ARTICLE IN PRESS

0344-0338/$ - se

doi:10.1016/j.pr

�CorrespondiE-mail addre

Pathology – Research and Practice 204 (2008) 655–662

www.elsevier.de/prp

ORIGINAL ARTICLE

Rabbit monoclonal antibodies show higher sensitivity than mouse

monoclonals for estrogen and progesterone receptor evaluation

in breast cancer by immunohistochemistry

Rafael Rocha, Cristiana Nunes, Gislene Rocha, Flavio Oliveira,Fernanda Sanches, Helenice Gobbi�

Department of Anatomic Pathology, Universidade Federal de Minas Gerais, Faculdade de Medicina da UFMG,

Av. Alfredo Balena, 190 Belo Horizonte, MG 30.130-100, Brazil

Received 4 July 2007; accepted 13 March 2008

Abstract

A novel generation of rabbit monoclonal antibodies for estrogen (ER) and progesterone (PR) receptor evaluation inbreast cancer by immunohistochemistry has been released recently. We compared the novel RabMab anti-ER andanti-PR antibodies with the mouse monoclonal antibodies using a tissue microarray of breast carcinomas. Twocylinders (2mm diameter) of formalin-fixed, paraffin-embedded tissue were obtained from 24 invasive breast cancersand were immunostained using anti-ER mouse (1D5 and 6F11) and rabbit antibodies (SP1 and B644), and anti-PRmouse (PgR312 and PgR636) and rabbit antibodies (SP2 and B645). The immunohistochemistry was evaluated byconsidering positive those tumors in which more than 10% of the tumor cell nuclei stained independently on thestaining intensity. Our results demonstrated that rabbit antibodies against ER have a similar staining patterncompared to the 6F11, but better than 1D5 from three different suppliers. The rabbit antibodies against PR (SP2 andB645) provide a stronger and sharper immunohistochemical signal compared to mouse antibodies (PgR636 andPgR312). Both ER and PR rabbit antibodies allow a lower cost per test because of higher working dilutions comparedto mouse antibodies using the same procedure. The novel rabbit antibodies against ER and PR are highly sensitive forimmunohistochemical testing of breast carcinomas.r 2008 Elsevier GmbH. All rights reserved.

Keywords: Monoclonal antibodies; Immunohistochemistry; Breast cancer

Introduction

Since 1990, pathology has played an important rolenot only in diagnosis but also by providing additionalinformation about prognostic and predictive molecularmarkers that aimed at the best breast cancer treatment[22,30].

e front matter r 2008 Elsevier GmbH. All rights reserved.

p.2008.03.010

ng author.

ss: hgobbi@medicina.ufmg.br (H. Gobbi).

Although predictive factors of therapy response havemore clinical value, the broadly validated predictivefactors for routine clinical use only are estrogen (ER)and progesterone receptors (PR), and HER2/neu [3].Since the release of monoclonal antibodies, scientistshave been evaluating ER and PR almost exclusively byimmunohistochemistry. The mouse monoclonal antibo-dies, as well as polyclonal antibodies, can be applied toformalin-fixed, paraffin-embedded tissues [2,6,8,18,26].Several different methods and different antibodies

ARTICLE IN PRESS

Table 1. Antibodies, clones, concentrations, and suppliers of

the rabbit and mouse monoclonal antibodies against estrogen

and progesterone receptors

R. Rocha et al. / Pathology – Research and Practice 204 (2008) 655–662656

against ER and PR have been used for immunohisto-chemistry testing. As a result, there is no universallyaccepted standard [15,23]. As with any laboratorymethod, technical details have a great effect on the finalresult [10,20,21].

European and American studies show a large varia-tion in the results of ER and PR evaluation byimmunohistochemistry [19,21,25]. An important pointis the choice of the most suitable primary antibody[19,27]. New studies result in the release of newantibodies every year. Although very useful in routineand random clinical tests, there is no recommendation inthe literature about which antibodies would be morespecific and sensitive for evaluating ER and PR byimmunohistochemistry [6,8]. However, the choice of thebest primary antibody has great importance in the finalresults of the immunohistochemistry testing, which caninfluence treatment decisions [21].

The most used antibodies for ER and PR evaluationby immunohistochemistry have been mouse monoclonalantibodies. Recently, a novel generation of rabbitmonoclonal antibodies has been developed [13]. Thetechnology to prepare these antibodies from a singlehybridoma allows the production of antibodies withhigh sensitivity and specificity according to the manu-facturers. These characteristics allow higher workingdilutions and, consequently, a lower cost per test.According to suppliers, the novel rabbit monoclonalantibodies could replace the mouse monoclonal anti-bodies because of their high specificity, lower cost, andfaster technical procedure, which would not require anantigen retrieval step [4,13,25].

In the present study, we compared the sensitivity andcost of each test using rabbit monoclonal antibodies andmouse monoclonal antibodies for ER and PR evalua-tion by immunohistochemistry in invasive breast carci-nomas.

Antibody Clone Dilution Supplier

Mouse monoclonal

Estrogen receptor 1D5 1:100 Dako, USA

Estrogen receptor 1D5 1:100 Cell marque, USA

Estrogen receptor 1D5 1:100 Biogenex, USA

Estrogen receptor 6F11 1:100 Novocastra (Leica),

UK

Progesterone

receptor

PgR

312

1:100 Novocastra (Leica),

UK

Progesterone

receptor

PgR

636

1:100 Dako, USA

Rabbit monoclonal

Estrogen receptor SP1 1:300 Lab vision, USA

Estrogen receptor B644 1:300 Cell marque, USA

Progesterone

receptor

SP2 1:300 Lab vision, USA

Progesterone

receptor

B645 1:300 Cell marque, USA

Materials and methods

Case selection

Twenty-four cases of invasive breast carcinomas,diagnosed between 1990 and 2005, were randomlyselected from the files of the Breast Pathology Labora-tory of the School of Medicine, Federal University ofMinas Gerais, Brazil. All original slides were reviewedto confirm the diagnosis and to select representativeareas of tumors. Two cylinders (2mm diameter) of eachtumor with representative areas of neoplasia wereselected from paraffin blocks to build a tissue micro-array (TMA). Two cylinders from previously testedtumors and whose results were positive and negativewere also included as internal controls for the TMA.

Sequential 5 mm sections were obtained and stained forhematoxylin and eosin (first and last sections) to confirmdiagnosis, and the interval sections were used for theimmunohistochemical study. Slides containing sectionsof a positive breast tumor were included in all batches asexternal control.

Immunohistochemical study

The sections were mounted on glass slides coated withsilane (3-aminopropyltriethoxysilane) and dried for30min at 37 1C. The sections were deparaffinized inxylene and rehydrated via a series of graded alcohols.Endogenous peroxidase activity was blocked by incu-bating the sections in a methanol bath containing 3%hydrogen peroxide for 20min, followed by washing indistilled water. All sections were initially submitted toheat-induced epitope retrieval using citrate buffer (pH6.0). After that, the primary antibody was applied andincubated for 90min at room temperature. Mousemonoclonal antibodies and rabbit monoclonal antibo-dies were used in order to evaluate ER and PR receptors(Table 1). Preliminary testing was performed in ourlaboratory to identify the best concentration for eachantibody and to choose the negative and positivecontrols using the dilution data supplied by themanufacturer as the starting point. After washing theprimary antibody with phosphate buffered saline (PBS),the slides were incubated with linking biotinylatedantibody (Super Sensitive Link, Label IHC DetectionSystem, BiogenexTM) for 20min. The sections wererinsed with PBS, followed by incubation with perox-idase-conjugated streptavidin complex for 20min (Super

ARTICLE IN PRESS

Table 2. Number of cases and percentage (in parentheses) classified according to the intensity of staining for each primary

antibody tested against estrogen receptor

Staining intensity (score) Rabbit Mouse

SP1 B644 1D5 6F11

Dako Cell Marque BioGenex

0 9 (37.5) 9 (37.5) 11 (45.8) 11 (45.8) 14 (58.4) 9 (37.5)

1 2 (8.4) 2 (8.4) 4 (16.7) 4 (16.7) 3 (12.5) 2 (8.4)

2 4 (16.7) 5 (20.8) 4 (16.7) 2 (8.4) 6 (25) 4 (16.7)

3 9 (37.4) 8 (33.3) 5 (20.8) 7 (29.1) 1 (4.1) 9 (37.4)

Total 24 (100) 24 (100) 24 (100) 24 (100) 24 (100) 24 (100)

Cases were considered positive from scores 1 to 3.

Intensity of staining ranged from: negative ¼ 0; to positive: weak ¼ 1, moderate ¼ 2, and strong ¼ 3.

Table 3. Number of cases and percentage (in parentheses)

classified according to the intensity of staining for each

primary antibody tested against progesterone receptor

Staining

intensity

(score)

Rabbit Mouse

SP2 B645 PgR 312 PgR636

0 10 (41.7) 9 (37.5) 12 (50) 12 (50)

1 3 (12.5) 3 (12.5) 1 (4.2) 1 (4.2)

2 5 (20.8) 3 (12.5) 3 (12.5) 3 (12.5)

3 6 (25) 9 (37.5) 8 (33.3) 8 (33.3)

Total 24 (100) 24 (100) 24 (100) 24 (100)

Cases were considered positive from scores 1 to 3.

Intensity of staining ranged from: negative ¼ 0; to positive: weak ¼ 1,

moderate ¼ 2, and strong ¼ 3.

R. Rocha et al. / Pathology – Research and Practice 204 (2008) 655–662 657

Sensitive Link, Label immunohistochemical DetectionSystem, BiogenexTM). Freshly prepared DAB solution(1 drop of 3,30-diaminobenzidine tetrahydrochloride for1ml of substrate, DAKOTM) was applied for 2min oneach section. DAB was removed by rinsing with distilledwater. The slides were counterstained with hematoxylin,dehydrated in ethanol, cleared in xylene, and mountedusing EntelanTM.

Immunostaining analysis

All sections submitted to immunohistochemistry werecoded, and the examiner was blinded to the antibodyused. The cut offs used were those most used in theliterature [1,8,9,12,16,29]. We considered positive thosetumors containing more than 10% of stained nucleiindependently on the intensity of staining. The moreintensely stained area of the two disks of each tumor wasconsidered in the analysis. The intensity of the reactionand the background were evaluated as negative (0),weak (1), moderate (2), and strong (3).

Statistical analysis

The Software SPSS, Version 12.0, was used for thestatistical analysis. Wilcoxon test was used to comparethe different groups of categorical variables. The Spear-man coefficient was used to evaluate the positive corr-elation among the categorical variables of the differentgroups.

Results

Estrogen receptor analysis

The percentage of positivity and overall result can beseen in Tables 2 and 3, and Fig. 1. ER receptor waspositive in 15 cases (62.5%), with variable staining

intensity among the different cases and antibodies used.In one case, there was no variation in the intensity ofstaining using the different clones, and tumor cellsshowed strong reactivity. In five cases, there werepositive reactions for some antibodies (including bothrabbit antibodies) and negative reactions for otherantibodies. In all these five cases, the reaction wasnegative for 1D5 (Biogenex). Two cases were negativefor 1D5 from three suppliers and positive for bothrabbit and mouse antibodies 6F11. Nine cases (37.5%)were negative for all antibodies tested.

There was a statistically significant difference whencomparing the clone 1D5 from different suppliers toboth rabbit antibodies which stained stronger (po0.05).However, no difference was observed between the stainsof both rabbit and the mouse antibody 6F11. Astatistically significant difference was observed whencomparing all the antibodies to the clone 1D5 (Biogen-ex), which showed weaker reactions than the otherantibodies (po0.02).

The highest correlation was obtained among the tworabbit clones (SP1 and B644) and the mouse clone 6F11,which showed stronger nuclear staining when compared

ARTICLE IN PRESS

Fig. 1. Estrogen receptor: sections stained for rabbit antibodies SP1 (A) and B644 (B) show stronger staining intensity compared to

mouse antibody 1D5 from three different suppliers (C, Dako; D, CellMarque; and E, Biogenex) and similar staining intensity to

6F11 (F). Sections stained for 1D5 (E, Biogenex) showed weaker staining intensity compared to all other anti-ER antibodies.

R. Rocha et al. / Pathology – Research and Practice 204 (2008) 655–662658

to the other mouse antibodies in the majority of cases(Fig. 1).

Time of storage of the paraffin blocks, which rangedfrom 1 to 15 years, had no effect on staining intensityand positivity.

Progesterone receptor analysis

Sixteen cases (66.7%) were positive for PR (Table 3).In eight of the 16 cases, there was no variation in theintensity of the staining using the different clones. Infour cases, there were positive reactions for rabbit (SP2and B645) but negative reactions for mouse (PgR 312and PgR636) antibodies. One case was positive for bothrabbit and negative for both mouse antibodies. Eightcases (33.4%) were negative for all antibodies tested.Two cases were negative for ER and positive for PR,and one case was negative for PR and positive for ER.

Sections that stained for both SP2 and B645 showedsignificantly stronger staining intensity than thosestained for both PR mouse antibodies (p ¼ 0.025).

There was a statistically significant agreement be-tween the ER and PR receptor positivity and negativityin the same cases (p ¼ 0.001). The negative control caseincluded in the study was confirmed to be negative withall ER and PR antibodies tested.

Cost of each test using different antibodies

The cost of each test (in US$) using differentantibodies is shown in Table 4.

Discussion

In the present study, we compared mouse monoclonalantibodies to novel rabbit monoclonal antibodies for

ARTICLE IN PRESS

Table 4. Price per test considering only the cost of the mouse

and rabbit monoclonal antibodies

Antibody type Clone Supplier Price per test

Rabbit anti-ER SP1 Lab Vision 0.46

B644 Cell Marque 0.33

Rabbit anti-PR SP2 Lab Vision 0.46

B645 Cell Marque 0.33

Mouse anti-ER 1D5 DAKO 1.56

1D5 Cell Marque NAa

1D5 Biogenex 2.01

6F11 Novocastra 0.98

Mouse anti-PR PgR 636 DAKO 1.96

PgR 312 Novocastra 1.07

aNot available. The antibody is no longer sold by the supplier.

R. Rocha et al. / Pathology – Research and Practice 204 (2008) 655–662 659

ER and PR testing in breast carcinomas using a TMA of24 breast tumors and the same immunohistochemicaldetection system. Although 24 carcinomas seem to be asmall sample, our results were relatively uniform in thisseries of randomly selected tumors from routinediagnostic pathology. After evaluating the results, weconcluded that the number of tumors was enough forthe purpose of comparison of sensitivity of the mostcommonly used antibodies against ER and PR clones.

There are few other studies in the literature compar-ing the new rabbit monoclonal antibodies with mouseantibodies [4,5,24]. Cano et al. [4] evaluated ER and PRreceptors in fine-needle aspirates and paraffin-embeddedsections from breast cancers using SP1 and SP2 rabbitantibodies. They found that the use of rabbit mono-clonal antibodies against ER and PR on alcohol-fixedsmears and paraffin sections provided several advan-tages, such as high sensitivity and specificity of thereaction, stronger immunostaining, shorter proceduretimes, and avoidance of antigen retrieval step. Rossiet al. [24] carried out a comparative study betweenrabbit clones against ER and PR receptors, Ki67, cyclinD1, CD3, CD5, CD23, and synaptophysin and classicmouse monoclonal antibodies against the same antigenson several tumor types, as well as normal tissues. Theyfound no significant differences in the percentage ofpositive cells and staining intensity. However, theauthors suggest that the rabbit antibodies appear tooffer increased sensitivity with no apparent loss ofspecificity and also allowed a higher working dilution.Cheuk et al. [7] compared the novel rabbit anticyclin D1to the mouse antibodies against cyclinD1 and consid-ered that the rabbit antibody SP4 showed superiorperformance over the mouse monoclonal antibody. Aconsistent immunostaining for cyclin D1 was readilyachieved when compared to mouse antibody DCS-6 in150 cases of lymphoproliferative lesions [15]. Cheanget al. [5] published a robust study that evaluatedimmunohistochemistry using the new rabbit antibody

SP1 and the mouse antibody 1D5. They evaluated therelationship to biochemical ER assay results and clinicaldata on survival and adjuvant systemic therapy. Theauthors detected 69.5% of positivity when using therabbit SP1 and 63.1% using the mouse antibody 1D5.Rabbit antibody SP1 was also a better independentprognostic factor than 1D5 in multivariate analysis,including age, tumor size, grade, and lymphovascularand nodal status. SP1 was considered by these authors[5] as an improved standard for ER immunohistochem-istry assessment in breast cancer compared to 1D5.

In our study, we added original information on thecomparison of rabbit and mouse antibodies includingother clones against ER and also against PR. The 6F11showed similar results to the SP1 and B644 and betterones than 1D5 from different suppliers. We used a semi-quantitative evaluation of the intensity of staining inorder to detect differences of sensitivity among 10different antibodies against ER and PR. This scoringsystem allowed us to distinguish the clones SP1, B644,and 6F11 as the most sensitive for detecting positivity.We also evaluated the cost per test of all antibodiesused, and this comparison showed that the rabbitantibodies presented the lowest cost per test due to thehigh working dilution they allowed.

We found that both ER rabbit antibodies stained withstronger intensity when compared to the mouse anti-body 1D5 from different suppliers but with similarintensity to the 6F11. A significant number of negativeresults were obtained using the 1D5 from Biogenex.These same cases were positive when stained with theother antibodies. The negative control was previouslytested using 6F11 and, despite binding different epitopesof ER, all other antibodies tested showed negativeresults for the negative control.

The difference of sensitivity or staining intensityamong different antibodies against the same proteinmay exist due to the different epitopes of the sameprotein bound for the specific antibody paratope. TheER gene consists of more than 140 kb of genomic DNAdivided into eight exons, being translated into a proteinwith six functionally discrete domains labeled A to F.The a-ER receptor is a protein of 67 kDa that can berecognized by different clones of antibodies which binddifferent epitopes. The most used mouse monoclonalantibody 1D5 binds an epitope localized in theN-terminus domain of a-ER receptor only. On theother hand, the novel rabbit monoclonal anti-ER SP1binds the C-terminal domain. The mouse monoclonalantibody against PR receptor PgR 636 binds the amino-terminal domain, which appears in the PR-A and also inthe PR-B, and the novel rabbit SP2 binds the epitope412–526aa.

Many studies reported their results of ER and PRanalysis based on a semiquantitative method consider-ing both intensity and percentage of positive nuclei.

ARTICLE IN PRESSR. Rocha et al. / Pathology – Research and Practice 204 (2008) 655–662660

However, other investigators excluded the intensity ofstaining, as it has been shown to have no relationship totherapy response or prognosis. There is a variation ofER scoring interpretation in the literature ranking from1% to 10%, whether considering the staining intensityor not [2,8,11,17,28]. Harvey et al. [12] demonstratedthat patients whose tumors presented 1% or more ER-positive cells, with moderate or stronger intensity,responded to anti-ER therapy.

We do not have a gold standard for evaluating ERand PR since we did not carry out a correlation with thepatient therapy responsiveness. However, we made acomparison between the signal obtained by IHC fromnovel rabbit monoclonal antibodies and the broadlyclinically validated mouse monoclonals, including clone1D5. The cut off of IHC positivity for ER and PR iscontroversial, and the literature shows therapy responsefrom a cut off of 410% of the stained nucleiindependently of the staining intensity (including weakstaining) [8] to 41% of the stained nuclei that stainedmoderate or stronger [12]. As we were analyzing andcomparing the staining intensity of different antibodies,it seemed more interesting to use the score 410% of thestained nuclei independent of the staining intensity; sowe could also emphasize the weak staining. We alsocarried out a re-evaluation of all cases using the Allred’sscore system [1], and no difference in positivity wasobserved.

Although the suppliers of the rabbit antibodiesadvocate that the total reaction time can be shorter byavoiding the epitope retrieval step, pretests performed inour laboratory showed that the working dilution was1:50 when antigen retrieval was not used, whereas adilution of 1:300 was achieved when using epitoperetrieval. Although the antigen retrieval step is moretime-consuming, the total cost of each test dropped byusing antibody dilutions six times less concentrated.Rossi et al. [24] used the rabbit monoclonal antibodySP1 10 times less concentrated than the 6F11 when usingantigen retrieval. These authors also described thatacceptable staining was obtained with SP1 even in theabsence of antigen retrieval, a condition in which 6F11sensitivity dropped dramatically [24]. SP2 was alsoshown to have important advantages, including itsreactivity even without heat-based antigen retrievaland the predominance of nuclear immunostaining withonly very low cytoplasmic signal [14].

Our results showed that in the comparison among thecosts of each test for each antibody, the rabbitmonoclonal antibodies presented lower cost per testdue to their higher working dilutions. However, if thedilution had been that recommended by the suppliers,the cost per test would be quite similar for the differentantibodies.

According to Huang et al. [13], the high affinity of theclone SP1 and its binding to a different epitope from

clone 1D5 would explain why antigen retrieval is notnecessary. The rabbit SP1 has appropriate tissuereactivity, with nuclear staining in epithelial tissues ofknown ER status showing an affinity eight times higherthan that of 1D5 and reactivity with the predicted bandon Western blotting [13]. Huang et al. [13] alsodemonstrated that the rabbit SP1, against ER, showedresults similar to mouse monoclonal 1D5 in all but sixout of the 61 samples. According to the authors, theclone SP1 may be more sensitive than clone 1D5 and hasthe same specificity as clone 1D5 in immunohistochem-istry [13]. However, our results confirmed that the costof each test dropped six times when the epitope retrievalwas used.

There are few studies comparing different mouseclones against PR and one study comparing rabbit andmouse clones anti-PR which compared the rabbit anti-PR SP2 to the mouse 1A6 [14]. These authors showedthat the antibody affinity of SP2 is 12 times higher thanthat of 1A6 [14]. In our study, we found a significantlybetter staining for the rabbit clones SP2 and B645 ascompared to the mouse clones PgR636 and PgR312.

There was a concordance between ER and PRstaining in the positive cases. However, we observedthat PR antibodies stained with stronger intensity thandid the ER antibodies in the same cases. In our study,low background and sharp nuclear staining wereobserved in all sections stained for the rabbit clonesagainst ER and PR (Fig. 1) and may reflect the highspecificity of the monoclonal antibodies, especially therabbit clones, whose reactions showed the least back-ground.

In conclusion, the new rabbit monoclonal antibodiesagainst ER (SP1 and B644) and PR (SP2 and B645) arehighly sensitive for immunohistochemical evaluation inbreast carcinomas. Our results demonstrate that rabbitantibodies against ER have similar staining patternscompared to mouse 6F11 but better ones than clone1D5 from three different suppliers using the sameimmunohistochemical protocol. The rabbit antibodiesagainst PR (SP2 and B645) provide a stronger andsharper immunohistochemical signal compared tomouse antibodies (Pgr636 and PgR312). Both ER andPR rabbit antibodies allow a lower cost per test becauseof higher working dilutions compared to mouse anti-bodies. Further studies are necessary to confirm if thehigh sensitivity correlates to therapeutic response.

Acknowledgments

Financial support: Fundacao de Amparo a Pesquisado Estado de Minas Gerais (FAPEMIG), ConselhoNacional de Desenvolvimento Cientıfico e Tecnologico(CNPq) and Comite de Aperfeicoamento de Pessoal de

ARTICLE IN PRESSR. Rocha et al. / Pathology – Research and Practice 204 (2008) 655–662 661

Nıvel Superior (CAPES). We are grateful to Sandra J.Olson, MS for the language review.

Financial disclosure: The authors do not have anycommercial or financial interests in any of the anti-bodies, reagents, or companies involved in this study.

References

[1] D.C. Allred, et al., Prognostic and predictive factors in

breast cancer by immunohistochemical analysis, Mod.

Pathol. 11 (1998) 155–168.

[2] J. Anderson, H.S. Poulsen, Immunohistochemical estro-

gen receptor determination in paraffin-embedded tissue:

prediction of response to hormonal treatment in advanced

breast cancer, Cancer 64 (1989) 1901–1908.

[3] D.M. Barnes, W.H. Harris, P. Smith, et al., Immunohis-

tochemical determination of oestrogen receptor: compar-

ison of different methods of assessment of staining and

correlation with clinical outcome of breast cancer

patients, Br. J. Cancer 74 (1996) 1445–1451.

[4] G. Cano, F. Milanezi, D. Leitao, et al., Estimation of

hormone receptor status in fine-needle aspirates and

paraffin-embedded sections from breast cancer using the

novel rabbit monoclonal antibodies SP1 and SP2, Diagn.

Cytopathol. 29 (2003) 207–211.

[5] M.C. Cheang, D.O. Treaba, C.H. Speers, et al., Im-

munohistochemical detection using the new rabbit mono-

clonal antibody SP1 of estrogen receptor in breast cancer

is superior to mouse monoclonal antibody 1D5 in

predicting survival, J. Clin. Oncol. 24 (2006) 5626–5628.

[6] G. Chebil, P. Bendahl, M. Ferno, Estrogen and progester-

one receptor assay in paraffin-embedded breast cancer–re-

producibility of assessment, Acta Oncol. 42 (2003) 43–47.

[7] W. Cheuk, K.O. Wong, C.S. Wong, et al., Consistent

immunostaining for cyclin D1 can be achieved on a

routine basis using a newly available rabbit monoclonal

antibody, Am. J. Surg. Pathol. 28 (2004) 801–807.

[8] P.L. Fitzgibbons, D.L. Page, D. Weaver, et al., Prog-

nostic factors in breast cancer. College of American

Pathologists consensus statement 1999, Arch. Pathol.

Lab. Med. 124 (2000) 966–978.

[9] H. Gobbi, W.D. Dupont, J.F. Simpson, et al., Breast

cancer risk associated with estrogen receptor expression in

epithelial hyperplasia lacking atypia and adjacent lobular

units, Int. J. Cancer 113 (2005) 857–859.

[10] A.P. Gouvea, J.R.M. Fernandes, S.J. Olson, et al., Her-2/

neu immunoreactivity in invasive mammary carcinomas:

a comparative study using monoclonal and polyclonal

antibodies including the Herceptest, J. Bras. Patol. 40

(2004) 47–52.

[11] W. Hanna, D.R. McReady, J.W. Chapman, et al., The

predictive value of ERICA in breast cancer recurrence: a

univariate and multivariate analysis, Mod. Pathol. 6

(1993) 748–754.

[12] J.N. Harvey, G.M. Clark, C.K. Osborne, et al., Estrogen

receptor status by immunohistochemistry is superior to

the ligand binding assay for predicting response to

adjuvant endocrine therapy in breast cancer, J. Clin.

Oncol. 17 (1999) 1474–1481.

[13] Z. Huang, W. Zhu, G. Szekeres, et al., Development of

new rabbit monoclonal antibody to estrogen receptor:

immunohistochemical assessment on formalin-fixed, par-

affin-embedded tissue sections, Appl. Immunohistochem.

Mol. Morphol. 13 (2005) 91–95.

[14] Z. Huang, W. Zhu, Y. Meng, et al., Development of new

rabbit monoclonal antibody to progesterone receptor

(clone SP2): no heat pretreatment but effective for

paraffin section immunohistochemistry, Appl. Immuno-

histochem. Mol. Morphol. 14 (2006) 229–233.

[15] T.W. Jacobs, J.E. Prioleau, I.E. Stillman, et al., Loss of

tumor marker-immunostaining intensity on stored paraf-

fin slides of breast cancer, J. Natl. Cancer Inst. 88 (1996)

1054–1059.

[16] B.K. McLaren, H. Gobbi, P.A. Schuyler, et al., Im-

munohistochemical expression of estrogen receptor in

enlarged lobular units with columnar alteration in benign

breast biopsies: a nested case–control study, Am. J. Surg.

Pathol. 29 (2000) 105–108.

[17] E.M. Pellicer, A. Sundblad, Evaluation antibodies to

estrogen receptors, Appl. Immunohistochem. 2 (1994) 141.

[18] L.P. Pertschuk, J.G. Feldman, Y.-D. Kim, et al., Estrogen

receptor immunocytochemistry in paraffin embedded

tissues with ER1D5 predicts breast cancer endocrine

response more accurately than H222Sp gamma in frozen

sections or cytosol-based ligand-binding assays, Cancer

77 (1996) 2514–2519.

[19] A. Rhodes, B. Jasani, D.M. Barnes, et al., Reliability of

immunohistochemical demonstration of oestrogen recep-

tors in routine practice: interlaboratory variance in the

sensitivity of detection and evaluation of scoring systems,

J. Clin. Pathol. 53 (2000) 125.

[20] A. Rhodes, B. Jasani, A.J. Balaton, et al., Immunohisto-

chemical demonstration of oestrogen and progesterone

receptors: correlation of standards achieved on ‘in house’

tumours with that achieved on external quality assessment

material in over 150 laboratories from 26 countries,

J. Clin. Pathol. 53 (2000) 292.

[21] A. Rhodes, B. Jasani, A.J. Balaton, et al., Study of

interlaboratory reliability and reproducibility of estrogen

and progesterone receptor assays in Europe, Am. J. Clin.

Pathol. 115 (2001) 44.

[22] A. Rhodes, B. Jasani, E. Anderson, et al., Evaluation of

HER-2/neu immunohistochemical assay sensitivity and

scoring on formalin fixed and paraffin processed cell lines

and breast tumours: a comparative study involving results

from laboratories in 21 countries, Am. J. Clin. Pathol. 118

(2002) 408–417.

[23] S. Romain, C.L. Bidron, P.M. Martin, H. Magdelenat,

On behalf of the EORTC receptor study group EORTC

receptor study group report: steroid receptor distribution

in 47, 892 breast cancers. A collaborative study of 7

European laboratories, Eur. J. Cancer 31A (1995)

411–417.

[24] S. Rossi, L. Laurino, A. Furlanetto, et al., A comparative

study between a novel category of immunoreagents and

the corresponding mouse monoclonal antibodies, Am. J.

Clin. Pathol. 124 (2005) 295–302.

[25] T. Rudiger, H. Hofler, H.H. Kreipe, et al., Qua-

lity assurance in immunohistochemistry: results of an

ARTICLE IN PRESSR. Rocha et al. / Pathology – Research and Practice 204 (2008) 655–662662

interlaboratory trial involving 172 pathologists, Am. J.

Surg. Pathol. 26 (2002) 873–882.

[26] P. Sannino, S. Shousha, Demonstration of estrogen

receptors in paraffin wax sections of breast carcinoma

using the monoclonal antibody 1D5 and microwave oven

processing, J. Clin. Pathol. 47 (1994) 90–92.

[27] T. Seidal, A.J. Balaton, H. Battifora, Interpretation and

quantification of immunostains, Am. J. Surg. Pathol. 25

(2001) 1204–1207.

[28] M. Stierer, H. Rosen, R. Weber, Immunohistochemical

and biochemical measurement of estrogen and progester-

one receptors in primary breast cancer: correlation of

histopathology and prognostic factors, Ann. Surg. 218

(1993) 13–21.

[29] F.A. Tavassoli, P. Devilee, World Health Organization

Classification of Tumours. Pathology and Genetics of

Tumours of the Breast and Female Genital Organs,

IARC Press, Lyon, 2003, p. 9.

[30] M.R. Wick, P.E. Swanson, Editorial: targeted con-

trols in clinical immunohistochemistry; a useful

approach to quality assurance, Am. J. Clin. Pathol. 117

(2002) 7.