Endocrine Aspects of Breast Cancer

A relationship between the endocrine system and breast cancer was first demonstrated by Beatson in 1896, when he obtained some benefit from oophorectomy in two patients with advanced disease. Until that time the treatment of breast cancer was limited to removal of the mamma and excision of superficial metastatic or recurrent nodules. The practice of castrating cows to maintain lactation after calving suggested to Beatson that the ovaries must be concerned in some way with the regulation of the physiological processes of the breast. Reporting his findings to the Medico-Chirurgical Society of Edinburgh, Beatson summed up: “The conclusion I draw from two cases I have brought under notice is this — that we must look in the female to the ovaries as the seat of the exciting cause of carcinoma, certainly of the mamma . . . .” Copyright Royal Medical Society. All rights reserved. The copyright is retained by the author and the Royal Medical Society, except where explicitly otherwise stated. Scans have been produced by the Digital Imaging Unit at Edinburgh University Library. Res Medica is supported by the University of Edinburgh’s Journal Hosting Service: http://journals.ed.ac.uk ISSN: 2051-7580 (Online) ISSN: 0482-3206 (Print) Res Medica is published by the Royal Medical Society, 5/5 Bristo Square, Edinburgh, EH8 9AL Res Medica, April 1967, 5(4): 19-21, 23-25 doi: 10.2218/resmedica.v5i4.505 ENDOCRINE ASPECTS OF BREAST


INTRODUCTION
A relationship between the endocrine system and breast cancer was first dem onstrated by Beatson in 1896, when he obtained som e benefit from oophorectom y in two patients with advanced disease.U n til th at tim e the treatm ent o f breast cancer was lim ited to removal of the m am m a and excision of superficial m etastatic or recurrent nodules.T h e practice of castrating cows to m aintain lactation after calving suggested to Beatson that the ovaries m ust be concerned in som e way w ith the regulation of the physiological processes of the breast.R epo rtin g his findings to the M edico-Chirurgical Society o f Edinbu rgh , Beatson sum m ed u p : " T h e conclusion I draw from two cases I have brought under notice is this -that we m ust look in the fem ale to the ovaries as th e seat of the exciting cause of carcinom a, certainly of the m am m a . . . ." H e further em phasised : 1. " T h a t there seems to be evidence of the ovaries and testicles having control in the human body over local proliferation o f epithelium " .
2. " T h a t the rem oval of th e tubes and ovaries has effect on the local prolifer ation of epithelium w hich occurs in carcinom a o f the m am m a . . . ." 3. " T h a t the effect is best seen in cases of carcinom a in young people, a class of case where local rem oval o f the disease is often unsatisfactory." W ith in the next decade, B o yd (1900) and L e tt (1905) confirm ed B eatso n 's observations and showed that abou t one-third of patients responded favourably to oophorectom y.It is interesting that no im provem ent in this figure has taken place since then, for th e Jo in t C o m m ittee of the A m erican M ed ical A ssociation on A blative Procedures in Breast C an cer reported that objective tum our regression occurred in 29 % o f wom en castrated before or during the m enopause (Taylor, 1962).
In the seventy years since B eatson's pioneer effort th e concept of horm one-dependence has becom e an accepted biological characteristic of certain tumours, particularly o f the breast and prostate.It had its origin in the classical observations of H uggins and his colleagues who dem onstrated that in dogs not only are the physiological processes of the prostate under endocrine control but that tumours of the glands regress after castration or in response to the adm inistration o f oestrogens (H uggins and C lark, 1940).It was subsequently shown that cancer of the hum an prostate could be in flu enced in a sim ilar way.
T h e developm ent o f our understanding, albeit im perfect, of the relationship between endocrine status and breast cancer is th e result o f a fruitful alliance of clinical practice, bio chem ical investigation and anim al experim ent.Som e of the m ost significant contributions in these fields are discussed below.

EXPERIMENTAL TUMOURS IN ANIMALS
A great deal o f research has been carried out on experim ental tumours, particularly in the U nited States and a volum inous literature has accum ulated.T h re e main lines of investigation have been followed.T h e breast is not the only organ to be affected.Oestrogen adm inistration has yielded lym phoid tumours in mice, chrom ophobe tumours of the anterior pituitary fibromyom ata of the uterus in guinea pigs, m alignant tumours of the cervix uteri, m alignant renal tumours in m ale golden hamsters and interstitial cell tumours of testes in mice.
A lthough oestrogens are the m ost potent, other hormones have been found also to induce tumours.M oon and his colleagues (1950), for exam ple, described lym phosarcom ata, adrenal cortical and ovarian tumours in rats receiving injections of pituitary growth hormone.
T h ese exam ples dem onstrate the carcino genic effects of the adm inistration of exogenous horm ones.T h e secretion of edogenous horm ones m ay be altered by suitable m anipulation of the endocrine system.E legan t experim ents which dem onstrate the interdependence of endocrine organs, and the oncogenic potential of interference with the feed-back mechanisms were reported by Biskind and B iskind (1944).T h e y excised both ovaries from rats and trans planted one of them into the spleen.Ovarian oestrogen was thus liberated directly into the portal circulation and inactivated by the liver.T h e y concluded that the resultant excessive production o f anterior pituitary gonadotro phins was responsible for the granulosa-cell tumours and luteom ata in the transplanted ovary.

The effects of horm ones on spontaneous m am m ary tum ours
Spontaneous fibroadenom ata occur in about fifty per cent of elderly fem ale rats o f the Sprague-Dawley strain.T h ese tumours arc readily transplanted into young rats and their experim ental value depends upon their responsiveness to alterations in the an im al's endocrine environm ent.
A biphasic effect on the growth o f the tum our has been observed during the adm inis tration o f oestrogen both in intact (M illar and N oble, 1954) and ovariectom ised rats (Huggins, e t a l., 1956).Sm all doses o f oestrogen prom ote, and large doses inhibit, the growth of the tumours.T h e effects of androgenic steroids have also been intensively investigated by H uggins and M ain zer (1957) and in general, they arc inhibitory.

Endocrine factors influ encing hyd rocarbon-induced m am m ary cancers
T h e two m ost useful m am m ary carcinogens are 3-M ethyleholanthrene and 9, 10-D im ethyl-1 , 2-Benzanthracene (D M B A ). T h e tumours appear within a few weeks in response to a single intragastric instillation or intravenous injection (H uggins, Briziarelli and Sutton, 1959).T h e y are adenocarcinom as histologically sim ilar to hum an tumours.T h e y are also horm one-dependent, and regression or disappearance of the tumours occurs after oophorectom y, after hypophysectom y and in response to the exhibition of exogenous steroids.
On the basis o f their growth characteristics the tumours can be classified into three bio logical types.Som e grow to a certain size, then regress spontaneously; som e remain static after an initial period of growth: and others continue to enlarge throughout their life-history (Young, ct a l, 1963).
M an y of these features are characteristic of the human disease, and the hydrocarbon-induced m am m ary tum our is, therefore, a unique experi mental m odel for intensive investigation of the endocrine factors concerned in the induction and m aintenance of breast cancer.
In clinical practice, alteration of the endocrine state of patients has been achieved either by the adm inistration of steroid hormones or by ablation of endocrine organs.Such measures have generally been lim ited to the treatm ent of advanced m alignant disease.
Since B eatson's original observations there have been m any confirm atory reports of the value in certain cases of removal of the ovaries.
Elimination of ovarian function has also been achieved by radiotherapy, and Douglas (1952) reported a favourable response in 20 per cent of patients with advanced disease.A major debate has now centred on the timing of castration, which may be done either as a prophylactic measure at an early stage in the disease or deferred until metastases have appeared.From an extensive review of the relevant literature Lewison (1962) concluded that therapeutic oophorectomy is an effective palliative procedure in about 25 per cent of patients and that there is a well defined trend towards prolonged life expectancy and improved survival rate from prophylactic castration whether achieved by surgery or by irradiation.
Oophorectomy is occasionally necessary in young women for the treatment of gynaecological disorders.Careful follow-up studies (Lilienfield, 1956;Hirayama and W ynder, 1962) indicate that in these women there is a reduced incidence of breast cancer in later years, particularly if the operation was done before the age of 37 years.Bilateral oophorectomy offers the greater protection, but even women who have had one ovary excised are significantly less prone to the subsequent development of cancer.These reports suggest that the ovary and its secretions are important in the induction as well as the maintenance of human breast cancer.

Adrenalectom y
Removal of the adrenal glands is designed to remove the main source of extra-ovarian sex hormones.It is usually combined with oophor ectomy even in post-menopausal women in whom secretion of ovarian hormones continues although the uterus atrophies.
In 1945 Huggins and Scott obtained tumour remission following bilateral adrenalectomy in a man with advanced prostatic cancer.Survival lasted only four months in spite of the admini stration of deoxycorticosterone and a high salt intake.By 1952 cortisone was readily available and Huggins and Bergenstal successfully operated on seven patients with breast cancer of whom three, including one man, responded favourably.Since this time, many reports have indicated a variable degree of success, but remission occurs in about one-third of patients who are submitted to adrenalectomy and oophorectomy.

H ypophysectom y
This operation was introduced by Lu ft and Olivecrona in 1953.They reported (1958) a total remission rate of 44 per cent -regression of tumour in 29 per cent, and arrest of disease in 15 per cent of patients.A comprehensive survey of the value of adrenalectomy and hypophysectomy has been conducted by the Joint Committee of the American Medical Assoc iation (19 61).
Th e results from twelve American Clinics (representing a total of 673 patients) demonstrate no difference between the two methods; there was a 31 per cent response in each.
A dm inistratio n of Horm ones 1. Oestrogens The isolation of a crystalline oestrogenic hormone by Doisy in 1929 heralded an explosive development of steroid biochemistry, leading eventually to the elaboration of synthetic oestrogens.T he value of these compounds in the treatment of advanced malignant disease was demonstrated by Haddow, W atkinson and Paterson (1944).Perhaps the most significant feature of these studies for the clinician is that for the first time, they enabled him to write a prescription for the treatment of cancer.This marks the origin of chemotherapy in malignant disease.
Oestrogen therapy is of greatest value in post menopausal women in whom tumour regression occurs in about 35 p er cent with a mean survival time of 27 months compared with 10 months in non-responsive patients (Council on Drugs of the American Medical Association, 1 960).

A ndroge ns
Loesser first demonstrated in 1939 that be tween 20 per cent and 25 per cent of patients respond to androgen administration, irrespective of menopausal state.Survival times recorded by the Council on Drugs (i960) were 19 months and less than 10 months in responsive and non-responsive patients respectively.Hirsutism, acne, deepening of the voice and increase in libido are displeasing side-effects which have dissuaded many clinicians, and patients from continuing treatment.Numerous steroid derivatives have been investigated by the Co-operative Breast Cancer Group (1964) in an attempt to find compounds which are non-virilising and yet retain their therapeutic effectiveness.Although some success has been achieved it is doubtful whether these compounds are as effective as esters of the naturally occurring androgens.

Other Hormones
High doses of corticosteroids are beneficial in castrated patients and are believed to act by inhibition of the adrenal cortex and consequent reduced production of oestrogen.Prednisolone, a synthetic corticosteroid, is more effective than cortisone and causes less water retention.
Progesterone appears to have little effect but a synthetic progestational compound, 9 αbromo-1 1-ketoprogesterone, studied by Goldenberg and Hayes (1959) produced a 20 per cent remission.Combinations of oestrogen and progesterone may be more effective.Stimul ated by the observation that hydrocarboninduced experimental tumours responded to administration of high doses of oestradiol and progesterone, Landau and his colleagues (1962) tried the combination in 15 patients, of whom 9 responded though only for a short time.Recently, Stoll (1967) has reported a 20 per cent response in patients receiving the oral contraceptive, Lyndiol, in which an oestrogen and progestin are combined.

Effectiveness of Treatm ent
It is evident from this brief outline that about one-third of patients respond to the administration of hormones or to endocrine ablative procedures.Adverse reactions to treat ment do occur, however.W ilson, Jessiman and Moore (1958) have noted exacerbation of metastatic disease following oophorectomy and adrenalectomy, and Cade ( 19 58) has reported similar effects from the administration of androgens and oestrogens.T h e results in human patients prompt comparison with the biphasie effects of oestrogens and androgens observed in experimental tumours, and em phasise our present ignorance of the funda mental relationships between hormones and growth.
M uch has been written and many opinions have been expressed on the value of certain clinical criteria for the selection of patients with metastatic disease for major endocrine surgery.In his comprehensive review of this subject Fairgrieve (1966) concludes that two categories of patient are most likely to benefit; those with a long " free-interval" , that is, in whom more than a year has elapsed between primary treatment and first recurrence, and those with skeletal or skin metastases.He emphasises the importance of assessing the extent and type of metastases.Slowly growing cutaneous or lymph node recurrences are perhaps best treated" by local surgery or radiotherapy.M ajor surgery is inappropriate in preterminal states when life expectancy is limited.
Between these two extremes are patients with metastatic disease which is progressive and/or symptomatic for whom endocrine ablative procedures should be reserved.
The main problem which besets the clinician is his inability to forecast for the individual patient whether or not a favourable reponse will be obtained.It is towards this end that the greatest efforts have been made in the last decade.

B I O C H E M IC A L IN V E S T IG A T IO N S
One of the most rewarding features of medical practice in recent years has been an awareness of the need for collaborative efforts by members of different disciplines, especially in assessing complex biological problems.Not least successful among the alliances so estab lished has been the team comprising endocrino logist, biochemist, surgeon and radiotherapist.Much pioneer work along these lines has been actively pursued in Edinburgh.
The principal approach of the steroid bio chemist has been the determination of steriod metabolites in urine and their relationship to clinical observations.

U rin ary Oestrogens
Animal experiment and clinical observation indicated that oestrogen played a major role in the induction and maintenance of breast cancer.T he elaboration of reliable methods for the analysis of urinary oestrogens (Brown, 1955 andBauld, 1956) provided the means for intensive collaborative activity in Edinburgh.Strong and his colleagues (1956) found no correlation between pre-treatment oestrogen levels and subsequent response to the adminis tration of stilboestrol or endocrine ablative operations.Subsequent reports from other centres have confirmed these observations.Attempts to correlate the clinical result with changes in urinary oestrogen levels occurring after ablative procedures similarly have been disappointing.Some patients improve in spite of continuing high levels of urinary oestrogen and others deteriorate when oestrogen excretion has fallen to negligible amounts (Brown, et al., 1959)-Several studies have indicated, however, that post-menopausal women with breast cancer differ from the normal population by an in creased excretion of oestrogen.T he physio logical significance of this finding remains obscure, and more detailed study is required to establish its possible relationship to the development of breast cancer.
Overall, the efforts to relate oestrogen excre tion to the course and prognosis of human breast cancer have been inconclusive.

U rin ary A ndrogens an d C orticosteroids
' T h e estim ation of urinary androgen and corticosteroid metabolites has proved the most hopeful approach so far.A major contribution in this field has resulted from the collaborative studies in London by G u y 's Hospital and the Imperial C an cer Research Fund.W orkers there have investigated patients w ith advanced disease, patients w ith early breast cancer and normal wom en.

A dvanced b reast cancer
Urine was collected before operation from patients treated by adrenalectom y or hypophysectom y and estim ations were made of gonado trophins, androgen, oestrogen, corticosteriods and progesterone m etabolites.B y retrospective analysis of the data it was found that patients w ho responded satisfactorily to th e operation were characterised by a relatively low excretion of corticosteroids and relatively high levels of urinary androgens.A reasonable distinction between the responsive and unresponsive clinical groups was obtained by expressing 17hydroxy corticosteroids (17-O H C S ) and 11deoxy-17-oxosteroids as a sim ple ratio.
A more sophisticated statistical approach led to the formulation of a discrim inant func tion w hich takes the form 80-80 [17-O H C S (mgm /24 hrs.)] + aetiocholanolone* (μ gs/24 hrs.) T h e design of the formula is such that a negative or positive num ber is obtained depend ing on the relative values of the steroids.By this means a clearer distinction between the clinical groups was obtained.A significant majority of patients w ith a " positive discrimin ant" responded to adrenalectom y or hypophysectom y whereas th e results of treatm ent in the " negative discrim inant" category were generally poor (Bulbrook, et al., 1962).
W h e n the discrim inant function is con sidered in association with clinical observations a group of poor-risk patients w ith advanced disease is revealed.Rem ission is obtained in on ly 2 per cen t of " discrim inant-negative" patients in whom the primary tum our was noted w ithin six years after the menopause and in whom recurrence had appeared within two years of primary treatment.cancer also obtain at an early stage of the disease.A negative discrim inant function is a rare finding in normal wom en under the age of 60 years but has been observed in about 50 per cent of patients w ith early breast cancer.In these patients the incide nce of recurrence is four times higher and the death rate is eight times greater than those with a positive dis crim inant (Bulbrook, et al., 1964).

P rospective stu d ies on norm al w om en
T h e preceding observations suggest that in som e women with breast cancer there is an abnorm ality in the secretion of androgens and corticosteriods.Prognosis is poor in these patients, perhaps because the tum our has be com e horm one-independent.T o determ ine whether the abnorm ality precedes the develop m ent of disease or results from the presence of th e tum our a prospective survey was set up in G uernsey in 1961.
A lm ost 5,000 healthy fem ale volunteers from this relatively closed population have provided one com plete collection of urine.T o date, breast cancer has developed in nineteen of the wom en.Analysis of the results suggests that the excretion of androgen and corticosteroid m etabolites was abnormal in a significant pro portion of the patients prior to the develop m ent of the disease (Bulbrook and Hayward, 1967).

CONCLUSION
T h e m anagem ent of breast cancer presents a form idable challenge to the clinician.Inten sive research on the relationship of the endocrine system to the induction and m ain tenance of mammary tumours has provided some insight into the nature of the disease.Some tumours appear to regress after alteration of the hormonal environm ent in w hich they have developed, others do not.It is not yet possible to determ ine the biological nature of a particular tum our at a sufficiently early stage in the disease to orientate the search for effective counter-measures.T here are indic ations however, that poor-risk patients, and even healthy w om en w ho are " susceptible" to the disease, m ight be identified by abnorm al ities of secretion of urinary steroids.It should be possible in the future to confirm these suggestions and to sim plify the measurements so that they are more generally available.T h ere after, attem pts to produce a more favourable endocrine environm ent m ight not only be possible, but acceptable in a society in which hormonal m odifications for fertility and birth control arc now com m onplace.

A C ase for Dissection
In an anonymous monograph in the Library entitled " An Answer to the Several Attacks which have appeared against the University of Edinburgh" the author includes in his refutation of the " calumnies which have been circulated against this University, particularly that lately published in the city of D ublin" an interesting little snippet on Edinburgh dissecting habits in the early 19th century: " Let us figure to ourselves a group of students assembled round a subject.One is cutting out an eye, another is making a prize of the heart, which he is proceeding to examine at his leisure, a third is grubbing in the entrails, and a fourth practising the insertion of the female catheter; the whole party brandishing their scalpels, singing and laughing, as merry as coblers (sic) over their awls.Suppose that the lecturer were to come in and to recognise in this one-eyed, mangled, heartless creature, the frail remains of the wife of his bosom, the sister, or daughter, whom he had cherished with the tenderest affection!How would he be able to bear such a sight?W ould he not start back with horror?W ould it not put all his philosophy at once to flight?"Perhaps they weren't calumnies after all . . .

1.
Induction of tum ours by horm ones O n ly three years after D oisy had isolated the first crystalline oestrogen, Lacassagne (19 32) induced m am m ary cancer in m ale m ice by repeated injections o f the horm one.M any workers since have confirmed that oestrogens may prom ote tum our growth in several species (N oble, 19 57).
2. E arly b re a st cancerT h ere are indications that the relationships originally dem onstrated in advanced breast *A etiocholan olon e is one o f the u rin ary androgen excretion products.