Acta Oncologica
ISSN: 0284-186X (Print) 1651-226X (Online) Journal homepage: http://www.tandfonline.com/loi/ionc20
Progestins in Breast Cancer Treatment: A Reveiw Steinar Lundgren To cite this article: Steinar Lundgren (1992) Progestins in Breast Cancer Treatment: A Reveiw, Acta Oncologica, 31:7, 709-722, DOI: 10.3109/02841869209083859 To link to this article: http://dx.doi.org/10.3109/02841869209083859
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 50
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ionc20 Download by: [Laurentian University]
Date: 10 March 2016, At: 06:19
A C T A O N C O L O G I C A Vol. 31 No. 7
1992
PROGESTINS IN BREAST CANCER TREATMENT A reveiw
Downloaded by [Laurentian University] at 06:19 10 March 2016
STEINARLUNDGREN
The two most widely used synthetic progestins in breast cancer treatment, medroxyprogesterone acetate (MPA) and megestrol acetate (MA), are reviewed with regard to pharmacological, endocrinological and clinical aspects. In high oral doses as second- or first-line endocrine therapy in advanced breast cancer, they give a similar response rate as tamoxifen (TAM) and aminoglutethimide(AG). The mechanism of action is probably complex. Considerable changes in serum levels of different hormones are induced by progestin treatment. The decrease of serum estrone sulfate (E,S) may be part of the therapeutic mechanism. Some studies suggest that the two drugs, M P A and MA, have a different mode of action, and possibly a low cross resistance. Randomized studies using the two progestins with a cross-over design may answer these questions. Further studies on the influence of progestin on different receptors and growth factors are warranted. To determine the most effective clinical dose of the two progestins, studies with increasing therapeutic doses are needed.
Endocrine therapy, the oldest form of systemic treatment of advanced breast cancer ( I ) , has an important place in breast cancer treatment. For many pre- and postmenopausal women with advanced breast cancer, endocrine treatment is currently the most effective and welltolerated treatment. The median duration of response after first-line treatment is 12-18 months, and for some patients a response duration as long as 3-10 years has been observed (2). Response to first-line treatment may also predict a further response to second- and third-line endocrine therapy. As adjuvant treatment, endocrine therapy has resulted in reduced numbers of recurrences and prolonged survival (3). A possible role of endocrine agents in preventing breast cancer has also been discussed (4-7). Improvements in endocrine therapy have been achieved by better ways of selecting patients likely to respond (i.e. steroid hormone receptor measurements, prior response to endocrine therapy), by new drugs with less side-effects and better understanding of the mechanism of action.
Received 19 August 1991. Accepted 26 August 1992. From the Department of Oncology, Regional and University Hospital of Trondheim, N-7006 Trondheim, Norway.
Concerning hormone receptors only estradiol receptor (ER) and progesterone receptor (PgR) assays have at present clinical importance. Further studies of other steroid receptors (androgen receptor, AR; glucocorticoid receptor, GcR) may add new information with regard to the mechanism of action of different hormonal agents used in breast cancer treatment. Information on receptor modulation and the influence on growth factors and their receptors by different endocrine therapies can possibly be used in new treatment strategies. Clinical and experimental data have emphasized the importance of estrogens as potent growth factors in breast cancers. The main strategy of endocrine treatment is therefore to reduce the amount of estrogen available to tumor cells. This can be achieved by ablation of hormone producing glands, or by adding agents which reduce the estrogen effect. For several reasons most of the ablative forms of treatment (oophorectomy, hypophysectomy and adrenalectomy) have been replaced by additive therapy. Additive treatment with androgens, estrogens or progesterone has, at present, no place in breast cancer treatment due to sideeffects of the two first mentioned and short duration of action of progesterone. Glucocorticoides play an important role as a symptom-relieving treatment of CNS metastases and of the lymphangitic form of lung metastasis, but 709
710
S. LUNDGREN
antifertility properties, but since Cole et al. (15) in 1971 PPOTH~LAMUS ) described the clinical use of TAM in patients with ad-
PERIPHERAL TISSUE
\
I
J
A
vanced breast cancer, the drug has been widely used as first-line endocrine treatment for postmenopausal women. In randomized trials, TAM has not given a higher response rate than other endocrine treatment modalities, such as progestins (16-19) or aminoglutethimide (20, 21) but it has in most cases a more favorable side-effect profile. Its exact mechanism of action is not known, but it may, at least partly, be mediated by competitive binding to the ER, thereby preventing endogenous estrogens from forming a complex (ER-E) that interacts with DNA to give the estrogenic effect. Antiestrogenic binding sites, different from ER, have also been found in human breast cancer biopsies, although no role of this binding for the therapeutic action of TAM has so far been shown (22, 23). The increase in the serum level of sex hormone binding globulin (SHBG)induced by TAM (24, 25) may possibly play a role by decreasing the free fraction of active estrogens available to the target cells. The usual dose of TAM is 20 mg given once daily and higher doses have not been shown to give better results (26, 27). Due to its low side-effect profile, TAM has also been used in adjuvant settings (28). Based on several randomized studies with adjuvant TAM treatment, a recent meta-analysis has shown a small but highly significant mortality reduction by TAM treatment (29). Due to the estrogenic effect of TAM, a possible carcinogenic influence on the endometrium has recently been discussed (30). New anti-estrogens have recently been introduced, possibly with lower or no estrogen agonist properties and more favorable pharmacokinetic properties (31, 32). Aromatase inhibitors. In postmenopausal women, estrogens are synthesized in peripheral tissues. Androgens from the adrenals (2/3) and ovaries (1/3) are transformed to estrogens by aromatization (33, 34) (Fig. 1). Aromatase inhibitors are drugs that can, at least partly, inhibit this synthesis pathway. The main drug in clinical practice so far is aminoglutethimide (AG) (35, 36). AG was introduced as an anticonvulsant but was soon found to suppress steroid synthesis in the adrenals. Due to the feed-back increase of ACTH, a concurrent use of corticosteroid is therefore mandatory to avoid adrenal stimulation (35). Probably more important is the inhibition of the conversion of androstenedione to estrogens in peripheral tissues, the aromatase inhibition (35). Recent investigations have also suggested a stimulating influence of AG on the metabolism of estrogens (37). The usual dose schedule is 250mg q.i.d. with glucocorticoid substitution (35, 36). In randomized studies similar response rates as for TAM (20, 21), MPA (38) and MA (39) have been reported, but with apparently more side-effects, at least during the first weeks,of treatment (39). AG initially causes a substantial drowsiness in approximately 40% of patients and 1/3 of patients have self-limiting morbilliform, macular-papular
Downloaded by [Laurentian University] at 06:19 10 March 2016
UE(2) I B--SHBG tk +
TARGET TISSUE
E-SHBG
Fig. 1. Production of estrogen and androgen in pre-( 1) and post-(2) menopausal women. Abbreviations used: E = estrogens, A = androgens, SHBG = sex hormone binding globulin, LH = lutenizing hormone, FSH = follicle stimulating hormone, ACTH = adrenalcorticotropine hormone.
give a low objective response rate. Only the additive forms of treatment in clinical use are discussed below. Some newer treatment modalities
The site of estrogen production is different in pre- and postmenopausal women (Fig. 1) and treatment strategies therefore depend on the menstrual status of the patient. Chemical castration. In premenopausal women, the estrogens are mainly produced in the ovaries under influence of gonadotropins from the pituitary gland (Fig. 1). The effect of oophorectomy or irradiation of the ovaries is well documented (2, 8). For the non-responding patients, however, i.e. about 2/3, these procedures cause an irreversible menopause without therapeutic gain. A reversible medical castration has recently become achievable by luteinizing hormone-releasing hormone (LH-RH) analogues, which interfere with the pituitary LH-RH receptors. Thus surgical or radiological castration can be replaced. by a reversible treatment procedure (8- 10). The LH-RH analogues are now best administrated by a subcutaneous depot injection given once monthly. Side-effects are seldomly reported, except for symptoms related to the induced menopause (8). LH-RH treatment has shown a low response rate in postmenopausal women with advanced breast cancer ( 11- 14), and is mainly used in premenopausal patients. Anti-estrogen. Antiestrogens may be defined as substances that antagonize estrogen action, and the drugs currently used are tamoxifen (TAM) and substances derived from TAM. They were originally designed for their
71 1
Downloaded by [Laurentian University] at 06:19 10 March 2016
PROGESTINS IN BREAST CANCER
skin rash (35, 36). Due to these side-effects more pure aromatase inhibitors have been searched for. 4-hydroxyandrostendione (4-OHA) is a potent new aromatase inhibitor, reducing serum E, levels and causing tumor regression in postmenopausal patients with advanced breast cancer. The drug has very few side-effects, and studies determining the optimal dose schedule and route of administration are in progress (36, 40,41). The substances CGS 16494A (Ciba-Geigy) has also been shown to be a very potent inhibitor of aromatase activity in vitro, effectively blocking the conversion of androgenic precursors to estrogens in human placental microsomes. It is more potent than AG in reducing the growth of DMBA-induced rat mammary tumors (42). Phase I1 trials with different doses have been started as second-line therapy in patients with metastatic breast cancer. Progestins. Progesterone was shown to have antiestrogen effects soon after its discovery in 1930 (43). When used as treatment of advanced breast cancer in 18 patients, an objective response was seen in 3 patients. It has, however, been rarely used due to the short duration of response (44, 45). To cope with this, several synthetic progestins have been synthesized. The two main drugs in clinical practice are medroxyprogesterone acetate (MPA) and megestrol acetate (MA) (Fig. 2). These drugs have been known since the 1960s (46-48), and the first clinical trial in breast cancer, using a low dose, showed a response rate of 17-18% (49). A breakthrough came when an Italian group published results of using MPA i.m. in high doses (50).
Progestins as natural anti-estrogens Effect on serum level of estrogens. The serum levels of steroid hormones depend on the rate of production and the metabolism of these hormones. They are transported in blood bound to different transport proteins. The most potent estrogen, estradiol (E,) and testosterone (T) are mostly bond to sex hormone binding globulin (SHBG), and most probably only the unbound fraction of the hormone is immediately available for metabolism and for responsive cells (51). In premenopausal women, estrogens are mainly produced in the ovaries, influenced by the gonadotropins (Fig. 1). The use of hypophysectomy or LH-RH analogues reduces the serum level of luteotropic hormone (LH) and thereby the production of estrogens (2, 8, 9). The hypothalamic-pituitary influence of progestins seen in postmenopausal patients (52-56) also occurs in premenopausal women. In postmenopausal women the estrogens are produced in peripheral tissues with androgenic precursors as substrate (Fig. 1). About 1/3 of the precursor androgens are produced in the ovaries (34), the rest in the adrenal glands (33). Serum levels of androgens are reduced to about 60% of basal levels during progestin treatment (52, 55-59), due to progestin action on the hypothalamic-pituitary-adrenal axis (56, 58). Only one
Wr bl YLDROXYPIOOESTEROWE ACETATE
cy
I
Cn, Cl NEOESTROL ACLTATC
Fig. 2. Chemical structure of a) progesterone, b) medroxyprogesterone acetate (MPA) and c) megestrol acetate (MA). The only difference between MPA and MA is the double bound between C6-, in MA.
study reports an increased androgen catabolism (60), but this is probably of less clinical importance due to the minor influence of oral high-dose progestin treatment on drug catabolism by the mixed function oxidase system (61). The two progestins, MPA and MA, were found to induce closely similar reduction of the androgen in one study (56), but a significantly lower serum level of androstenedione was found during MPA treatment compared to MA treatment in another study (62). Estradiol ( E,) and estrone (E, ) both decreased by about 30% during progestin treatment (52, 55-57, 59). No significant difference between the two progestins was observed with regard to influence on serum levels of E, (62,63), and a significant difference in E, was found in one study (63), but not in another (62). The progestins have been reported to induce a substantial decrease in the serum level of SHBG (56, 57, 59, 63), which must be taken into consideration when the serum level of E2 is discussed, as only free E, is thought to be available for cellular uptake. This might be explained by differences between the two progestins with regard to their androgenic and/or glucocorticoid effect on hepatic protein synthesis. The different effect on SHBG cannot be
712
[
S. LUNDGREN
SHY-,
E,-SHBG eLooD
I
L SURROUNDING TISSUES
El
L
TMORCELL
J
(E,) production. Abbreviations used: A = androstenedione, T = testosterone, El = estrone, E, = estradiol, E,S = estrone sulfate, SHBG =sex hormone binding globulin. a = aromatase, dh = dehydrogenase, s = sulfatase. st = sulfotransferase. Downloaded by [Laurentian University] at 06:19 10 March 2016
Fig. 3. Pathways of intracellular estradiol
explained by the differences in serum concentrations of progestins (56,63). Breast cancer cells have been shown to transform estrone sulfate (El S) to more active estrogens (Fig. 3) (64-67). The serum level of E, S is about 20 times higher than that of E2 in postmenopausal women (37,63), and a 30% reduction of E, S (63) might be of importance for the action of progestins. The difference in reduction of El S in the circulation promoted by MPA and MA (63) has been proved to be significant (68). This could be explained, at least partly, by differences in serum levels of the two progestins used (63, 68). Whether the reduction of EIS found during progestin treatment is due to a decrease in production or an increase in metabolism requires further studies. The antiestrogen effect of progestins may be caused by a decrease of the amount of estrogen reaching the breast cancer cell. Whether the progestin-induced changes in serum levels of relevant steroid hormones cause similar alterations inside tumor cells is not known. Effect on cellular uptake and intracellular metabolism of estradiol. The different steroid hormones most probably enter the cell by passive diffusion. Intratumoral El and E, concentrations in postmenopausal breast cancer patients were found to be 10 times higher than the serum levels (69), suggesting local estrogen production (70) or an active uptake mechanism. The enzymes necessary for E, production from different steroid hormones have been found in breast cancer cells (Fig. 3) (64-67, 69-71). By changing the amount or activity of these enzymes an antiestrogenic effect may be achieved by reducing the amount of available E,. The influence of progestins on different enzymes has mainly been investigated in endometrial tissue (72, 73). The increase in activity of arylsulfotransferase and estradiol dehydrogenase favors the transformation of E2 to El and E I S (Fig. 3). No influence of progestin on the aromatase activity of human breast cancer cells has been observed, but a stimulation of aromatase activity in stroma1 cells of the rabbit endometrium has been reported (74). Whether progestins cause a net antiestrogenic effect
on the intracellular metabolism of androgenic and estrogenic hormones or not remains to be shown. Eflect on receptor levels. E, is the most potent estrogen, having the highest binding affinity for ER. Estrogenic activity results from an interaction of the complex E-ER with DNA. An anti-estrogenic effect can be obtained by an interference with this interaction. No binding of progestins to ER has been reported. The regulation of receptor synthesis in human breast cancer cells in vivo is far from clear. TAM has been found to increase the PgR level in one study (75) but not in another (76), depending on the time of the second biopsy. The influence of progestins on steroid receptor levels has been investigated mainly in endometrial tissues (77), where a reduction of the ER amount was found, thereby inducing a relative insensitivity of the tissue to endogenous estrogens. The possibility that tumor regression in breast cancer by progestin treatment proceeds through the same mechanisms as those postulated for endometrial cancer is at least partly challenged by two papers (78, 79). The question if whether a difference in effect on receptor levels between the two progestins MPA and MA exists, as indicated in one study (78), requires further studies. Both progestins reduced PgR significantly using a DCC method for measuring PgR levels (78). This might be due to progestin binding to PgR, thereby mediating the antiestrogenic effect after interacting with DNA. The method used for PgR determination (78) only detects unoccupied receptors, and does not discriminate between occupation by exogenous progestins and a down-regulation of receptors. By using monoclonal antibodies against PgR this may be clarified. The antiestrogenic effect of progestin may also be exerted by binding to or influencing other steroid receptors (78, 80, 81). Furthermore, effects not related to the receptor-mediated pathway cannot be excluded (82, 83). Effects on growth factors. The estrogenic effect on the cellular level is believed to be regulated by different growth factors acting by autocrine and/or paracrine mechanisms (84). The antiestrogenic effects of progestins could include interference with these factors. Interesting findings have been reported from experiments with cell cultures of breast cancer cells (85, 86). New knowledge is needed in order to evaluate the importance of these changes for the action of progestins. Treatment stragety in relation to pharmacology and patient compliance
Treatment strategy is based on clinical experience and, if available, pharmacokinetic data on the specific drug to use. To judge from early clinical experience, progestins give, in advanced breast cancer, a similar response rate as other endocrine modalities provided that high-dose (1 000 mg daily) and intramuscular administration is used (50). Later randomized studies have shown that similar clinical re-
713
Downloaded by [Laurentian University] at 06:19 10 March 2016
PROGESTINS IN BREAST CANCER
sponse can be obtained with 500 mg daily i.m. (87, 88). Pharmacokinetic studies have shown that MPA can be absorbed from the GI tract, and therefore be used orally. The oral dose that gave similar serum levels as 500 mg i.m. daily was found to be about 1 000 mg daily (89). Two later randomized studies showed a similar antitumor effect by i.m. (500 or 1 000 mg daily) and oral route (900 mg daily) (90, 91). Thus, i.m. administration could safely be replaced by oral administration. The optimal oral dose is yet to be found. Clinical pharmacokinetic data on progestins are conflicting, as certain investigators have reported a direct association between drug bioavailability and objective (92) or subjective (93) response to progestins, whereas others have not found such association (57, 94, 95). No difference in MPA serum levels has been observed between patients with and without side-effects (94, 95). For studies of the pharmacokinetic properties of MPA and MA the methods of serum measurements are crucial. The most commonly used method, radioimmunoassay (RIA) (96- 102), gives serum levels comparable to other methods (103-105) when extraction with a low polar solvent is done before RIA to prevent a great overestimation of progestins (96, 106). MA, in doses of 160- 180 mg daily, shows similar response rates as MPA and due to its long half life (107, 108) this drug can be used once daily (39, 109- 11 1). No serious interactions between progestins and other drugs have been reported. Only a possible decrease in warfarin catabolism (61), if proven significant, may be of
clinical importance when these drugs are to be combined. When a combination of AG and MPA or MA are given, the serum levels of both progestins are suppressed (68, 112), probably due to an increased catabolism of the progestins induced by AG. MPA and MA have been compared in two randomized studies and similar side-effect profile was observed (62, 1 1 1). However, in some clinical studies reviewed below, a small difference in side-effects between these two progestins was seen (Table 1). The most frequent side-effect is weight gain, abserved in 27% and 18% for MPA and MA respectively. A minor, but important, effect to inform patients about to prevent anxiety, is the vaginal discharge that occurs about 1-2 weeks after stopping progestin treatment. Fatal side-effects have seldom been reported but by using a combination of MPA and radiotherapy fatal sideeffects were reported in one study (132). In adjuvantly MPA-treated patients, however, no such side-effects were seen (133, 134). Only 1-4% of the patients treated in these reports (Table 1) had to withdraw the drug due to sideeffects. Although 2-3 times higher serum levels of MA compared to MPA were detected, (56, 63, 95) a slightly higher incidence of side-effects caused by MPA is seen in Table 1, probably reflecting differences in androgenic or glucocorticoid effects, as seen in the influence on serum levels of SHBG (56, 63). Using higher doses of MA (200- 1 600 mg/day) weight gain, increased appetite and edema were observed in 82%, 15% and 26% of the patients respectively (129, 135, 136).
Table 1 Reported side-effects in patients with advanced breast cancer treated with oral medroxyprogesterone acetate ( M P A ) and megestrol acetate ( M A )
Side-effects
Cushing-like symptoms Vaginal spotting/bleeding Increased appetite Dyspnoe CNS problems GI problems Diabetes, hyperglycemia Sweating Tremor Thrombophlebitis Embolism Hirutism, acne, hairloss, hoarseness Hypertension Edema Weight gain Therapy withdrawal due to side-effects
MPA ( > 500- 1 400 mg/day) n = 805
47 38 42 34 42 32 5 40 45 9
I
6 5 5 4 5 4
ISSN: 0284-186X (Print) 1651-226X (Online) Journal homepage: http://www.tandfonline.com/loi/ionc20
Progestins in Breast Cancer Treatment: A Reveiw Steinar Lundgren To cite this article: Steinar Lundgren (1992) Progestins in Breast Cancer Treatment: A Reveiw, Acta Oncologica, 31:7, 709-722, DOI: 10.3109/02841869209083859 To link to this article: http://dx.doi.org/10.3109/02841869209083859
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 50
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ionc20 Download by: [Laurentian University]
Date: 10 March 2016, At: 06:19
A C T A O N C O L O G I C A Vol. 31 No. 7
1992
PROGESTINS IN BREAST CANCER TREATMENT A reveiw
Downloaded by [Laurentian University] at 06:19 10 March 2016
STEINARLUNDGREN
The two most widely used synthetic progestins in breast cancer treatment, medroxyprogesterone acetate (MPA) and megestrol acetate (MA), are reviewed with regard to pharmacological, endocrinological and clinical aspects. In high oral doses as second- or first-line endocrine therapy in advanced breast cancer, they give a similar response rate as tamoxifen (TAM) and aminoglutethimide(AG). The mechanism of action is probably complex. Considerable changes in serum levels of different hormones are induced by progestin treatment. The decrease of serum estrone sulfate (E,S) may be part of the therapeutic mechanism. Some studies suggest that the two drugs, M P A and MA, have a different mode of action, and possibly a low cross resistance. Randomized studies using the two progestins with a cross-over design may answer these questions. Further studies on the influence of progestin on different receptors and growth factors are warranted. To determine the most effective clinical dose of the two progestins, studies with increasing therapeutic doses are needed.
Endocrine therapy, the oldest form of systemic treatment of advanced breast cancer ( I ) , has an important place in breast cancer treatment. For many pre- and postmenopausal women with advanced breast cancer, endocrine treatment is currently the most effective and welltolerated treatment. The median duration of response after first-line treatment is 12-18 months, and for some patients a response duration as long as 3-10 years has been observed (2). Response to first-line treatment may also predict a further response to second- and third-line endocrine therapy. As adjuvant treatment, endocrine therapy has resulted in reduced numbers of recurrences and prolonged survival (3). A possible role of endocrine agents in preventing breast cancer has also been discussed (4-7). Improvements in endocrine therapy have been achieved by better ways of selecting patients likely to respond (i.e. steroid hormone receptor measurements, prior response to endocrine therapy), by new drugs with less side-effects and better understanding of the mechanism of action.
Received 19 August 1991. Accepted 26 August 1992. From the Department of Oncology, Regional and University Hospital of Trondheim, N-7006 Trondheim, Norway.
Concerning hormone receptors only estradiol receptor (ER) and progesterone receptor (PgR) assays have at present clinical importance. Further studies of other steroid receptors (androgen receptor, AR; glucocorticoid receptor, GcR) may add new information with regard to the mechanism of action of different hormonal agents used in breast cancer treatment. Information on receptor modulation and the influence on growth factors and their receptors by different endocrine therapies can possibly be used in new treatment strategies. Clinical and experimental data have emphasized the importance of estrogens as potent growth factors in breast cancers. The main strategy of endocrine treatment is therefore to reduce the amount of estrogen available to tumor cells. This can be achieved by ablation of hormone producing glands, or by adding agents which reduce the estrogen effect. For several reasons most of the ablative forms of treatment (oophorectomy, hypophysectomy and adrenalectomy) have been replaced by additive therapy. Additive treatment with androgens, estrogens or progesterone has, at present, no place in breast cancer treatment due to sideeffects of the two first mentioned and short duration of action of progesterone. Glucocorticoides play an important role as a symptom-relieving treatment of CNS metastases and of the lymphangitic form of lung metastasis, but 709
710
S. LUNDGREN
antifertility properties, but since Cole et al. (15) in 1971 PPOTH~LAMUS ) described the clinical use of TAM in patients with ad-
PERIPHERAL TISSUE
\
I
J
A
vanced breast cancer, the drug has been widely used as first-line endocrine treatment for postmenopausal women. In randomized trials, TAM has not given a higher response rate than other endocrine treatment modalities, such as progestins (16-19) or aminoglutethimide (20, 21) but it has in most cases a more favorable side-effect profile. Its exact mechanism of action is not known, but it may, at least partly, be mediated by competitive binding to the ER, thereby preventing endogenous estrogens from forming a complex (ER-E) that interacts with DNA to give the estrogenic effect. Antiestrogenic binding sites, different from ER, have also been found in human breast cancer biopsies, although no role of this binding for the therapeutic action of TAM has so far been shown (22, 23). The increase in the serum level of sex hormone binding globulin (SHBG)induced by TAM (24, 25) may possibly play a role by decreasing the free fraction of active estrogens available to the target cells. The usual dose of TAM is 20 mg given once daily and higher doses have not been shown to give better results (26, 27). Due to its low side-effect profile, TAM has also been used in adjuvant settings (28). Based on several randomized studies with adjuvant TAM treatment, a recent meta-analysis has shown a small but highly significant mortality reduction by TAM treatment (29). Due to the estrogenic effect of TAM, a possible carcinogenic influence on the endometrium has recently been discussed (30). New anti-estrogens have recently been introduced, possibly with lower or no estrogen agonist properties and more favorable pharmacokinetic properties (31, 32). Aromatase inhibitors. In postmenopausal women, estrogens are synthesized in peripheral tissues. Androgens from the adrenals (2/3) and ovaries (1/3) are transformed to estrogens by aromatization (33, 34) (Fig. 1). Aromatase inhibitors are drugs that can, at least partly, inhibit this synthesis pathway. The main drug in clinical practice so far is aminoglutethimide (AG) (35, 36). AG was introduced as an anticonvulsant but was soon found to suppress steroid synthesis in the adrenals. Due to the feed-back increase of ACTH, a concurrent use of corticosteroid is therefore mandatory to avoid adrenal stimulation (35). Probably more important is the inhibition of the conversion of androstenedione to estrogens in peripheral tissues, the aromatase inhibition (35). Recent investigations have also suggested a stimulating influence of AG on the metabolism of estrogens (37). The usual dose schedule is 250mg q.i.d. with glucocorticoid substitution (35, 36). In randomized studies similar response rates as for TAM (20, 21), MPA (38) and MA (39) have been reported, but with apparently more side-effects, at least during the first weeks,of treatment (39). AG initially causes a substantial drowsiness in approximately 40% of patients and 1/3 of patients have self-limiting morbilliform, macular-papular
Downloaded by [Laurentian University] at 06:19 10 March 2016
UE(2) I B--SHBG tk +
TARGET TISSUE
E-SHBG
Fig. 1. Production of estrogen and androgen in pre-( 1) and post-(2) menopausal women. Abbreviations used: E = estrogens, A = androgens, SHBG = sex hormone binding globulin, LH = lutenizing hormone, FSH = follicle stimulating hormone, ACTH = adrenalcorticotropine hormone.
give a low objective response rate. Only the additive forms of treatment in clinical use are discussed below. Some newer treatment modalities
The site of estrogen production is different in pre- and postmenopausal women (Fig. 1) and treatment strategies therefore depend on the menstrual status of the patient. Chemical castration. In premenopausal women, the estrogens are mainly produced in the ovaries under influence of gonadotropins from the pituitary gland (Fig. 1). The effect of oophorectomy or irradiation of the ovaries is well documented (2, 8). For the non-responding patients, however, i.e. about 2/3, these procedures cause an irreversible menopause without therapeutic gain. A reversible medical castration has recently become achievable by luteinizing hormone-releasing hormone (LH-RH) analogues, which interfere with the pituitary LH-RH receptors. Thus surgical or radiological castration can be replaced. by a reversible treatment procedure (8- 10). The LH-RH analogues are now best administrated by a subcutaneous depot injection given once monthly. Side-effects are seldomly reported, except for symptoms related to the induced menopause (8). LH-RH treatment has shown a low response rate in postmenopausal women with advanced breast cancer ( 11- 14), and is mainly used in premenopausal patients. Anti-estrogen. Antiestrogens may be defined as substances that antagonize estrogen action, and the drugs currently used are tamoxifen (TAM) and substances derived from TAM. They were originally designed for their
71 1
Downloaded by [Laurentian University] at 06:19 10 March 2016
PROGESTINS IN BREAST CANCER
skin rash (35, 36). Due to these side-effects more pure aromatase inhibitors have been searched for. 4-hydroxyandrostendione (4-OHA) is a potent new aromatase inhibitor, reducing serum E, levels and causing tumor regression in postmenopausal patients with advanced breast cancer. The drug has very few side-effects, and studies determining the optimal dose schedule and route of administration are in progress (36, 40,41). The substances CGS 16494A (Ciba-Geigy) has also been shown to be a very potent inhibitor of aromatase activity in vitro, effectively blocking the conversion of androgenic precursors to estrogens in human placental microsomes. It is more potent than AG in reducing the growth of DMBA-induced rat mammary tumors (42). Phase I1 trials with different doses have been started as second-line therapy in patients with metastatic breast cancer. Progestins. Progesterone was shown to have antiestrogen effects soon after its discovery in 1930 (43). When used as treatment of advanced breast cancer in 18 patients, an objective response was seen in 3 patients. It has, however, been rarely used due to the short duration of response (44, 45). To cope with this, several synthetic progestins have been synthesized. The two main drugs in clinical practice are medroxyprogesterone acetate (MPA) and megestrol acetate (MA) (Fig. 2). These drugs have been known since the 1960s (46-48), and the first clinical trial in breast cancer, using a low dose, showed a response rate of 17-18% (49). A breakthrough came when an Italian group published results of using MPA i.m. in high doses (50).
Progestins as natural anti-estrogens Effect on serum level of estrogens. The serum levels of steroid hormones depend on the rate of production and the metabolism of these hormones. They are transported in blood bound to different transport proteins. The most potent estrogen, estradiol (E,) and testosterone (T) are mostly bond to sex hormone binding globulin (SHBG), and most probably only the unbound fraction of the hormone is immediately available for metabolism and for responsive cells (51). In premenopausal women, estrogens are mainly produced in the ovaries, influenced by the gonadotropins (Fig. 1). The use of hypophysectomy or LH-RH analogues reduces the serum level of luteotropic hormone (LH) and thereby the production of estrogens (2, 8, 9). The hypothalamic-pituitary influence of progestins seen in postmenopausal patients (52-56) also occurs in premenopausal women. In postmenopausal women the estrogens are produced in peripheral tissues with androgenic precursors as substrate (Fig. 1). About 1/3 of the precursor androgens are produced in the ovaries (34), the rest in the adrenal glands (33). Serum levels of androgens are reduced to about 60% of basal levels during progestin treatment (52, 55-59), due to progestin action on the hypothalamic-pituitary-adrenal axis (56, 58). Only one
Wr bl YLDROXYPIOOESTEROWE ACETATE
cy
I
Cn, Cl NEOESTROL ACLTATC
Fig. 2. Chemical structure of a) progesterone, b) medroxyprogesterone acetate (MPA) and c) megestrol acetate (MA). The only difference between MPA and MA is the double bound between C6-, in MA.
study reports an increased androgen catabolism (60), but this is probably of less clinical importance due to the minor influence of oral high-dose progestin treatment on drug catabolism by the mixed function oxidase system (61). The two progestins, MPA and MA, were found to induce closely similar reduction of the androgen in one study (56), but a significantly lower serum level of androstenedione was found during MPA treatment compared to MA treatment in another study (62). Estradiol ( E,) and estrone (E, ) both decreased by about 30% during progestin treatment (52, 55-57, 59). No significant difference between the two progestins was observed with regard to influence on serum levels of E, (62,63), and a significant difference in E, was found in one study (63), but not in another (62). The progestins have been reported to induce a substantial decrease in the serum level of SHBG (56, 57, 59, 63), which must be taken into consideration when the serum level of E2 is discussed, as only free E, is thought to be available for cellular uptake. This might be explained by differences between the two progestins with regard to their androgenic and/or glucocorticoid effect on hepatic protein synthesis. The different effect on SHBG cannot be
712
[
S. LUNDGREN
SHY-,
E,-SHBG eLooD
I
L SURROUNDING TISSUES
El
L
TMORCELL
J
(E,) production. Abbreviations used: A = androstenedione, T = testosterone, El = estrone, E, = estradiol, E,S = estrone sulfate, SHBG =sex hormone binding globulin. a = aromatase, dh = dehydrogenase, s = sulfatase. st = sulfotransferase. Downloaded by [Laurentian University] at 06:19 10 March 2016
Fig. 3. Pathways of intracellular estradiol
explained by the differences in serum concentrations of progestins (56,63). Breast cancer cells have been shown to transform estrone sulfate (El S) to more active estrogens (Fig. 3) (64-67). The serum level of E, S is about 20 times higher than that of E2 in postmenopausal women (37,63), and a 30% reduction of E, S (63) might be of importance for the action of progestins. The difference in reduction of El S in the circulation promoted by MPA and MA (63) has been proved to be significant (68). This could be explained, at least partly, by differences in serum levels of the two progestins used (63, 68). Whether the reduction of EIS found during progestin treatment is due to a decrease in production or an increase in metabolism requires further studies. The antiestrogen effect of progestins may be caused by a decrease of the amount of estrogen reaching the breast cancer cell. Whether the progestin-induced changes in serum levels of relevant steroid hormones cause similar alterations inside tumor cells is not known. Effect on cellular uptake and intracellular metabolism of estradiol. The different steroid hormones most probably enter the cell by passive diffusion. Intratumoral El and E, concentrations in postmenopausal breast cancer patients were found to be 10 times higher than the serum levels (69), suggesting local estrogen production (70) or an active uptake mechanism. The enzymes necessary for E, production from different steroid hormones have been found in breast cancer cells (Fig. 3) (64-67, 69-71). By changing the amount or activity of these enzymes an antiestrogenic effect may be achieved by reducing the amount of available E,. The influence of progestins on different enzymes has mainly been investigated in endometrial tissue (72, 73). The increase in activity of arylsulfotransferase and estradiol dehydrogenase favors the transformation of E2 to El and E I S (Fig. 3). No influence of progestin on the aromatase activity of human breast cancer cells has been observed, but a stimulation of aromatase activity in stroma1 cells of the rabbit endometrium has been reported (74). Whether progestins cause a net antiestrogenic effect
on the intracellular metabolism of androgenic and estrogenic hormones or not remains to be shown. Eflect on receptor levels. E, is the most potent estrogen, having the highest binding affinity for ER. Estrogenic activity results from an interaction of the complex E-ER with DNA. An anti-estrogenic effect can be obtained by an interference with this interaction. No binding of progestins to ER has been reported. The regulation of receptor synthesis in human breast cancer cells in vivo is far from clear. TAM has been found to increase the PgR level in one study (75) but not in another (76), depending on the time of the second biopsy. The influence of progestins on steroid receptor levels has been investigated mainly in endometrial tissues (77), where a reduction of the ER amount was found, thereby inducing a relative insensitivity of the tissue to endogenous estrogens. The possibility that tumor regression in breast cancer by progestin treatment proceeds through the same mechanisms as those postulated for endometrial cancer is at least partly challenged by two papers (78, 79). The question if whether a difference in effect on receptor levels between the two progestins MPA and MA exists, as indicated in one study (78), requires further studies. Both progestins reduced PgR significantly using a DCC method for measuring PgR levels (78). This might be due to progestin binding to PgR, thereby mediating the antiestrogenic effect after interacting with DNA. The method used for PgR determination (78) only detects unoccupied receptors, and does not discriminate between occupation by exogenous progestins and a down-regulation of receptors. By using monoclonal antibodies against PgR this may be clarified. The antiestrogenic effect of progestin may also be exerted by binding to or influencing other steroid receptors (78, 80, 81). Furthermore, effects not related to the receptor-mediated pathway cannot be excluded (82, 83). Effects on growth factors. The estrogenic effect on the cellular level is believed to be regulated by different growth factors acting by autocrine and/or paracrine mechanisms (84). The antiestrogenic effects of progestins could include interference with these factors. Interesting findings have been reported from experiments with cell cultures of breast cancer cells (85, 86). New knowledge is needed in order to evaluate the importance of these changes for the action of progestins. Treatment stragety in relation to pharmacology and patient compliance
Treatment strategy is based on clinical experience and, if available, pharmacokinetic data on the specific drug to use. To judge from early clinical experience, progestins give, in advanced breast cancer, a similar response rate as other endocrine modalities provided that high-dose (1 000 mg daily) and intramuscular administration is used (50). Later randomized studies have shown that similar clinical re-
713
Downloaded by [Laurentian University] at 06:19 10 March 2016
PROGESTINS IN BREAST CANCER
sponse can be obtained with 500 mg daily i.m. (87, 88). Pharmacokinetic studies have shown that MPA can be absorbed from the GI tract, and therefore be used orally. The oral dose that gave similar serum levels as 500 mg i.m. daily was found to be about 1 000 mg daily (89). Two later randomized studies showed a similar antitumor effect by i.m. (500 or 1 000 mg daily) and oral route (900 mg daily) (90, 91). Thus, i.m. administration could safely be replaced by oral administration. The optimal oral dose is yet to be found. Clinical pharmacokinetic data on progestins are conflicting, as certain investigators have reported a direct association between drug bioavailability and objective (92) or subjective (93) response to progestins, whereas others have not found such association (57, 94, 95). No difference in MPA serum levels has been observed between patients with and without side-effects (94, 95). For studies of the pharmacokinetic properties of MPA and MA the methods of serum measurements are crucial. The most commonly used method, radioimmunoassay (RIA) (96- 102), gives serum levels comparable to other methods (103-105) when extraction with a low polar solvent is done before RIA to prevent a great overestimation of progestins (96, 106). MA, in doses of 160- 180 mg daily, shows similar response rates as MPA and due to its long half life (107, 108) this drug can be used once daily (39, 109- 11 1). No serious interactions between progestins and other drugs have been reported. Only a possible decrease in warfarin catabolism (61), if proven significant, may be of
clinical importance when these drugs are to be combined. When a combination of AG and MPA or MA are given, the serum levels of both progestins are suppressed (68, 112), probably due to an increased catabolism of the progestins induced by AG. MPA and MA have been compared in two randomized studies and similar side-effect profile was observed (62, 1 1 1). However, in some clinical studies reviewed below, a small difference in side-effects between these two progestins was seen (Table 1). The most frequent side-effect is weight gain, abserved in 27% and 18% for MPA and MA respectively. A minor, but important, effect to inform patients about to prevent anxiety, is the vaginal discharge that occurs about 1-2 weeks after stopping progestin treatment. Fatal side-effects have seldom been reported but by using a combination of MPA and radiotherapy fatal sideeffects were reported in one study (132). In adjuvantly MPA-treated patients, however, no such side-effects were seen (133, 134). Only 1-4% of the patients treated in these reports (Table 1) had to withdraw the drug due to sideeffects. Although 2-3 times higher serum levels of MA compared to MPA were detected, (56, 63, 95) a slightly higher incidence of side-effects caused by MPA is seen in Table 1, probably reflecting differences in androgenic or glucocorticoid effects, as seen in the influence on serum levels of SHBG (56, 63). Using higher doses of MA (200- 1 600 mg/day) weight gain, increased appetite and edema were observed in 82%, 15% and 26% of the patients respectively (129, 135, 136).
Table 1 Reported side-effects in patients with advanced breast cancer treated with oral medroxyprogesterone acetate ( M P A ) and megestrol acetate ( M A )
Side-effects
Cushing-like symptoms Vaginal spotting/bleeding Increased appetite Dyspnoe CNS problems GI problems Diabetes, hyperglycemia Sweating Tremor Thrombophlebitis Embolism Hirutism, acne, hairloss, hoarseness Hypertension Edema Weight gain Therapy withdrawal due to side-effects
MPA ( > 500- 1 400 mg/day) n = 805
47 38 42 34 42 32 5 40 45 9
I
6 5 5 4 5 4
