Although many studies have been performed to develop a method of efficient controlled ovarian stimulation (COS) in infertile women with diminished ovarian reserve who are scheduled for in vitro fertilization (IVF) treatment, the management for them remains challenging. Several COS regimes have been developed, including the gonadotrophin-releasing hormone (GnRH) agonist low-dose long protocol [1,2], GnRH agonist flare-up regimes [3,4], and GnRH antagonist protocols [5,6,7]. However, none of these protocols have been particularly effective in improving the ovarian response of poor responders. Therefore, hormonal manipulation that aim to augment follicular recruitment and coordinate subsequent antral follicle growth during COS has been used in poor responders. Such hormonal manipulation involves the use of aromatase inhibitors.

Indeed, several preliminary studies showed that when the aromatase inhibitor letrozole is added to a COS protocol in poor responders, the ovarian response to follicle stimulating hormone (FSH) improves and total doses of gonadotropin required are reduced [8,9,10,11].

Therefore, the present study was performed to evaluate whether letrozole incorporated in a GnRH antagonist multiple dose protocol (MDP) would improve the ovarian response to COS and IVF results in poor responders who underwent IVF/intracytoplasmic sperm injection (ICSI).

The two groups did not differ with respect to the ages of the patients and their spouses, duration of infertility, BMI, the proportion of nullipara, antral follicle count (AFC), basal endocrine profile and IVF/ICSI indications (Table 1).

In the letrozole group, 1 out of 46 cycles initiated (2.2%) was cancelled before ET, because no oocytes were obtained despite a follicular aspiration for oocyte retrieval. In the control group, 2 out of 57 cycles initiated (3.5%) was cancelled before oocyte retrieval due to no follicular development. There was no difference in cycle cancellation rate between the two groups (Table 2). Compared with the control group, the letrozole group had a significantly shorter duration of rhFSH administration for COS (P<0.001) and required significantly less total dose of rhFSH (P<0.001) (Table 2). The number of oocytes retrieved was significantly higher in the letrozole group of 5.3±2.0, compared with 4.3±1.9 in the control group (P<0.020). However, there were no significant differences in the numbers of mature oocytes, fertilized oocytes and grade I or II embryos between the two groups (Table 2). The clinical pregnancy rates per cycle initiated and per cycle ET, embryo implantation rate, miscarriage rate, and live birth rate per cycle initiated were comparable in the two groups (Table 2). No patients reported any systemic or local adverse effects attributed to the use of letrozole.

Approximately 9% to 30% of infertile women that undergo assisted reproductive technologies (ARTs) are poor responders with diminished ovarian reserve (DOR). Poor ovarian response to COS is still a major concern in IVF/ICSI treatment. Especially, the growing tendency of women to postpone childbearing until their late 30's or early 40's significantly decreases their chance of achieving a pregnancy and increases the incidence of poor ovarian response to COS. The most important objective in managing poor responders whose IVF cycles fail repeatedly is to increase ovarian sensitivity to gonadotropin. This is likely to improve pregnancy outcomes and reduce the high cost of IVF overall.

Recently, aromatase inhibitors were found to elevate follicular sensitivity to gonadotropin effectively, which makes them useful in the management of infertility [8,9,10,11,13]. Aromatase inhibitors inhibit the enzyme by competitively binding to the heme of the cytochrome P450 subunit. This blocks androgen conversion into estrogen, thus causing a temporary accumulation of intraovarian androgens [13]. Androgen promotes the initiation of primordial follicle growth and increases the number of growing preantral and small antral follicles in the primate ovary [14,15]. Howles et al. [16] also suggested that, in women with DOR who undergo ART treatment, boosting intraovarian androgens may increase the number of small antral follicle available to enter the recruitment stage as well as the process of follicle recruitment itself. In addition, androgen accumulation in the follicle may also stimulate insulin-like growth factor I, which may synergize with FSH to promote folliculogenesis [17,18]. Kim et al. [19] and Kim [20] investigated whether androgen treatment affects ovarian features. Their study demonstrated that both 3 weeks and 4 weeks transdermal testosterone gel pretreatment result in the significant increase of AFC and significant decreases of mean follicular diameter and resistance index value of ovarian stromal artery, compared with baseline values.

In 2002, Mitwally and Casper [8] reported the results of an observational cohort study that included women with unexplained infertility who were undergoing COS and intrauterine insemination. The subjects were divided into three groups that were assigned to receive the FSH protocol alone (FSH only group), letrozole plus the FSH protocol (letrozole/FSH group), or clomiphene citrate (CC) plus the FSH protocol (CC/FSH group). The letrozole/FSH and CC/FSH groups required significantly lower total FSH doses than the FSH only group. However, pregnancy rates were significantly higher in the letrozole/FSH and FSH only groups than in the CC/FSH group. Thus, they concluded that aromatase inhibition with letrozole reduces the FSH dose required for COS and does not have the undesirable antiestrogenic effects that are frequently seen in women who take CC [8]. The success of letrozole in increasing ovarian sensitivity to gonadotropins encouraged further exploration of the ability of letrozole to improve the response to ovarian stimulation in poor responders. Recent studies again found that the addition of letrozole improved the ovarian response to FSH and reduced the gonadotropin dose needed for COS in poor responders [9,10,11]. Goswami et al. [11] performed a randomized controlled single-blind trial to evaluate the effect of the adjunctive use of letrozole in 38 poor responders. In their study, subjects were randomly assigned to the GnRH agonist long protocol or the letrozole plus FSH protocol during IVF cycles. Their study showed that the letrozole/FSH group had comparable pregnancy outcomes with lower total dose of FSH required, compared with the GnRH agonist long protocol group [11].

In the present study, letrozole was incorporated in GnRH antagonist MDP during IVF/ICSI cycles and its efficacy was compared with the standard GnRH antagonist MDP. The addition of letrozole yielded comparable pregnancy outcome with significantly fewer dose and days of rhFSH required even in GnRH antagonist MDP. Although the criteria for GnRH antagonist initiation were same in both letrozole and control groups, duration of GnRH antagonist administration was significantly shorter in the letrozole group than in the control group. Letrozole increases ovarian sensitivity to gonadotropins, thereby improving ovarian response to rhFSH and significantly reducing the duration of rhFSH administration. This action of letrozole may contribute to decrease the duration of GnRH antagonist use in the letrozole group. Moreover, letrozole supplementation resulted in a significant increase of the number of oocytes retrieved. When considering that GnRH antagonist protocol is recently recognized as one of promising COS regimen for poor responders, our results are encouraging. Studies comparing letrozole/GnRH antagonist protocol with standard GnRH antagonist protocol are very limited. Only one randomized controlled trial (RCT) by Ozmen et al. [21] was found. The results of their study were similar to those of our retrospective cohort study. In both studies, the total dose of gondotropin required for COS was significantly lower in the letrozole group and clinical pregnancy rate per ET cycle was comparable between the letrozole and control groups. In the study by Ozmen et al. [21], rhFSH was administered with a fixed dose (450 IU/day) in both study and control groups. However, rhFSH was administered with a flexible manner in our study. In the present study, the starting dose of rhFSH was 250 IU/day which was much smaller than the daily dose of rhFSH used in the RCT by Ozmen et al. [21] and the dose of rhFSH was adjusted according to ovarian response, every 3 to 4 days. The total dose of rhFSH used in each letrozole and control groups in our study appeared to be lower, when compared with those in the study by Ozmen et al. [21]. It may be caused by the difference of gonadotropin administration method between their RCT and our study.

Our results suggest that GnRH antagonist MDP in which letrozole is incorporated may be more effective than standard GnRH antagonist MDP in poor responders undergoing IVF/ICSI. However, our study has a limitation to evaluate the efficacy of letrozole addition due to a small number of sample available and its retrospective nature. In the near future, well-designed prospective randomized trials are required to confirm the efficacy of letrozole as an adjuvant to gonadotropin in COS for poor responders.