Letrozole has been used with increasing frequency in infertility therapy. Letrozole blocks the conversion of C-19 androgens to C-18 estrogens by competitively inhibiting the enzyme, aromatase (cytochrome P-450 19), which is an essential step in estrogen biosynthesis in the ovary and other tissues. The use of letrozole results in increased androgens and decreased estrogens in tissues throughout the body [1, 2]. Letrozole has proven to be a valuable drug in the treatment of estrogen receptor–positive breast cancers which constitute about two-thirds of all breast cancers [3, 4].
In this mini-review, we examine four applications of letrozole to treat infertility that have appeared in the medical literature. The focus of this review is on how these therapies utilize the pharmacological properties and physiological ovarian consequences of letrozole. The initial use of letrozole in infertility mimics the long-term use of clomiphene citrate in infertility therapy. Clomiphene citrate was used to increase FSH secretion from the pituitary by blocking a negative estradiol feedback loop by binding to estradiol receptors in the hypothalamus. Contemporary applications of letrozole to infertility use the same negative feedback loop to increase FSH secretion, but do so by decreasing estradiol production from the ovary and other tissues.
Infertility applications of letrozole are quite different from the primary medical application of letrozole to treat breast cancer. Aromatase inhibitors are used in post-menopausal women with breast cancer, who have a (relatively) long duration of planned treatment. The impact of letrozole use in a setting of rapid endocrine shifts requires a more subtle assessment of its impact. Letrozole effects follicle development, follicle growth, and follicle atresia. These effects on a given ovarian follicle are complex, because they vary depending on where that follicle is in its growth trajectory and on what other drugs are being used at the same time. Theoretically, using letrozole solely to enhance FSH secretion is to only utilize a portion of its impact on the ovary. Little information is presently available on whether or not it is clinically useful to try to incorporate non-FSH impacts of letrozole on the ovary into infertility therapies.
The goal of this paper is to first review clinically important pharmacological aspects of letrozole. We will then review follicular physiology and endocrine activity as a follicle changes from a pre-antral to a pre-ovulatory follicle. We will further review some of the impacts of the endocrinological changes induced by androgens on granulosa cell apoptosis and antral follicle atresia. Finally, we will look at four selected applications of letrozole (see Table 1) from the infertility literature to see how these applications use letrozole or how they could more fully utilize the impact of letrozole on ovarian physiology.
First-generation aromatase inhibitors have been available since the 1970s, but, although they blocked estrogen biosynthesis, they also inhibited the biosynthesis of cortisol, aldosterone, and thyroid hormone . An additional problem was liver toxicity. Second-generation aromatase inhibitors were safer and more potent, but were still not adequately selective in inhibiting estrogen production. Third-generation aromatase inhibitors became available in the 1990s. They were created utilizing new molecular modeling techniques designed to enable better binding to the active site in aromatase. Letrozole was one of the molecules created that met the goal of high potency and selectivity. Letrozole and anastrozole are clinically available non-steroidal (i.e., type II) third-generation aromatase inhibitors.
Letrozole has 99.9% bioavailability after oral administration. It has a single dose terminal half-life of 42 h, but a steady-state concentration is not achieved for 2 to 6 weeks (steady-state half-life is 118 h) [1, 15]. This is thought to be because letrozole is metabolized into inactive metabolites by two cytochrome P-450 enzymes: 3A4 and 2A6. Cytochrome P-450 2A6 has a much higher affinity (and is a more efficient metabolizer) to letrozole than 3A4, but becomes saturated at a letrozole dose of about 2.5 mg [1, 15].
Estrogen suppression by letrozole has been looked at in many in vitro settings, for example, human placental tissue, hamster ovarian tissue, various cancer cell lines, human breast tissue, human adipose fibroblasts, rodent cells, and particulate breast cancer tissue . In these settings, letrozole is highly efficient in suppressing estrogen production. Consequently, letrozole is an effective drug for the treatment of post-menopausal women with breast cancer [4, 15]. Letrozole was not designed for short-term use in pre-menopausal women and there is relatively limited direct research on how it impacts the ovarian follicle throughout its development.
Aromatase is a member of the cytochrome P450 superfamily which contains over 480 members. Aromatase is found in tissues throughout the body and the regulation of its expression is by tissue-specific promoters and activators. In the ovary, cAMP and gonadotropins regulate aromatase expression . Significant aromatase expression does not occur in granulosa cells in follicles until they are about 9 mm in diameter and starts to markedly accelerate in 10-mm follicles [17,18,19]. However, where ever aromatase occurs, it is highly specific for binding androgens, and, by design, letrozole. If testosterone docks with an aromatase molecule in ovarian tissue, it is rapidly converted to estradiol, which has only a weak affinity for it and the estradiol molecule is quickly released into the cytoplasm. The aromatase binding site is then free to competitively dock with either an androgen or a letrozole molecule. Letrozole circulates with 55% bound to albumen and the impact of its pharmacology is that aromatase activity is decreased by more than 99% (at a dose of 2.5 mg) [1, 15, 16]. In the pre-menopausal woman treated for infertility with letrozole, estradiol can still be produced during the follicular phase of the menstrual cycle, either because of a reduction in letrozole availability due to letrozole’s half-life, the treatment regimen, or because of enhanced aromatase expression primarily due to increased FSH and FSH receptor availability.
Androgen production is not directly increased by letrozole. In the pre-menopausal woman, intra-ovarian androgen levels are increased, primarily because less androgen substrate is used for estrogen biosynthesis. Ovarian androgen production occurs in the theca (and interstitial) cells in response to LH and, in the large follicle, is augmented by IGF-1. However, in the early follicle, oocyte-derived growth differentiation factor 9 (GFD9) is essential for theca cell differentiation and androgen production [20, 21]. Follicles with poor theca cell development do not develop further.
Follicular development in women is a complex event that is challenging to understand at the molecular level. To understand how letrozole affects ovarian physiology, current nomenclature does not fully capture a follicle’s changing molecular endocrine environment. The most widely used nomenclature describing follicle growth divides follicles into primordial, primary, secondary, pre-antral, antral (or tertiary), and pre-ovulatory stages based on unequivocally defined morphological characteristics (Fig. 1). This characterization is not optimal in applying ovarian physiology to infertility applications since most cycle manipulations used occur in the late pre-antral and the antral follicle stages. Gougeon’s characterization of follicle growth in humans defines eight stages of development based on the number of granulosa cells in follicles which also correlates to a follicle’s diameter (Fig. 2) . Gougeon’s characterization is more helpful in understanding infertility interventions in humans, since human follicle diameters are easy to obtain, but they do not directly facilitate specific application of research done in other species (which have different follicle diameters and granulosa cell numbers) to humans; information transfer would require a more physiological characterization of follicle growth. An example of this type of observation, limited to FSH’s significance for follicle growth, is given by Orisaka et al.  (Fig. 3). FSH receptors are detectable on granulosa cells in the early pre-antral follicle [17, 23] and thus can respond to FSH several months prior to ovulation. However, the time at which the follicle transitions from being FSH responsive to FSH dependent is clinically important and occurs in all mammals. In the human, this occurs when follicles are between 4 and 6 mm in diameter (Gougeon class 5 to 6). As will be discussed, the transition from peak androgen dependence also occurs around this same time [18, 24].
Beneficial effects of androgens on small follicles
Studies consistently show that androgens play important positive roles in pre-antral to early antral follicle development [12, 25,26,27,28,29]. In this setting, they promote granulosa cell mitosis [24, 25, 30], increase FSH receptor expression and FSH receptor protein [21, 26, 30, 31], and prevent granulosa cell apoptosis . Although androgens are critical hormones in the pre-antral to early antral follicle’s stage of development, they are not acting alone in promoting follicle growth and development. For example, acting in conjunction with FSH and modulated by IGF-1 and GDF9 (positively) and AMH and BMP15 (negatively), androgens promote granulosa cell proliferation  (Fig. 4).
Peak numbers of androgen receptors are found in granulosa cells of human antral follicles between 3 and 6 mm in diameter [18, 19, 24]. Androgen receptors become fewer in granulosa cells of larger antral follicles and are significantly reduced in granulosa cells of early pre-ovulatory follicles [18, 19, 24]. The antrums of human small antral follicles less than 7 mm in diameter are primarily androgenic [18, 32]. During this period of increasing responsiveness to androgen actions through their receptor, the human antral follicle has increased its granulosa cell mass from a few thousand cells to about ten million cells [22, 33]. The follicles, which continue to grow and differentiate to become pre-ovulatory, will need to continue mitosis at this pace to create a granulosa cell mass of about 50 million cells approximately 1 week later. Classical steroid androgen hormone signaling with receptor activation in the cytoplasm, transfer of the activated receptor into the nucleus, and binding of the activated receptor to a gene promoter is less frequent after the follicle grows out of this small antral follicle stage and is almost non-existent in the last stages of follicle growth. The mechanism of granulosa cell mitosis and follicle growth must change from a primarily androgen-driven system.
FSH and estrogen’s actions on granulosa cells become important for extending follicle growth past the early antral follicle transition point [27, 34, 35]. Nevertheless, FSH receptor expression decreases markedly per granulosa cell in human follicles 5–6 mm in diameter and continues to decrease as the follicle diameter increases [18, 19]. In fact, Jeppesen found a 250-fold decrease in FSH receptor expression per granulosa cell from small antral to pre-ovulatory follicles before ovulation . For a dominant (or rescued) follicle to be able to maintain a healthy FSH receptor population to respond adequately to FSH in producing estrogen, it must increase the number of its granulosa cells. FSH induces aromatase expression in granulosa cells and that expression starts to markedly increase in the human 11–13-mm follicle, with a concomitant increase in estradiol, and continues to increase as follicle diameter expands to pre-ovulatory follicle size . More granulosa cells in a follicle lead to more FSH receptors per follicle with increased estrogen production resulting in more mitosis [27, 33]. Estrogen is also essential for LH receptor expression in the granulosa cell, which only starts to be expressed on granulosa cells in an antral follicle greater than 10 mm in diameter and is required for eventual ovulation .
Most human atretic follicles have diameters in the 1 to 10 mm in range (Gougeon class 5, 6, and 7) with most atresia occurring in the smaller antral follicles . Gougeon found that the doubling time for granulosa cell mitosis decreased by 50% (from 10 to 5 days) in a 5-mm-diameter human follicle compared with granulosa cell division in slightly smaller follicles . McNatty et al. found that healthy antral follicles contained more than twice as many granulosa cells as atretic follicles of the same size and that granulosa cell number was correlated with the estradiol level in the antral fluid . This suggests that follicles 6–12 mm are at one of the sensitive stages in their destiny, possibly because of this increased requirement for accelerated mitosis. As receptor-mediated androgen benefits taper off, a follicle must be able to use FSH receptor–mediated benefits to continue to grow. This requires healthy follicles to accelerate granulosa cell mitosis, which, as the follicle enlarges, is primarily a consequence of FSH [16, 35, 36].
FSH activity is essential to enable follicles to develop so that they contain a fertilizable oocyte. Estrogen production is essential for granulosa cells to develop LH receptors eventually enabling spontaneous ovulation . FSH works though its G-coupled cell surface receptor on granulosa cells to activate protein kinase A (via cAMP) and subsequently a pathway that leads to phosphorylation and activation of protein kinase B (Akt) [34, 38]. The Akt pathway is a necessary (but not sufficient) activator of many FSH target genes including the LH receptor and aromatase. The Akt pathway works by phosphorylation of many proteins (at least fifty have been identified) resulting in either stimulation or inhibition of various processes . Some of these pathways also require activation of other pathways to achieve a particular action. In particular, Akt phosphorylates members of the forkhead box O (FOXO) family to stimulate mitosis and inhibit apoptosis (by inhibiting expression of a proapoptotic protein) [39,40,41].
Although androgens are essential for pre-antral and early antral follicle development, the early follicular unit works to restrain androgen production from the theca cells with factors from both the granulosa cells and the oocyte . Androgen excess secondary to letrozole treatment will be limited by the amount of estrogen that the follicular unit is trying to produce (related to the amount of androgen that theca cells are producing). Clinically, it is an open question as to how effective letrozole can be in elevating intra-follicular androgens in these early follicles and if that elevation can be utilized beneficially by the follicle.
Granulosa cell apoptosis and antral follicle atresia
Apoptosis is a cellular event, whereas atresia is a follicular event. Cellular apoptosis is characterized by loss of cell volume (cytoplasmic condensation), nuclear pyknosis (resulting from margination of the chromatin and redistribution against the nuclear membrane), and cytoplasmic blebbing (formation of membrane-bound vesicles containing intact cytoplasmic organelles). These vesicles or apoptotic bodies are phagocytized by neighboring cells so that cell death and the cell’s elimination from the follicle occur without causing an inflammatory response. A hallmark of apoptosis is an endonuclease-mediated cleavage of the cell’s DNA in fragments that are multiples of 185 base pairs [42,43,44].
The morphological and histological findings of an antral follicle in the process of undergoing atresia include an increase in the number of pyknotic nuclei (apoptotic bodies) scattered in the membrana granulosa, granulosa cell detachment from the basal lamina which folds in on itself as the follicle shrinks. Penetration of the basal lamina (without degradation) by macrophages into the follicular antrum, cellular hypertrophy of the theca interna, the appearance of cellular debris in the antrum, disruption of the oocyte-cumulous cell connections, and eventually meiosis-like morphological changes of the oocyte followed by oocyte fragmentation [45, 46]. Note that pre-antral follicles utilize a different pathway to atresia and have a different histology [47, 48].
Antral stage follicular atresia begins with apoptosis in granulosa cells . Healthy follicles may contain some apoptotic granulosa cells and apoptotic granulosa cells in large antral follicles do not impair or predict oocyte quality [44, 50, 51]. However, when the number or proportion of apoptotic granulosa cells exceeds a certain threshold, follicles become irreversibly atretic. Clinically, there are varying degrees of follicular atresia and in early stages, the process of becoming atretic can be suspended, most commonly by preventing further granulosa cell apoptosis by FSH rescue.
Granulosa cell apoptosis is a non-pathological process that is complexly regulated. There are two major pathways to granulosa cell apoptosis: a cell surface pathway and a mitochondrial pathway. Triggering either of these pathways involves the expression of one or more of a number of “death” inducing factors that are not suitably balanced by a number of “survival” or anti-apoptotic factors. Figures 5 and 6 illustrate these types of interactions, taking place in a granulosa cell, involved in maintaining a healthy follicle or in producing an atretic one [42, 43, 52]. Both experimentally and clinically exogenous FSH is a highly effective survival factor for granulosa cells and, in the late antral and pre-ovulatory follicles, enables follicles to avoid atresia.
To recapitulate, androgens are the primary anti-apoptotic factor in pre-antral and early antral follicles. Androgens indirectly are anti-apoptotic by increasing FSH receptors and through stimulation of mitosis. In addition, androgens directly inhibit apoptosis in granulosa cells by production of microRNA-125b which suppresses proapoptotic protein production . FSH is the primary anti-apoptotic factor in larger follicles.
Harmful effects of androgens on granulosa cells
High levels of androgens or exposure to androgens during late follicular development can promote follicular atresia [53,54,55]. Large atretic follicles have long been recognized to contain high androgen to estrogen ratios . Female to male transsexuals demonstrate a 3.5-fold increase in atretic follicles in their ovaries . Available data suggests that androgen levels are not increased in atretic follicles during spontaneous menstrual cycling; rather, estrogen levels are markedly decreased within an atretic follicle, resulting in a high androgen to estrogen ratio .
Chen et al. used a rat model to study the impact of dihydrotestosterone (DHT), an androgen that cannot be converted into estrogen, on large antral/pre-ovulatory follicles . They attempted to model androgen levels consistent with those of women with polycystic ovarian syndrome (PCOS). In one short-duration experiment, they cultured granulosa cells from young rats. FSH treatment significantly enhanced cellular proliferation, but DHT suppressed this enhancement of proliferation with an accumulation of granulosa cells arrested in the G2/M phase of the cell cycle. DHT was found to inhibit FSH-induced phosphorylation of Akt. Phosphorylated Akt plays an essential role in granulosa cell survival and proliferation . DHT treatment also promoted PTEN (phosphatase and tensin homolog deleted on chromosome 10) expression, which is involved in cell cycle arrest . PPARɣ, one of the transcriptional factors upregulating PTEN expression , was also enhanced by DHT treatment. The increased mRNA expression of PTEN after DHT treatment was suppressed significantly by silencing RNA (siRNA) of PPARɣ. Since large follicles contain few or no androgen receptors, this impact of androgens must utilize a non-genomic (non-androgen receptor) pathway .
Chen et al. also looked at long-term in vivo effects of androgen treatment with DHT in large rat follicles . Again, DHT caused a cellular proliferation in large follicles to be reduced, phosphorylated Akt to be suppressed, and PTEN and PPARɣ to be enhanced. Thus, they found the same impact of diminished granulosa cell proliferation by the same mechanism in both short-term granulosa cell culture with DHT and long-term DHT treatment of intact animals on large antral follicles.
However, granulosa cells from follicles from intact animals were also evaluated for apoptosis and no differences in the incidence of apoptosis by TUNEL (terminal deoxy-UTP nick end labeling) staining in follicles from DHT-treated and untreated animals were found. Nevertheless, these long-term DHT-treated animals had significantly fewer mature and ovulated follicles. The treated animals also had longer and fewer estrous cycles than controls. It is unclear exactly why apoptosis was not found in these large follicles with slowed development. TUNEL staining does not detect all apoptosis in granulosa cells . However, since there were fewer ovulated follicles, this experiment suggests that follicles were lost from the ovulatory pool. Vendola et al. treated rhesus monkeys with DHT or testosterone and found that the number of small follicles increased threefold, but large antral follicles were similar to untreated controls . Slow-growing follicles could remain viable and non-ovulatory by becoming cysts . Follicles in long-term androgen-treated animal PCO models, which do not proceed to ovulate, often become cystic and eventually die using a pathway other than apoptosis . Noorafshan et al. looked at the impact of letrozole treatment (21 days) in rats and found that granulosa cell numbers were decreased by 56% and there was a fivefold increase in the volume of ovarian cysts compared with controls . Paixão et al. reviewed animal models from diverse species, using sundry androgens in short- and long-term androgen exposures and with androgen exposures at different times in the animal’s development . Most studies of these models found an increase in both atretic and cystic follicles, but with some models having either atretic or cystic follicles dominant. In humans with severe PCOS, naturally occurring elevated androgens only rarely produce follicles larger than 10 mm with most follicles arresting below 7 mm and yet surviving for months .
In summary, high levels of androgens appear to directly slow follicle development in larger follicles by diminishing the ability of the granulosa cell mass in a follicle to proliferate enough to maintain an adequate FSH receptor population. As this process of mitosis slows, so does aromatase production, and resulting estradiol production. Slowing the growth of the membrana granulosa of a follicle decreases its access to apoptotic survival factors. If this decrease in mitosis is not enough to make the follicle atretic, it otherwise decreases the ability of that follicle to ever spontaneously ovulate. Cystic follicles may still contain non-apoptotic oocytes that could possibly remain useful in an IVF setting.