The hormonal dependency of breast cancer was first recognized more than a century ago. Although it has yet to be proven that estrogen is directly responsible for the initiation of breast tumors, it is clear from epidemiologic evidence, from "prevention" studies using the antiestrogen tamoxifen(Drug information on tamoxifen) (Nolvadex), and from the clinical impact of hormonal manipulation[4,5] that estrogen is a significant factor in the maintenance and progression of established tumors.
Estrogen is produced by aromatization of androgens. In premenopausal women, androgens are synthesized from cholesterol by the adrenals and the ovaries in roughly equivalent proportions. Approximately 60% of estrogens(Drug information on estrogens) in premenopausal women are synthesized in the aromatase-rich cytoplasm of the granulosa cells of the ovaries. Aromatization by the cycling ovary is regulated by follicle-stimulating hormone that is regulated, in turn, by estrogen in a negative feedback loop. The remaining 40% of estrogens in premenopausal women are synthesized in the peripheral tissues, particularly in fat.
At menopause, ovarian production of both estrogen and the estrogen precursor androstenedione ceases, so that most of the circulating estrogen in postmenopausal women derives from the peripheral conversion of adrenal androgens. Circulating estrogen levels in postmenopausal women are approximately 20% of those of premenopausal women, and they achieve a steady-state concentration in the absence of cyclical ovarian function.
Although circulating levels of estrogens are relatively low in postmenopausal women, aromatase expression is maintained in breast tissue after menopause. Estrogen levels in the breast tissue of postmenopausal women are thus significantly higher than those detected in plasma, and may be as high as the plasma levels in premenopausal women.[7,8]
Although the exact site of aromatase production in breast cancer tissues has not yet been determined, both immunocytochemistry and in situ hybridization techniques have demonstrated aromatase enzyme and mRNA expression in the epithelial cells of the terminal ductal lobular units and the surrounding stromal cells of the normal human breast. Tumor cells may produce aromatase themselves or they may produce cytokines that induce tumor-stromal-cell expression of aromatase. Importantly, breast cancer tissues that retain aromatase expression may be able to function in an autocrine fashion by producing their own growth factor.[11-13]
The functional significance of tumor aromatase has not been well defined but is suggested by several lines of evidence. Aromatase activity is frequently found to be much higher in tumor tissue than in surrounding benign tissue from the same breast, supporting a role for aromatase activity in the emergence of the malignant phenotype.[14,15] Studies of tumor aromatase levels and known prognostic factors, such as tumor cell proliferative activity or lymph node involvement, have yielded conflicting results. No clear correlation between the level of tumor aroma-tase activity and the biological behavior of the tumor has yet been demonstrated.[14,16,17]
Studies examining the relationship between aromatase expression and estrogen- and progesterone(Drug information on progesterone)-receptor positivity have also been inconsistent.[16,18,19] Notably, two small studies have suggested a correlation between tumor aromatase activity and response to aromatase inhibition therapy with aminoglutethimide (Cytadren).[20,21]
There are two general categories of aromatase inhibitors: (1) the nonsteroidal inhibitors, which bind competitively with aromatase, and (2) the steroidal inhibitors, which bind irreversibly (see Table 1).
First- and Second-Generation Aromatase Inhibitors
The first aromatase inhibitor with documented antitumor efficacy was the nonsteroidal agent aminoglutethimide. Although its use as second- or third-line endocrine therapy achieved response rates of 20% to 40%, the drug was associated with problematic effects. Aminoglutethimide inhibits the production of other adrenal steroids, including cortisol, and therefore must be taken with hydrocortisone(Drug information on hydrocortisone). A high incidence of skin rash and fatigue also made the drug difficult for many patients to tolerate. Other early aromatase inhibitors, such as fadrozole (CGS 16949A) and the parenterally administered formestane(Drug information on formestane) (4-OHA), demonstrated antitumor activity and fewer adverse effects than aminoglutethimide, but they have now been supplanted by the third-generation inhibitors described below.
Third-Generation Aromatase Inhibitors
The current generation of nonsteroidal inhibitors includes anastrozole(Drug information on anastrozole) (Arimidex), letrozole (Femara), and vorozole (Rivizor), all of which are administered orally as a once-daily dose. The development of vorozole has been terminated, so it will not be discussed below. The only registered steroidal inhibitor of the current generation is exemestane(Drug information on exemestane) (Aromasin).
Relative Potency of Aromatase Inhibitors
The in vivo potency of aromatase inhibitors is defined by their ability to suppress both aromatase activity and plasma estrogen levels. In vivo aromatase activity is assessed by radioimmunoassay of urinary estrogens following administration of radiolabeled androstenedione. Plasma endogenous estrogens are usually measured with highly sensitive radioimmunoassays after separation with high-performance liquid chromatography.
While the early aromatase inhibitors inhibited aromatization by approximately 90% in postmenopausal women, the third-generation aromatase inhibitors are far more potent, suppressing aromatization by approximately 98%. When radioimmunoassays are used to assess estrogen suppression, they generally correlate with the degree of aromatization suppression observed (see Table 2).
The randomized clinical studies of letrozole(Drug information on letrozole) and vorozole vs aminoglutethimide have demonstrated that the improvement in aromatase inhibition provided by the third-generation inhibitors is clinically meaningful, but the clinical relevance of any differences between members of the third generation is less clear. While most aromatization studies are not randomized studiesso that any comparison of their results must be interpreted with cautionone small (n = 12) randomized, crossover study has compared anastrozole to letrozole. This study demonstrated that letrozole is a more potent aromatase inhibitor than anastrozole (aromatization suppression rates were > 99.1% vs 97%, P = .003, with confirmatory estrogen suppression data).
The clinical relevance of this small difference, demonstrated at a level of inhibition that is so close to complete, remains uncertain. Equally uncertain is the clinical relevance of exemestane’s irreversible binding to aromatase, compared with the competitive, reversible binding of the nonsteroidal agents.