Oral dose is absorbed fast. Serum levels of zuclomiphene peak 6 h after oral intake. It is metabolized in liver and excreted in intestine. Final serum levels of zu- and enclomiphene are determined by its metabolism and weight of the women. It is generally agreed that enclomiphene is the more potent is omer and it is rapidly cleared from serum, making it theoreticlly more appealing as an ovulogen [4]. Females with higher BMI have larger plasma volume and thus have lower serum levels of clomiphene. Both isomers have long half-life. Serum half-life of zuclomiphene is 14.2–33.4 days and enclomiphene about 2.5–11.8 days [5].

Even after discontinuation of the drug, 50 % of dose can be detected in serum after 5 days, and its metabolites can be found up to 6 weeks. Zuclomiphene tends to accumulate over consecutive cycles with 50 % rise in serum level in next cycle [6].

Clomiphene acts as a competitive antagonist to 17 ß oestradiol at the level of nuclear receptor complex as the binding of clomiphene to the oestrogen receptor is more prolonged than oestrogen [7], which results in a pseudo-hypo-oestrogenic state. However, in hypo-oestrogenic condition clomiphene may show oestrogen agonistic action. Its primary site of action is hypothalamus as an oestrogen antagonist and secondarily at pituitary as a weak agonist, increasing the sensitivity of pituitary to GnRH. Serum LH levels start rising 2–3 days after admnistration by its direct action on pituitary [8].

Anovulatory PCOS patients have high GnRH pulse frequency. In these patients CC increases gonadotropin secretion by increasing GnRH pulse amplitude [9], while in ovulatory patients with normal GnRH pulse it acts by increasing pulse frequency [10]. It causes moderate rise in gonadotropin levels. In anovulatory patients gonadotropins rise by 50 % [7]. LH surge usually occurs 5–12 days after last clomiphene tablet.

CC can be started any day from cycle day 2 to 5 [11]. One study has reported lower ovulation rates when CC was started immediately after spontaneous or induced bleed [12].

Usual starting dose is 50 mg/day for 5 days in women less than 50 kg, while in women with weight more than 75 kg,100 mg/day can be started. Approximately 46–52 % patients ovulate with 50 mg/day, 21–22 % ovulate with 100 mg/day and 8–12 % will need 150 mg/day [13]. Dose higher than 100 mg/day is not approved by US FDA. Patients hypersensitive to CC can be started with 25 mg/day. Doses higher than 150 mg/day are not used due to high LH level and anti-oestrogenic action. Approximately 15–20 % anovulatory patients do not ovulate with these doses [14].

Lower ovulation rates are seen in women with higher BMI, higher free androgen index, insulin resistance and large ovarian volume [15]. There is no reliable predictor of ovarian response to CC [16, 17]. Failure to ovulate with CC for 6 months is termed as clomiphene resistance. An overall cumulative pregnancy rate of 55–73 % is seen in patients treated with CC [18]. Despite high ovulation rates, PR is low, and this is due to anti-oestrogenic action of CC on endometrium causing several dysfunctions and making cervical mucus thick and hostile for sperm penetration [19–21].

Lower pregnancy rates are seen in older women, hyperandrogenic PCOS with severe cycle disorder [22] and obese women. The landmark study by Legro et al. reported 28 % LBR in women with less than 30 kg/m² BMI vs. 16 % LBR in women with more than 35 kg/m [23].

Up to 71–87.5 % of pregnancies occur in first three cycles. The cumulative PR among anovulatory patients who ovulate with CC 50 mg/day, 100 mg/day and 150 mg/day for 3 months are 50 %, 45 % and 33 %, respectively, and further at 6 months are 62 %, 66 % and 38 %, respectively [18]. Since most of the pregnancies occur in first three to six cycles, treatment beyond six cycles is not recommended. Failure to conceive despite ovulation with clomiphene citrate is termed as CC failure.

CC is the most common protocol for IUI and results in 5–7 % pregnancy rate per cycle even after seven cycles [27]. Prior to using IVF, IUI with clomiphene ovarian stimulation is relatively cheap, and many couples will conceive and not require IVF [28].

Anovulatory patients do not need IUI [29], unless cervical or male factor is abnormal. However, for patients with unexplained infertility, CC with IUI is superior to expectant management. But CC without IUI is not superior to expectant management. A recent Cochrane review found that CC was not superior to expectant management or placebo for live birth (odds ratio [OR] 0.79, 95 % CI 0.45–1.38; p = 0.41) or for clinical pregnancy per woman randomized both with intrauterine insemination (IUI) (OR 2.40, 95 % CI 0.70–8.19; p = 0.16), without IUI (OR 1.03, 95 % CI 0.64–1.66; p = 0.91) and without IUI but using human chorionic gonadotropin (hCG) (OR 1.66, 95 % CI 0.56–4.80; p = 0.35) [30].

LH surge monitoring or ovulation trigger is needed to time IUI. Traditionally, ovulation trigger is usually given with urinary hCG 5,000 IU s/c once the follicle reaches 18 mm and IUI performed 36 h post trigger or after ovulation is detected on ultrasound. A recent study challenged this practice and showed that ovulation trigger at follicular diameter of 24 mm was associated with thicker endometrial thickness (9 mm) and higher probability of pregnancy [31].

The recent interest in mild/minimal stimulation and low-cost IVF has made CC an important weapon in the arsenal of a reproductive medicine practioner. It is started from day 2–3 of periods and can be either given for 5 days as in OI with addition of antagonist taking care of premature LH surge or till a day prior to trigger injection. The latter, apart from follicular growth, helps prevent premature LH surge due to its antagonistic action on the oestrogen receptor in the hypothalamus strongly inhibiting the positive feedback of the rising oestradiol level [32].

The largest study done till date has been of 43,433 cycles from Japan. CC was started from day 3 at 50 mg/day and was continued until the day before maturation trigger. If the ultrasound on day 8 suggested good number of follicles growing, 150 units of HMG or FSH were added. Emergency oocyte retrieval due to the premature LH surge was required in 3.5 % of cases. The ovulation rate was between 2 and 3 % confirming the efficacy of CC in preventing premature LH surge. The oocyte retrieval rate and embryo cleavage rate were 83 and 64 % respectively. The live birth rate in the study was 11.1 % [33].

A recent meta-analysis exploring the efficacy of CC-antagonist protocol vs. the conventional non-CC protocol evaluated seven trials with total of 702 women. There was no significant difference in the parameters of live birth (p = 0.26), clinical pregnancy (p = 0.12) and number of oocytes retrieved. Importantly, significant reduction in OHSS (1/216 = 0.5 % vs. 9/217 = 4.1 %, p = 0.01), consumption of gonadotropins and duration of COH were seen [34].

In our centre, we have been using the continuous CC protocol by starting CC on day 2 till the day prior to trigger. We add FSH 75–150 IU from day 4 of CC to recruit more follicles. Of 440 cycles with this protocol, the clinical pregnancy rate per started cycle is 29.8 % (unpublished data).

In case there is clomiphene resistance, the woman can be put on extended regimes with clomiphene. However, adverse effects of clomiphene on the endometrium must be kept in mind.

Side effects are mostly caused due to pesudo-hypo-oestrogenic environment. About 64–78 % women complained of mood swings, while 10 % women complained of hot flushes during the course of medication. Breast tenderness, nausea and pelvic discomfort are noted in 2–5 % women. Visual disturbances in the form of blurring of vision, scotoma, diplopia and light sensitivity are rare (<2 %). Visual symptoms are usually reversible but warrant discontinuation of further treatment with CC. Some clinicians suggest lower dose in next cycle. Optic neuropathy is very rare. Mild OHSS is found in 13 % patients, but severe OHSS is rare [37].

As clomiphene is metabolized in liver, it is contraindicated in patients with chronic liver disease. Functional ovarian cysts may become larger with clomiphene. History of blurring of vision or scotoma with use of clomiphene contradicts its use. Although it does not increase the risk of congenital anomalies [38], still it should be avoided during pregnancy. There is no cause/effect relationship between any OI agent and invasive ovarian neoplasia even when used for more than 12 months [39].

Letrozole is discussed in detail as it is the most researched of the aromatase group. Letrozole is 100 % bioavailable following oral administration and has a terminal half-life of around 45 h.

The rate-limiting step in oestrogen synthesis is the conversion of androgens (androstenedione and testosterone) into oestrogens (oestrone and oestradiol, respectively) and is catalyzed by aromatase or oestrogen synthetase enzyme.

Aromatase is a microsomal cytochrome P450 hemoprotein-containing enzyme (a product of the CYP19 gene). This enzyme is expressed in various tissues in the human body including ovaries, uterus, brain, breast and adipose tissues. Aromatase inhibitors selectively target the aromatase enzyme, which is the last in the cascade of steroidogenesis, thus depleting the oestrogen levels in target tissues.

Due to the presence of the aromatase enzyme in brain as well as ovaries, it is likely that the fertility effects of AIs are the result of dual actions at central and peripheral levels.

At the central level, lowered circulating oestrogens as a result of blocked oestrogen synthesis in brain and other tissues would release the hypothalamus and/or pituitary from the oestrogen-negative feedback on the production and release of gonadotropins. This action differs immensely from that of clomiphene, as there is no depletion of oestrogen receptors. This leads to an increase in gonadotropin secretion, which stimulates the growth of the ovarian follicles. It also increases the level of activins, which further stimulates synthesis of FSH by a direct action on the gonadotropes [45].

At the peripheral level, aromatase inhibition leads to temporary accumulation of intraovarian androgens since the conversion of the androgen to oestrogens is blocked by inhibition of the aromatase enzyme. This can lead to an increase in responsiveness of the ovarian follicle to FSH as a result of either a direct action of testosterone on the augmentation of FSH receptor expression [46] or indirectly by increasing the IGF-1 levels [47].

The chances of having a monofollicular growth is higher with AIs compared to clomiphene as the increasing oestradiol level towards mid and late follicular phase reduces the FSH levels, thus allowing only the follicles with highest number of FSH receptors to sustain growth while the others undergo atresia. In case of clomiphene, the longer half-life of clomiphene allows the oestrogen receptors in the hypothalamus and pituitary to be blocked for longer duration leading to growth of more than one follicles in many women.

The oestradiol level per growing follicle is 40–60 % lesser in cycles where AIs are utilized for ovulation induction or COS. This is in line with the reduced functioning of the intraovarian machinery responsible for converting androgens to oestrogens. Theoretically, these reduced levels can be of help during IVF cycles as supraphysiological steroid levels are the major factors leading to advancement of the endometrial maturity. Though appealing, it has not been proven yet.

Ovulation induction with letrozole is shown to have an ovulation rate of 70–84 % and a pregnancy rate of 20–27 % per cycle [48].

The indication most widely studied for letrozole usage has been that of PCOS, including ones with clomiphene resistance. There has been no formal dose-finding study for OI or COS. In fact, the doses used have been directly extrapolated from that used for breast cancer. The commonest dose of letrozole used is 2.5 mg for 5 days from day 2 to 3 of spontaneous or withdrawal bleed. Though dosage schedules of 5 and 7.5 mg have also been used, they did not incur any significant advantage in terms of pregnancy rates, though the total number of follicles growing were higher with higher dosage [49].

As for the dosage, different length of letrozole supplementation has been studied as well. Extended letrozole therapy for 10 days was tried in 218 patients who had previously failed clomiphene citrate at 100 mg for 5 days. They were randomized to receive either 5 mg of letrozole for 5 days or 2.5 mg for 10 days, both starting on day 1 of the menstrual cycle. Ovulation rates were similar at 65.7 % for the extended versus 61.8 % for the short course while a mean of three follicles more than18 mm were seen in the extended regimen compared to 1.8 in the short regimen. Pregnancy rates with the short and extended regimens were 12.4 % and 17.4 % respectively [50].

AIs with added gonadotropins can be used if more than one follicle are required to be grown as in case of unexplained infertility. This obviously will have attendant risk of OHSS and multifoetal gestation, so it should be used cautiously. Letrozole and clomiphene have been tried in unexplained infertility in many studies, the former at dosage between 2.5 and 7.5 mg/day and the latter at 100-mg/day dosage. A recent meta-analysis showed equivalence in terms of pregnancy rate in between both ovulogens, though the numbers of growing follicles were lesser with letrozole [51].

There have been many studies done in PCOS comparing clomiphene with letrozole, and the results have been mixed. But in the recent Cochrane review on the role of aromatase inhibitors in anovulatory PCOS comprising of 26 RCTs (5,560 women), it was shown that letrozole when compared to clomiphene had significantly better live birth rate (OR 1.63, 95 % CI 1.31–2.03, n = 1,783) and clinical pregnancy rate (OR 1.32, 95 % CI 1.09–1.60, n = 2,066) [52]. The reviewers advised caution in interpreting the results, as the quality of evidence was low. Nevertheless, it means that letrozole is at least as effective, if not better than clomiphene in this group of women. In clomiphene-resistant PCOS, when compared with placebo, letrozole was shown to have 33.3 % ovulation rate compared with nil in the placebo group [53].