TIPS-April 1983 man? Secondly, how far can such models be used to predict the clinical utility of potential anti-allergic compounds? The former question is clearly pertinent since a number of compounds in therapeutic use, including plasma substitutes, solubilizing agents for certain drugs, and anti-histamines can act as histamine liberators in particular animal models. Of the species tested, the guinea-pig is perhaps the animal which most closely resembles man in its response to such agents 2. Preliminary studies in this animal, together with investigations on isolated human mast cells, may then give some indication of whether a compound is likely to act as a histamine liberator in man. However, given the diversity of mast cells even within a single species, the ultimate criterion for this designation can at present only be based on carefully controlled trials on human subjects. The heterogeneous response of mast cells to particular drugs renders the determination of anti-allergic activity most difficult. This problem is exemplified, for example, by the drug cromoglycate which is active in man against mastocytes in some areas (lung, nose and conjunctiva) but not others (skin, basophils). This observation is of fundamental clinical importance since it cannot be assumed that a drug active against mast cells in one site will necessarily he effectiv~against mast cells in another site. If a similar selectivity is observed with
167 other cromoglycate-like drugs, then no single compound can act as a panacea for all allergic conditions. The availability of isolated mast cells from defined locations (lung, nose, skin, conjunctiva, and bowel) would then provide a series of test systems for the screening of drugs directed against specific inflammatory problems. In this context, the use of dispersed cells rather than chopped tissues offers distinct advantages since problems of diffusion, nonspecific adsorption, access of reagents and masking of receptor sites are avoided. Intense effort is now being directed towards the isolation of mast cells from different tissues and such work should facilitate the development of new and more effective anti-allergic agents.
Reading list 1 Kazinuerczak, W. and Dmmant, B (1978) Prog. Allergy 24, 295-365 2 Pearce, F. L. (1982) Klimsche Wochenschr 60, 954-957 3 Befus, A D., Pearce, F L , Gauldie, J , Horse.. wood, P. and Bmnenstock, J. (1982)J lmraunol. 128, 2475-2480
A. B. G. Lansdown Life Science Research, Stock, Essex, CM4 9PE, U.K.
Prolactin is a particularly versatile hormone which has now been identified in a wide range of species throughout the animal kingdom. In mammals, prolactin is a polypeptide hormone with close structural similarities to growth hormone. With advances in protein chemistry in the early 1970s, human prolactin was isolated and since then has become a subject of intense clitiical interest and controversy. Not least amongst these controversies is the part played by prolactin in the genesis of human breast cancer. Clearly much of the concern over the involvement of prolactin in breast cancer
467-530, Sprmger-Verlag,Berlin,Heidelbergand New York
Acknowledgements Work from the author's laboratories was supported by grants from Fisons Pharmaceuticals Ltd, the MRC, NATO, SERC, Wellcome Trust and World University Service.
Prolactin secretion and mammary cancer Introduction
4 Pearce, F L., Befus, A. D., Gauldie, J. and Bienenstock, J (1982) J. lmmunol. 128, 2481-2486 5 Pearce, F. L and Ennis, M. (1980) Agents Acnons 10, 124-131 6 Enms, M. and Pearce, F L (1980)Eur J Pharmacol. 66, 33%345 7 Enms, M. (1982)AgentsActtons 12, 29~32 8 Church, M K , Coleman, J. W., Holgate, S. T , Pan, G. J.-K and Welch, M. T. (1981) Br. J. Pharmacol. 74, 979..-980P 9 Pearce, F. L , Behrendt, H , Blum, U., PobleteFreundt, G., Pult, P., Stang-Voss, C. and Sehmutzler, W (1977)Agents Actions 7, 45-56 10 Gadand, L G.,Green, A F. andHodson, H. F. (1978) m Handbook o f Experimental Pharmacology (Born, G. V. R., Farah, A., Herken, H and Welch, A D , eds), Vol. 50/I1, pp.
derives from experimental studies in laboratory animals, notably the Sprague--Dawley rat. These studies have lead to the general conclusion that prolactin levels higher than normal in the circulation predispose to breast cancer. The true relevance of this concept to the human situation is far from clear, however, since it is now well known that prolactin subserves different functions in different species. Thus, in the rat and some other species, prolactin acts in osmoregulation, post-natal growth, reproductive function, differentiation of the integument and mammary glands, and in lactation. In humans, the exact function of prolactin is still not clear. A recent Lancet
Fred Pearce obtained his B.Sc. and Ph.D. at University College London in 1967 and 1971 respecnvely. He is currently Lecturer m Chemistry and Tutor to medicinal chemistry students
editorial stated that apart from its obvious role in lactation, we are no clearer now about what, if anything, prolactin does in humansL Hyperprolactinaemia is a problem in clinical medicine and the accompanying comphcations in women include breast enlargement with or without abnormal secretory activity (galactorrhoea), and menstrual failure (amenorrhoea). In men, gynecomastia and galactorrhoea are recognized complications. The regulation of prolactin release through the so-called prolactin inhibiting factor (PIP) is well documented 2. The PIP is secreted at the hypothalamus and acts on the pituitary by way of the hypothalamicpituitary portal vasculature. This inhibitor is now recognized to be largely, if not entirely, dopamine. Thus those drugs which block dopamine synthesis or dopamine receptors in the pituitary, will lead to increased prolactin secretion and possibly the sequalae noted above. Hyperprolactinaemia results from trans-section or obstruction of the portal blood vessels. In contrast, dopamine agonists like hromo-
~c~1983,Elsevte~S4:~ncePubhshersINV , Amsterdam XXXX- XXXX/83/$OI00
168 criptine (CB-154), apomorphine and ergot conditioning agents. It may be that in the alkaloids, markedly suppress prolactin sec- promoting stage, cells altered constitutionretion, may lead to cell hypoplasia, and ally by a mutagen undergo a form of ultimately regression of the symptoms of 'awakening' process and an ability to selfhyperprolactinaemia. perpetuate. In the absence of a promoting Hyperprolactinaemia as a potential com- situation, transformed cells remain dormant ponent in human breast cancer was high- and may even regress. Promotors seem to lighted several years ago in relation to the exhibit weak or negligible ability to invoke use of reserpine, a drug known to deplete mutagenic changes by themselves but are dopamine reserves and to enhance protactin capable of inducing RNA and DNA replicasecretion. Although some supportive evi- tion. Evidence favours prolactin as a prodence for an association between reserpine moting agent. It seems to have no proven use and breast cancer was provided by ability to precipitate neoplastic changes but studies in Finland and the USA, subse- is known to stimulate DNA and RNA synquent large scale investigations involving thesis in cells in vitro. 12 000 patients have failed to confn'm this concern. Nowadays nobody seriously con- Experimental studies in animals relating siders reserpine to be a potential cause of hyperprolactinaemia to mammary human breast cancer. carcinogenesis Nevertheless, there is still real concern The action of prolactin in mammary carabout the role of hormones, including cinogenesis has been examined in a large prolactin, in breast cancer and colloquia are number of experimental studies, those still being organized on the subjecP. The using rodents being most relevant in the long-term effects of neuroleptic and other present context. Broadly, they fall into two drugs causing chronic hyperprolactinaemia main categories, those in which tumour are still under surveillance particularly in incidence has been observed in relation to view of the fact that several of these drugs natural or experimentally induced hyperhave been associated with mammary can- prolactinaemia, and those in which animals cers in laboratory animals. have been treated with potent carcinogens It is my intention to consider here the such as 7,12-dimethylbenzanthracene general question: is there good evidence (DMBA), diethylstilboestrol or 3that prolactin per se is a primary cause of methylcholanthrene (3-MCA) and in breast cancer in any species including which tumour induction growth rates have humans, or that in some way it creates a • been plotted in relation to circulating levels favourable environment for the growth of of prolactin and other hormones. The neoplasias induced through other factors? It experiments, some involving elaborate is notable that as clinical assays for prolac- microsurgery, point to the general conclutin have been improved in recent years, so sion that 'spontaneous' tumours are more observations have become more accurate frequent and grow more rapidly in animals and hence more reliable than those con- with inherently higher prolactin levels or ducted earlier. where prolactin secretion is enhanced through physiological or experimental The two-stage carcinogenesis concept means 5. From the numerous studies in animals, In rodents, spontaneous mammary prolactin has been identified as a tumour tumours are usually benign adenofibromas. initiator, promotor and conditioning agent. They occur in up to 80% of female animals. It is necessary, therefore, before discussing They are more prevalent in older and multhe various views on the action of prolactin tiparous animals and in those strains with in carcinogenesis, to discuss briefly what is higher prolactin levels. currently understood by the term 'carHyperprolactinaemia has been induced cinogenesis'. in rodents in a variety of ways: pregnancy, Berenblum proposed that carcinogenesis grafting supplementary pituitaries, is not a simple all-or-none response in administration of exogenous prolactin, which cells are transformed from a normal lesions in the hypothalamus and hypophysto a neoplastic state with a propensity for ial stalk to impair the dopaminergic influuncontrolled growth, but is a multi-stage ence, and dosage with dopamine antagonprocess 4. Two stages at least are clearly ists. In each case, the tumour incidence was recognized, initiation and promotion. I n i - directly related to the circulating prolactin tiation supposedly involves a somatic mutalevels. In contrast, if animals were subtion in susceptible cells by environmental or jected to a hypoprolactinaemia treatment endogenous factors. The change is irreversfrom early in life, then tumour incidence ible but may not progress further with cell was generally low. proliferation and frank tumour formation in These experiments suggest that prolactin the absence of favourable conditions. Such actually induces tumour formation in the an environment is created by promoting or absence of other obvious agents. However,
TIPS - April 1983
no evidence is available, as far as I am aware, to show that a critical level of prolactinacmia is necessary for tumour formation to occur, or that prolactin is mutagenic in vitro. It is held that initial neoplastic transformation arises at an early stage through the action of viruses or other environmental agents and that prolactin promotes tumour growth. Tumours induced by carcinogens differ in the important respect that most are malignant tumours. Many are strongly prolactin dependent. In the absence of prolactin they may regress. The time when hyperprolactinaemia is induced through drug action or microsurgely has a marked influence on tumour formation by DMBA and 3-MCA. High prolactin favours tumour growth, but if hyperprolactinaemia is induced by stress, drugs, surgery, etc., prior to carcinogen treatment, tumoar growth is inhibited. This may be due to the fact that cells in the ductal epithelium which are sensitive to carcinogens, are suppressed by prolactin and that carcinogen resistant cells of the mammary alveoli are activated. Similarly, if a carcinogen is given to lactating animals when DNA synthesis is low in sensitive cells and when alveolar cells are secreting, tumorigenesis is minimal. The carcinogen is voided in the milk. There is evidence that prolactin per se is insufficient to sustain the growth of DMBA or 3-MCA induced turnouts and that oestrogen is also necessary. Oestrogen stimulates prolactin secretion in the pituitary but prolactin is known to activate oestrogen receptors in the mammary gland. Limited evidence is available to show that oestrogens alone allow the growth of mammary tumours in hypopbysectomized rodents. It would thus seem that in these tumours prolactin facilitates oestrogen binding. In summary, tumour growth in rodents depends upon adequate circulating levels of prolactin and other hormones, notably oestrogen. Although prolactin undoubtedly fulfds a central role in rodent carcinogenesis it is still unclear whether it should be regarded as a true promotor, like croton oil, in two stage carcinogenesis, or whether it acts in some way to 'condition' susceptible cells to the action of environmental carcinogens.
Hyperprolaetinaemia in humans The role o f endogenous and disease processes Prolactin is commonly classified as a 'stress hormone' is humans and elevated prolactin levels are seen in a variety of situations including hospital interviews, gynecological examination, endometrial biopsy and venepuncture. The prolactin rises are highly variable according to the
71PS - April 1983 patient and type of stress involved; they are transitory and normally decrease as the stress subsides without untoward sequelae. In the human population, prolactin levels are usually higher in women than men but they show a clear diurnal rhythm 2. Prolactin levels are higher in the afternoon and evening than in the morning. Peak levels are seen during sleep. Nursing is a powerful stimulant for prolactin release in women, and this is no doubt mediated by cutaneous sensations at the breast and nipple. Diseases such as depression, schizophrenia and Huntington's chorea have occasionally been associated with higher-thannormal prolacdn levels but the reports tend to he inconsistent. It is important to note that elevated prolactin levels recorded in patients treated for depression and attributed to the action of the drugs may in part he associated with the disease state". Pituitary tumours Prolacdn synthesis occurs in the lactotroph cells of the anterior pituitary. In pregnancy and during lactation this tissue hypertrophies and this phenomenon is accompanied by breast development and lactation. Foci of lactotroph cells may proliferate through unknown causes in the form of a benign tumour or 'prolactinoma' and be accompanied by galactorrhoea and amenorrhoea in women, and gynecomastia and impotence in men. The diagnosis of prolactinomas on radiological evidence alone is often difficult. These tumours, often quite minute, may occur in up to 30% of clinically normal patients with unexplained secondary amenorrhoea 7. Estimates vary greatly. In the aetiology of prolactinomas, it is tempting to speculate that oestrogens as used in oral contraceptives are contributory. They are known to promote prolactin secretion by a direct action on the pituitary gland, and to cause prolactin-cell hyperplasia. The incidence of prolactinomas in oral contraceptive users varies greatly and inconsistencies emphasize large gaps in our knowledge. The literature does indicate that whatever their causation, prolactinomas are frequently diagnosed on the basis of hyperprolactinaemia and abnormal breast secretions. The symptoms resolve following hypophysectomy to remove the tumour or treatment with dopaminergic drugs to suppress the hyperprolactinaemia. For longterm therapy, the latter could prove prohibitively expensive. Bromocriptine treatment does lead to a significant reduction in prolactinoma size without undesirable side effects. Drug therapy Antipsychotic and centrally acting drugs
169 which inhibit dopaminergic pathways cause hyperprolactinaemia in humans and experimental animals. This effect is specific and is not seen with agents or metabolites not active at the hypothalamic level. Gruen" believed that responses in humans to dopamine antagonists were proportional to their antipsychotic activity and clinical potency. The better known prolactin secretogogues are shown in Table I. The neuroleptic drugs tend to be more active hyperprolactinaemic agents and have been widely examined. For many drugs, there seems to he a reasonably good correlation between the circulating prolactin levels and blood levels of active drugs, but many exceptions to this are seen in the literature. The magnitude of prolactin release varies considerably according to the drug, its daily dose and the patient. Women exhibit a more marked response than men. Prolactin release tends to increase with drug dose until reserves in the pituitary are exhausted; at this stage, prolactin release reflects its rate of synthesis. Some workers believe that no state of tolerance is achieved and that high blood levels are maintained in response to a particular drug dose. Thus a drug dose equivalent to 600 mg of chlorpromazine is believed to achieve maximal disinhibition of prolactin secretion in most patients. This disinhibition declines after each dose of drug but new transitional peaks result with each additional dose. Although in the eyes of some psychiatrists and endocrinologists, chlorpromazine is not truly representative of prolactin seeretogogues, it appears to have been more widely examined than other drugs. Even with this drug, wide variations in response are seen.
With the various groups of drugs used in clinical trials, considerable variability in prolactin response has been seen depending upon whether the drug was given once or over a number of weeks. Most studies have recorded increased prolactin levels after multiple doses of neuroleptic drugs, but on occasions single or short-term dosage lead to a similar or higher level of prolactin followed by a decline to normal after several days. Clearly, the ability of the pituitary gland to secrete prolactin in response to drug therapy depends upon its capacity to store and synthesize the hormone. The prolactin content in the adult human pituitary was estimated to be 44-220 btg/gland, although this estimate may be low in the light of more accurate assays available now. Rubin and Hays8 calculated that 0.5 mg of haloperidol given intravenously or intramuscularly could liberate up to 450/zg of prolactin (mostly stored hormone) in the
initial 2 h after dosing. Thus, any variability of prolactin secretion seen would reflect individual differences in the ability of the pituitary to store hormone.
Mammary gland changes in humans in relation to hyperprolactinaemia Galactorrhoea and related disorders Hyperprolactinaemla is a relatively common chnical problem occurring in more than 25% of women with secondary amenorrhoea. Amenorrhoea, and galactorrhoea are well known clinical sequalae of the problem irrespective of its underlying cause, prolactinoma, psychoses or chronic drug therapy. In culture and probably in situ, prolactin exhibits an ability to promote RNA-synthesis and to transform mammary epithelial cells from an intermediate, resting phase to a secretory profile with casein and a-lactalbumcn synthesis. This transformation regresses in the presence of progesterone. It is suggested on experimental grounds that prolactin exerts its effects by a direct action on mammary cell nuclei. This remains to be confirmed. Although high prolactin levels are a cause of galactorrhoea and breast enlargement in women, they are not regularly seen in patients with pituitary tumours or with chronic drug therapy x°. On the other hand, a surge of prolactin secretion accompanies human pregnancy and with the combined stimulus of prolactin, oestrogens and progesterone, further breast development occurs and lactation commences. After parturition in humans, the levels of circulating oestrogen and progesterone decline but high levels of prolactin persist through lactation. Incidences vary as to the frequency of galactorrhoea and amenorrhoea as a side effect of drug therapy but the most prevalent form of galactorrhoea is deemed to he that associated with regular menses ~ I he majority of these patients (85%) exhibit prolactin levels within the normal range. Hyperprolactinaemia is present in most patients with galactorrhoea with accompanying amenorrhoea, but by itself may not he adequate to miuate the process; this is not established. In most cases reported where h)poprolactinaemia v, as induced through bromocriptine therapy or withdrawal of the offending drugs, symptoms of galactorrhoea regressed and reproductive function was restored. In only one report of patients, subject to chronic phenothlazme therapy. were the breast secretions abnonnaP. The expressed milk was richer in-fat and protein content than normal human milk and resembled colostrum. "Ihe sigmficance of this observation is not clear.
170 Gynecomastia Gynecomastia or enlargement of the breast in males is a common disorder and has been associated with neuroleptic and hyperprolactinaemic drugs. However, prolactin levels are within normal limits in most patients with an idiopathic gynecomastia and it is well known that enlarged breasts do not develop in most subjects with hyperprolactinaemia. Gynecomastia is reputed to be more common in men with hypogonadism and would seem to be a consequence of a hormonal imbalance. Androgen levels are frequently lower than normal whereas serum oestrogens may be elevated resulting in an increased oestrogen-androgen ratio. Thus breast enlargement is a manifestation of feminization in Klinefelter's syndrome. The incidence of gynecomastia due to drugs is not clear but when it does occur, it is probably secondary to hypogonadism rather than due to a direct effect of the drug or to the effects of the high levels of prolactin on mammary gland. Gynecomastia attributable to hyperprolactinaemic drugs tends to be a transitory change regressing when the use of offending drugs is discontinued or hormone balances are restored by bromocriptine or testosterone therapy. This regression may only be partial, however, since not all patients treated with hyperprolactinaemic drugs exhibited higher-than-normal prolactin levels. In general, men show a significantly lower response to most drugs known to cause hyperprolactinaemia in women.
Hyperprolactinaemia and m a m m a r y cancer in humans The hormone dependence of some human breast turnouts has been known for many years and as more accurate assays have been developed so the scientific and clinical interest has increased. Prolactin dependent tumours account for at least 30% of all human mammary tumours and these are known to regress with ovariectomy, adrenalectomy or hypophysectomy. Although evidence reviewed above suggests that endogenously mediated hyperprolactinaemia or high circulating prolactin due to some drugs favours tumour growth in animals, evidence that prolactin plays a primary role in the induction or growth rate of human tumours is inconclusive. Contraceptive steroids are a known cause of high prolactin levels in humans but epidemiological evidence so far indicates a
TIPS - A p r i l 1983 TABLE I Prolactin secretogogues
Phenothlazines Chlorpromazme Fluoroperazme Perphenazine Ttuondazine
H~receptor antagonists Cimetidine Serotonerglc agents L-Tryptophan
Other agents Reserpme Imlpramme Diphenylbutylpipendlnes Oestrogens
Butyrophenones Haloperidol Pimozide Thioxanthenes Ttuothixene
Dlbenzoxazepmes Loxapin
important role for prolactin in idiopathic and carcinogen-induced rodent tumours. However, apart from the inherently high prolactinaemia in high risk breast cancer patients, there is no conclusive information to show that prolactin contributes significandy to the genesis of human mammary tumours. There is no consistent evidence that patients who develop prolactinsecreting pituitary tumours, or galactorrhoea and breast enlargement through long-term therapy with neuroleptic or other drugs, are more susceptible than other people to develop tumours of the breast.
Reading list lower incidence of breast cancer in these women. Evidence that other drugs and conditions which lead to increased prolactin secretion are related to breast cancer in women or men is also unconvincing. Phenothiazine tranquillizers have been used extensively in treating mental patients for many years and no greater risk of breast cancer has been demonstrated in this group of patients than in the normal population. Where galactorrhoea, breast enlargement and gynecomastia have been reported, no evidence has been detected of neoplastic changes and the prolactin-related breast responses have, in general, regressed when the drug has been withdrawn, although only on rare occasions have breast biopsies been taken. Although there seems to be a paucity of substantive information from human studies that induced hyperprolactinaemia either predisposes to or enhances mammary cancer, Henderson and Pike ~ noted that in six comparisons of breast cancer patients, five showed higher prolactin than matched controls. Also a consistently elevated prolactinaemia was seen in the daughters of breast cancer patients. Oestrogen levels were also elevated. In humans, the risk of breast cancer is greatest with the first fullterm delivery and decreases with subsequent pregnancies. Prolactin levels also decline with parity unlike the situation in rodents. Although prolactin may participate in breast cancer formation in high risk families, this does not appear to be the case in the general population. Prolactinomas and long-term drug therapy leading to high circulating prolactin levels have not been 'associated with an increased breast cancer incidence.
Condnsion The evidence discussed here points to an
I EdttoriaI(1979)Lancet,4August, 234-235 2 Franz, A. G (1978) New Engl. J. Med. 298, 201-207 3 Henderson, B . E . and Pike, M . C . (1981) in
Banbury Report Hormones and Breast Cancer
4 5 6 7 8 9
l0
(Pike, M. C., Siiteri, P K. and Welsch, C. W., eds), Vol. 8, 115-130 Berenblum, I. (1980) Br. J Cancer 41,490--493 Welsch, C. W and Nagasawa, H (1977) Cancer Res. 37, 951-963 Green, P. H. (1978)Med. Clin. NorthAm. 62, 409-424 Reichlin, S. (1979) New Engl. 3. Med. 300, 313-315 Rubm, R. T. and Hayes, S. E. (1979) Psychopharmacology 6 l, 17-24 Beaumont, P. J. V., Gelder, M . G . , Friesen, H. G andHarris, G. W (1974)Br.J. Psychiatry 124, 413-419 Schyve, P. M , Smlthlme, F. and Meltzer, H. Y. (1978)Arch. Gen. Psychiatry 35, 1291-1301
Dr A. B. G. Lansdown graduated with his Ph.D. from London University in 1969 and has done research at the British Industrial Biological Research Association and the Medtcal Research Council into pathological mechanisms of drug toxicity, dermal irritancy and abnormal foetal
development. He was pathologist with the Wyeth Institute of Medical Research from 1977--1982 and became Deputy Head of Pathology for Life Sctence Research in August 1982.