Mechanisms of resistance to aromatase inhibitors

Aromatase inhibitors are rapidly becoming the first choice for hormonal treatment of steroid receptor positive breast cancer in postmenopausal women. An understanding of the resistance mechanisms to these agents is, therefore, important for the appropriate delivery of treatment to responsive patients and the rational development of new agents targeted at the resistance pathways. De novo resistance appears to be a quantitative rather than qualitative phenomenon with virtually all oestrogen receptor positive tumours showing an anti-proliferative response to the aromatase inhibitor anastrozole. While the expression of type 1 growth factor receptors reduces response to tamoxifen this appears to have little detrimental effect on response to aromatase inhibitors. Studies of acquired resistance in vitro have indicated that acquisition of hypersensitivity to oestrogenic stimulation is a key mechanism that is dependent on enhanced cross-talk of growth factor and oestrogen signaling pathways. Collection of resistant biopsy tissues from patients is important to determine if this mechanism is clinically relevant.     

Aromatase inhibitors: assessment of biochemical efficacy measured by total body aromatase inhibition and tissue estrogen suppression

The implementation of aromatase inhibitors for treatment of early and metastatic breast cancer has been one of the major improvements in endocrine therapy of breast cancer. Measurement of endocrine effects of aromatase inhibition in vivo has been a major tool in the process of evaluating novel compounds. Biochemical efficacy of aromatase inhibitors in vivo may be determined from their effects on “total body aromatization” as well changes in plasma and tissue estrogen levels. Due to high sensitivity, tracer methods allowing calculation of whole body aromatase inhibition are still considered the gold standard. The method developed by our group in collaboration with the Royal Marsden Hospital and the results of this joint program are summarized and discussed. These studies allowed classification of the different aromatase inhibitors and their optimal dosage, selecting the best compounds for clinical evaluation. In vivo total body aromatase assessment is a work-consuming method, allowing such studies to be conducted in a limited number of patients only. In contrast, plasma estrogen measurement is a cruder but simpler method, allowing screening of larger groups of patients. As plasma estrogens arise through passive diffusion of estrogens synthesized in different body compartments, plasma estrogens, as well as total body aromatase assessment, present a rough estimate of total body tissue estrogen production, and changes associated with treatment with aromatase inhibitors reflect the effects on tissue estrogen production in general. However, plasma estrogen levels do not correlate to breast cancer tissue estrogen levels. This is due to the endocrine autonomy of breast cancer tissue with significant local estrogen production in some tumors. Thus, direct measurement of intratumor estrogens is demanded to evaluate the effects of aromatase inhibitors in malignant target tissues. Our group has developed a highly sensitive HPLC-RIA for the simultaneous measurement of estrone, estradiol, and estrone sulfate in malignant breast tissue samples, and we are currently using this method to assess alterations in intratumor estrogen levels during treatment with different aromatase inhibitors.     

High-performance liquid chromatography of the aromatase inhibitor, letrozole, and its metabolite in biological fluids with automated liquid-solid extraction and fluorescence detection

An analytical method for the determination of letrozole (CGS 20 267) in plasma and of letrozole and its metabolite, CGP 44 645, in urine is described. Automated liquid-solid extraction of compounds from plasma and urine was performed on disposable 100-mg C₈ columns using the ASPEC system. The separation was achieved on an ODS Hypersil C₁₈ column using acetonitrile-phosphate buffer, pH 7, as the mobile phase at a flow-rate of 1.5 ml/min. A fluorescence detector was used for the quantitation. The excitation and emission wavelengths were 230 and 295 nm, respectively. The limits of quantitation (LOQ) of letrozole in plasma and in urine were 1.40 nmol/l (0.4 ng/ml) and 2.80 nmol/l, respectively. The respective mean recoveries and coefficient of variation (C.V.) were 96.5% (9.8%) in plasma and 104% (7.7%) in urine. The LOQ of CGP 44 645 in urine was 8.54 nmol/l (2 ng/ml). The mean recovery was 108% (6.3%). The compounds were well separated from co-extracted endogenous components and no interferences were observed at the retention times of compounds. The sensitivity of this method for letrozole in plasma should be sufficient for kinetic studies in humans with single doses of 0.5 mg and possibly less.     

Biochemical and pharmacological development of steroidal inhibitors of aromatase

Androstenedione analogs containing 7 alpha-substituents have proven to be potent inhibitors of aromatase both in vitro and in vivo. Several of these agents have exhibited higher affinity for the enzyme complex than the substrate. In order to examine further the interaction(s) of 7-substituted steroids with aromatase, biochemical and pharmacological studies were performed on 7 alpha-thiosubstituted androstenediones and 7-substituted 4,6-androstadiene-3,17-diones. Potent inhibition of aromatase activity in human placental microsomes has been observed with several new 7 alpha-thiosubstituted androstenediones. 7-Benzyl- and 7-phenethyl-4,6-androstadiene-3,17-diones effectively inhibited microsomal aromatase, with apparent Kis ranging from 61 to 174 nM. On the other hand, 7-phenyl-4,6-androstadiene-3,17-dione exhibited poor activity, with an apparent Ki of 1.42 microM. Similar inhibitory activity was observed with reconstituted, purified cytochrome P450Arom preparations. Additionally, these agents were evaluated for inhibition of aromatase activity in two human carcinoma cell lines, the MCF-7 human mammary cancer line and the JAr choriocarcinoma line. The 7 alpha-thiosubstituted androstenediones and 7-substituted 4,6-androstadiene-3,17-diones produced dose-dependent inhibitions of aromatase activity in the cell cultures. The most effective inhibitors were the 7 alpha-substituted androstenediones, with EC50 values ranging from 7.3 to 105 nM. Finally, the JAr cell culture system exhibited prolonged inhibition of aromatase activity following exposure to 7 alpha-APTADD, suggesting enzyme inactivation by this inhibitor. Thus, these agents are effective aromatase inhibitors, and the results encourage further development of this group of medicinal agents for the treatment of estrogen-dependent mammary carcinoma.     

Overview of adjuvant trials of aromatase inhibitors in early breast cancer

The third-generation aromatase inhibitors are an important class of drugs for use in adjuvant therapy for postmenopausal women with resected estrogen receptor positive breast cancer. Multiple large prospective randomized trials have established their value in this setting and provided guidance for their use in clinical management. This review will outline the trials that have provided evidence on the value of the aromatase inhibitors in the adjuvant setting as well as the ongoing trials that will expand our knowledge of how to use them most effectively.     

Breast cancer tissue estrogens and their manipulation with aromatase inhibitors and inactivators

Despite the dramatic fall in plasma estrogen levels at menopause, only minor differences in breast tissue estrogen levels have been reported comparing pre- and postmenopausal women. Thus, postmenopausal breast tissue has the ability to maintain concentrations of estrone (E₁) and estradiol (E₂) that are 2–10- and 10–20-fold higher than the corresponding plasma estrogen levels. This finding may be explained by uptake of estrogens from the circulation and/or local estrogen production. Local aromatase activity in breast tissue seems to be of crucial importance for the local estrogen production in some patients while uptake from the circulation may be more important in other patients. Beside aromatase, breast tissue expresses estrogen sulfotransferase and sulfatase as well as dehydrogenase activity, allowing estrogen storage and release in the cells as well as conversions between estrone and estradiol. The activity of the enzyme network in breast cancer tissue is modified by a variety of factors like growth factors and cytokines. Aromatase inhibitors have been used for more than two decades in the treatment of postmenopausal metastatic breast cancer and are currently investigated in the adjuvant treatment and even prevention of breast cancer. Novel aromatase inhibitors and inactivators have been shown to suppress plasma estrogen levels effectively in postmenopausal breast cancer patients. However, knowledge about the influence of these drugs on estrogen levels in breast cancer tissue is limited. Using a novel HPLC-RIA method developed for the determination of breast tissue estrogen concentrations, we measured tissue E₁, E₂ and estrone sulfate (E₁S) levels in postmenopausal breast cancer patients before and during treatment with anastrozole. Our findings revealed high breast tumor tissue estrogen concentrations that were effectively decreased by anastrozole. While E₁S was the dominating estrogen fraction in the plasma, estradiol was the estrogen fraction with the highest concentration in tumor tissue. Moreover, plasma estrogen levels did not correlate with tissue estrogen concentrations. The overall experience with aromatase inhibitors and inactivators concerning their influences on breast tissue estrogen concentrations is summarized.     

Endocrine effects of aromatase inhibitors and inactivators in vivo: review of data and method limitations

The so-called “third-generation” aromatase inhibitors/inactivators have become standard first-line endocrine therapy for postmenopausal women in the metastatic setting. In addition, these compounds, administered as monotherapy or in sequence with tamoxifen, are likely to become standard adjuvant therapy in most countries in the near future. In contrast to the SERMs, aromatase inhibitors may be assessed for their biochemical efficacy in vivo either by measuring their ability to suppress plasma and tissue estrogen levels or, alternatively, by measuring their ability to inhibit the conversion of tracer-labelled androstenedione into estrone. While contemporary methods for estrogen measurement (with the exception of estrone sulphate) lack the sensitivity to measure plasma estrogen levels during treatment with the most potent compounds, in vivo aromatase inhibition can be determined with a much better sensitivity. Thus, in a joint program conducted by the Royal Marsden Hospital, London and our team in Bergen, we were able to reveal profound differences between first- and second-generation aromatase inhibitors, causing 50–90% aromatase inhibition, and the three third-generation compounds, causing >98% inhibition of total body aromatization.     

Binding characteristics of aromatase inhibitors and phytoestrogens to human aromatase

We have evaluated the binding characteristics of three steroidal inhibitors [4-hydroxyandrostene-dione (4-OHA), 7-(4′-amino)phenylthio-1,4-androstadiene-3,17-dione (7-APTADD), and bridge (2,19-methyleneoxy) androstene-3,17-dione (MDL 101,003)], four nonsteroidal inhibitors [aminoglutethimide (AG), CGS 20267, ICI D1033, and vorozole (R83842)], and two flavone phytoestrogens (chrysin, and 7,8-dihydroxyflavone) to aromatase through a combination of computer modeling and inhibitory profile studies on the wild-type and six aromatase mutants (I133Y, P308F, D309A, T310S, I395F, and I474Y). We have generated two aromatase models based on the x-ray structures of cytochrome P450-cam and cytochrome P450bm3, respectively. A major difference between the cytochrome P450cam-based and cytochrome P450bm3-based models is in the predicted lengths of helices F and G. In the cytochrome P450cam-based model, helices F and G lie antiparallel and extend across the active-site face of the molecule from one edge to the center, so that the carboxyl-terminal residues of helix F and the N-terminal residues of helix G make a major contribution to the structure of the active site. In the cytochrome P450bm3-based model, both helices are longer and so extend almost all the way across the active-site face of the molecule. Considering the size of the androgen substrate, we evaluated our results mainly based on the cytochrome P450cam model. The mutations involved in this study are thought to be at or near the proposed active site pocket. The inhibitory profile analysis has produced very interesting results and provided a molecular basis as to how seven aromatase inhibitors with different structures bind to the active site of aromatase. Furthermore, the investigation reveals that phytoestrogens bind to the active site of aromatase in a different orientation from that in the estrogen receptor.     

Predictors of response to aromatase inhibitors

Aromatase inhibitors are now considered to be part of the endocrine treatment for most hormone receptor-positive breast cancer in post-menopausal women for both early and advanced disease. Despite the impressive efficacy of these agents, up to 50% of treated patients exhibit de novo or intrinsic resistance to aromatase inhibitors and hence identification of response predictors is essential to allow treatment to be directed towards responsive populations and for alternative or additional therapies to be offered to resistant patients. Emerging data seem to suggest a role for the conventional tumour markers of oestrogen receptor and progesterone receptor as possible predictors of response but, particularly in the adjuvant setting, the extent to which these are useful has not been fully elucidated. Data from both the neo-adjuvant and advanced disease settings suggest that response to aromatase inhibitors does not appear to be adversely affected by HER-2 overexpression. Within neo-adjuvant aromatase inhibitor studies, the proliferation marker Ki67 has shown a significant correlation with relapse-free survival, suggesting a role in prediction for measurement of Ki67 and other dynamic markers of response. Analysis of multiple gene expression changes over a short treatment period may also have potential clinical utility for prediction of response.     

Evaluation of plasma and tissue estrogen suppression with third-generation aromatase inhibitors: Of relevance to clinical understanding?

Development of aromatase inhibition and aromatase inhibitors as a therapeutic strategy was initiated through two different pathways. The one pathway went through systematic exploration of aromatase substrate analogues for enzyme inhibitions, subsequently leading to the development of steroidal agents for clinical use. The second involved clinical observation with an unsuccessful anti-epileptic compound named aminoglutethimide, attempting to achieve a “medical adrenalectomy”. Endocrine studies on patients treated with aminoglutethimide lead to direct assessment of in vivo aromatase inhibition in patients on treatment, thus identifying a novel therapeutic strategy. As such, both research programs represent different examples of pioneering translational work leading towards a successful therapeutic strategy. Subsequent studies with respect to total aromatase inhibition have led to successful development of more potent strategies. Most importantly, these studies have revealed a correlation between aromatase inhibition and clinical outcome. Ongoing studies exploring tissue estrogen levels as well as gene expression profiles on therapy may further improve this important therapeutic area.     

DCIS and aromatase inhibitors

In patients with hormone receptor positive DCIS tamoxifen reduces recurrence rates by almost 50%. Few data are available with aromatase inhibitors from randomised studies. In the ATAC study there were three DCIS lesions in the anastrozole arm and four in the tamoxifen arm in the women with ER positive invasive cancer. In the MA17 study which randomised patients to up to 5 years of letrozole or placebo there was only one DCIS event in the contralateral breast in patients taking letrozole and five on placebo. There were also four patients in this study who had DCIS in the conserved breast on placebo and none in the letrozole treated group. The few clinical data that are available therefore suggest the aromatase inhibitors are likely to be effective in DCIS. A histological review of a study of 206 postmenopausal women with invasive oestrogen receptor positive breast cancer who were randomised as part of a 14 day preoperative study to receive 2.5 mg of letrozole or 1 mg of anastrozole identified 27 patients with 28 pairs of tumours in whom there was sufficient ER positive DCIS in invasive cancer in the initial core biopsy and in the subsequent surgery specimen, to evaluate for PgR activity and proliferation. Within the DCIS both aromatase inhibitors significantly reduced PgR expression and both drugs also produced a significant fall in proliferation. There was a moderate degree of agreement between the fall in PgR in both the invasive cancer and DCIS (Kappa = 0.5; p = 0.0013) and between the fall in proliferation and between the invasive and in situ components (correlation coefficient = 0.68; p < 0.001). This study has shown significant effects of aromatase inhibitors on DCIS indicating that these agents are therapeutically active in this condition.     

Contribution of aromatase to the deoxyribonucleic acid synthesis of MCF-7 human breast cancer cells and its suppression by aromatase inhibitors.

We have studied the effects of various steroids on DNA synthesis in MCF-7 human breast carcinoma cells, which have aromatase activity and which exert an oestrogen receptor-mediated growth, to assess the significance of intracellular aromatase on growth stimulation as well as inhibition by aromatase inhibitors. The cells were cultured for 96 h in phenol red-free medium containing 10% charcoal-treated fetal bovine serum and test reagents and pulse-labelled with [3H]thymidine. Physiological concentrations of oestradiol, oestrone, testosterone (T) and androstenedione (AD) stimulated thymidine incorporation. However, oestrone-sulphate and dihydrotestosterone (DHT) only stimulated at concentrations greater than the physiological levels. T and DHT stimulation was blocked by tamoxifen, but not by cyproterone acetate, suggesting that the stimulation was mediated via the oestrogen receptor but not by the androgen receptor. Stimulation by T and AD was reduced by aminoglutethimide and 14 alpha-hydroxy-4-androstene-3,6,17-trione, both of which inhibit aromatase activity, however, stimulation by nonaromatizable DHT was not reduced by the inhibitors, suggesting that androgens were converted by the intracellular aromatase to oestrogens which stimulated the thymidine incorporation. It is suggested that intracellular aromatase significantly contributes to the stimulation of DNA synthesis and that aromatase inhibitors suppress the stimulation.     

Structure-function studies of aromatase and its inhibitors: a progress report

The utilization of computer modeling, site-directed mutagenesis, inhibition kinetic analysis and reaction metabolite analysis allows us to better understand the structure–function relationship between aromatase and its inhibitors. Our results have helped in determining how steroidal and nonsteriodal aromatase inhibitors bind to the active site of the enzyme. This information has also aided in the understanding of the reaction mechanism of aromatase. Furthermore, our structure–function studies of aromatase have generated important information for predicting how environmental chemicals interact with the enzyme. During the last 2 years, a new aromatase computer model based on the X-ray structure of rabbit cytochrome P450 2C5 has been generated and used to evaluate the results obtained from new aromatase mutants produced in this laboratory. In addition, we have succeeded in the expression and purification of functionally active aromatase using an Escherichia coli expression method. The catalytic properties of this recombinant aromatase are similar to those properties exhibited by the human placental aromatase preparation and the mammalian cell-expressed enzyme. The E. coli expressed aromatase will be very useful for further structure–function studies of aromatase. Our laboratory has also evaluated the growth-inhibiting activity of aromatase inhibitors in estrogen receptor-positive breast cancer using three-dimensional cell cultures of aromatase-over expressing MCF-7 and T-47D cell lines (i.e. MCF-7aro and T-47Daro). Our results demonstrate that these three-dimensional cultures are valuable approaches to assess the growth-inhibiting activity of aromatase inhibitors. Finally, we have identified several phytochemicals to be potent inhibitors of aromatase. To demonstrate the impact of the phytochemicals on estrogen formation in vivo, we showed that the intake of anti-aromatase chemicals from red wine was capable of suppressing MCF-7aro-mediated tumor formation in nude mice and aromatase-induced hyperplasia in a transgenic mouse model in which aromatase is over-expressed in the mammary tissue.     

Binding features of steroidal and nonsteroidal inhibitors

Aromatase is the rate-limiting enzyme in estrogen biosynthesis. As a cytochrome P450, it utilizes electrons from NADPH-cytochrome P450 reductase (CPR) to produce estrogen from androgen. Estrogen is a key factor in the promotion of hormone-dependent breast cancer growth. Aromatase inhibitors (AIs) are drugs that block estrogen synthesis, and are widely used to treat estrogen-dependent breast cancer. Structure-function experiments have been performed to study how CPR and AIs interact with aromatase to further the understanding of how these drugs elicit their effects. Our studies have revealed a strong interaction between aromatase and CPR, and that the residue K108 is situated in a region important to the interaction of aromatase with CPR. The published X-ray structure of aromatase indicates that the F221, W224 and M374 residues are located in the active site. Our site-directed mutagenesis experiments confirm their importance in the binding of the androgen substrate as well as AIs, but these residues interact differently with steroidal inhibitors (exemestane) and non-steroidal inhibitors (letrozole and anastrozole). Furthermore, our results predict that the residue W224 also participates in the mechanism-based inhibition of exemestane, as time-dependent inhibition is eliminated with mutation on this residue. Together with previous research from our laboratory, this study confirms that W224, E302, D309 and S478 are important active site residues involved in the suicide mechanism of exemestane against aromatase.     

The intratumoral aromatase model: studies with aromatase inhibitors and antiestrogens

Aromatase inhibitors have now been approved as first-line treatment options for hormone-dependent advanced breast cancer. When compared to tamoxifen, these aromatase inhibitors provide significant survival and tolerability advantages. However, the optimal use of an aromatase inhibitor and tamoxifen remains to be established. To date, the intratumoral aromatase xenograft model has proved accurate in predicting the outcome of clinical trials. Utilizing this model, we performed long-term studies with tamoxifen and letrozole to determine time to disease progression with each of the treatment regimens. Aromatase-transfected MCF-7Ca human breast cancer cells were grown as tumor xenografts in female ovariectomized athymic nude mice in which androstenedione was converted to estrogen and stimulated tumor growth. When tumor volumes were approximately 300 mm³, the animals were grouped for continued supplementation with androstenedione only (control) or for treatment with letrozole 10 μg per day (long-term), tamoxifen 100 μg per day (long-term), letrozole alternating to tamoxifen (4-week rotation), tamoxifen alternating to letrozole (4-week rotation), or a combination of the two drugs. Tumors of control mice had doubled in volume in 3–4 weeks. In mice treated with tamoxifen and the combination, tumor doubling time was significantly shorter (16 and 18 weeks, respectively) than with letrozole (34 weeks). Furthermore, alternating letrozole and tamoxifen treatment every 4 weeks was less effective than letrozole alone. Tumors doubled in 17–18 weeks when the starting treatment was tamoxifen and in 22 weeks when the starting treatment was letrozole. Tumors progressing on tamoxifen remained sensitive to second-line therapy with letrozole (10 μg per day). However, when mice with letrozole-resistant tumors were switched to antiestrogen treatment, tumors did not respond to tamoxifen (100 μg per day) or faslodex (1 mg per day). This suggests that advanced breast cancers treated with letrozole may be insensitive to subsequent second-line hormonal agents. Thus, although letrozole was determined to be an effective second-line treatment option for tumors progressing on tamoxifen, antiestrogen therapy does not appear to be effective for tumors progressing on letrozole. However, response to second-line treatment was observed in a model where tumors that had progressed on letrozole were transplanted to new mice. These tumors had been allowed to grow in the presence of supplemented androstenedione but absence of letrozole. This suggests that resistance to letrozole may be reversible, allowing tumors to respond to subsequent antiestrogens and letrozole.     

Evidence for the pharmacological relevance of benzodiazepine receptors to anticonvulsant activity

Each of a series of benzodiazepines was found to be effective in preventing convulsions evoked by intermittent photic stimulation of epileptic chickens. There was a high correlation between the anticonvulsant potencies (mean effective dosages) and the affinity of the agents for the putative benzodiazepine receptor as measured by displacement of [3H]diazepam from binding sites on chicken synaptosomal membranes. This correlation in a genetic model of epilepsy provides further evidence that benzodiazepines exert their anticonvulsant effects by interacting with the benzodiazepine receptor.     

Synthesis and biochemistry of fluorescent aromatase inhibitors

Effective aromatase inhibitors have been developed that contain aryl functionalities at the 7 alpha-position of the steroid nucleus. The exact interactions of 7 alpha-substituted androstenediones with the active site of aromatase is unknown. Fluorescent derivatives may provide a useful spectroscopic method for examining the binding of these inhibitors to the microsomal complex and purified aromatase protein. Dinitrophenyl, dansyl, and naphthyl derivatives of 7 alpha-(4'-amino)phenylthio-4-androstene-3,17-dione and androstenedione were synthesized as potential fluorescent agents. An in vitro assay with human placental microsomes was used to evaluate aromatase inhibitory properties. These fluorescent compounds were effective competitive inhibitors and have apparent Ki values ranging from 24.1 to 86.7 nM.     

Preparation and pharmacological profile of 7-(alpha-azolylbenzyl)-1H-indoles and indolines as new aromatase inhibitors

Aromatase (P450 arom) is a target of pharmacological interest for the treatment of breast cancer. New series of 7-(alpha-azolylbenzyl)-1H-indoles and indolines were synthesized as non-steroidal inhibitors of P450 arom. Selectivity was studied towards P450 17alpha enzyme. The most active compound, 1-ethyl-7-[(imidazol-1-yl)(4-chlorophenyl)methyl]-1H-indole 12c exhibited promising relative potency (rp) of 336 (rp of aminoglutethimide=1) and most of the described azoles were active and selective towards P450 arom.