amyloid; CB, citrate buffer; FA, formic acid; G, good; HPtau, hyperphosphorylated tau; P, poor; RT, room temperature.



Copyright @

2006 by the Am

erican Association of N

europathologists, Inc. Unauthorized reproduction of this article is prohibited.

TABLE 3. Silver Stainings: Assessments of Alzheimer Disease-Related Hallmark Lesions at the Tissue Microarray Section Level

Bielschowsky Silver Impregnation/Neuritic Plaques Gallyas Silver Impregnation/Neurofibrillary Tangles

AssessmentCode*

Qualityof Staining

PrimaryAssessment†(mean T

standard error)

Reassessment†(mean T

standard error)WilcoxonTest‡

AbsoluteAgreement§

(%)Assessment

Code*Quality

of Staining


standard error)


standard error)Wilcoxon

Test‡

AbsoluteAgreement§

(%)

6 G 3.3 T 0.3 3.1 T 0.3 84 1 G 2.7 T 0.3 2.3 T 0.3 50

4 G 3.0 T 0.3 3.1 T 0.3 81 11 G 2.6 T 0.3 2.0 T 0.3 * 55

10 G 2.6 T 0.3 3.1 T 0.3 * 61 5 G 2.3 T 0.3 2.4 T 0.3 71

16 G 2.5 T 0.3 3.0 T 0.3 * 52 16 G 2.3 T 0.3 2.6 T 0.3 63

2 G 2.3 T 0.3 2.9 T 0.3 * 66 17 G 2.3 T 0.3 1.5 T 0.3 * 50

17 A 2.8 T 0.3 2.4 T 0.3 * 83 2 G 1.9 T 0.3 1.8 T 0.3 68

3 A 2.6 T 0.4 1.9 T 0.3 68 12 A 2.1 T 0.2 1.9 T 0.3 23

9 A 2.3 T 0.3 2.5 T 0.3 * 50 8 A 1.5 T 0.3 1.4 T 0.3 71

12 P 3.6 T 0.2 0.0 T 0.0 * 6 14 P 2.6 T 0.3 1.7 T 0.3 * 63

8 P 2.7 T 0.3 2.5 T 0.3 73 6 P 2.3 T 0.3 0.7 T 0.2 * 26

1 P 2.6 T 0.3 1.2 T 0.3 * 34 13 P 2.0 T 0.2 1.8 T 0.3 56

11 P 2.3 T 0.3 2.1 T 0.4 80 10 P 1.8 T 0.3 2.0 T 0.3 72

14 P 2.2 T 0.3 1.4 T 0.3 * 61 4 P 1.4 T 0.3 1.3 T 0.3 74

13 P 1.6 T 0.3 1.1 T 0.3 * 50 15 P 1.4 T 0.3 1.4 T 0.3 55

5 P 1.2 T 0.5 2.0 T 0.4 * 55

K-W test/allstainingsk

p = 0.00 (15) p = 0.00 (15) K-W test/allstainingsk

p = 0.01 (14) p = 0.00 (14)

K-W test/good andacceptablestainingsk

p = 0.05 (8) p = 0.02 (8) K-W test/good andacceptablestainingsk

p = 0.13 (8) p = 0.07 (8)

*, See Table 2.†, Mean values of semiquantitative scores of lesions T standard error of mean in the primary assessment (done by the BrainNet Europe participants) and reassessment (done by the coordinating center) of lesions in the tissue

microarray section, including 32 core samples.‡, For statistics of the comparison of results between primary and reassessments, Wilcoxon signed rank test was used (p G 0.05).§, Absolute agreement between primary and reassessments (%) i.e. the proportion of core samples assessed equally in the primary and reassessments.k, For statistics of the comparison of results between primary assessments and between reassessments, Kruskal-Wallis test was used (p G 0.05). Number of cases in parentheses.A, acceptable; G, good; K-W, Kruskal-Wallis; P, poor.

Alafuzoffet

al

JNeuropatholExp

Neurol�

Volume65,Number8,August

2006

�2006America

nAsso

ciatio

nofNeuropathologists,

Inc.

746

Copyright @

2006 by the Am



TABLE 4. Silver-Stained Tissue Microarray Sections of Good/Acceptable Staining Quality: Assessments of Alzheimer Disease-Related Hallmark Lesions in Core Samples Consisted Mostly of

Grey Matter (75%Y100%)

Bielschowsky Silver/Neuritic Plaques Gallyas Silver/Neurofibrillary Tangles Gallyas Silver/Neuropil Threads/Neurites

Core

Sample

Number

of Assessments Range*

Most Frequent

Score†Agreement‡

(%)

Absolute

Agreement§

(%)

Number

of

Assessments Range*

Most

Frequent


(%)

Absolute

Agreement§

(%)

Number

of

Assessments Range¬

Most

Frequent


(%)

Absolute

Agreement§

(%)

1A 8 3Y4 4 88 88 8 2Y4 3 50 25 7 0Y1 1 86 86

1D 8 1Y4 1, 2 38 25 8 1Y4 4 50 50 7 0Y1 1 86 86

1F 8 3Y4 4 88 88 8 4 4 100 100 7 1 1 100 100

1H 8 0Y4 1Y3 25 25 8 3Y4 4 75 63 7 1 1 100 100

2H 8 4 4 100 100 7 2Y4 4 71 14 6 1 1 100 100

3A 7 0Y4 2Y4 25 50 8 1Y3 1 50 50 7 0Y1 1 57 57

3E 8 0 0 100 88 8 0Y1 0 88 88 6 0 0 100 83

4A 8 1Y4 4 38 38 7 0Y4 1 43 29 6 0Y1 1 83 67

5A 7 4 4 100 100 8 0Y4 4 75 63 7 1 1 100 100

5B 8 3Y4 4 88 88 8 0Y4 4 75 75 7 0Y1 1 86 86

5E 7 3Y4 4 71 71 7 0Y3 1, 3 43 57 6 1 1 100 100

6D 6 0Y4 0, 4 33 17 7 0Y1 0 86 86 5 0Y1 1 60 40

6F 8 2Y4 4 75 63 8 3Y4 4 75 75 7 1 1 100 100

*, Range of scores (0Y4) within the primary assessments.†, The most frequent score within primary assessments.‡, Proportion of the most frequent score value within the primary assessments of each individual core sample.§, The proportion of equally assessed core samples between primary and reassessments.k, Range of scores (yes/no) within the primary assessments.

JNeuropatholExp

Neurol�


2006

Interla

boratory

Assessm

ents

ofAD

�2006America

nAsso

ciatio


Inc.

747


assessed. To enable the comparison of the staining results,only those core samples were included in which Q75% of thetissue was available for analysis (Fig. 1). In addition to loss ordamage of cores during the staining procedures, some coresamples were excluded because there was not enoughmaterial in the recipient block. In total, assessments of 10 ofthe initial 42 core samples (3D, 3G, 4E, 4G, 4H, 5C, 5D, 6A,6C, and 6E) were excluded (Fig. 2).

The quality of staining of most of the core samples in aTMA section was estimated on a 3-step scale (good, accept-able, or poor) under light microscopy at 25� to 100� originalmagnification. Briefly, the quality of staining was assessed asgood when the staining clearly labeled the lesion in most ofthe core samples in a TMA section (Fig. 2). The staining wasacceptable when the lesions were detectable, but countingwas laborious as a result of partial staining of the lesions orexcessive background staining. The staining was assessed aspoor when the lesions were not at all detectable or they weredetectable but not stained as required (Fig. 2).

Reassessment of all TMA sections for each stainingwas carried out in one sitting by 2 evaluators following thegiven assessment instructions (Fig. 2). During the reassess-ment, strict criteria were applied and only those lesions thatfulfilled the required staining criteria were included. Thereassessment was carried out in the following order:Bielschowsky, Gallyas, IHC/AA, and lastly IHC/HPtau witha few days between the viewings.

The results are reported as mean values of the semi-quantitative scores of lesions of the 32 core samples in a TMAsection (at the TMA section level) for both the primary (eachBNE participant) and the reassessments (coordinating center).Moreover, to analyze the variability in the assessmentswithout the interference of staining quality, only TMA

sections having good or acceptable staining quality wereselected in a core-to-core comparison (at the core samplelevel). However, because the nature of the core sample (i.e.grey matter/white matter) might have influenced the results,only those core samples that were constructed mostly ofgrey matter (75Y100%) were included in this core-to-coreassessment.

Additionally, to evaluate the impact of the assessmentsof various lesions (NPs and NFTs) on the neuropathologicdiagnosis of AD, a comparison of diagnoses using differentprotocols (CERAD, Braak and Braak, NIA-Reagan) wascarried out on selected representative core samples withgood staining quality as described previously (4Y6).

Statistical Analysis and PhotographyFor statistical analyses, SPSS program for Windows

(version 11.5) was used. The statistical difference incomparability of the shipped sections was estimated by thenonparametric Kruskal-Wallis (K-W) test. The statisticaldifference in both primary assessments and in reassessmentsat the TMA section level was estimated by the nonpara-metric K-W test. The agreement in the assessments oflesions between the primary assessment and the reassess-ment was estimated using the nonparametric Wilcoxon test(scores ranging from 0Y4) or Fisher exact test (scoresranging from 0Y1). In addition, the value of absoluteagreement (%) was calculated, that is, the proportion ofcore samples assessed equally in the primary and reassess-ments. At the core sample level, the proportion of the mostfrequent score was calculated and given as a percentage ofagreement. In addition, a value for the absolute agreement(%) between the primary and reassessments was calcu-lated. Digital images were taken using a Leica DM4000 B

FIGURE 4. Gallyas silver stain. Core sample(F1) assessment code 1 (A, C) and 10 (B,D). Original magnifications: 50� and 400�,respectively. Scale bars = 100 Km and 10Km, respectively. (B, D) Note the difficultyin assessing neurofibrillary tangles and neu-ropil threads in the overstained section.



Copyright @

2006 by the Am



TABLE 5. Assessments of Neuropil Threads/Neurites in Silver and Immunostainings at the Tissue Microarray Section Level

Gallyas Silver Impregnation/Neuropil Threads/Neurites IHC/Hyperphosphorylated Tau/Neuropil Threads/Neurites

AssessmentCode*

Quality ofStaining

PrimaryAssessment†Yes (%)

Reassessment†Yes (%)

FisherExact Test‡

AbsoluteAgreement§ (%)

AssessmentCode*

Quality ofStaining

PrimaryAssessment†Yes (%)

Reassessment†Yes (%)

FisherExact Test‡

AbsoluteAgreement§ (%)

5 G 87 71 91 15 G 94 88 * 94

11 G 86 69 * 83 4 G 93 86 * 93

1 G 81 75 * 81 1 G 91 97 94

2 G 68 71 * 77 10 G 90 90 * 94

17 G 37 68 * 67 17 G 88 88 * 100

8 A 71 72 * 77 14 G 88 88 * 100

12 A 61 65 * 77 16 G 88 91 * 97

13 P 88 66 * 78 12 G 88 90 * 94

14 P 83 60 * 77 2 G 87 87 * 94

6 P 81 45 * 65 5 G 84 88 * 97

15 P 81 42 * 61 8 G 80 87 * 93

10 P 75 88 * 88 11 A 90 83 * 93

4 P 65 65 * 87 13 A 88 84 * 97

7 A 87 84 * 97

9 A 85 85 * 100

3 A 67 68 63

6 P 47 72 * 75

*, See Table 2.†, Proportion of core samples with positive (yes) assessments in the primary assessments (done by BrainNet Europe participants) and in the reassessments (done by the coordinating center) of lesions in the tissue microarray

section, including 32 core samples.‡, Agreement between the primary and reassessments is presented by Fisher exact test (p G 0.05).§, Absolute agreement between primary and reassessments (%) i.e. the proportion of core samples assessed equally in the primary and reassessments.A, acceptable; G, good; IHC, immunohistochemistry; P, poor.

JNeuropatholExp

Neurol�


2006

Interla

boratory

Assessm

ents

ofAD

�2006America

nAsso

ciatio


Inc.

749


microscope equipped with a Leica DFC 320 digital camera(Leica Microsystems Wetzlar Ltd., Heerbrugg, Germany).

RESULTSThe semiquantitative assessments of IHC/HPtau-

labeled NFTs in the 6 stained consecutive TMA sectionsdid not differ significantly (K-W test, p = 0.9); mean values(T standard error) ranged from 3.2 T 0.2 to 2.8 T 0.3,indicating that the number of lesions did not change in asignificant manner over the depth of the tissue.

Bielschowsky Silver ImpregnationDetails regarding the staining results are given in

Figure 3 and Table 3. Bielschowsky staining was done in 15of 17 centers. The staining intensity of both the lesions andthe background varied extensively not only between dif-ferent BNE centers, but also within a single TMA section.According to the strict assessment criteria, the staining ofmost of the core samples in a TMA section was assessed asbeing good or acceptable in 8 of 15 stainings (53%). Onlyonce did the Bielschowsky method stain all core sampleswithin the TMA section evenly. Core samples 1C, 1D, 6B, and6D were usually understained, whereas core samples 3A, 3B,4C, and 4D were usually overstained. During the reassessment,it was noted that the NPs in each core sample were diffuselydispersed and no overlapping or confluent lesions were noted.

At the TMA section level, when all assessments wereincluded (Table 3), the primary assessment and the reassessment

of NPs differed significantly (K-W test, p G 0.05), with meanvalues (T standard error) ranging from 3.6 T 0.2 to 1.2 T 0.5 andfrom 3.1 T 0.3 to 0, respectively. However, when only the good/acceptable stainings were included, the primary assessmentsdid not differ significantly (K-W test, p = 0.05), whereas thedifference in the reassessments was still significant (K-W test,p = 0.02). The reassessments differed significantly (Wilcoxon,p G 0.05) from the primary assessments in 63% of the good/acceptable stainings (5 of 8). Finally, absolute agreementbetween the primary and reassessment in the good/acceptablestainings ranged from 50% to 84%, with good staining results(Q75%) in 3 of 8 assessments.

The core-to-core comparison of the good/acceptablestainings (Table 4), including only the 13 core samplescomposed mostly of grey matter (75Y100%), revealed thatthere was still some variability in the assessments of NPs. Theassessment ranged from zero to the highest score value of 4 in3 of 13 core samples (23%). One of these core samples wasrepeatedly overstained (3A) and the other understained (6D).The score value of 4 was the most frequent assessment in62% of the core samples (8 of 13). The agreement in theassessments of these 13 core samples ranged from 25% to100%, being good (Q75%) in 7 of the core samples (54%). Itis notable that in 38% of the core samples (5 of 13), theagreement was less than 50%. The absolute agreementbetween the primary and reassessments for the selected 13core samples ranged from 17% to 100%, being good (Q75%)in only 46% of the assessments (6 of 13).

FIGURE 5. Immunohistochemistry/AAstaining. The same core sample (1A) in(AYC) original magnification: 50�(scale bar = 100 Km) and in (DYF),original magnification: 400� (scale bar= 10 Km) stained used an immunohis-tochemical technique with antibodiesdirected to beta-amyloid (AA), assess-ment codes 17, 15, and 4, respectively.Note the difference in labeling of theprotein. (GYI) Staining of 3 core sam-ples within the same tissue microarraysection, assessment code 4. Originalmagnification in panels (G, H) (coresamples 3D and 5B, respectively) is50�; scale bar = 100 Km. In panel (I)(1G), original magnification: 400�;scale bar = 10 Km. Note the variablelabeling of the lesions in different coresand within the same core.




Gallyas Silver ImpregnationDetails regarding the staining results are seen in Figure 4

and Table 3. Gallyas silver impregnation was carried out by 14of the 17 BNE centers. Within most of the TMA sections, thestaining of both lesions and the background in core sampleswas rather uniform. However, when comparing sections stainedby different BNE centers, the staining of lesions and thebackground varied. According to the strict assessment criteria,the Gallyas staining (i.e. staining of most of the core samplesin a TMA section) was good or acceptable in 57% of the cases(8 of 14). The poor stainings were either too dark or too lightwith almost no detectable staining. Another common problemwith Gallyas staining was the high amount of grainy silverprecipitate diffusely spread over the core sample.

At the TMA section level, when all assessments wereincluded (Table 3), both the primary and reassessments ofGallyas silver-stained NFTs differed significantly (K-W test,p G 0.05), with the mean values (T standard error) ranging from2.7 T 0.3 to 1.4 T 0.3 and from 2.6 T 0.3 to 0.7 T 0.2, respec-tively. However, when only good/acceptable stainings wereevaluated, there were no significant differences between theprimaries or the reassessments (K-W test, p 9 0.05). Compar-ison of the primary and reassessments of NFTs indicated onlyminimal differences and no statistically significant differences(Wilcoxon, p 9 0.05) in most of the assessments (6 of 8); how-ever, the absolute agreement still ranged from 23% to 71%.

The core-to-core comparison of good/acceptable stain-ings (Table 4), including only the 13 core samples composedmostly of grey matter (75Y100%), revealed that theevaluators still disagreed in the assessment of NFTs. In23% of the selected core samples (3 of 13), the primaryassessments ranged from zero to the highest score value of 4.The score value of 4 was the most frequent assessment in54% of the core samples (7 of 13). The agreement betweenprimary assessments ranged from 43% to 100%, being good(Q75%) in 54% of the core samples (7 of 13). Finally, theabsolute agreement between the primary and reassessmentsof NFTs ranged from 14% to 100%, being good (Q75%) inonly 38% of the selected core samples (5 of 13). Moreover,in 15% of these core samples (2 of 13), the absolute agree-ment was no better than 25%.

The assessment of Gallyas silver-stained NTs wascarried out by 13 of 17 centers (Table 5). At the TMA sectionlevel, when all 13 stainings were included, NTs were seen in37% to 88% of the core samples in primary assessments fromand in 42% to 88% in reassessments. When only good/acceptable stainings (7 of 13) were evaluated, the range wassimilar (37Y87% and 65Y75%, respectively). In most of thesecases, the primary assessments were significantly related tothe reassessments (Fisher exact test, p G 0.05). The absoluteagreement between primary and reassessments, when onlygood/acceptable stainings were included, ranged from 67% to91%, being good (Q75%) in most of the assessments (6 of 7).

The core-to-core comparison of good/acceptable stain-ings (Table 4), including only the 13 core samples composedmostly of grey matter (75Y100%), revealed that in 46% of thecore samples (6 of 13), the evaluators either noted NTs or didnot detect them at all. The agreement within the assessments

ranged from 57% to 100%, being good (975%) in 11 of 13core samples (85%). Similarly, the absolute agreementbetween the primary and reassessments ranged from 40% to100%, being good (Q75%) in 10 of 13 core samples (77%).

Immunohistochemistry/Beta-AmyloidDetails regarding the IHC/AA staining results are given

in Figure 5 and Table 6. IHC/AA staining was carried out byall BNE centers (n = 17). There was considerable variation inthe staining protocols, for example, 3 different antibodies and15 different pretreatment and dilution combinations wereused (Table 2). According to the strict assessment criteria, theIHC/AA staining of most of the core samples in a TMAsection was good or acceptable in 71% of the stainings (12 of17). In 3 of the good stainings (assessment codes 2, 12, and17 in Table 2), the intensity of the lesions was quite uniformbetween various core samples within the TMA section. In theother good stainings, some of the core samples were lessintensively stained, but the AA-labeled aggregates were stilldetectable at the 100�magnification. Three of the participantsfailed completely with the staining. During the reassessment,it was noted that no confluent stained aggregates were noticedand all aggregates were larger than a single cell.

At the TMA section level (Table 6), both primary andreassessments of IHC/AA-labeled plaques differed signifi-cantly (K-W test, p G 0.05), with the mean values (T standarderror) ranging from 3.8 T 0.1 to 0.6 T 0.1 and from 3.7 T 0.2to 0.2 T 0.1, respectively. When only good/acceptablestainings were included, the primary and reassessments stilldiffered significantly (K-W test, p G 0.05), with the meanvalues (T standard error) ranging from 3.8 T 0.1 to 1.6 T 0.2and from 3.7 T 0.2 to 2.2 T 0.3, respectively. When comparingthe primary with the reassessments including only good/acceptable stainings, the results differed significantly in 5of 12 assessments (Wilcoxon, p G 0.05). Furthermore, theabsolute agreement between the primary and the reassess-ments ranged from 36% to 100%, being good (Q75%) in only6 of 12 cases (50%).

The core-to-core comparison of good/acceptable stain-ings (Table 7), including only the 13 core samples composedmostly of grey matter (75Y100%), revealed that the primaryassessments of IHC/AA-labeled plaques still varied. In 23% ofthese selected core samples (3 of 13), the assessments rangedfrom zero to the highest score value of 4, whereas the scorevalue of 4 was the most frequent assessment in 77% of thecore samples (10 of 13). The agreement in the primary as-sessments was good (975%) in 8 of 13 core samples (62%).The absolute agreement between primary and reassessmentsof IHC/AA aggregates, in turn, ranged from 55% to 92% andwas good (Q75%) in 62% of the core samples (8 of 13).

Immunohistochemistry/Hyperphosphorylated Tau

All BNE centers carried out the IHC/HPtau staining(Fig. 6; Table 6). Again, there was considerable variation inthe staining protocols; for example, 2 antibodies were usedand some of the participants used some kind of pretreatment,although most centers did not (Table 2). According to strict



Copyright @

2006 by the Am



TABLE 6. Immunohistochemical Stainings: Assessments of Alzheimer Disease-Related Hallmark Lesions at the Tissue Microarray Section Level

Immunohistochemistry/Beta-Amyloid/Plaques Immunohistochemistry/Hyperphosphorylated Tau/Neurofibrillary Tangles

AssessmentCode*

Quality ofStaining


standard error)



AbsoluteAgreement§

(%)Assessment

Code*Quality ofStaining


standard error)



AbsoluteAgreement§

(%)

17 G 3.8 T 0.1 3.7 T 0.2 100 12 G 3.4 T 0.2 3.0 T 0.3 * 71

2 G 3.4 T 0.2 3.4 T 0.2 90 10 G 3.3 T 0.2 3.3 T 0.2 93

12 G 3.4 T 0.2 3.5 T 0.2 84 16 G 3.3 T 0.2 3.4 T 0.2 88

1 G 3.3 T 0.2 3.1 T 0.3 * 83 17 G 3.3 T 0.2 3.2 T 0.3 88

10 G 3.3 T 0.2 3.5 T 0.2 * 73 2 G 3.2 T 0.3 3.2 T 0.3 100

16 G 3.0 T 0.2 2.9 T 0.2 87 14 G 3.2 T 0.3 3.2 T 0.3 91

5 G 3.0 T 0.2 3.3 T 0.2 * 63 1 G 3.2 T 0.3 3.1 T 0.3 88

3 G 2.7 T 0.2 3.2 T 0.2 * 60 4 G 3.2 T 0.3 3.3 T 0.3 86

15 A 3.6 T 0.2 3.5 T 0.2 87 5 G 3.1 T 0.3 3.3 T 0.3 72

7 A 3.3 T 0.2 2.8 T 0.3 59 8 G 3.0 T 0.3 3.0 T 0.3 83

4 A 2.8 T 0.2 2.8 T 0.2 66 15 G 3.0 T 0.3 3.1 T 0.3 72

8 A 1.6 T 0.2 2.2 T 0.3 * 36 7 A 3.2 T 0.3 3.2 T 0.3 83

11 P 3.5 T 0.2 2.6 T 0.3 * 41 9 A 3.2 T 0.3 2.6 T 0.4 * 78

6 P 3.2 T 0.2 2.9 T 0.2 * 69 3 A 3.1 T 0.2 2.7 T 0.3 * 63

14 P 1.2 T 0.3 0.4 T 0.2 * 47 11 A 2.9 T 0.3 2.6 T 0.3 69

9 P 0.9 T 0.3 0.6 T 0.2 69 13 A 2.7 T 0.3 2.9 T 0.3 * 66

13 P 0.6 T 0.1 0.2 T 0.1 * 59 6 P 2.5 T 0.3 2.0 T 0.3 * 50

K-W test/allstainingsk

p = 0.00 (17) p = 0.00 (17) K-W test/allstainingsk

p = 0.04 (17) p = 0.00 (17)

K-W test/good andacceptablestainingsk

p = 0.00 (12) p = 0.00 (12) K-W test/good andacceptablestainingsk

p = 0.24 (16) p = 0.26 (16)

*, See Table 2.†, Mean values of semiquantitative scores of lesions T standard error of mean in the primary assessment (done by the BrainNet Europe participants) and reassessment (done by the coordinating center) of lesions in the tissue

microarray section, including 32 core samples.‡, For statistics of the comparison of results between primary and reassessments, Wilcoxon signed rank test was used (p G 0.05).§, Absolute agreement between primary and reassessments (%) i.e. the proportion of core samples assessed equally in the primary and reassessments.k, For statistics of the comparison of results between primary assessments and between reassessments, Kruskal-Wallis test was used (p G 0.05). Number of cases in parentheses.A, acceptable; G, good; K-W, Kruskal-Wallis; P, poor.

Alafuzoffet

al

JNeuropatholExp

Neurol�


2006

�2006America

nAsso

ciatio


Inc.

752

Copyright @

2006 by the Am



TABLE 7. Immunostained Tissue Microarray Sections of Good/Acceptable Staining Quality: Assessments of Alzheimer Disease-Related Hallmark Lesions From Core Samples Consisted

Mostly of Grey Matter (75Y100%)

Core

Sample

Immunohistochemistry/Beta-Amyloid/Plaques

Immunohistochemistry/Hyperphosphorylated Tau/

Neurofibrillary Tangles

Immunohistochemistry/Hyperphosphorylated Tau/Neuropil

Threads/Neurites

Number of

Assessments Range*

Most Frequent


(%)

Absolute

Agreement§

(%)

Number

of

Assessments Range*

Most

Frequent


(%)

Absolute

Agreement§

(%)

Number

of

Assessments Range¬

Most

Frequent


(%)

Absolute

Agreement§

(%)

1A 10 1Y4 4 80 80 15 2Y4 4 93 87 15 1 1 100 100

1D 12 1Y4 4 83 92 16 3Y4 4 94 94 16 0Y1 1 94 94

1F 12 1Y4 4 75 67 16 4 4 100 100 16 1 1 100 100

1H 11 1Y4 4 64 55 16 3Y4 4 94 94 16 1 1 100 100

2H 12 3Y4 3, 4 50 58 16 3Y4 4 94 94 16 1 1 100 100

3A 12 1Y4 1, 3 33 67 15 3Y4 4 73 87 15 1 1 100 100

3E 11 0Y4 1 27 73 16 0 0 100 100 16 0Y1 0 81 69

4A 12 1Y4 4 92 92 15 3Y4 4 87 80 15 1 1 100 100

5A 12 0Y4 4 83 75 15 3Y4 4 93 80 15 1 1 100 93

5B 12 0Y4 4 92 92 16 3Y4 4 94 94 16 1 1 100 94

5E 12 2Y4 4 92 92 14 3Y4 4 93 79 14 1 1 100 93

6D 12 2Y4 4 58 83 14 0Y4 2 43 29 14 0Y1 1 93 93

6F 12 1Y4 4 75 83 16 4 4 100 94 16 1 1 100 88

*, Range of scores (0Y4) within the primary assessments.†, The most frequent score within primary assessments.‡, Proportion of the most frequent score value within the primary assessments of each individual core sample.§, The proportion of equally assessed core samples between primary and reassessments.k, Range of scores (yes/no) within the primary assessments.

JNeuropatholExp

Neurol�


2006

Interla

boratory

Assessm

ents

ofAD

�2006America

nAsso

ciatio


Inc.

753


assessment criteria, the IHC/HPtau staining (i.e. staining ofmost of the core samples in a TMA section) was good oracceptable in as many as 94% of the stainings (16 of 17). Inthese good specimens, the staining intensity of the lesionswas uniform between different core samples within the TMAsection and there was no disturbing background staining. In4 of the acceptable stainings (assessment codes 3, 7, 9, and 13in Table 2), the individual core samples were variably stainedwithin the TMA section, whereas in one of the acceptablestainings (assessment code 11 in Table 2), the quality ofimmunostaining was good but the absence of counterstainingcomplicated the assessment of the TMA section.

When good/acceptable stainings were included (16 of17), the distribution of the mean values (T standard error)of primary and reassessments of IHC/HPtau-labeled NFTs(3.4 T 0.2 to 2.7 T 0.3 and 3.4 T 0.2 to 2.6 T 0.3, respectively) atthe TMA section level was rather uniform and the results didnot differ statistically significantly (K-W test, p 9 0.05) (Table6). It was noteworthy that in only 4 of 16 good/acceptablestainings (25%) did the primary and reassessments differsignificantly (Wilcoxon, p G 0.05). Furthermore, the absoluteagreement between the primary and reassessments rangedfrom 63% to 100%, being good (Q75%) in 63% of cases(10 of 16).

The core-to-core comparison of good/acceptable stain-ings (Table 7), including only the 13 core samples composedmostly of grey matter (75Y100%), indicated that somedisagreement was still noted in the primary assessments ofIHC/HPtau-labeled NFTs. In 62% of these selected coresamples (8 of 13), the evaluators were unable to decidebetween the 2 highest scores (3 or 4), and this accounted formost of the disagreement. However, in only one of the 13core samples did the primary assessments range from zero tothe highest score of 4. The highest score value of 4 was themost frequent assessment in as many as 85% of the coresamples (11 of 13). The agreement between the primaryassessments ranged from 43% to 100% and was good(Q75%) in 11 of 13 core samples (85%). Lastly, the absoluteagreement between the primary and reassessments of HPtau-

labeled NFTs ranged from 29% to 100%, being good(Q75%) in as many as 92% of the core samples (12 of 13).

The assessment of IHC/HPtau-labeled NTs was carriedout by all BNE centers (n = 17). At the TMA section levelwhen all 17 stainings were included, the IHC/HPtau-labeledNTs were seen in primary assessments in 47% to 94%, andin reassessment, the range was from 68% to 97% of the coresamples (Table 5). When only the 16 good/acceptablestainings were evaluated, NTs were seen in 67% to 94% ofthe core samples in primary assessments and in 68% to 97%of the core samples in reassessments. In most of these cases,the primary assessments were significantly related to thereassessments (Fisher exact test, p G 0.05). The absoluteagreement between the primary and reassessments, includingthe 16 good/acceptable stainings, ranged from 63% to 100%and it was good (Q75%) in 94% of the assessments.

The core-to-core comparison of good/acceptable stain-ings (Table 7), including only the 13 core samples composedmostly of grey matter (75Y100%), revealed that NTs wereeither not noted or not detected in 3 of 13 core samples(23%). The agreement within the primary assessmentsranged from 81% to 100% and the absolute agreement was100% in 6 of 13 core samples (46%).

Neuropathologic Diagnosis ofAlzheimer Disease

In Figure 7, neuropathologic diagnoses of AD follow-ing different protocols (i.e. CERAD, Braak and Braak, andNIA-Reagan) were given for the 13 core samples (Tables 4,7) composed mostly of grey matter (75Y100%) and that wereof good/acceptable staining quality. Six to 8 assessmentswere available for different core samples. Each core samplewas obtained from a demented subject with cognitive impair-ment. Following the CERAD protocol in 5 of 13 (38%) cases,Braak and Braak staging in 3 of 13 cases (23%), and the NIA-Reagan protocol in 7 of 13 cases (54%), full agreement indiagnosis was obtained. When the results obtained using IHC/HPtau stainings were included in all core samples except 2 (3Eand 6D), moderate to numerous labeled NFTs were seen,

FIGURE 6. Immunohistochemistry/HPtaustaining. The same core sample (1A) atoriginal magnification: 50� (scale bar =100 Km) in panels (AYC) and at originalmagnification: 400� (scale bar = 10 Km)in panels (DYF) immunohistochemicallystained using antibodies directed to hyper-phosphorylated tau (HPtau), assessmentcodes 4, 9, and 6, respectively. Note thedifference in visualization of neurofibrillarytangles, neuropil threads, and neurites.




indicating that these 11 cores were from subjects withdementia, likely being AD. Core sample 3E lacked any AD-associated lesions and in core sample 6D, using the additionalIHC/HPtau method, moderate or numerous NFTs were seenwith either high (50%) or intermediate (50%) likelihood.

DISCUSSIONThe neuropathologic diagnosis of AD includes both the

semiquantitative assessment of NPs and/or NFTs (1, 4, 6)and/or the evaluation of the regional distribution of NFTs (5).The variability existing in the visualization of hallmarklesions such as NPs and NFTs may, however, constitutesa severe obstacle. The Gallyas silver stain, as recommendedfor use by Braak and Braak (5), is selective for NFTs andNTs and is less suitable for NPs (13). It is sensitive to labora-tory protocols and is known to be capricious (3, 13). TheBielschowsky stain, which is recommended for use in as-sessing NPs (4), is also sensitive to laboratory routines andconditions such as ambient temperature (3, 11). Further-more, irrespective of which silver staining technique wasused, staining results have been shown to be influenced by thestorage time in formalin (14). In addition, IHC techniques tovisualize both NFTs and NPs have been recommended recently

(6). However, IHC staining results can also be influenced bythe methodology and the antibody used (15Y20). The presentinterlaboratory study was designed to assess the quality ofstaining and to evaluate the level of agreement in assessingNPs and NFTs in postmortem brain tissue using both silverimpregnation and IHC methods.

In a previous interlaboratory comparison of assessmentof AD lesions, the reported high interrater variability amongthe participating 24 neuropathologists was believed to beattributable to the lack of specific guidelines for selectingassessment fields with each neuropathologist assessing adifferent neuroanatomic region (21). Our previous pilot studyyielded similar results (data not published), and to avoidproblems related to neuroanatomy, we used the 2-mm TMAtechnique (7, 8) for construction of the section to be assessed.Therefore, each participant evaluated the same region, withonly minor variations resulting from the sectioning.

Because no systematic analysis regarding the influence ofpresectioning protocol on the stainings have been published,each center supplied their own tissue for the construction of theTMA block. Surprisingly, the material was indeed heteroge-neous for postmortem delay and fixatives and fixation time, andthis variability in the presectioning protocols was noted toinfluence the staining results. This was clearly notable at the

FIGURE 7. Impact of the assessments on the neuropathologic diagnosis of Alzheimer disease. Thirteen representative coresamples (Q75% grey matter, good/acceptable staining quality) were assessed by 6 to 8 evaluators. Each sample was from apatient with dementia. Different diagnostic protocols (4Y6) were followed for comparison.




TMA section level where the individual core samples processedby the same center stained differently. This is in line with aprevious report (14) indicating that the fixation time can becrucial, particularly influencing the results obtained with silverstains. Similar results have also been reported for immunohis-tology (19). Additionally, there are probably also still unknowndifferences between individual cases that influence the stainingresults. It is noteworthy that there is no information addressingwhy specific structures are labeled with silver stains; thus, theinfluence of presectioning protocol on these stainings is diffi-cult to analyze. Regarding the immunohistochemical stainings,more knowledge is available and systematic analysis of thepreservation of some antigens and retrieval has been done.However, based on our results, it appears that good stainingcan be obtained irrespective of postmortem delay (range,3Y120 hours), fixation time (range, 1 day to 420 days), andstorage time (range, 1 day to 5 years). One should, however, becautious regarding the fixative used. Our results indicated thatthe use of formic acid in fixation solution seemed to influencethe IHC/AT8 staining by diminishing the labeling. The reasonswhy and how the presectioning protocol is so critical liebeyond the scope of this article, but our results indicate thatthese steps are of importance and a more systematic evaluationof these issues is currently underway.

It was noted that among the participating BNE centers,there was a variation in the quality of stain, especially withregard to the silver stains (approximately 45% poor quality). Asseen in Figure 3, the silver-labeled NPs were visualizeddifferently in different stainings, and different core samplesstained differently within the same TMA section. Similarly, asseen in Figure 4, various results were seen when Gallyasstaining was used for visualization of NFTs. Variation in thestaining quality was also noted in the IHC/AA stain (approx-imately 29% poor quality). As seen in Figure 5, the IHC/AA-labeled protein aggregates were visualized differently indifferent stainings and different core samples stained differ-ently within the same TMA section. In contrast, the IHC/HPtau staining quality was quite uniform (approximately 6%poor quality), although some variability was noted (Fig. 6).Although the majority of participants used identical stainingprotocols, the variable staining quality with using silvertechniques might also be attributable to the proficiency of theBNE center performing the technique. Some centers use thestaining routinely, whereas others only rarely.

The primary assessment of lesions in the TMA sectionsvaried significantly. This is partly explained by the methodo-logical aspects discussed here. However, some level of agree-ment was reached when the poor stainings were excluded forall stains except the IHC/AA stain. The extensive variabilityin visualization of IHC/AA aggregates might be the result ofthe many different types of pretreatment used in IHC/AA

staining by the participating BNE centers. This is in line witha previous study (15) that demonstrated that the IHC/AA

staining is especially sensitive to the pretreatment conditions.However, the variability in the staining protocols does notalone explain the unevenness in the IHC/AA staining quality.The precisely same protocol yielded both good as well aspoor quality of staining. This might be the result of theproficiency of the BNE center performing the techniques, but

the results might also be influenced by factors such as buffersand detection kits used. Obviously, not only the presectioningprotocol, but also the way in which the IHC staining isconducted is a major obstacle to reliably visualize the IHC/AA-labeled aggregates. Further refinement of the method-ology, including both use of a panel of antibodies and a strictpretreatment protocol, should be undertaken.

When comparing results assessing good/acceptable silverstains, the primary and reassessments differed. It should benoted, however, that the good/acceptable stained TMA sectionsalso included some poorly stained (i.e. over- and understained)core samples that might have influenced the results. In theBielschowsky silver stain, the significant difference seen in mostof the assessments was attributed to the strictness of the criteriaused during the reassessment; thus, although a plaque wasperceivable, it was not counted if silver-stained, thickenedneurites were not seen. Moreover, some variance was noted afterrepeated assessments by the same primary and reassessor,indicating that some difficulties were encountered in identifica-tion of plaque subtypes such as overstained diffuse plaquescounted as neuritic plaques and vice versa. In agreement withprevious reports (3, 13, 21), there are both methodological andassessment-related pitfalls associated with the Bielschowskysilver-impregnation technique, making this method susceptibleto serious errors. In the Gallyas silver-stained sections, thesignificant difference between the primary and reassessmentsof NFTs was less striking. However, the absolute agreement(i.e. the proportion of equally assessed core samples) was lessthan 75%. This was probably the result of different staining ofthe NFTs in core samples, although the staining of the TMAsection was assessed as being rather uniform. Some of thecores were overstained and some were understained, with bothaffecting the identification as well as the rating of NFTs.However, the absolute agreement regarding the assessment ofNTs was much better. It should, however, be noted that theGallyas stain was performed by only 14 of 17 participants andin 6 of 14 stains, and the quality of staining was assessed asbeing poor, indicating that this technique is both difficult tostandardize and capricious.

The most reproducible staining techniques proved tobe IHC/HPtau in terms of both the staining quality and theassessments. In most cases, the major disagreement in theassessments was related to selecting between the 2 highestscore values (range, 0Y4), whereas alternatives ranging fromzero to 4 were awarded only in one core sample. This coresample (6D) originated from a tissue block that had beenfixed in 4% formaldehyde including formic acid and it waspoorly stained (pale) in most of the stainings. Two centersused formic acid pretreatment, and it was noted that with themonoclonal antibody (AT8), the staining result was poor,whereas when applying the polyclonal antibody, the stainingresult was acceptable. These results are in contrast to aprevious report (17) indicating that the formic acid pretreat-ment enhanced the labeling specifically of neuropil threads.However, these results are not fully comparable because alldetails regarding presectioning and postsectioning protocolsare not available. This highlights the importance of reportingmethodological details when capricious immunolabelingprocedures are used.




A current Medline search reveals that 3 major protocolsare used for the neuropathologic diagnosis of AD. Theseinclude the CERAD protocol based on semiquantitativeassessment of NPs, the Braak and Braak staging based on theregional distribution of NFTs, and the NIA-Reagan protocolcombining the previous ones (4Y6). When only the good/acceptable stained sections and the 13 representative cores(Q75% grey matter; 6Y8 evaluations) were included in 35% ofthe core samples, the same diagnosis of definite AD wasgiven. It is noteworthy that 3 core samples were assessed asbeing either not from AD cases by 6% of evaluators or fromdefinite AD cases by 14% of evaluators. Based on thevariability in assessment of Bielschowsky-stained NPs, wecannot recommend the use of this protocol in an interlabor-atory setting. Similar diagnostic variability was noted follow-ing the Braak and Braak protocol. The NIA-Reagan protocolcombines the CERAD and Braak and Braak protocols, sothere is a high likelihood of AD in cases with numerous NPsand isocortical NFTs. Regarding our 13 core samples, in 25%of the received assessments, the NIA-Reagan protocol (whenstrictly followed) could not be applied because the CERADand Braak and Braak assessment results were not consistentwith each other. In some cases with CERAD-definite AD, noNFTs were noted (plaque-predominant) and vice versa(tangle-predominant). However, when IHC/HPtau method-ology was applied, in 11 of the 13 core samples, all evaluatorsdetected numerous NFTs, indicating a high likelihood of AD.

In conclusion, some variability in the staining quality ofsilver techniques was noted, although almost all the partic-ipants followed the same staining procedure. This led toundesirable variation in the assessments of AD-related lesionsand consequently neuropathologic diagnoses; therefore, thesestainings are not recommended for use in interlaboratorysettings. The variability found with respect to IHC/AA

labeling is attributed to extensive methodological differencesinfluencing the rating and, therefore, further methodologicaladjustments are required. Our study indicated that there is onemethod (the IHC/HPtau labeling of NFTs and NTs) thatyields almost uniform quality of staining and appropriateassessment of lesions, although the IHC/HPtau methodologywas not uniform. These findings suggest that the IHC/HPtaulabeling can be recommended for assessing AD lesions,especially in interlaboratory settings. This method is consis-tent with the recommendations in http://www.ICDNS.org,which emphasize the use of immunocytochemistry in ADdiagnostics. Furthermore, our findings point out the need for aneuropathologic diagnostic protocol of AD that is primarilybased on the IHC/HPtau methodology.

Based on the number of poor stainings and thevariability in the assessments of AD lesions found in thisstudy, it can be further concluded that there is a need forcontinual quality control and standardization of methodology.

ACKNOWLEDGMENTSThe authors thank Frances Carnie, Tarja Kauppinen,

Maria Kemerli, M. Kooreman, Tanja Treutlein, A. Van denBerg, and J. Wouda as well as the other laboratorytechnicians of the individual brain banks for their skillful

technical assistance and Vesa Kiviniemi for his assistance instatistics. MBG is grateful to the Parkinson’s DiseaseSociety Tissue Bank at Imperial College London, funded bythe Parkinson’s Disease Society of the U.K., registeredcharity 948776. The study has been authorized by the EthicsCommittee of Kuopio University Hospital.

REFERENCES1. Khachaturian ZS. Diagnosis of Alzheimer’s disease. Arch Neurol 1985;

42:1097Y1052. Yamamoto T, Hirano A. A comparative study of modified Bielschowsky,

Bodian and thioflavin S stains on Alzheimer’s neurofibrillary tangles.Neuropathol Appl Neurobiol 1986;12:3Y9

3. Lamy C, Duyckaerts C, Delaere P, et al. Comparison of seven stainingmethods for senile plaques and neurofibrillary tangles in a prospectiveseries of 15 elderly patients. Neuropathol Appl Neurobiol 1989;15:563Y78

4. Mirra SS, Heyman A, McKeel D, et al. The Consortium to Establish aRegistry for Alzheimer’s Disease (CERAD). Part II. Standardization ofthe neuropathologic assessment of Alzheimer’s disease. Neurology 1991;41:479Y86

5. Braak H, Braak E. Neuropathological stageing of Alzheimer-relatedchanges. Acta Neuropathol 1991;82:239Y59

6. Consensus recommendations for the postmortem diagnosis of Alzheimer’sdisease. The National Institute on Aging, and Reagan Institute WorkingGroup on Diagnostic Criteria for the Neuropathological Assessment ofAlzheimer’s Disease. Neurobiol Aging 1997;18:S1Y2

7. Kallioniemi OP, Wagner U, Kononen J, et al. Tissue microarraytechnology for high-throughput molecular profiling of cancer. Hum MolGenet 2001;10:657Y62

8. Kononen J, Bubendorf L, Kallioniemi A, et al. Tissue microarrays forhigh-throughput molecular profiling of tumor specimens. Nat Med1998;4:844Y47

9. Kauppinen T, Martikainen P, Alafuzoff I. Human postmortem braintissue and 2-mm tissue microarrays. Appl Immunohistochem MolMorphol, in press

10. Wang H, Wang H, Zhang W, et al. Tissue microarrays: Applications inneuropathology research, diagnosis, and education. Brain Pathol 2002;12:95Y107

11. Litchfield S, Nagy Z. New temperature modification makes theBielschowsky silver stain reproducible. Acta Neuropathol 2001;101:17Y21

12. Lowe J. Techniques in neuropathology. In: Bancroft JD, Stevens A, eds.Theory and Practice of Histological Techniques. Philadelphia: Churchill-Livingstone, 1996:341Y81

13. Cullen KM, Halliday GM, Cartwright H, et al. Improved selectivity andsensitivity in the visualization of neurofibrillary tangles, plaques andneuropil threads. Neurodegeneration 1996;5:177Y87

14. Wilcock GK, Matthews SM, Moss T. Comparison of three silver stainsfor demonstrating neurofibrillary tangles and neuritic plaques in braintissue stored for long periods. Acta Neuropathol 1990;79:566Y68

15. Kraszpulski M, Soininen H, Riekkinen P Sr, et al. Pitfalls in thequantitative estimation of beta-amyloid immunoreactivity in humanbrain tissue. Histochem Cell Biol 1998;110:439Y45

16. Wester K, Wahlund E, Sundstrom C, et al. Paraffin section storage andimmunohistochemistry. Effects of time, temperature, fixation, and re-trieval protocol with emphasis on p53 protein and MIB1 antigen. ApplImmunohistochem Mol Morphol 2000;8:61Y70

17. Cummings BJ, Mason AJ, Kim RC, et al. Optimization of techniques forthe maximal detection and quantification of Alzheimer’s-related neuro-pathology with digital imaging. Neurobiol Aging 2002;23:161Y70

18. D’Andrea MR, Nagele RG, Wang HY, et al. Consistent immunohisto-chemical detection of intracellular beta-amyloid42 in pyramidal neuronsof Alzheimer’s disease entorhinal cortex. Neurosci Lett 2002;333:163Y66

19. Leong AS. Pitfalls in diagnostic immunohistology. Adv Anat Pathol2004;11:86Y93

20. Croisier E, Mres DE, Deprez M, et al. Comparative study of commerciallyavailable anti->-synuclein antibodies. Neuropathol Appl Neurobiol 2006;32:351Y56

21. Mirra SS, Gearing M, McKeel DW Jr, et al. Interlaboratory comparisonof neuropathology assessments in Alzheimer’s disease: A study of theConsortium to Establish a Registry for Alzheimer’s Disease (CERAD).J Neuropathol Exp Neurol 1994;53:303Y15



https://www.researchgate.net/publication/8612513_Leong_AS-Y_Pitfalls_in_diagnostic_immunohistology?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/8612513_Leong_AS-Y_Pitfalls_in_diagnostic_immunohistology?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/21463347_Comparison_of_three_silver_stains_for_demonstrating_neurofibrillary_tangles_and_neuritic_plaques_in_brain_tissue_stored_for_long_periods_Acta_Neuropathol?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==



https://www.researchgate.net/publication/225577233_Pitfalls_in_the_quantitative_estimation_of_b-amyloid_immunoreactivity_in_human_brain_tissue?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==



https://www.researchgate.net/publication/12075719_Tissue_microarray_technology_for_high-throughput_molecular_profiling_of_cancer?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==



https://www.researchgate.net/publication/12136009_Litchfield_S_Nagy_ZNew_temperature_modification_makes_the_Bielschowsky_silver_stain_reproducible_Acta_Neuropathol_10117-21?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/12136009_Litchfield_S_Nagy_ZNew_temperature_modification_makes_the_Bielschowsky_silver_stain_reproducible_Acta_Neuropathol_10117-21?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/20672593_Comparison_of_seven_staining_methods_for_senile_plaques_and_neurofibrillary_tangles_in_a_prospective_series_of?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==



https://www.researchgate.net/publication/6856450_Human_Postmortem_Brain_Tissue_and_2-mm_Tissue_Microarrays?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==



https://www.researchgate.net/publication/14386472_Improved_Selectivity_and_Sensitivity_in_the_Visualization_of_Neurofibrillary_Tangles_Plaques_and_Neuropil_Threads?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==



https://www.researchgate.net/publication/20291167_Diagnosis_of_Alzheimer's_Disease?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/20291167_Diagnosis_of_Alzheimer's_Disease?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/284688338_Neuropathological_stageing_of_Alzheimer-related_neurofibrillary_changes?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/284688338_Neuropathological_stageing_of_Alzheimer-related_neurofibrillary_changes?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==

https://www.researchgate.net/publication/13898158_The_National_Institute_on_Aging_and_Reagan_Institute_Working_Group_on_Diagnostic_Criteria_for_the_Neuropathological_Assessment_of_Alzheimer's_Disease?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==




https://www.researchgate.net/publication/295805049_Erratum_Interlaboratory_comparison_of_neuropathology_assessments_in_Alzheimer's_disease_A_study_of_the_consortium_to_establish_a_registry_for_Alzheimer's_disease_CERAD_Journal_of_Neuropathology_and_Ex?el=1_x_8&enrichId=rgreq-2b371ec5-2943-4f9f-b689-73c27d7708e4&enrichSource=Y292ZXJQYWdlOzY4OTE0NDM7QVM6OTkzNjMzODA1MzEyMDdAMTQwMDcwMTQ4NDQ0Mw==



Interlaboratory Comparison of Assessments of Alzheimer Disease-Related Lesions

Documents