endocrine-responsive pancreatic carcinoma: steroid binding ... · sex-steroid dependent; in this...

[CANCER RESEARCH 46, 2276-2281, May 1986]

Endocrine-responsive Pancreatic Carcinoma: Steroid Binding and CytotoxicityStudies in Human Tumor Cell Lines1

Chris Benz,2 Charlene Hollander, and Becky Miller

Cancer Research Institute, University of California, San Francisco, San Francisco, California 94143

ABSTRACT

We have begun to investÃgatethe steroid responsiveness of pancreaticcancer by comparing human (MiaPaCa, Colo-357, RWP-1, RWP-2) androdent (AR42J) pancreatic tumor cell lines with cultured estrogen receptor-positive breast cancer cells (MCF-7, T47-D). The four human pancreatic tumors contain measurable levels of specific est radio! bindingsites with dissociation constants (A"d)that range from 1 to 9 UM, in

contrast to the higher-affinity binding sites measured in the breast cancercells (Ad < l UM). Growth of one pancreatic tumor line (MiaPaCa) isstimulated 40% above control by exposure to nanomolar concentrationsof fstradiol, suggesting that the estrogen receptor in these cells isfunctioning like that in MCF-7 and T47-D cells. Glucocorticoids (dexa-methasone, hydrocortisone) and androgen (fluoxymesterone) stimulateproliferation of Colo-357 cells by as much as 30%. Paradoxically, gln-cocorticoids inhibit AR42J cells to less than 50% of control growth.Micromolar exposures of estrogen (17/3-estradiol), antiestrogen (tamox-ift-n), antiandrogen (dehydroxyflutamide), progestins (progesterone,K5(12(1.medroxyprogesterone acetate), and inhibitors of steroid-metabolizing enzymes (170-N,N-diethylcarbainyl-4-inethyl-4-aza-5a-androstan-3-one, danazol) impair growth of these pancreatic tumors to varyingdegrees, and with little relationship to estrogen receptor content. Ingeneral, progestins are slightly more growth inhibiting to these pancreatictumor lines than the other endocrine agents tested, including tamoxifen.Only the RWP-2 cells appear completely resistant to steroidal therapy,showing less than 25% growth inhibition with exposure to therapeuticconcentrations (<2.5 Â«tM)of these agents. Colo-357, MiaPaCa, andAR42J cells are most responsive to these endocrine agents, and theiroverall pattern of sensitivity suggests that the steroid-dependent growth-inhibitory mechanisms of some pancreatic carcinomas may involve bothreceptor antagonism and direct inhibition of steroidal oxidoreductases.170-W,JV-Diethylcarbamyl-4-methyl-4-aza-5a-androstan-3-one, a potentinhibitor of 5a-reductase with minimal affinity for androgen receptor,inhibits growth of Colo-357 cells to less than 40% of control and alsoinhibits AR42J and MiaPaCa cells. Dehydroxyflutamide, a potent androgen receptor antagonist with no direct influence on 5a-reductase activity,inhibits growth of MiaPaCa and AR42J cells but has no affect on Colo-357 growth. These studies emphasize the need for further preclinical andclinical investigations to determine whether endocrine therapy can be ofany benefit in the treatment of pancreatic cancer.

INTRODUCTION

Steroidal and nonsteroidal endocrine agents are important inthe therapy and control of advanced breast, endometrial, andprostate cancers. For breast cancer in particular, the presenceof specific estrogen binding in fresh biopsy specimens determines which tumors respond to endocrine therapy. ER3 has

been found in a variety of other human cancers includingsarcomas, meningiomas, melanomas, renal cell and ovariancarcinomas, as well as hepatomas, colorectal, and pancreatictumors (1, 2). However, in very few of these other neoplasmshas the potential benefit of endocrine therapy ever been ade-

ReceivedS/21/8S;revised11/27/8S;accepted1/16/86.1Supported by Grant CH-235B from the American Cancer Society and Grants

CA 36773 and CA 36769 from the National Cancer Institute.2To whom requests for reprints should be addressed.1The abbreviations used are: ER, estrogen receptor, IDU, concentration re

sulting in 25% growth inhibition; PCS, fetal calf serum; ITS, serum-free supplement containing insulin, transferrin, and selenium; IXC, dextran-coated charcoal; dFLUT, dehydroxyflutamide; DM.AA. 17,Â¡,V.\ tlicthylciirbumoyM-methyl-4-aza-5a-androstan-3-one; DES, diethylstilbestrol.

quately assessed or correlated with receptor content (3-6).In the United States, non-islet cell carcinoma of the pancreas

has been increasing at the alarming rate of 15% per year andrepresents the fourth leading cause of cancer mortality (7-9).The overall response rate of pancreatic cancer to single agentor combination chemotherapy is less than 40%; thus, it isimperative that we investigate new approaches for the treatmentof this lethal cancer. Recently, there have been reports describing estrogen binding in normal apocrine cells from mammalianpancreatic tissue (10-13). In eight of nine human pancreaticcarcinoma tissue specimens, Greenway et al. measured highlevels of high affinity ER associated also with aromatase and5a-reductase activity, two microsomally associated enzymesfound only in sex steroid-dependent tissues (14-16). We havebegun to investigate the possibility that pancreatic cancer issex-steroid dependent; in this paper, we compare cultured human breast with rodent and human pancreatic tumor cell linesfor estrogen binding and sensitivity to cytotoxic endocrineagents.

MATERIALS AND METHODS

The rat pancreatic (AR42J). human pancreatic (Coio-357, MiaPaCa,RWP-1, RWP-2), and human breast (MCF-7, T47-D) cancer cell linesare maintained as monolayer stock cultures in 5% CO, incubators at37V, using RPMI 1640 medium (Gibco Labs, Grand Island, NY)supplemented with 5-10% PCS (HyClone Labs, Logan, UT). Thisserum has been commercially analyzed and found to contain the following average concentrations of endogenous steroids: Cortisol, 380 +192ng/dl; dehydroepiandrosterone, 34.3 Â±9 ng/dl; androstenedione, 10.0Â±3 ng/dl; progesterone, 14.3 Â±11 ng/dl; testosterone, 13.7 Â±4 ng/dl;17,i i'stnidiol, 11.1 Â±2 pg/ml; and estrone, 22.7 Â±7 pg/ml. The ER-positive MCF-7 and T47-D cells are additionally supplemented withinsulin (0.2 lU/ml; Sigma, St. Louis, MO), and their characteristicshave previously been described in detail (17, 18). All stock cultures arepassaged weekly no more than 25 times before being replaced by cellstocks from medium containing 10% dimethyl sulfoxide and 15% PCSfrozen in liquid .V.

The Colo-357 cell line was derived from a metastatic deposit of awell-differentiated mucin-secreting pancreatic adenocarcinoma resectedfrom a post menopausa 1woman (19). The MiaPaCa cell line was derivedfrom an undifferentiated epithelial carcinoma occurring in the bodyand tail of the pancreas in a 65-yr-old man (20). The RWP-1 and RWP-2 cell lines were independently derived from two distinct, moderatelywell-differentiated pancreatic adenocarcinomas and generously provided by M. D. Turner (21). The AR42J cell line was originally derivedfrom a hyperplastic nodule of the exocrine pancreas from a rat treatedwith azaserine and is known to be glucocorticoid sensitive (22).

The commercially obtained endocrine agents used in this study are17,i-est radiili, diethylstilbestrol, progesterone, fluoxymesterone, dexa-methasone, hydrocortisone, aminoglutethimide (Sigma), promegestone(R5020); and [2,4,6,7-'H]estradiol at 115 Ci/mmol (New England

Nuclear, Boston, MA). The following compounds were generouslyprovided by their manufacturers: tamoxifen citrate (Stuart Pharmaceuticals, Wilmington, DE); danazol (Sterling-Winthrop, Rensselaer, NY);medroxyprogesterone acetate (Upjohn, Kalamazoo, MI); llÃŸ-NJV-di-ethylcarbamoyl-4-methyl-4-aza-5Â«-androstan-3-one (Merck Sharp andDohme, Railway, NJ); and dehydroxyflutamide (Schering, Bloomfield,NJ).

ER is biochemically characterized using Scatchard analysis (23) of

2276

Association for Cancer Research. by guest on August 27, 2020. Copyright 1986 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

https://bloodcancerdiscov.aacrjournals.org

ENDOCRINE-RESPONSIVE PANCREATIC CARCINOMA

specific estradici binding measured by the whole cell assay of Shadeand Brooks (24). For monolayer cultures of breast cancer cell lines, thiswhole cell assay yields results similar to those obtained from assay ofbroken cell preparations (25).

For assessment of cytotoxicity or growth inhibition, cells (15-100 xIO3) are plated into triplicate 25-cnr sterile plastic flasks (Costar,Cambridge, MA) and allowed to attach for 24-36 h before addition ofspecific endocrine agents. In experiments using variable amounts ofFCS or serum-free, chemically defined medium (ITS; CollaborativeResearch, Lexington, MA), stock cultures in 75-cm2 flasks are first

exposed to selected medium conditions for 48 h prior to trypsinizationand passage into 25-cnr flasks. To permit cell attachment in serum-free ITS medium, flasks are precoated with a baked film of DCC-stripped serum (26). Cells grown in medium containing 5% FCS, 10%IX '( '-strum, or ITS have greater doubling times than those grown in

5-10% FCS, but they remain viable for at least 14 days as judged bydye exclusion, surface attachment, morphological appearance, and repeat plating efficiency. All endocrine agents are added from ethanol-con taming stock solutions to a final etHanoi concentration <0.1%;control flasks contain an equivalent ethanol concentration which doesnot inhibit culture growth. Exposure to exogenous agents is continuous,and flasks are harvested at 5-7 days when control culture growth isstill logarithmic. Cells are counted on a Coulter Model ZB1 counter(Hialeah, FL); in selected experiments, cell colonies were enumeratedon a Biotran II (New Brunswick Scientific, Edison, NJ) by a previouslydescribed technique (27). Results of growth inhibition studies are reported as: (a) mean (Â±SD)number of cells counted; (b) mean (Â±SD)percentage of control number of cells or colonies counted; or (c) ID25.

RESULTS

The two ER-positive human breast cancer cell lines, MCF-7and T47-D, have long been used to study mechanisms underlying receptor-mediated cytotoxicity by endocrine agents suchas tamoxifen. Fig. \A shows the dose-dependent growth inhibition of these two cell lines by tamoxifen, an antiestrogenbelieved to act via ER. The number of MCF-7 and T47-D ER

100

â€¢MCF-7

T470

0.1 0.5 1.0

TAMOXIFEN (uMi

Â§ 100U* 80

60

40

20

00.1 0.5 1.0

PROGESTERONE (UMÂ¡

10

Fig. 1. Sensitivity of two estrogen receptor-positive human breast cancer celllines (MCF-7, T47-D) to tamoxifen (A) and progesterone (B). Triplicate 25-cm2plastic flasks were seeded with 15 x Id' cells in medium supplemented with

insulin (0.2 lU/ml) and 10% fetal calf serum. After 24 h, tamoxifen and progesterone were added at the concentrations shown above, and cells were counted Sdays later, at which time control flasks contained a mean of 7.1 x 10s cells

(Â±9%).Results are given in percentage of control growth (Â±SD).

binding sites and their affinity for estradiol (K,Â¡)were determined to be 35 and 37 fmol/10' cells, 0.2 and 1.0 DM, respec

tively. Under these experimental conditions, it is believed thatthe greater sensitivity of MCF-7 cells for tamoxifen is explainedin part by its more avidly binding ER, although previous studiesusing different assay techniques suggest that MCF-7 cells mayalso contain more ER binding sites than T47-D cells (28, 29).Both these cell lines are also sensitive in vitro to progestins(Fig. IB), which may act via steroid receptors or by directinteraction with microsomally associated steroid binding enzymes (30-35). Previous analysis has shown that these T45-Dcells contain 50-fold greater levels of cytosolic progesteronereceptor than the MCF-7 cells (28). Although standard therapeutic doses of these endocrine agents usually achieve 0.8-2.5pM serum concentrations of the parent compound, some high-dose regimens produce serum levels as high as 30 Â¿tM(36-38).Most investigators believe that continuous in vitro exposure ofthese cell lines to >5 /Â¿Mtamoxifen in serum-containing medium produces nonspecific toxicity, not mediated via ER.

The human pancreatic carcinoma cell line, Colo-357, wasfirst shown to contain steroid binding sites by a flow cytometrictechnique using a fluoresceinated estradiol conjugate; this binding was partially displaced by DES at concentrations known tosaturate high affinity ER (39). To determine whether these cellsmight exhibit a pattern of endocrine sensitivity similar to thebreast cancer cell lines, the Colo-357 cells were exposed to avariety of steroidal and nonsteroidal endocrine agents. In thesestudies, glucocorticoids such as dexamethasone and hydrocortisone, at concentrations ranging from 0.01-10 MM, actuallystimulated Colo-357 clonogenic growth by 5-30% and decreased culture doubling time from 28-20 h. Fluoxymesterone,an androgen agonist, had somewhat less stimulatory effect (upto 15%) on Colo-357 growth over this same concentrationrange. In contrast, tamoxifen, estradiol, progesterone, and dan-azol showed progressively more potent in vitro growth-inhibiting activity (Fig. 2). In particular, progesterone and danazolwere substantially more effective against Colo-357 than either

P,08â€žt.,oâ€ž,

0 2 4 6 8 10

Fig. 2. Sensitivity of the human pancreatic carcinoma cell line, Colo-357, tofour different endocrine agents. Cells were plated into tissue culture flasks withmedium containing 10% FCS, as described in Fig. 1. Varying concentrations ofeither tamoxifen, estradiol, progesterone, or danazol were added at 24 h later,and cell colonies were counted 5 days later using an automated colony counter(27). As shown, results are recorded as the percentage of control (Â±SD)colonygrowth.

2277




MCF-7 or T47-D cells. Danazol is both an antiprogestin andan antiandrogen known to inhibit a wide variety of steroidaloxidoreductases (40); it has little if any effect on the breastcancer cells. While some experiments have shown that l MMtamoxifen can inhibit Colo-357 growth up to 15% (variabilitydepending on serum content), these cells are not nearly assensitive to antiestrogens as are the MCF-7 and T47-D cellswhen tested under identical conditions.

These endocrine agents were also compared for growth-inhibitory activity against four other pancreatic carcinoma celllines, three of human origin (MiaPaCa, RWP-1, RWP-2) andone of rat origin (AR42J). In Table 1, the four human cell linesare listed in order of endocrine responsiveness; three of thesecell lines show growth inhibition >25% by at least one agent ata continuous exposure dose <2.5 MM.RWP-2 cells, relativelyresistant to all four compounds, have also been shown to lackDES-displaceable binding by flow cytometry using the previously mentioned fluoresceinated conjugate (39). These data alsodemonstrate that the MiaPaCa, Colo-357, and RWP-1 cellshave a pattern of sensitivity that is unlike the breast cancercells, with the single most potent endocrine agent varying fromprogesterone (MiaPaCa) to danazol (Colo-357) to tamoxifen(RWP-1). Medroxyprogesterone acetate and R5020 are twoother progestins tested against MiaPaCa and Colo-357 cells.In general, the growth-inhibiting potency of these progestins iscomparable to that of progesterone itself, with 2.5-10 MMR5020 reducing Colo-357 growth from 25% to less than 0.01 %

of control.One important difference between the rat and human pan

creatic carcinoma cells is in their sensitivity to the antiandrogen,dFLUT. This avid receptor-binding metabolite of flutamide hasno reported ability to inhibit 5Â«-reductase. The breast cancercells T47-D and MCF-7, which are known to contain androgenreceptor, are maximally inhibited 20-30% by micromolar concentrations of dFLUT. Of the five pancreatic carcinoma celllines, AR42J cells are most sensitive to dFLUT with 40%growth inhibition produced after 1 MM exposure. The lesssensitive MiaPaCa cells are inhibited to a similar degree as theT47-D and MCF-7 cells. The Colo-357 cells, on the other hand,show no growth inhibition after exposure to 5 MMdFLUT (96%control growth). Table 2 compares the sensitivity of AR42J cellsto a given dose (2 MM)of progesterone, tamoxifen, danazol,dehydroxyflutamide, or dexamethasone.

Previous studies in MCF-7 and T47-D cells have shown thatvarying the content of serum in culture medium alters theobserved potency of antiestrogens (41). The endogenous steroid

Table I Comparative sensitivity of four human pancreatic carcinoma cell lines toselected endocrine agents

ID25(MM)"

CelllinesMiaPaCa

Colo-357RWP-1RWP-2Progesterone0.8

2.55.0

>10.0Estradici10.0

6.0>10.0

10.0Tamoxifen1.0

10.02.5

4.0Danazol10.0

0.5>10.0>10.0

* Dose required to produce at least 25% growth inhibition by continuousexposure in serum-containing medium; mean values from two or more experiments.

content of fetal calf serum as well as nonspecific serum proteinbinding effects can each contribute to reduce in vitro activity oftamoxifen. Fig. 3 illustrates with MiaPaCa cells the effect ofincreasing serum concentrations on the antiproliferative activityof two different doses of tamoxifen. In chemically definedmedium containing no serum or steroids, 1 MM tamoxifeninhibits MiaPaCa growth by 40%, while 5 MMtamoxifen prevents all cell growth. With increasing serum concentrations,the differential effectiveness of these two doses is substantiallyreduced until all antiproliferative activity is abolished in culturescontaining 10% serum. Similar studies in serum-free mediumwith the less sensitive RWP-1 and RWP-2 cells show that,under these conditions, tamoxifen concentrations < 1 //Minhibitthe growth (>25%) of RWP-1 but not RWP-2 cells. To dissociate the ER-competing effects of endogenous steroids from thenonspecific protein binding of tamoxifen in serum-containingcultures, other experiments were performed using DCC-stripped serum. Fig. 4 demonstrates that addition of 1 DMestradici to MiaPaCa cultures under these conditions enhancescell growth by 40%, comparable to the observed degree ofgrowth inhibition produced by 1 MMtamoxifen in estradiol-supplemented medium. This dose of tamoxifen is no moregrowth inhibiting in the absence of estradiol. These resultssuggest that the reduced potency of tamoxifen observed inMiaPaCa cultures containing higher serum concentrations islargely a result of increased nonspecific tamoxifen binding toserum proteins. In addition, the data indicate that estradiolalone can directly stimulate growth of MiaPaCa cells, similarto the known effect of estradiol on ER-positive breast cancer

cells (28).Since studies in MCF-7 and T47-D breast cancer cells have

shown that the presence of high-affinity binding ER is essentialfor growth stimulation by nanomolar concentrations of estradiol, the four human pancreatic tumor lines were analyzed forER using a standard whole-cell binding assay (24). Fig. 5 showsspecific estradiol binding curves and Scatchard plots for theMiaPaCa, Colo-357, RWP-1, and RWP-2 cell lines, includingthe calculated values for number of binding sites per cell andtheir binding affinity (Ad). From these data, it appears thatthese cell lines all possess some specific binding sites for estradiol; however, by comparison with MCF-7 and T47-D cells,only the MiaPaCa cell receptors have similar high affinitybinding (KÂ¿< l HM). Moreover, binding affinity rather thannumber of binding sites, as measured by this technique, appearsto provide the better correlate with tamoxifen sensitivity inthese four human pancreatic tumor lines (Table 1) and in thetwo breast cancer lines (Fig. 1).

The disconcerting pattern of endocrine sensitivity observedin the pancreatic tumor lines, even under conditions controlledfor serum content, led to the consideration that endocrinemechanisms other than receptor binding might be important inthese cells. The fact that danazol as well as progestins directlyinteract with microsomal oxidoreductases, which are importantin the intracellular conversion of steroids (30, 42), promptedfurther study of other enzyme-inhibiting drugs. The aromataseinhibitor, aminoglutethimide, produced no significant growthinhibition of the pancreatic tumor cells at concentrations <10

Table 2 Sensitivity of the rat pancreatic carcinoma AR42J to selected endocrine agents

Progesterone Estradiol Tamoxifen Danazol dFLUT Dexamethasone

% (mean Â±SD)Control growth"

29 Â±7 66Â± 13 87 Â±3 58 Â±4 61 Â±6 14 Â±2

Â°Triplicate flasks were seeded at 3 x IO4 cells in medium containing 10% fetal calf serum, and a 2 MMdose of endocrine agent was added at 24 h; cells werecounted on Day 6 when control flasks (100 Â±5%) reached a mean cell density of 1.8 x 10s cells.

2278




A.

\

Fig. 5. Specific estradici binding and resultant Scatchard determination of ER capacity (fmol/10* cells) and affinity (A'j) in Mia-

PaCa U). Colo-357 (A), RWP-1 (Q, andRWP-2 (/>) human pancreatic tumor cell lines.Using the whole cell binding assay of Shafieand Brooks (24), 2 x 10* adherent cells inserum-free (ITS) medium were exposed to theindicated concentrations of (3H]estradiol (Â£2)(138 Ci/mmol) Â±100-fold excess DES for l hat 37*C to determine specific binding (cpm, ortniol/W cells). Specific bound/free (Â«//â€¢")ra

tios of radiolabeled steroid are given on thevertical axis of the Scatchard plots in units ofIO-3 ml/10Â«cells.

\100 ./

/M.'

0246tmol bound/IO*cells

0 Z 4 6 8 10 12(mol bound/to6cells

D.

e 12 i6 20 2Â« zafmol bound/io'ceits Imol bound/10 celts

Fig. 6. The effect of a Sa-reductase inhibitor, DMAA, on culture growth ofRWP-2, MiaPaCa, AR42J, and Colo-357 pancreatic carcinoma cell lines. Asdescribed in Fig. 3, cells were plated into triplicate flasks using F( 'S supplementÂ«!

medium, and the indicated concentrations of DMAA were added after cellattachment at 24 h. Cells were counted on Day 6 when control flasks were stillin logarithmic growth. Points, mean percentage of control number of cells counted;bars, SD.

accordance with the results of Greenway et al. in resected tumortissue (14), agree well with the range of values reported forestrogen binding in normal pancreatic tissue (11-13). If, as hasbeen mentioned, these cell lines express variable amounts of anendogenous accessory factor (49), the measured estradici binding constants for a given protein receptor might vary as muchas 10-fold. Other steroid receptors have not yet been studied inthese cells and might yield interesting correlations in light ofthe growth sensitivity observed with exposure to glucocorti-coids, progestins, and antiandrogen in MiaPaCa, Colo-357, andAR42J cells.

Steroid analogues with poor receptor binding affinity maystill bind and inactivate key microsomal oxidoreductases, suchas aromatase or Sa-reductase, resulting in growth inhibition ofa sex-steroid-dependent tumor. Greenway et al. (16) demonstrated the presence of aromatase and Sa-reductase activity inseveral human pancreatic neoplasms. Further studies showedgrowth inhibition of human pancreatic tumor xenografts innude mice with in vivo treatment by the antiandrogen, cypro-

terone acetate (15). We found that aminoglutethimide retardedthe /// vitro growth of breast cancer cells but did not significantlyaffect growth of the pancreatic tumor lines. Inhibitors of Sa-reductase, an enzyme that converts testosterone to dihydrotes-tosterone, are being studied for their potential clinical use inother endocrine-sensitive cancers, particularly prostate cancer.DMAA is a potent Sa-reductase inhibitor which binds to an-drogen receptor with 1000-fold less affinity than dihydro tester-one (51). This steroid analogue directly inhibited growth of thepancreatic tumor lines in a pattern similar to danazol andprogestins, with Colo-357 AR42J, and MiaPaCa cells showingthe greatest sensitivity. Progesterone, a strong competitor forboth androgen and glucocorticoid receptors, also inhibits Sa-reductase activity, albeit with less than one-tenth the potencyof DMAA (30). Some investigators believe that progesteroneinhibits steroid-dependent tumor growth by interfering withone or more of these microsomal oxidoreductases (31, 33). Themultiplicity of possible growth-inhibiting mechanisms available

to progestins in cells with mixed steroid responsiveness mightexplain why these agents appear more effective than tamoxifenagainst three of the five endocrine-sensitive pancreatic tumorlines. The pattern of endocrine sensitivity of these most responsive pancreatic tumor lines (Colo-357, MiaPaCa, AR42J) suggests that the steroid-dependent growth-inhibitory mechanismspresent in some pancreatic cancers involve both receptor antagonism and direct inhibition of steroidal oxidoreductases. Theweak receptor antagonist DMAA effectively inhibits growth ofColo-357 cells, with less activity against AR42J and MiaPaCacells. The potent receptor antagonist dehydroxyflutamide, withno effect on Sa-reductase activity, inhibits MiaPaCa and AR42Jcell growth but does not affect Colo-357 growth.

British and Swedish investigators have begun trials withtamoxifen in the treatment of pancreatic cancer (15, 52). Preliminary results from one small uncontrolled study indicate thata third of patients may respond to tamoxifen with prolongationof survival (52). Such a response rate, if substantiated, wouldbe comparable to the effectiveness of endocrine therapy inunselected patients with breast cancer. Progestins, which havebeen shown to produce clinical regression of hepatocellularcarcinomas, have not been clinically tested against pancreaticcarcinomas (53). We are currently comparing the effectiveness

2280




of tamoxifen, progestins, and danazol in patients with unre-sectable pancreatic carcinoma. Further preclinical studies are 26also necessary to determine whether combinations of endocrineor chemoendocrine therapy might be of potential clinical benefitin the treatment of this lethal cancer (54).

ACKNOWLEDGMENTS 28.

We thank Israel Wiznitzer and Constance Benz for their technicalassistance.

REFERENCES 30.

1. Stedman, K., Moore, G.. and Morgan, R. Estrogen receptor proteins indiverse human tumors. Arch. Surg., 115:244-248, 1980. 31.

2. Sica, V., Noia, E., Contieri, E., Bova, R., Masucci, M. T., Medici, N.,Perrillo, A., Weise, A., Molinari, A. M., and Puca, G. A. Estradici andprogesterone receptors in malignant gastrointestinal tumors. Cancer Res.,4*4670-4674,1984. 32.

3. Leake, R. E., Laing, L., Calman, K. C., and Macbeth, F. R. Estrogen receptorsand an tÂ¡estrogentherapy in selected human solid tumors. Cancer Treat. Rep.,6*797-799,1980. 33.

4. Karakousis, C. P., Lopez, R. E., Bhakoo, H. S., Rosen, F., Moore, R., andCarlson, M. Estrogen and progesterone receptors and tamoxifen in malignantmelanoma. Cancer Treat. Rep., 64:818-827,1980. 34.

5. Lazo, J. S., Schwartz, P. E., MacLusky, N. J., Labaree, D. C., and Eisenfeld,A. J. Antiproliferative actions of tamoxifen to human ovarian carcinomas invitro. Cancer Res., 44: 2266-2271,1984.

6. Panettiere, F. J., Groppe, C. W., Athens, J. W., Costanzi, J. J., Bonnet, J. 35.D., Saiki, J. H., and Chen, T. T. Tamoxifen therapy and colorectal adeno-ca reinoma : a Southwest Oncology Group study. Cancer Treat. Rep., 69:113-114,1985. 36.

7. Levin, D. L., Connelly, R. R., and Devesa, S. S. Demographic characteristicsof cancer of the pancreas: mortality, incidence, and survival. Cancer (I'hi la.).47:1456-1468, 1981.

8. Ihse, I., and Isakson, G. Pancreatic carcinoma: diagnosis and treatment. Clin. 37.Gastroenterol, 13:961-984, 1984.

9. Whittington, R., Solin, L., Mohiuddin, M., Cantor, R. I., Rosato, F. E.,Bierman, W. A., Weiss, S. M., and Pajak, T. F. Multimodality therapy of 38.localized unresectable pancreatic adenocarcinoma. Cancer (Phila.), 5* 1991-1998, 1984.

10. Rosenthal, H. E., and Sandberg, A. A. Estrogen binding proteins in ratpancreas. J. Steroid Biochem., 9:1133-1139,1978. 39.

11. Boctor. A. M., Band, P., and Grossman, A. Analysis"of binding of [-'Ill

esi rudiol to the cytoso! fraction of rat pancreas: comparison with sites in thecytosol of uterus. Endocrinology, 113: 453-462, 1983.

12. Pousette, A., Carlstrom, K., Skoldefors, H., Wilking, N., and Theve, N. O. 40.Purification and partial characterization of a 17/3-estradiol binding macro-molecule in the human pancreas. Cancer Res., 42:633-637, 1982. 41.

13. Pousette, A., Appelgren, l... Carlstrom, K., Skold, H., Wilking, N., andTheve, N. O. Purification and characterization of an estradici-17/3 bindingmacromolecule in rat pancreas. Acta Chem. Scarni. B, 3ti: 407-414, 1984. 42.

14. Greenway, H., Igbal, M. J., Johnson, P. J., and Williams, R. Oestrogenreceptor proteins in malignant and fetal pancreas. Br. Med. J., 283: 751-753, 1981. 43.

15. Greenway, B., Duke, D., I'yni, B., Igbal, M. J., Johnson, P. J., and Williams,R. The control of human pancreatic adenocarcinoma xenografts in nude 44.mice by hormone therapy. Br. J. Surg., 69: 595-597, 1982.

16. Greenway, B., Igbal, M. J., Johnson, P. J., and Williams, R. Abnormal serum 45.sex-steroid levels in patients with pancreatic carcinoma. Br. Med. J., 286:93-95, 1983. 46.

17. Soule, H., Vasquez, J., Long, A., Albert, A., and Brennan, M. A human cellline from a pleural effusion derived from a breast carcinoma. J. Nati. CancerInst., 51: 1409-1416,1973. 47.

18. Keydar, I., Chen, L., Karby, S., Weiss, F. R., Delarea, J., Radu, M., Chaicik,S., and Brenner, H. J. Establishment and characterization of a cell line ofhuman breast carcinoma origin. Kur. J. Cancer, 75:659-670, 1979. 48.

19. Morgan, R. T., Woods, L. K., Moore, G. E., Quinn, L. A., MeGravran, L.,and Gordon, S. G. Human cell line (Colo-357) of metastatic pancreaticadenocarcinoma. Int. J. Cancer, 25:591-598,1980. 49.

20. Yunis, A. A., Arimura, G. K., and Russin, D. J. Human pancreatic carcinoma(MiaPaCa-2) in continuous culture: sensitivity to asparaginase. Int. J. Cancer,19:128-135, 1977.

21. Dexter, D., Matook, G., Meitner, P., Bogaars, H., Jolly, G., Turner, M., and 50.Calabresi, P. Establishment and characterization of two human cancer celllines tumorigenic in athymic mice. Cancer Res., 42: 2705-2714, 1982. 51.

22. Logsdon, C., Moessner, J., Williams, J., and Goldfine, I. Glucocorticoidsincrease amylase niRN'A levels, secretory organelles, and secretion in pan

creatic acinar AR42J cells. J. Cell Biol., 100:1200-1208, 1985. 52.23. Scatenarci, G. The attractions of proteins for small molecules and ions. Ann.

NY Acad. Sci., 51:660-672, 1959.24. Shafie, S., and Brooks, S. Effect of prolactin on growth and the estrogen 53.

receptor level of breast cancer cells (MCF-7). Cancer Res., 37: 792-799,1977. 54.

25. Taylor, C. M., Blanchard, B., and /.ava. D. R. A simple method to determinewhole cell uptake of radiolabeled oestrogen and progesterone and their

2281

subcellular localization in breast cancer cell lines in monolayer culture. J.Steroid Biochem., 20:1083-1088, 1984.Eckert, B., and Katzenellenbogen, B. Effects of estrogens and antiestrogenson estrogen receptor dynamics and the induction of progesterone receptor inMCF-7 human breast cancer cells. Cancer Res., 42: 139-144, 1983.( 'adman, E., and Benz, C. Cytotoxicity of several chemotherapeutic agentsin a human pancreatic cancer cell line (Colo-357). Cancer Treat. Rep., 68:429-430, 1984.Benz, C., Cadman, E., Gwin, J., Wu, T., Amara, J., Eisenfeld, A., andDannies, P. Tamoxifen and 5-fluorouracil in breast cancer cytotoxic syner-gism in vitro. Cancer Res., 43: 5298-5303, 1983.Horwitz, K. B., Zava, D. T., Thilagar, A. K., Jensen, E. M., and McGuire,W. L. Steroid receptor analyses of nine human breast cancer cell lines.Cancer Res., 38:2434-2437, 1978.Liang, T., Brooks, J. R., Cheung, A., Reynolds, G. F., and Rasmusson, G.H. 4-Azasteroids as inhibitors of 5Â«reducÃase.Prog. Cancer Res. 1her., 31:497-505,1984.Teulings, F. A. G., van Gilse, H. A., Henkelman, M. S., Portengen, H., andAlexieva-Figusch, J. Estrogen, androgen, glucocorticoid, and progesteronereceptors in progestin-induced regression of human breast cancer. CancerRes., Â¥0:2557-2561, 1980.Sadler, S. E., Mailer, J. L., and Cooper, D. M. F. Progesterone inhibition ofxenopus oocyte adenylate cyclase is not mediated via the Bordetella pertussistoxin substrate. Mol. Pharmacol., 261: 526-531, 1984.Evans, R. W., and Leavitt, W. W. Progesterone inhibition of uterine nuclearestrogen receptor dependence on RNA and protein synthesis. Proc. Nati.Acad. Sci. USA, 77: 5856-5860, 1980.Horwitz, K. B., and Freidenberg, G. R. Growth inhibition and increase ofinsulin receptors in ant Â¡estrogenresistant 147-1\â€žhuman breast cancer cellsby progestins: implications for endocrine therapies. Cancer Res., 45: 167-173, 1985.Maclndoe, J. H., Woods, G. R., and Lee, F. J. The specific binding ofandrogens and the subsequent dilution of androgen-receptor complexeswithin MCF-7 human breast cancer cells. Steroids, 38:439-452, 1981.Blossey, H. C., Bartsch, H. H., Kanne, D., Koebberling, J., and Nagel, G. A.The pharmacokinetics of high-dose medroxyprogesterone acetate (MDA) inthe therapy of advanced breast cancer. Cancer Chemother. Pharmacol., 8:77-81, 1982.Fabian, C., Sternson, L., and Barnett, M. Clinical pharmacology of tamoxifenin patients with breast cancer: comparison of traditional and loading doseschedules. Cancer Treat. Rep., 64: 765-773,1980.Jordan, V. C, Bain, R. R., Brown, R. R., Gosden, B., and Santos, M. A.Determination and pharmacology of a new hydroxylated metabolite of tamoxifen observed in patient sera during therapy for advanced breast cancer.Cancer Res., 43:1446-1450, 1983.Benz, C., Wiznitzer, I., and Lee, S. Flow cytometric analysis of fluorescentestrogen binding in cancer cell suspensions. In: L. Pertschuk and S. Lee(eds.), Localization of Putative Steroid Receptors: Experimental Systems.Boca Raton, FL: CRC Press, pp. 95-110,1985.Madanes, A. E., and Farber, M. Danazol. Ann. Intern. Med., 96: 625-630,1982.Wiznitzer, I., and Benz, C. Comparative effects of tamoxifen and LY156758on cultured human breast carcinomas. Proc. Am. Assoc. Cancer Res., 25:815, 1984.Barbieri, R. L., Osathanondh, R., Canick, J. A., Stillman, R. J., and Kenneth,J. R. Danazol inhibits human adrenal 21 and 11 beta hydroxylation in vitro.Steroids, 35:251-263,1980.Johnson, L. R. Effects of gastrointestinal hormones on pancreatic growth.Cancer (Phila.), 47:1640-1645, 1981.Ullberg, S., and Bengtsson, G. Autoradiographic studies with natural oestro-gens. Acta Endocrino!. (Copenhagen), 43: 73-86, 1963.Sandberg, A. A., and Rosenthal, H. E. Steroid receptors in exocrine glands;the pancreas and prostate. J. Steroid Biochem., / /: 293-299,1979.Clark, J. H., Hardin, J. W., Unchurch, S., and Eirksson, H. Heterogeneityof estrogen binding sites in the cytosol of the rat uterus. J. Biol. Chem., 253:7630-7634, 1978.Panko, W. B., Watson, C. S., and Clark, J. H. The presence of a second,specific estrogen binding site in human breast cancer. J. Steroid Biochem.,M-1311-1316, 1981.

Boctor, A. M., Band, P., and Grossman, A. Requirement for an accessoryfactor for binding of [3H]estradiol to protein in the cytosol fraction of ratpancreas. Proc. Nati. Acad. Sci. USA, 78: 5648-5651, 1981.Boctor, A. M., Band, P., and Grossman, A. Specific binding of | 'I I [esiradiol

to the cytosol of rat pancreas: alteration of the apparent number of bindingsites by an endogenous factor and oligopeptide derivatives. J. SteroidBiochem., 18: 245-252, 1983.Howell, M. A. The association between colorectal cancer and breast cancer.J. Chron. Dis., 29: 243-261, 1976.Liang, T., and Heiss, C. E. Inhibition of 5<.-reducÃase,receptor binding, andnuclear uptake of androgens in the prostate by a 4-methyl-4-aza-steroid. J.Biol. Chem., 256:7998-8005, 1981.Theve, N. O., Pousette, A., and Carlstrom, K. Adenocarcinoma of thepancreasâ€”a hormone sensitive tumor? A preliminary report on nolvadextreatment. Clin. Oncol., 9: 193-197, 1983.Friedman, M. A., Demanes, D. J., and Hoffman, P. G., Jr. Hepatomas,hormone receptors, and therapy. Am. J. Med., 73: 362-366, 1982.Redding, T. W., and Schally, A. V. Inhibition of growth of pancreaticcarcinomas in animal models by analogs of hypothalamic hormones. Proc.Nati. Acad. Sci. USA, 81: 248-252, 1984.



endocrine-responsive pancreatic carcinoma: steroid binding ... · sex-steroid dependent; in this...

Documents