A large proportion of unexplained infertility is thought to be related to alterations in ovulatory function that may not produce detectable hormonal changes [2] . The use of exogenous gonadotropins to achieve controlled ovarian hyperstimulation has been widely demonstrated [3, 12, 13]. Historically, ovulation induction for infertility was reserved for treatment of anovulatory women. It was expanded to treat suspected luteal phase defects and unexplained infertility in the 1980s. The rationale for use of ovulation induction in women with unexplained infertility, who by definition have regular ovulatory menstrual cycles, is to augment the probability of pregnancy by targeting subtle defects in follicle development that cannot be elucidated by a standard infertility workup. Increasing follicle stimulating hormone (FSH) stimulation would promote granulosa cell development and might enhance both oocyte health and the number of oocytes ovulated. Oral ovulation induction agents are typically administered on days 3 to 7 of the menstrual cycle . When used in conjunction with IUI to increase the density of motile sperm available to these oocytes, the likelihood of pregnancy may be further increased. Over the years, oral medications such as clomiphene citrate (CC) and aromatase inhibitors (AI) or injectable gonadotropin treatment have been used. Comparisons of the efficacy of oral agents with different types of gonadotropins in IUI programs have led to conflicting results in the literature [14–19].

CC is a nonsteroidal triphenylethylene derivative with both estrogenic and antiestrogenic effects with well-established efficacy in ovulation induction. The antiestrogenic effects of CC create the perception of an estrogen deficit at the level of the hypothalamus, triggering an increase in GnRH and gonadotropins with stimulation of follicular growth [20–23]. Up to 15–20 % of patients with unexplained infertility fail to conceive after stimulation with CC [24]. This may be attributable to prolonged estrogen receptor depletion in the endometrium and the cervix, which may lead to endometrial thinning and poor cervical mucus in 15–50 % of patients [24–26]. Additional potential negative effects on uterine blood flow [27], embryo development [28], and the coagulation system [29] have been proposed. Based on this, it has been argued that CC use may be associated with lower pregnancy rates and increased incidence of miscarriage. However, while some studies have suggested that fecundity may relate to endometrial thickness , others have failed to demonstrate any significant correlation [30, 31] . Combining CC with gonadotropins results in high estrogen and progesterone levels, which are hypothesized to neutralize the adverse effects on estrogen responsive tissues, if they exist. However, the antiestrogenic effects of CC on the endometrium may counterbalance the benefit of increased endogenous FSH stimulation. AI , on the other hand, lead to a transient decline in estrogen production and do not cause estrogen receptor downregulation like CC. Therefore, these compounds stimulate gonadotropin production and follicular growth but do not interfere with estrogen levels in peripheral tissues such as the endometrium and cervix. Based on this fundamental difference and the experience gained with CC, AI have been examined for the purpose of ovulation induction in unexplained infertility [32] .

Letrozole is a third-generation AI that was originally developed for the treatment of breast cancer. Aromatase is a cytochrome P-450 hemoprotein-containing enzyme complex that catalyzes the conversion of androstenedione and testosterone into estrogens [33, 34]. Letrozole blocks estrogen production by its selective, competitive, reversible inhibition of the aromatase enzyme. These compounds have a high oral bioavailability and relatively long half-life of 45 h [35]. AI have been shown to decrease estradiol production, with approximately 50 % diminution in the amount of estradiol per mature follicles in peripheral blood on the day of human chorionic gonadotropin (HCG) administration [22, 32, 36]. As the luteinizing hormone (LH) surge is induced by a late follicular rise in estradiol concentrations that feeds back positively on the hypothalamic–pituitary axis [37], it is anticipated that letrozole delays the rise in LH. Letrozole is thought to not only lower estradiol but also the follicular proteins that antagonize the LH surge [22]. Studies have shown that in spite of estradiol levels per maturing follicles being nearly half the level seen in other ovarian stimulation regimens, a spontaneous LH surge occurs in up to 50 % of patients receiving letrozole with FSH , with markedly reduced LH triggering levels of estradiol [22]. This indicates that letrozole may induce ovarian effects beyond merely reducing estrogen production; however, these effects are not well elucidated .

A preliminary study by Mitwally and Casper [32] illustrated the utility of letrozole , an AI as an ovulation induction agent in patients with polycystic ovarian syndrome (PCOS) and ovulatory infertility. Letrozole was given orally in a dose of 2.5 mg on days 3–7 after menses. In the PCOS group, ovulation was achieved in 9 of 12 cycles and pregnancy was achieved in three patients. A cohort study by Cortinez et al. [38] demonstrated the feasibility of letrozole in treating unexplained infertility, in that it was shown to induce moderate ovarian hyperstimulation in ovulatory infertile patients with estrogen levels similar to spontaneous cycles and higher midluteal progesterone , leading to both a normal endometrial histology and development of pinopodes , which are viewed as markers of endometrial receptivity. According to the prospective trial published in 2002, patients were randomized to receive either CC or letrozole and it compared the effect of either treatment to the outcome of the natural cycle that immediately preceded either treatment cycle [39]. In both groups, one follicle reached maturity during the menstrual cycle preceding the intervention. The number of growing follicles was 2.2 in the CC group, which was slightly higher than the finding of 1.7 in the AI group, both of which were statistically greater than the preceding natural cycle. Like CC, AI increase the stimulus to recruit follicles as well as the ovulatory surge as evidenced by a larger number of follicles reaching maturity. This study provided the basis for the use of AI for inducing controlled ovarian hyperstimulation in women who ovulate regularly .

Baysoy et al. [40] conducted a study in which 80 patients with unexplained infertility were randomized to receive letrozole or human menopausal gonadotropin (HMG). Low estradiol concentrations and small numbers of mature follicles were obtained at the time of the LH surge in the letrozole group. Despite this, the pregnancy rate per cycle was highest in the letrozole group, although this difference was not statistically significant. In addition, the financial burden was significantly higher in the HMG stimulation cases and no injections were required in the letrozole group . The findings of this study shed light on letrozole as a simple, convenient, and cost-effective treatment regimen in ovarian stimulation for IUI .

When administered in the early follicular phase, AI induce a hypoestrogenic state that releases the hypothalamic–pituitary axis from estrogenic negative feedback, which in turn increases FSH secretion and the development of ovarian follicles. Peripherally, inhibition of the aromatase-mediated conversion of androgens into estrogens may lead to temporary accumulation of androgens. Androgens have been shown to increase follicular sensitivity to FSH through amplification of the FSH receptor gene expression either directly or through other mediators such as the insulin-like growth factor system [41, 42]. As such, aromatase inhibition and estradiol suppression may allow for ovarian stimulation to be achieved with lower doses of FSH, improve ovarian response to FSH in poor responders, terminate the positive-feedback loop, and improve ovarian response to stimulation. In addition, aromatase inhibition may result in a better implantation rate and lower estrogen concentrations compared to CC [43]. AI , such as letrozole and anastrozole, have been tested successfully as alternative ovulation-inducing agents with a much lower average cost per cycle than FSH [32]. A low-cost adjuvant may eliminate the need for costly high doses of gonadotropins. Healey et al. [44] retrospectively analyzed results in patients using letrozole and FSH versus FSH alone in IUI cycles . The authors concluded that the addition of letrozole to gonadotropins decreased gonadotropin requirements and increased the number of preovulatory follicles, and despite thinner endometrial development, no effect on pregnancy rate was seen. Lack of endometrial development may have been attributable to the daily-administered dose of letrozole being twice as high as typical dose administered in other studies. Mitwally and Casper [36] examined whether adding letrozole to FSH for ovarian stimulation in patients undergoing ovulation induction with IUI could improve outcomes in patients that had previous FSH stimulated cycles demonstrating a poor response. This was an observational cohort study performed on 12 patients by using 2.5 mg letrozole daily and FSH ranging from 50 to 225 IU per day. Patients in the combined regimen group demonstrated improved ovarian response compared with FSH alone group as evidenced by a reduced gonadotropin dose requirement as well as a higher number of mature follicles. The cost benefit of letrozole was further proven in a study by Bedaiwy et al. [45] in which patients with anovulation, male factor infertility, unexplained infertility, or endometriosis underwent IUI cycles with FSH with and without letrozole as an oral adjunct. The FSH dose required for ovarian stimulation was shown to be significantly lower when letrozole was used. Although a significantly higher number of follicles greater than 16 mm size and endometrial thickness at the day of HCG administration was observed in the group receiving FSH alone, pregnancy rate was comparable in both groups. Furthermore, the IUI cancellation rate, an expression of excessive or poor response, was significantly lower with combined FSH and letrozole treatment .

CC is the standard first-line treatment for ovulation induction for many decades . Some women (~ 15 %) are resistant to even maximal doses (250 mg/day for 5–7 days) or demonstrate a suboptimal response to CC and require additional treatment using gonadotropins. There are clinical scenarios in which AIs may demonstrate an advantage over ovulation induction with CC. AI may be a better choice when a limited number of follicles are required [46]. Due to significantly lower estradiol levels, letrozole may be considered for empirical ovulation induction in infertile patients with estrogen-dependent neoplasms. As no hypothalamic–pituitary downregulation occurs during the late follicular phase, no adverse effects on peripheral targets such as the endometrium are expected. Letrozole treated patients have been shown to have endometrial thickness with an estrogenic triple-line pattern [22, 32, 36, 47, 48]. Peripheral effects on the endometrium have traditionally been hypothesized to contribute to lower pregnancy rates and higher miscarriage rates in patients stimulated with CC [49] . Controlled ovarian hyperstimulation with AI may avoid the theorized undesirable effects of CC on the endometrium. Letrozole is also thought to have a more favorable side-effect profile compared with CC, likely because of the difference in pharmacodynamics between the two drugs. AI have a reduced half-life (40–48 h) compared with clomiphene [50]. Clomiphene has been hypothesized to result in prolonged central estrogen receptor depletion because of its greater half-life, leading to supraphysiologic levels of estrogen without central suppression of FSH as the normal estrogen receptor-mediated feedback mechanisms are blocked. Persistent CC mediated downregulation of estrogen receptors may also result in bothersome symptoms such as hot flushes, premenstrual symptoms, and visual scotomata in addition to poor cervical mucus and thin endometrium. Vasomotor symptoms are also reported in patients using letrozole for ovulation induction , along with other idiosyncratic side effects, which may include mild headache or muscle and joint pain. These side effects have also been reported in long-term breast cancer studies of letrozole [51] .

Mitwally and Casper [32] evaluated the efficacy of AI for ovulation induction in women that had previously failed CC treatment with unsuccessful ovulation or suboptimal endometrial thickness . This study included anovulatory women with PCOS and ovulatory women with other causes of infertility including male factor, endometriosis, or unexplained infertility. A higher proportion of PCOS patients (75 %) were ovulatory with letrozole in comparison to their previous cycle with CC (44 %). The mean number of mature follicles in letrozole and CC-stimulated cycles was similar in the ovulatory infertility patients. Pregnancy rates were 25 and 10 % in patients with PCOS and ovulatory infertility, respectively. The mean endometrial thickness at the time of HCG administration was significantly higher in both patient groups compared with their previous CC-stimulated cycle. Although limited by a small sample size and selection bias in the patients who failed CC treatment were used as their own controls, the findings suggested that letrozole may be beneficial in patients with failed CC cycles and who demonstrate a thin endometrium with CC use. The endometrium sparing effect of letrozole was also demonstrated in a trial by Bayar et al. [30] in which 46 patients with unexplained infertility, early stage endometriosis and borderline male factor infertility were randomized to receive letrozole or CC for ovulation induction . The letrozole group was seen to have a significantly lower median estradiol level on the day of HCG administration. The rates of ovulation and median endometrial thickness were comparable in both groups. However, in six of the CC cycles the endometrium was found to be less than 5 mm, in contrast to patients undergoing letrozole stimulation who all demonstrated endometrial thickness greater than 5 mm .

Despite a number of studies that suggest letrozole has a positive influence on ovarian response and the endometrium, there are many conflicting reports in the literature showing no evidence of a beneficial effect of letrozole in ovarian stimulation cycles . In a pilot study, Fatemi et al. [43] compared the efficacy of 100 mg of CC versus 2.5 mg of letrozole daily in patients undergoing IUI and demonstrated lower estradiol concentration and follicle count in the AI group. Badawy et al. [52] conducted a randomized controlled trial in which 412 infertile women with unexplained infertility, undergoing IUI, were randomized to pretreatment with 100 mg of CC daily or 5 mg of letrozole daily for 5 days starting on day 3 of menses. Both groups had similar endometrial thickness before treatment and at the time of HCG administration, but differed in that serum estradiol and progesterone concentrations were statistically significantly higher in the CC group. The total number of follicles during stimulation was also significantly greater in the CC group. There was no statistically significant difference in pregnancy rates between both groups. Of note, there were two twin pregnancies achieved in the CC group . No higher order pregnancies or cases of ovarian hyperstimulation syndrome occurred in either group. The results of this study provided evidence that the antiestrogenic effects of CC did not adversely affect outcomes in the majority of treated women. The findings failed to demonstrate superiority of either letrozole or CC for inducing ovulation and achieving pregnancy in women with unexplained infertility. Furthermore, the rate of pregnancy loss after ovarian stimulation , with either AI or CC was not higher than after spontaneous pregnancy. Similarly, a randomized trial by Al-Fozan et al. [53] comparing 7.5 mg of letrozole with 100 mg of CC per day demonstrated similar ovarian response, endometrial thickness , and pregnancy rates after IUI. However, the patients treated with letrozole were seen to have a lower miscarriage that was not statistically significant.

Recent meta-analyses reviewed six randomized controlled studies comparing ovulation induction with letrozole versus CC in patients with PCOS [54, 55]. The reviews concluded that the pregnancy and live birth rates were comparable in both treatment groups. Of note, letrozole use was associated with significantly fewer mature follicles and significantly lower estrogen concentrations per cycle [54]. The multiple pregnancy rates were not significantly different between letrozole and CC treated subjects. The majority of studies was limited by small sample sizes, with the exception of the study by Badawy et al. [52] which included more than 50 % of the study subjects analyzed in the meta-analysis, therefore having a large, disproportionate impact on all of the analyzed outcomes .

When gonadotropins are used alone for ovarian hyperstimulation the cost of treatment, incidence of multiple pregnancies, and probability of significant ovarian hyperstimulation syndrome are increased [16–18]. As such, combining oral ovulation induction agents with gonadotropin therapy to decrease the gonadotropin dose required for optimum stimulation has been explored as a method of ovarian induction [26, 56–58]. In a randomized trial by Badawy et al. [59], letrozole was shown to have no advantage over CC when either agent was combined with FSH in the treatment of unexplained female infertility . No difference between either regimens with respect to endometrial thickness or pregnancy rates was identified. However, the total number of follicles and serum estradiol and progesterone concentrations were significantly greater in the CC + gonadotropin group. Supraphysiological estradiol is thought to contribute to dyssynchrony between the development of the endometrium and early embryo development which can hinder implantation [60]. In conjunction with having higher estradiol levels, the CC group exhibited a nonstatistically increase in the rate of multiple follicles without a corresponding significant increase in multiple pregnancy or miscarriage rates. In a prospective nonrandomized study, Mitwally and Casper compared three ovulation induction regimens in women with unexplained infertility, consisting of letrozole combined with FSH, CC combined with FSH, or FSH alone [22]. The authors concluded that similar to CC, aromatase inhibition with letrozole reduced the required FSH dose for ovulation induction without the undesirable antiestrogenic effects sometimes observed with CC. Endometrial thickness was significantly lower in the CC group compared with the other treatment groups. The pregnancy rate was also significantly lower in the combined CC-FSH group (10.5 %) as compared to the combined letrozole-FSH group (19.1 %) and the FSH-only group (18.7 %). Aromatase inhibition allowed a reduction in FSH dose similar to that seen with CC while maintaining the high pregnancy rate encountered with FSH-only regimens. The observation of a thicker endometrium in the letrozole-FSH group compared to the CC-FSH group supports the inference that the effect of CC on peripheral tissues can significantly contribute to a decrease in pregnancy rates. Estradiol levels were markedly lower in letrozole-FSH group as compared to both FSH-only and CC-FSH groups [22].

These findings have been confirmed by other studies, including a randomized blinded clinical trial by Barroso et al. [61] in which letrozole and CC were compared as adjuncts to FSH for the purpose of controlled ovarian hyperstimulation in women with unexplained infertility. Letrozole with FSH was shown to achieve similar number of mature follicles compared with CC with FSH, with lower peak serum estradiol levels, which were associated with increased endometrial thickness. The total dose of recombinant FSH was similar between groups. However, in this study, the differential effects of letrozole and CC on the endometrium did not translate into a significant difference in pregnancy rate between both groups, with pregnancy rates of 23.8 and 20 % in letrozole/FSH and CC/FSH groups, respectively. This discrepancy may be attributable to a relatively small sample size resulting in poor statistical power for assessing this secondary outcome. There was no difference in the miscarriage rate or the proportion of multiple pregnancies in the groups treated with letrozole and CC.

Despite several studies that have shown letrozole to be efficacious for ovarian stimulation , there is a paucity of studies examining the optimal dose. When deciding on the ideal oral ovulation induction agent to use alone or as an adjunct to gonadotropin stimulation with IUI, it is important to consider that the cost of letrozole per cycle is much higher than CC, especially when higher doses of letrozole are required. Through the inhibition of aromatase activity, letrozole prevents the conversion of androgens to estrogen, thereby increasing testosterone concentrations [62]. Although there is data to suggest that testosterone may increase follicular FSH-receptor expression in primates and promote follicular growth by amplifying FSH effects [41, 42, 62]. It has been hypothesized that higher concentrations of testosterone are produced with higher doses of letrozole, potentially causing an adverse effect on the endometrium. Most investigators have studied letrozole using a 2.5 mg dose for duration of 5 days during early stimulation [63–65], whereas others have used a range from 2.5 to 7 mg [54, 64]. In a study by Al-Fadhli et al. [66], patients assigned to receive 5 mg of letrozole produced an overall greater number of follicles, with higher pregnancy rates as compared to subjects that received 2.5 mg.

Badawy et al. [67] conducted a parallel-group randomized controlled trial comparing the efficacy of letrozole at doses of 2.5, 5, and 7.5 mg in 179 women undergoing ovulation induction and timed intercourse for treatment of unexplained infertility. The demographics of study groups were similar except for the fact that patients receiving the 7.5 mg dose tended to have a significantly longer duration of prior infertility compared with the 2.5 mg group. There was no significant difference between the three groups regarding the pretreatment endometrial thickness, pregnancy, or miscarriage rates. The findings showed a significant dose-dependent increase in the total number of follicles greater than 14 mm on the day of HCG administration, with the most follicles yielded following the 7.5 mg dose. The number of days needed to achieve a mature follicle also showed a significant dose-dependent decrease with increasing doses of letrozole . The increase in the number of mature follicles was not paralleled by a similar increase in pregnancy rate between the three groups. This is in contrast with the findings of Al-Fadhli et al. [66], who found no significant difference in the days needed for stimulation (11.4 versus 11.7 days) for patients receiving 2.5 versus 5 mg of letrozole, but demonstrated that a higher dosage was associated with a higher pregnancy rate. The mean endometrial thickness achieved was 7.5 (+ /− 0.3) mm in the group receiving the 2.5 mg dose and 7.8 (+ /− 0.3) mm in the group receiving the 5 mg dose. One might assume that the increased concentrations of midluteal progesterone and endometrial thickness in patients receiving higher doses of letrozole would optimize the luteal phase and lower miscarriage rates. Conversely, a slightly higher miscarriage rate was observed with the use of the 5 and 7.5 mg doses compared with the 2.5 mg dose, although this finding was not statistically significant. These conflicting findings from various studies may be attributed to recruitment of different patient populations.