Significance of Aromatase Activity in Human Breast Cancer1

[CANCER RESEARCH (SUPPL.) 42. 3365s-3368s. August 1982] 0008-5472/82/0042-OOOOS02.00 Significance of Aromatase Activity in Human Breast Cancer1

W. R. Miller, R. A. Hawkins, and A. P. M. Forrest

University Department of Clinical Surgery, Medical School, Teviot Place, Edinburgh EH8 9AG, Scotland

Abstract 3H]testosterone (8.9 Ci/mmol). The reaction was stopped by the ad dition of methanol, and 500 Â¿igofnonradioactive estradiol were added The significance of in vitro aromatization of [7a-3H]testoster- to monitor losses during purification and characterization. Estradiol one to estradici by human breast cancer has been investigated was extracted and purified by thin-layer chromatography (14, 15). by correlation with (a) estrogen receptor activity and (b) clinical Characterization was by chemical derivative formation (14) and was response to endocrine therapy. Evidence for estradiol synthe based on Chromatographie behavior of parent and derivatized estradiol sis was obtained in 66 and estrogen receptor activity in 81 of being identical with that of authentic steroids and by the maintenance 110 tumors. Whereas most estrogen receptor-positive tumors of consistent specific radioactivity throughout the procedures. Synthe synthesized estradiol, the majority of receptor-negative tumors sis was determined by measuring radioactivity in the purified estradiol fraction. Conversions in excess of 0.02% of the original precursor are did not. This tendency for aromatization to be associated with estrogen receptor-positive tumors was statistically significant detectable. Estrogen Receptors. Concentration of estrogen receptor was de (p < 0.005). Mean level of receptor was also significantly termined by saturation analysis (5). Tumor cytosol was incubated higher in tumors with aromatization than in tumors without overnight at 4Â°with 17/?-[2,4,6,7-3H]estradiol and varying amounts of estradiol synthesis (p < 0.001 ). Forty patients with advanced nonradioactive 17/Ã®-estradiol. Separation of free and bound steroid breast cancer have been treated by endocrine therapy. There was by addition of dextran-coated charcoal; the bound fraction was was a significant trend for tumors with aromatization to be measured by liquid scintillation counting. Concentration of receptors associated with response to treatment (p < 0.05), but the was determined by Scatchard analysis (23). Activities in excess of 5 correlation was not absolute and may simply reflect the asso fmol/mg cytosol protein were designated receptor positive (8). ciation between aromatase activity and estrogen receptors. Clinical Response. Forty patients with advanced breast cancer were Within the small subgroup of patients treated with aminoglu- treated with endocrine procedures including oophorectomy, adrenal ectomy, hypophysectomy, tamoxifen, diethylstilbestrol, and aminoglutethimide or adrenalectomy, tumors with high aromatase activ tethimide. ity responded whereas those without aromatization did not. For the purpose of this study, objective response to treatment was Tumor estrogen biosynthesis may therefore be of clinical sig graded independently by 2 members of the Department of Clinical nificance in selecting patients for treatments which remove Surgery who did not know the results of the biochemical studies. sources of precursor for aromatization or inhibit aromatase Patients were classified as having a response if they showed evidence activity itself. of tumor regression during therapy. Results Introduction Evidence for estradiol synthesis from testosterone was ob Many groups (4, 7, 14, 24) have now demonstrated that tained in 66 of the human breast cancers examined (60%). some breast cancers synthesize estrogen from androgen pre Level of conversion is shown in Chart 1 and ranged from 0.02 cursors in vitro. It is not known whether this potential for to 0.5%. aromatization is merely an example of tumor differentiation or Estrogen receptor activity was detected in 81 tumors (74%), whether it is of significance to the need of the tumor for and the relationship between the presence of receptors and estrogens. In the latter event, correlations between tumor aro potential for aromatization is shown in Table 1. Most estrogen matase and hormonal sensitivity would be expected. receptor-positive tumors (68%) synthesized estradiol, whereas In this study, we have related tumor aromatase activity to only the minority of receptor-negative tumors (38%) did so. estrogen receptors and, in a small group of patients with This tendency for aromatization to be associated with estrogen advanced breast cancer, to response to endocrine treatment. receptor-positive tumors was statistically significant (p < 0.005). Materials and Methods Tumors with aromatizing capacity were not only more likely to have estrogen receptor activity but, in addition, the mean Tumors. Breast cancer tissue was obtained from 110 patients. Of level of receptor of these tumors was significantly higher than these, 24 were premenopausal, 4 were menopausal (less than 5 years that in tumors lacking estradiol synthesis (Chart 2). The con since the last menstrual period), and 82 were postmenopausal. verse was not so; amounts of estrogen synthesized by tumors Tumor was obtained at mastectomy or by biopsy and immediately placed on ice. Following removal of tissue for histopathological diag with receptors did not differ significantly from those without nosis, the remainder was assayed for estrogen synthesis (aromatase receptors. In tumors with both aromatizing and estrogen recep activity) and estrogen receptor activity. tor activity, no significant quantitative relationship was found Estrogen Synthesis. A portion of each tumor (0.5 g) was finely (Chart 3). minced and incubated for 2 hr at 37Â°in Krebs-Ringer phosphate buffer, The relationship between tumor potential for aromatization pH 7.4, containing an NADPH-generating system and 22.5 /iCi [7a- and clinical response of 40 patients with advanced breast ' Presented at the Conference "Aromatase: New Perspectives for Breast cancer treated with endocrine therapy is presented in Chart 4. Cancer," December 6 to 9, 1981, Key Biscayne, Fla. Supported by grants from There was a statistically significant trend for aromatization to the Cancer Research Campaign and the CIBA-Geigy Corp. be associated with response, but the correlation was not ab-

AUGUST 1982 3365s

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1982 American Association for Cancer Research. W. R. Miller et al. 05-1 . 820 estradiol '783 700- conversion estrogen receptor 03- tmol "mg600-

cytosol protein

500- 02-

Â¿00-

0 1- 300- s t â€¢¿ 200- â€¢¿

Chart 1. Level of estradiol synthesis in tumors with aromatization. _Ã„- 100- Table 1 Ã¡ Correlation between potential for aromatization and presence of estrogen receptor X* = 7.99; p - 0.005.

Estrogen syn- Without syn estradiol synthesis no synthesis thesis thesis Chart 2. Level of estrogen receptor in tumors with and without estradiol Estrogen receptor positive 55 26 synthesis. The difference between groups is significant by Wilcoxon rank test Estrogen receptor negative 11 18 (p< 0.001). solute. This effect was largely accounted for by the association comparisons of adipose tissue and breast cancer have always of aromatase with estrogen receptors, all tumors with aroma- shown higher biosynthesis of estrogen in tumor (3, 22). Fur tase activity responding to treatment also being estrogen re thermore, using the methods of the present studies, biosyn thesis of estrogens from dehydroepiandrosterone, A"-andros- ceptor positive. It was also of interest to examine the relationship between tenedione, and testosterone was consistently below detectable tumor aromatization and response to specific therapies which levels in adipose tissue. The large mass of fat in the body may might be active against aromatization. The results are pre compensate for its low synthetic activity and make it the major sented in Table 2, both for adrenalectomy, which removes the source of circulating estrogen in postmenopausal women. However, "on site" tumor production of estrogen may be more major source of C-19 steroid pecursors in postmenopausal women, and for the administration of aminoglutethimide, which important for the growth of the breast cancer. also inhibits tumor aromatase (2). It has been calculated that the MCF-7 cell line of human Although numbers are small, tumors with the greatest in vitro breast cancer might synthesize sufficient intracellular estrogen conversion to estradiol were those which responded while to stimulate estrogen-mediated events (11 ). In the present those without aromatization failed to do so. studies, similar calculations indicate that 0.5 to 12.5 pmol are formed during incubation. This is sufficient to half-saturate estrogen receptor sites in the majority of breast cancers. Discussion However, it should be noted that endogenous levels of testos Our report that human breast cancer may synthesize estro terone in breast tumors are likely to be considerably lower than gen from C-19 steroid precursors (13) has been confirmed by are those used in our incubations (20). On the other hand, in several groups (1, 4, 7, 24). When compared with ovarian the breast there are high concentrations of other C-19 steroid production in premenopausal women, the levels of estrogen precursors (18) which breast tumors may convert to testoster synthesized by a tumor may appear insignificant. However, in one or metabolize directly to estrogen (1, 7, 17, 19). postmenopausal women with breast cancer, a tumor may rep In addition to these theoretical considerations, it is important resent a major estrogen-synthesizing organ. Using identical to relate the results from in vitro incubations to estrogen recep conditions of in vitro incubation, we have found that breast tor status which has already been established to be of clinical cancers may have higher aromatase activity than do adrenal value (12). cortex, liver, or fat.2 The finding of a positive correlation between aromatase and Fat is generally assumed to be the major site of peripheral estrogen receptor activity confirms our previous report (17) but conversion of androgens to estrogens (10,21 ). However, direct is not in agreement with others (3, 9, 24). However, in these latter studies, smaller numbers were reported. This may be â€¢¿W.R. Miller. Unpublished observations. important, since the relationship between aromatase and re-

3366s CANCER RESEARCH VOL. 42

800 -i be a positive relationship between aromatase activity and clin estrogen ical response, this was neither additive to, nor independent of, receptor frnp]____ 700- estrogen receptor activity. The relationship between tumor mg cytosol aromatase activity and endocrine responsiveness may be best protein 600 shown by studying those forms of treatment which deprive the tumor of its precursor C-19 steroids or directly inhibit tumor 500- aromatase. These include adrenalectomy and aminogluteth imide, and it is encouraging to note that tumors with highest 400- aromatase activity appeared to be more responsive to these therapeutic methods. However, the numbers of patients studied 300- are small, and it is now important to establish a prospective study to determine the relationships between aromatase activ 200- ity, steroid receptor, and tumor sensitivity to endocrine treat 100- ment. â€¢¿li' It is difficult to attribute a physiological role for estrogen synthesis in tumors which lack estrogen receptor activity, al 01 02 03 05 though it has been suggested (1 ) that these tumors may appear estradiol synthesis ('/Â«conversion| hormone independent by virtue of their de novo synthesis. Chart 3. Relationship between level of estrogen receptors and percentage of Further studies are required to elucidate this. conversion to estradiol in tumors with both aromatase and receptor activities. One can conclude that estrogen-synthetic activity can be demonstrated in approximately half of all breast cancers and in Â¿0patients with advanced breast cancer many is likely to be of sufficient magnitude to induce estrogen- 13 responders to endocrine therapy (32 5%) stimulated events. Although aromatase is correlated with estro

AROMATASE . gen receptor activity, the relationship is not absolute. However, +-ve - ve we have some evidence that estrogen synthesis may be an important influence on the hormonal sensitivity of a tumor, response 1) / 24 2/16 p<0-05 rate especially to regimens such as adrenalectomy and aminoglu (Â¿6%) ( 12 5V.) tethimide. /X /X E jR * ve E2R - ve EjR *ve E2R -ve

11/21 0/3 2/6 0/10 Acknowledgments ( 52% ) ( OVo ) ( 33 3% ) I 0% ) The authors thank D. Lee and R. Steele who assessed the clinical response of Chart 4. Relationship between tumor aromatization, estrogen receptors, and the patients and J. Telford and S. Murphy for their skilled technical assistance. response to endocrine therapy. References Table 2 Relationship between tumor aromatization and response to either 1. Abul-Hajj, Y. J. Metabolism of dehydroepiandrosterone by hormone depend aminoglutethimide or adrenalectomy ent and hormone independent human breast carcinoma. Steroids, 24.488- 500, 1975. 2. Abul-Hajj, Y. J. Inhibition of androgen aromatization in human breast cancer. gen re J. Steroid Biochem., J3.1395-1400, 1980. ceptorResponse+ 3. Abul-Hajj, Y. J., Iverson, R., and Kiang, D. T. Aromatization of androgens by AminoglutethimideAdrenalectomyPatientE.M.K. Yes+ human breast cancer. Steroids, 33.205-222, 1979. S.J. Yes+ 4. Adams, J. B., and Li, K. Biosynthesis of 17/3-oestradiol in human breast S.B.E.6LT."I.H.J.B.J.R.Aroma-fase3(%)0.180.090.080.03NDCNDNDNDEstroStatic+ carcinoma tissue and a novel method for its characterization. Br. J. Cancer, Mixed+ 3).'429-433, 1975. NoNoNoNo 5. Hawkins, R. A., Hill, A., and Freedman, B. A simple method for the deter mination of oestrogen receptor concentrations in breast tumors and other tissues. Clin. Chem. Acta, 64.203-210. 1975. 6. Jensen, E. V. The D. R. Edwards Memorial Lecture. In: K. Griffiths and C. G. Pierrepoint (eds), Some Aspects of the Aetiology and Biochemistry of Prostate Cancer, pp. 151-166. Cardiff, Wales: Alpha Omega Publishing " Aromatase activity expressed as percentage of conversion of [7a-3H]testos- Ltd.. 1970. terone to 17/3-estradiol. 7. Jones, D., Cameron, E. H. D., Griffiths. K., Gleave, E. N., and Forrest, A. P. b In combination with tamoxifen. M. Steroid metabolism by human breast tumors. Biochem. J.. f 76.919- 0 ND, not detectable, i.e., <0.02%. 921, 1970. 8. King, R. J. B., and Roberts, M. M. The use of steroid receptor assays in determining response to endocrine therapy: a summary of the clinical data. ceptors is not absolute and the trend may not be statistically In: R. J. B. King (ed.). Steroid Receptor Assays in Human Breast Tumors: significant when small numbers of tumors are investigated (16). Methodological and Clinical Aspects, pp. 1-6. Cardiff. Wales: Alpha Omega Our finding is logical in that tumors which synthesize estrogen Publishing, Ltd., 1979. 9. Li, K., and Adams, J. B. Aromatization of testosterone and oestrogen would be expected to possess the mechanism (6) by which it receptor levels in human breast cancer. J. Steroid Biochem., 14:269-272, gains access to the nucleus and exerts its effect. 1981. We postulated that tumors possessing both estrogen-syn 10. MacDonald. P. C., Grodin, J. M., Edman, C. D., Vellios, F., and Siiteri, P. K. Effect of obesity on conversion of androstenedione to estrone in postmen- thetic and receptor activity were more likely to be sensitive to opausal women with and without endometrial cancer. Am. J. Obstet. Gyne- deprivation of hormones than were those without one or both col., 730:448-455, 1978. 11. Maclndoe, J. H. Estradiol formation from testosterone by continuously of these properties (17). The clinical studies that we now report cultured human breast cancer cells. J. Clin. Endocrinol. Metab., 49.272- do not totally support this view. Although there did appear to 277, 1979.

AUGUST 1982 3367s

12. McGuire, W. L., Carbone, P. P.. Sears, M. E., and Esher, G. C. Estrogen 18. Miller, W. R.. Humenuik, V., and Kelly, R. W. Dehydroepiandrosterone receptors Â¡nhuman breast cancer: an overview. In: W. L. McGuire, P. P. sulphate in breast secretions. J. Steroid Biochem., Ã•3.145-151, 1980. Carbone, and E. P. Vollmer (eds.). Estrogen Receptors in Human Breast 19. Miller, W. R., McDonald, D., Forrest, A. P. M., and Shivas, A. A. Metabolism Cancer, pp. 1-7. New York: Raven Press, 1975. of androgens by human breast tissue. Lancet, 2:912-913, 1973. 13. Miller, W. R., and Forrest, A. P. M. Oestradiol synthesis by a human breast 20. Millington. D., Jenner, D. A., Jones, T., and Griffiths, K. Endogenous steroid carcinoma. Lancet, 2.866-869, 1974. concentrations in human breast tumors determined by high resolution mass 14. Miller, W. R., and Forrest. A. P. M. Oestradiol synthesis from C19 steroids fragmentography. Biochem. J.. 739.473-475, 1974. by human breast cancers. Br. J. Cancer. 33.116-118. 1976. 21. Nimrod, A., and Ryan, K. J. Aromatization of androgens by human abdominal 15. Miller. W. R., Forrest, A. P. M.. and Hamilton, T. Steroid metabolism by and breast fat tissue. J. Clin. Endocrinol. Metab., 40.367, 1975. human breast and rat mammary carcinomata. Steroids, 23.379-395, 1974. 22. Perel, E., Wilkens, D., and Killinger, D. W. The conversion of androstenedi- 16. Miller, W. R., Hawkins, P. A., and Forrest, A. P. M. Oestrogen biosynthfisis one to estrone, estradici and testosterone in breast tissue. J. Steroid and oestrogen receptors in human breast cancer. Cancer Treat. Rep., Biochem., i 3.89-94, 1980. 63.1153, 1979. 23. Scatchard, G. The attraction of protein for small molecules and ions. Ann. 17. Miller, W. R., Hawkins, R. A., and Forrest, A. P. M. Steroid metabolism and N. Y. Acad. Sci., 57.660-672, 1949. oestrogen receptors in human breast carcinomas. Eur. J. Cancer, ) 7.913- 24. Varella, R. M., and Dao, T. L. Estrogen synthesis and estrogen binding by 917, 1981. human mammary tumors. Cancer Res., 38:2439-2443, 1978.

3368s CANCER RESEARCH VOL. 42

W. R. Miller, R. A. Hawkins and A. P. M. Forrest

Cancer Res 1982;42:3365s-3368s.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/42/8_Supplement/3365s

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/42/8_Supplement/3365s. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.