distribution, frequency, and quantitative analysis of …...(cancer research 39, i447-1454, may...

(CANCER RESEARCH 39, i447-1454, May 19791

ABSTRACT

The distribution and frequency of steroid hormone receptorsare described in 329 patients with breast cancer. The distribution of each of the steroid hormone receptors is unimodalwitha progressiveincreaseintheproportionof patientspositiveat lower receptor values. Receptor values expressed as fmol/mg cytoplasmic protein are well correlated with values expressed as fmol/mg breast tumor. Estrogen receptor waspositive in 53% of the patients; progesterone receptor waspositive in 38% of the patients; glucocorticoid receptor waspositive in 52% of the patients; and androgen receptor waspositive in 31% of the patients. The type of tissue assayed didnot affect steroid hormone receptor positivity. For primarytumors, there was no correlation between steroid hormonereceptor positivity and location of the tumor in the breast, sizeof the tumor, or extent of the disease. Each of the steroidhormone receptors was positively associated with each of theother steroid hormone receptors. Estrogen receptor was correlated with menopausal status and axillary nodal status, butthese correlations did not exist for the other steroid hormonereceptors.Estrogenreceptorwasnot correlatedwith ageafteradjustment for menopausal status. The other steroid hormonereceptors were not correlated with age.

INTRODUCTION

Recently, greater understanding of the mechanism by whichsteroid hormones evoke specific phenotypic effects in targettissues has evolved. The hormones bind to specific receptorproteins located in the cytoplasm, and these hormone-receptorcomplexes translocate to the nucleus. In the nucleus, thesehormone-receptor complexes interact with chromatin leadingto qualitative and quantitative changes in hybridizable andtranslatable mRNA (14, 15).

Steroid hormone receptors are found in a variety of cancersincluding human breast cancer (14, 15, 17, 24). The distribution of ER2 in a breast cancer population is important since itplays a major role in the selection of patients for endocrinetherapy.

I To whom requests for reprints should be addressed, at Building 10, Room6B02, Medicine Branch, National Cancer Institute, NIH, Bethesda, Md. 2001 4.

2 The abbreviations used are: ER, estrogen receptor; PR, progesterone re

ceptor; AR, androgen receptor; GR, glucocorticoid receptor; R5020, 17,20-dimethyl-i 9-nor-4,9-pregnadiene-3,20-dione; Ri 881 , 17fl-hydroxy-1 7a-methylestra-4,9,i 1-triene-3,one; TED buffer, 0.01 PATris-HCI:0.15 M EDTA:0.5 mi,idithiothreitol, pH 7.4; DCC, dextran-coated charcoal: SSBG, sex steroid-bindingglobulin.

Received July 27. 1978: accepted January 25. 1979.

The present study describes the distribution and frequencyof ER, PR, AR, and GA in a large series of patients with primaryand metastatic breast cancer. The frequency of receptor positivity is also described as a function of tumor location, tumorsize, axillary nodal status,and extent of disease.Eachof thesteroid hormone receptors is analyzed quantitatively as a funclion of the other steroid hormone receptors. Their associationsand lack of associations with age and menopausal status aredescribed.

MATERIALSAND METHODS

Nonradioactive steroids 17f1-estradiol, 5a-dihydrotestosterone, dexamethasone, hydrocortisone, and progesterone werepurchased from Steraloids, Inc. (Pawling, N. J.). The nonradioactive compounds R5020 and Al 881 were gifts from Dr. JeanPierre Raynaud,Centre de Recherches,Roussel Uclaf (Romainville, France). The nonradioactive steroids were stored atâ€”20Â°as iO'@ M solutions in absolute ethanol. The labeledsteroids [2,4,6,7(n)-3Hjestradiol (-â€˜-90to 100 Ci/mmol), 5a-dihydro[1 ,2 ,4, 5,6,7(n)-3H]testosterone (â€”130 Ci/mmol),[1 ,2,6,7(n)-3H]progesterone (-â€˜-90to 100 Ci/mmol), and[1 ,2(n)-3H]dexamethasone (--25 Ci/mmol) were purchasedfrom Amersham-Searle Corp. (Arlington Heights, Ill.) and wereused without further purification. [3H]R5020 (86 Ci/mmol) waspurchased from New England Nuclear (Boston, Mass.). Theradioactive compounds, packaged in benzene:ethanol, weredried down under nitrogen and were stored at a concentrationof 5 x 10_6 Min absolute ethanol at _200. [3HjDexamethasonewas stored in ethanolat a concentrationof 10@ M.Protaminesulfate (histone free, Grade I from salmon) was purchasedfrom Sigma Chemical Co. (St. Louis, Mo.), as was charcoal(activated, untreated powder) and dextran (clinical grade; average molecular weight, @230,000).

Tissue Storage

Tumor specimens obtained at the Clinical Center, NIH, weretrimmed of fat and debris and, after suitable pieces weresubmitted to pathology, were transported on ice to the laboratory. After identifying data were recorded, the specimens werefrozen in liquid nitrogen and stored in a â€”70Â°freezer. Tissuespecimens from other institutions were sent on dry ice asrapidly after excisionas possibleand were storedas above.

Preparation of Cytosol

The frozen tissues were weighed and were pulverized to apowderwith a Thermovactissuepulverizer(ThermovacIndus

MAY 1979 1447

Distribution, Frequency, and Quantitative Analysis of Estrogen,Progesterone, Androgen, and Glucocorticoid Receptors inHuman Breast Cancer

Joseph C. Allegra, Marc E. Lippman,1 E. Brad Thompson, Richard Simon, Audrey Barlock, Linda Green,Karen K. Huff, Hoan My T. Do, and Susan C. Aitken

Medicine Branch (J. C. A., M. E. L., A. B., K. K. H., H. M. T. D., S. C. A.] and Laboratory of Biochemistry (E. B. T.], and Biometric Research Branch, ClinicalOncology Program (R. S.], National Cancer Institute, Bethesda, Maryland 20014, and Department of Oncology, Howard University, Washington, D. C. 20059(L. G.]

on March 5, 2020. © 1979 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

J. C. A!legra et a!.

tries Corp., Copiague, N. V.). The tissue was allowed to thawon ice in TED buffer, 1:2 (w/v). All subsequent procedureswere carriedout at 0 to 4Â°.Thetissuewas homogenizedwitha PolytronP-i 0-SThomogenizer(BrinkmannInstrumentsInc.,Westbury, N. Y.), using three 10-sec bursts interspersed with20-sec cooling periods. The homogenate was centrifuged at104,000 x g for 60 mm, and the supernatant (cytosoL) wascarefully removed with a Pasteur pipetto avoid the floating lipidlayer. The cytosol protein was estimated by measuring the260:280 n@absorbanceratio (7) andwas dilutedto a proteinconcentration of 1 to 2 mg/mI. Later, the protein concentrationwasdeterminedby the methodof Lowryet a!. (13).

Steroid Receptor Assays

The DCC-competitiveprotein bindingassaywas performedessentiallyas describedelsewhere(16). Briefly, this involvedan overnight incubation (at 0Â°)of 100 @lof cytosol with 100 @tlof labeled steroid prepared in TED buffer. A parallel incubationof cytosol with labeled steroid plus an excess of unlabeledsteroid or competitor compound was done to determine nonspecific binding. At the end of the incubation period, 1 ml of0.25% DCC solution (freshly prepared by making a 1:10 dilulion from a stock solution of 2.5% activated charcoal, 0.25%dextran, and 0.01 M Tris, pH 8.0, in TED buffer) was added.After vigorousmixingandstandingon ice for 15 mm,the tubeswere centrifugedat 1300 x g for 10 mm,except PRassays,which were centrifuged immediately.An 0.8-mI aliquotof thesupernatant was counted in 8 ml of aqueous counting scintillant(Amersham-Searle) or Aquasol (New England Nuclear) in aPackard liquid scintillation counter (efficiency for 3H, @40%).

The protamine sulfate assay used was our modification (10)of the techniqueof Chamnesset a!. (4), basedon the observation of the precipitation of steroid receptor by Steggles andKing (23). Cytosol (100 @d)is mixed with 200 @dof protaminesulfate solution (1.5 mg/mI in TED buffer) in 12- x 75-mm testtubes for 5 mm, followed by centrifugation at 1300 x g for 10mm. The precipitated receptor is in the form of a tightly adher

ent film on the bottom of the tube, allowing the supernatantcontaining the plasma transport proteins to be removed bysuction. The radioactive steroid solutions with or without excess unlabeled competitor were prepared in TED buffer, and200 @lwere added to each tube. After incubation overnight at40, the tubes were washed with three 1 .5-mi portions of cold

TEDbufferand decantedby inversion.Theboundsteroidwaseluted with two 1-ml portions of absolute ethanol. The ethanolwas decantedinto liquid scintillationvials, evaporatedto dryness, and counted as above.

Quantity of tissue available dictated the number of receptorsassayed, with ER given the highest priority. Whenever possible,binding to each receptor was examined by Scatchard plotanalysis (22). The calculations of binding data were accomplished utilizing computer-assisted methods (1).

ERAssay. ERwasdeterminedbyeitherthe protaminesulfateor DCC technique, both of which have been shown to giveessentially the same results in the absence of SSBG (10). Toblock any binding to SSBG in the DCC method, 108 M 5a-dihydrotestosterone was added to both the competed anduncompeted assay tubes. For Scatchard plot analysis, a set ofradioactive estradiol solutions was prepared with parallel solutions containing unlabeled estradiol at a final concentrationof 106 M.

PRAssay.PR'swereassayedonlybytheDCCmethodwith[3H]progesterone and unlabeled progesterone or with [3H]-R5020 and unlabeled R5020 when this synthetic progestinbecame commercially available. Hydrocortisone (5 x 1O@M)was addedto all PRassaytubes to block R5020 and progesterone binding to the GR and corticosteroid-binding globulin(11).

ARAssay. MostAR assayswere performedbythe protaminesulfate technique utilizing [3H1-5a-dihydrotestosterone and thehomologous unlabeled steroid as competitor. Recently, a synthetic androgen, Al 881 , has been shown to bind strongly tothe AR and the PR (18) with little affinity for SSBG (3). Therefore, the accurate determination of specific AR has becomepossible by the DCC method with the use of [3H]-5a-dihydrotestosterone and Ri 881 as the unlabeled competitor. The DCCmethod routinely gave lower dissociation constants. Ampledocumentation of the appropriateness for using this ligand forARstudieshasbeenpublished(19).

GRAssay.GA wasassayedbythe DCCmethodutilizingasingle saturating concentration of [3Hjdexamethasone of 5 x10_8 M and unlabeled homologous competitor of 5 x 10_6 Mto determine nonspecific binding.

Statistical Analyses

A positive ER, PR, or AR was taken to be equal to or greaterthan 10 fmol of [3H]estradiol, R5020, or dihydrotestosteronebinding per mg of cytoplasmic protein. A positive GR was takento be equal to or greater than 25 fmol of [3Hjdexamethasonebinding per mg of cytoplasmic protein. These cutoff valueswerechosenarbitrarily.

Comparison of assay values among 2 or more groups ofpatients was performed, using the Kruskal-Wallis nonparametric analysis of variance (9). Though results are often presentedas proportionpositive, the statistical significancetest utilizedthe actual values of the receptor assays. This test requires nodistributional assumptions about the data as would a t test.

Association among 2 quantitative variables was evaluatedusing the rank correlation coefficient of Kendall (5). It shouldbe noted that all of the data contribute to the Kendall rankcorrelations and the significance test depends in no way uponthe cutoff used nor upon the fact that the data are pictoriallypresented on a log scale. The Kendall rank correlation dependsonly on ranks, not upon the values of the receptor concentrations. This is also true of all other significance tests used. Thismethod was used in order that the significance of an association could be tested without assuming a linear relationship orbivariate normal sampling distributions. The rank correlationcoefficient does not, however, have the same interpretation asthe usual Pearson correlation R in the sense that R2 is theproportion of total variance explained by the linear association.All significance levels are of the 2-sided type.

RESULTS

Three hundred twenty-nine patients with breast cancer hadone or more steroid hormone receptor assays performed ontissue from either their primary tumor or from a metastasis. Thecharacteristics of these patients are shown in Table 1. Themedian age of the patients was 52.5 years. Twenty-seven %were premenopausal, and 73% were postmenopausal. Nodalstatus, location of the primary tumor, and location and size of

CANCERRESEARCHVOL. 391448



Steroid Hormone Receptors in Human Breast Cancer

Table1 the largest primary tumor nodule are listed. Fifty-two % of theCharacteristics of the patients whose tumors were assayed for patients had no evidence of macroscopic disease after their

steroid hormone receptors mastectomy was performed; 45% had metastatic disease atNo.of patients 329 the time of the assay, and the final 3% had locally unresectable

Ageof patients <40 15% disease.40-49 22% In Chart 1, the frequency distributions of the ER, PR, AR,50-59 38% and GR are plotted. The panels suggest that there is a cluster60-69 16% of patients with very small receptor concentrations and many>70 9% patients with larger but variable concentrations. The apparentMedian 52.5 yr . . . . . .

unimodality of some of these distributions IS not evidence forMenopausalstatus Pre 27% the nonexistence of 2 populations of tumor cells (negative and

Post 73% positive). Obviously, the number of potential peaks is influenced by grouping interval, measurement variability in each

Nodalstatusat diagnosis 0 45% range of receptor concentration, and scale of measurement.>4 29% Any unimodal distribution can be converted to a bimodal dis

tribution by an order-preserving transformation. Consequently,Location of primary tumor Left breast 48% modality cannot be used as a sole basis for inferring the

Rightbreast 4@ number of underlying populations unless one is certain of the

biologically appropriate scale of measurement.Locationof the largestprimary Upper-outer 47% In Chart 2, steroid hormone receptor values are shown,

tumor nodule Upper-inner 13% expressed as either fmol/mg cytoplasmic protein or fmol/mgLower-outer 15% breast tumor. There is a strong correlation between steroidLower@@'nner 21% hormone receptor values expressed in each manner. For ER,

the Kendall rank correlation coefficient is 0.80, and this corSizeof the largesttumornodule 0-2 cm 42% relation is significant at the p < 0.0001 level. This same

2-5 cm 39% correlation exists for the PR, AR, and GA, and data for all 3@ cm 17% receptorshaveKendallrankcorrelationcoefficientsgreater

ixe than 0.80 when receptor values expressed as fmol/mg cytoDiseasestatus Noevidenceof disease 52% plasmic protein are plotted against receptor values expressed

Local 3% as fmol/mg breast tumor. This indicates that expression ofMetastatic 45% data in each manner is equivalent and, most important, that the

â€˜@[A Bâ€˜Â°Fâ€¢ 9D@at

@7o 10

@5o at

@:20

10 10

C â€˜ 10'20 â€˜ 30.40 â€˜ 50'*t@ 70-at â€90-iOcVâ€•100150 200@250@ > XXI â€I 10.20 30.40 atlO@ 7&*t 90.100 @â€œ[email protected] 20.31 40.50 @070 80.90 lat.2at 250.@ 1.10 @@30 4050 60-70 80.90 150.@0 250300

ER tfmoi/mg CVTOP*.ASMIC PROTEINI PR ttmot/mg CYTOPLASMIC PROTEIN)

90 c D

at

â€˜I, at

z@50 50

@40 40@

@30

@30 â€¢

10 â€¢ 10 â€¢ â€¢ â€¢ â€¢ â€¢

I I I I t e e S C@ //@ â€¢ I I I I I I I@@ f@ i f â€¢ ,,,@ r@

>1@@ l03020.3030@Â°505050@e070@atat@l00 150200 200.250 @1 10.20 30.40 5050 7050 90-100 110120 130-140 â€œi50200 2@0300. . 1@10 3030 [email protected] 80.70 50.50 too-no 120130 140 150 200.250

AR )[email protected] PROTEIN) GE 1rnol/rng CYTOPt.ASM)C PROTEIN)

Chart 1. The distribution of the steroid hormone receptors. A, ER; B, PR; C, AR: 0, GR.

MAY 1979 1449



No. of assays%positiveER32853PR17638GA18752AR20831

J. C. Allegra et a!.

270W A

240

.@o

I: .U

@ 120@

@8o -.40@ ..

0 ,,@@@:@ ._:@__@@.. .@_@ ... ..

(78) 0.0 0.8 1.6 2.4 3.2 4.0 4.8 5.6 6.4 7.2 8.0 8.8 9.6 10.411.212.012.8PR fmol/mg BreastTumor)

160

@80

40

n(42)0 2 4 6 8 10 12 14 16 18 20 22

GR[ tmoiimg BreastTumor]

- Chart 2. The relationship between steroid hormone receptor values expressed as fmol/mg cytoplasmic protein and fmol/mg breast tumor. A, ER; B, PR; C, AR;

0, GR. Numbers in parentheses, number of patients represented by a data point.

0 2 4 6 8 10 12 14 16 i8@@ 22 24ER lfmoVmg Breest Tumorl

C @, D

@ .@â€¢.....

(73)@@ . [email protected] 0.6 1.2 1.8 2.4 3.0 3.6 4.2 4.8 54 6.0 66 7.2 7.8 &4 9.0 9.6

AR (fmol/mg BreestTumorI

manner of expression of the data does not significantly influence the classification of tumors as positive or negative.

Table 2 lists the incidence of steroid hormone receptorpositivity among our patients with breast cancer, using thecutoff valuesdefinedin â€˜â€˜Materialsand Methods.â€˜â€˜TheERwaspositive in 53% of the tumors and the PR was positive in 38%.Fifty-two % of the tumors were positive for GR and 31% werepositivefor AR.

In Table 3, steroid hormonereceptor positivity is analyzedas a function of lateralityof the primarytumor. No correlationwas found betweenany of the receptorsand lateralityof thetumor. The numbers of patients with steroid hormone receptorassays and bilateral tumors are too few to analyze at thepresenttime.

In Tables 4 and 5, steroid hormonereceptor positivity ispresented as a function of location and size of the largestprimary tumor nodule. No correlation was found between thepresence or absence of any steroid hormonereceptor andeither location or size of the primary tumor.

In Table 6, steroid hormonereceptor positivity is analyzedas a function of extent of disease. Fifty-six % of patients withno evidence of disease after their mastectomy was performedhad a positive ER compared to 47% of patients who had theassay performedon a metastasis.This trend toward a lowerfrequencyof ERpositivityin metastatictumorswas not statis

Table2Frequency of steroid hormone receptor positivity in patients with

breast cancer

tically significant, nor was it evident for the other 3 steroidhormone receptors.

Table7 lists the frequencyof the steroidhormonereceptorsas a function of the type of tissue assayed. Breast, skin, node,or soft tissue containing metastatic breast cancer appeared tohave about an equal frequency of receptor positivity. Unfortunately, we have at present too few specimens from sites suchas lung, liver, or bone marrow to be able to comment on thefrequency of receptor positivity at these sites.

Chart 3 illustrates the quantitative relationship of each steroidreceptor versus each of the other steroid hormone receptors.For each pairing, the Kendall rank correlation coefficient ispositive, and all of the p values are highly significant. Theseanalysesdemonstratethat as steroidhormonereceptorvaluesincrease there is a corresponding increase in other steroidhormone receptor values. For example, 38% of all patients had

1450 CANCERRESEARCHVOL. 39

210

.@180

@ 150

U 1@

(7;

100

90

.â€¢s@ :... ;@:.,?:@;,@ â€¢



Steroid hormone receptor positivity as a function of laterality of theprimarytumorNo.positiveERPR

ARGRLeftbreast

RightbreastBilateral77/151

(51)085/157(54)

7/9 (78)35/80(44)

32/99 (32)28/86(33) 27/100(27)

2/3 (67) 3/4 (75)41/80

(51)50/99 (51)

4/4 (100)

nodule

BreastquadrantNo.

positiveERPRARGAUpper-outer

Upper-innerLower-outerLower-innerAreolar

a Numbers in34/68(50)a

7/25(29)10/1 8 (56) 4/1 0 (40)

9/21 (43) 3/11 (27)5/7 (71) 3/4 (75)

18/30 (60) 8/19 (42)

parentheses, percentages.12/41

(29)4/1 4 (29)3/13(23)2/4 (50)8/22 (36)16/39

4/1 29/124/4

1 1 /19(41)

(33)(75)

(100)(58)

Steroid hormone receptor positivity as a functiolargest primary tumor nodulen

of the sizeoftheSize

of No. positiveprimarytumor ER PRARGR0â€”2cm

43/84(51)8 11/39(28) 17/552â€”5cm 46/78(59) 22/45(49) 13/53>5cm 17/34(50) 7/14(50) 7/20Fixed 2/4 (50) 1/3 (33) 2/2

tumor(31)

(25)(35)

(100)22/48

25/487/162/2(46)

(52)(44)

(100)a

Numbers in parentheses, percentages.


Table3

a Numbers in parentheses, percentages.

Table4Steroid hormone receptor positivity as a function of location of the largest primary tumor

Table5 Table 10 shows the analysis of ER as a function of axillarynodal status. All ER assays were performed on tissue at thetime of mastectomy. Patients with ER-positive tumors have amore favorable axillary nodal status at the time of mastectomy.The ER-positive group contains a greater proportion of nodenegative patients. The Wilcoxon rank test (adjusted for ties andcontinuity) was used to compare the distributions, and thedistributions do differ significantly (p < 0.05). There was noassociation between the other steroid hormone receptors andaxillary nodal status.

DISCUSSION

In this report, we have presented data analyzing possiblecorrelations of all 4 steroid hormone receptors (estrogen, progesterone, androgen, and glucocorticoid) with multiple patientcharacteristics such as location and size of the breast tumor,nodal status, extent of disease, type of tissue assayed, patientage, and menopausal status. The frequency, distribution, andquantitative analysis of each of the steroid hormone receptorshas been presented along with an analysis of different meansof data expression. Several points are noteworthy.

First, all 4 steroid hormone receptors are present in ourpopulation of breast cancer patients. ER was positive in 53%of the tumor specimens, and PR was positive in 38%. Thesefrequencies are in good agreement with those reported byother authors (15, 17). Our frequency of 52% positivity for theGA is in good agreement with the 51% reported by Teulingsand Van Gilse (24). The 31% frequency for AR positivity is alsoin agreement with the reported frequencies for dihydrotestosterone binding published by Maass et a!. (14).

Second, the distribution of ER's as well as the distributionsof values for PR, AR and GA appeared unimodal. The proportion of patients increased toward lower receptor values. Thishas also been reported by Leclercq et a!. (8) for the ER. Also,it is of interest that different methods of expressing receptorvalues do not alter the distribution of receptor values. There isa very high correlation between any steroid hormone receptorvalue expressed as either fmol/mg cytoplasmic protein or

a positive PR, but 52% of ER-positive patients were positivefor PR. Only a very small percentage of patients possessedtumors which were ER negative but PR positive. Although therelationships among receptor values were statistically significant, they were not particularly strong correlations.

Table 8 shows analysis of each of the receptors as a functionof age. No correlation with age existed for PR, AR, or GA, butthere appeared to be a significant association between ER andage. The ER was positive in 44% of patients less than 40 yearsof age and in 69% of patients older than 70.

Chart 4 shows a more detailed analysis of the relationship ofER with age. In A, when ER is plotted versus age, the Kendallrank correlation coefficient is 0. 11 with a p value of less than0.005@Note, however, that when patients are separated intopremenopausal (B) and postmenopausal (C) subsets, receptorvalues no longer correlate significantly with age. This suggeststhat ER is actually correlating with menopausal status and thatmenopausal status correlates with age. This is demonstratedin Chart 5. In this chart, ER values are plotted in premenopausalpatients and postmenopausal patients. The median ER value issignificantly higher in the postmenopausal patients comparedwith the premenopausal patients (p < 0.02). Table 9 lists thepercentage of patient tumor samples positive for each of thesteroid hormone receptors as a function of menopausal status.There are no statistically significant associations betweeneither PR, AR, or GA and menopausal status, although thetrends are all toward postmenopausal patients having highervalues.

MAY 1979 1451



No.positiveExtent

of disease ER PR ARGANo

evidenceof disease 96/1 72 (56)8 34/93 (37) 38/1 20 (32) 54/1 12 (48)Metastaticdisease 68/145 (47) 30/80 (38) 23/83 (28) 40/72(56)a

Numbers in parentheses, percentages.Table

7Steroidhormone receptor positivity as a function of type of tissueassayedNo.

positiveTissue

assayed ER PR ARGABreast

109/202 (54)8 39/1 09 (36) 45/1 44 (31) 65/1 32 (49)Skin, soft tissue 34/74 (46) 16/36 (44) 12/35 (34) 15/27 (56)Node 17/35 (49) 8/21 (38) 7/22 (32) 15/23 (65)Other 7/17 (41) 2/10 (20) 0/7 (0) 2/5(40)a


Table6Steroid hormone receptor positivity as a function of extent of disease

a.@@ Con@[email protected]

32

3.6Ip( a@t

3.2

A B

0.0 0.2 OA 0.6 0.8 10 1.2 14 16 ii 2.0 2.2 2.4 2.6 28 3.0 3.2 3.4

LOGI 1+@ Rscspi@I

3.6@ C

(PS 0.0120132

j@e

04

0.0 0.2 04 as 0.8 1.0 1.2 14 1.6 1.0 2.0 2.2 2.4 26 2.8 3.0

LOG I1+M&ogatRateÃ§*orj

3.0 kat@ @,*[email protected] DlP<0.000ll

3.2

12J

J2:4@

@z0.}1.0rnb. . ......+ 1.2 . . . .

â€” a â€” .

@0J_@ @. .

a4

nfl .(50).

I

0.0 02 0.4 0.5 0.5 1.0 12 1.4 1.5 L5 LU 22 14 15 iS 3.0 32 3.4

LOG (1+ Gk@100ar@00iJRecepior)

34â€¢ @*c@[email protected] E1p'a@l

3.2

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.0 1.0 2.0 22 2.4 2.6 2.8 10LOG I1+An&ogenRecepiori

3. x@[email protected]@-a17 F3. (p<0.lXISl

[email protected]

i@11

@ 12

0.4

@2J

hA12.0

t@ 1.6

@ 12

@ 0.0

04

aCL

@!@@ N@@@@@

11.0 0.2 04 0.6 0.81.012 1A 1.6 1@ 2.0 2.2 2.4 2.6 2.8 10 3.2 14 0.0 0.2 0.4 0.0 0.8 1.0 12 1.4 16 18 2.0 2.2 2.4 2.6 2.0 10

LOG I1*Gk.cocw&oidRecesorl I-os [i+snciogennecepsor]

Chart 3. The relationship between the quantitative amounts of steroid hormone receptor. A, ER and PA; B, ER and AR: C, ER and GA: 0, PR and AR: E, PA andGA: F, GA and AR. Numbers in parentheses, number of patients represented by a data point. The units for receptor concentration are fmol/mg cytoplasmic protein.


J. C. Allegra et a!.

2.1

J7.4L2.c

I â€˜@+:. 12

@0j

OA

an

@::@rÂ°@

-â€˜an@@

I

,(z)l â€¢



Steroid hormone receptor positivity as a function ofageNo.positiveAgeER

PRARGA<40

40-4950â€”5960â€”69>7022/50

(44)4 8/26(31) 13/37(35)34/71 (48) 12/33(36) 10/44(23)66/124(53) 28/67(42) 23/75(31)31/53 (58) 12/33(36) 11/33(33)18/26 (69) 4/1 4 (29) 6/1 5 (40)14/30(47)

20/36(56)34/66(52)16/34(47)1 1 /1 7(65)a

Numbers in parentheses,percentages.

Steroidhormonereceptorpositivityas a functionof menopausaIstatusNo.

positiveMenopausal

status ER PRARGRPre

36/80 (45)4 12/44 (27) 14/52(27)21/46(46)Post

122/214(57) 44/116(38)43/131(33)66/119(56)a



fmol/mg breast tumor. Of more importance is the fact thatalternate methods of data expression do not alter the classifi

16@ Rankcorrelation-Oil

3.2@ A

0.)) ,, 23 31@ 47 00 @3 71 79

AGEATASSAY

3.6@ KendeeRankcorreletionO.06

3.2@ C

cation of values as positive or negative; that is, patients positiveby one method of expression are also positive by the other.

Third, no correlation was found between any of the steroidhormone receptors and laterality of the breast tumor, locationand size of the primary tumor, extent of disease, or type oftissue assayed. This lack of correlation has been previouslydescribed by McGuire et a!. (15) for the ER but not for PR, AR,or GA. Pollow et a!. (21) reported a lack of correlation betweenPR valuesand tumorhistology.

Fourth, none of the steroid hormone receptors correlatedwith age. There was an increase in ER values with increasingage but this correlation did not exist within the pre- and postmenopausal subsets. There was a strong correlation notedbetween ER values and menopausal status. Neither PR, AR,nor GA was significantly associated with menopausal status.Leclercq et a!. (8) have reported a similar association betweenER values and menopausal status. They find higher ER valuesin postmenopausal women. Kiang and Kennedy (6) report ERpositivity in 61% of postmenopausal patients as compared to

I@â€” C @_,.@ .@ .@ .@

Table8

16@ Re,*cor@elato@-a00 Btp > .781

32

z3 31 30 47 00 @3 71 79AGEAT ASSAY

3.6

3.2

@ 2.8

@ 2.4C

I 2.0W

@ 1.6

@ 1.2

0.8

0.4

a

Â£

Ia

a

. 4.

/, @. .@@ @â€”@â€”!@ f@

23 31 30 47 00 63 71 79

AGEAT ASSAY ..

Chart 4. The correlation between age and ER value. A, all patients; B, premenopausal: C. postmenopausal. The units for ER concentration are fmol/mgcytoplasmic protein.

Premenopausal Postmenopausal

Chart 5. The correlation between ER's and menopausal status. The units forEA concentration are fmol/mg cytoplasmic protein.

Table 9

MAY 1979 1453



Estrogen receptor status as a function of axillary nodalstatusNo.of positive

axillary nodes ER positiveERnegative0

52(61)81â€”3 13(15)4â€”10 10(12)>10 10(12)30(41)

23(31)13(18)

8(11)a


J. C. A!!egra et a!.

2. Barnes, D. M., Ribeird, G. G., and Skinner, L. G. Two methods for measurement of oestradiol 17$ and progesterone receptors in human breast cancerand correlation with response to treatment. Eur. J. Cancer, 13: 1133â€”1143,1977.

3. Bonne, C., and Aaynaud, J. P. Methyltrienolone, a specific ligand for cellularandrogen receptors. Steroids, 26: 227â€”232,1975.

4. Chamness, G. C.. Huff, K., and McGuire. W. L. Protamine precipitatedestrogen receptorâ€”a solid phase ligand exchange assay. Steroids, 25:627â€”635,1975.

5. Kendall, M. G. Rank Correlation Methods; New York: Hafner Publishing Co.,Inc.,1962.

6. Kiang, 0. T., and Kennedy, B. J. Factors affecting estrogen receptors inbreast cancer. Cancer (Phila.), 40: 1571â€”1576, 1977.

7. Layne, E. Spectrophotometric and turbidimetric methods for measuringproteins. Methods Enzymol., 3: 451â€”454,1957.

8. Leclercq, G., Heuson, J. C., Deboel, M. C., and Mattheiem, W. H. Oestrogenreceptors in breast cancer. A changing concept. Br. Med. J., 1: 185â€”187.1975.

9. Lehman, E. H. Non-parametrics: Statistical Methods Based on Ranks, p.208. San Francisco: Holden-Day, Inc., 1975.

10. Lippman, M. E., and Huff, K. K. A demonstration of androgen and estrogenreceptors in a human breast cancer using a new protamine sulfate assay.Cancer (Phila.). 38: 868â€”874,1976.

11. Lippman, M. E.. Huff. K.. Bolan, G.. and Neifeld, J. V. Interaction of A5020with progesterone and glucocorticoid receptors in human breast cancer andperipheral blood lymphocytes in vitro. In: E. E. Banlieu, W. L. McGuire, andJ. P. Raynaud (eds.), Aoussel Workshop on Application of A5020 to theDetection of Progesterone Aeceptors, pp. 193â€”210. New York: RavenPress, 1977.

12. Liskowski, L., Rose, D. P., Dondlinger, T., and Olenick, J. S. The determination of progesterone receptors in breast cancer and their relationship toestrogen receptors. Clin. Chim. Acta, 71: 309â€”318,1976.

13. Lowry, 0. H.. Rosebrough, N. J., Farr, A. L., and Randall, A. J. Proteinmeasurement with the Folin phenol reagent. J. Biol. Chem., 193: 265â€”275,1951.

14. Maass, H., Engel, B., Trams, G., Nowakowski, H., and Stolzenbach, G.Steroid hormone receptors in human breast cancer and the clinical significance. J. Steroid Biochem., 6: 743â€”749,1975.

15. McGuire, W. L., Carbone, P. P., and volimer, E. P. (eds.). Estrogen receptorsin Human Breast Cancer. New York: Raven Press, 1974.

16. McGuire, W. L., and DeLaGarza,M. Improved sensitivity in the measurementof estrogen receptor in human breast cancer. J. Clin. Endocrinol. Metab.,36: 548-552, 1973.

17. McGuire, W. L., Aaynaud, J. P., and Baulieu, E. E. (eds.). ProgesteroneAeceptors in Normal and Neoplastic Tissues. New York: RavenPress, 1977.

18. Menon, M., Tananis, C. E., Hicks, L. L., Hawkins, E. F., McLaughlin, M. G.,and Walsh, P. C. Characterization of the binding of a potent syntheticandrogen, methyltrienolone, to human tissues. J. Clin. Invest., 61: 150â€”162,1978.

19. Ojasoo, T., and Raynaud, J. P. Unique steroid congeners for receptorstudies. Cancer Res., 38: 4186-4198. 1978.

20. Pichon, M. F., and Milgrom, E. Characterization and assay of progesteronereceptor in human mammary carcinoma. Cancer Res., 37: 464â€”471, 1977.

21. Pollow, K., Sinnecker. R., Gollwitzer, M. S., Boquoi, E., and Pollow, B.Binding of [3HJprogesterone to normal and neoplastic tissue samples fromtumor bearing breast. J. Mol. Med., 2: 69â€”82,1977.

22. Scatchard, G. The attraction of proteins for small molecules and ions. Ann.N. Y. Acad. Sci., 51: 660-672, 1949.

23. Steggles, A. W., and King, A. J. B. The use of protamine to study 6,7-3H-oestradiol-1 7$ binding in rat uterus. Biochem. J., 118: 695â€”701, 1970.

24. Teulings, F. A. G., and van Gilse, H. A. Demonstration of glucocorticoidreceptors in human mammarycarcinomas. Horm. Res. (Basel), 8: 107â€”116,1977.

25. Walt, A. J., Singhakowinta, A.. Brooks, S. C., and Cortez, A. The surgicalimplications of estrophile protein estimations in carcinoma of the breast.Surgery, 80: 506-512, 1976.

Table 10


29% in premenopausal patients, and Walt et a!. (25) notedmedian ER values in postmenopausal patients which wereapproximately 2-fold higher than in premenopusal patients.Multiple authors have reported lack of correlation of ER withage but a strong correlation with menopausal status (15). Withregard to PR, Liskowski et a!. (12), Pichon and Milgrom (20),Pollow et a!. (21), and Barnes et a!. (2) have reported nocorrelation between PR and age, and Pichon and Milgrom (20)were also not able to correlate PR with menopausal status inagreement with @urdata. No other data analyzing either AR orGA in breast cancer specimens as a function of age or menopausal status have been reported thus far.

Fifth, ER status was correlated with axillary nodal status,with the ER-posjtive group containing a high proportion ofnode-negative patients.

Finally, it is interesting that a quantitative analysis of steroidhormone receptor values demonstrates correlations among allof the receptors. Plotting values of any one receptor versusany other receptor results in a positive Kendall rank correlationwhich is highly significant. As receptor values increase for onereceptor, there is a corresponding increase in values for theother receptors. This quantitative relationship between ER andPR values has been previously described by McGuire et a!.(1 7), Liskowski et a!. (1 2), and Pichon and Milgrom (20).

Barnes et a!. (2) were not able to demonstrate a correlationbetween ER and PR values.

ACKNOWLEDGMENTSWe wish to thank the many collaborating physicians who submitted clinical

material for analysis and allowed examination of their patients records, includingDr. Alford, Dr. Blom, Dr. Borssuck, Dr. Camalier, Dr. Casey, Dr. Egan, Dr.Funderburk, Dr. Gold, Dr. Hiltabidle, Dr. Isaacson, Dr. Jonas, Dr. Jones, Dr.Kiernan, Dr. Kim, Dr. Kirson, Dr. Kison, Dr. Kleinberg, Dr. MacDonald, Dr.Mitchell, Dr. Moore, Dr. Palmer, Dr. Perlin, Dr. Schein, Dr. Scully, Dr. Thistlethwaite, Dr. Tsangaris, Dr. Watkins. and Dr. Wherry.

We would also like to thank C. Baughman and v. Tinley for their assistance inreviewing patient records at the George Washington University Medical Center,M. Allegra for her assistance in reviewing the records at Georgetown University,and P. De Steffano for assistance in data management and computerization.

We would also like to thank Dr. D. C. Tormey (formerly of the Medicine Branch)and Dr. A. C. Young and the members of the Medicine Branch, National CancerInstitute, for their assistance with this study.

REFERENCES1. Aitken, S. C., and Lippman, M. E. A simple computer program for quantitation

and Scatchard analysis of steroid receptor proteins. J. Steroid Biochem., 8:77â€”94,1977.



1979;39:1447-1454. Cancer Res Joseph C. Allegra, Marc E. Lippman, E. Brad Thompson, et al. Human Breast CancerProgesterone, Androgen, and Glucocorticoid Receptors in Distribution, Frequency, and Quantitative Analysis of Estrogen,

Updated version

http://cancerres.aacrjournals.org/content/39/5/1447

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/39/5/1447To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



distribution, frequency, and quantitative analysis of …...(cancer research 39, i447-1454, may...

Documents