Meta-analysis of letrozole versus clomiphene citrate in polycystic ovary syndrome

Reproductive BioMedicine Online (2011) 23, 91 96 www.sciencedirect.com www.rbmonline.com REVIEW Meta-analysis of letrozole versus clomiphene citrate in polycystic ovary syndrome Donghong He, Fengyan Jiang * Department of Gynaecology, First Affiliated Hospital, Guangxi Medical University, Nanning, Guangxi 530021, PR China * Corresponding author. E-mail address: drjiangfengyan@yahoo.com.cn (F Jiang). Fengyan Jiang: Graduated from Guangxi Medical University and was awarded Master Degree in 1994. Now she is a tutor for master of Guangxi Medical University and a chief physician of Department of gynaecology, First Affiliated Hospital, Guangxi Medical University, Nanning, Guangxi, PR China. Specific research interests is Gynecological Endocrinology. Abstract The aim of this study was to systematically compare the clinical efficacy and safety of letrozole with clomiphene citrate for ovulation induction in women with polycystic ovary syndrome (PCOS). The Cochrane Central Register of Controlled Trials, Pub- Med, EMbase, CBMdisc and CNKI were searched for eligible randomized controlled trials (RCT) comparing letrozole with clomiphene citrate in PCOS patients. Two reviewers independently extracted information and evaluated methodological quality according to the Cochrane Handbook 5.0. Meta-analysis was performed with the fixed-effects model or random-effects model according to the heterogeneity. Six eligible RCT involving 841 patients were included. Letrozole was associated with a number of lower mature follicles per cycle (standardized mean difference (SMD) 1.41; 95% confidence intervales (CI) 1.54 to 1.28; P < 0.00001) compared with clomiphene citrate. There were no significant differences in pregnancy rate (relative risk (RR) 0.97; 95% CI 0.79 to 1.18), abortion rate (RR 1.38; 95% CI 0.48 to 3.96) and multiple pregnancy rate (RR 0.34; 95% CI 0.07 to 1.72) between the two groups. The evidence from ovulation rates was not enough to support either letrozole or clomiphene citrate. In conclusion, letrozole is as effective as clomiphene citrate for ovulation induction in patients with PCOS. RBMOnline ª 2011, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. KEYWORDS: clomiphene citrate, letrozole, meta-analysis, polycystic ovary syndrome Introduction Polycystic ovary syndrome (PCOS) is an endocrinopathy in women of reproductive age, with morbidity ranging from 5% to 10% (Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2008). The primary clinical manifestation is irregular or absent menstrual cycles and infertility. Ovulation induction is recognized as an essential treatment procedure for PCOS. The traditional ovulation induction agent is clomiphene citrate which is highly effective in selected patients. However, it has some adverse effects, especially the anti-oestrogenic effects on the endometrium and cervix mucus that could prevent pregnancy. Although exogenous gonadotrophin is recommended as second-line intervention when clomiphene citrate fails, it is associated with increased chances of multiple pregnancy and ovarian 1472-6483/$ - see front matter ª 2011, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.rbmo.2011.03.024

92 D He, F Jiang hyperstimulation syndrome (OHSS), and intense monitoring of ovarian response is required. Therefore, a more effective and safe ovulatory agent is needed. In 2001, an aromatase inhibitor letrozole, which was used to treat and prevent breast cancer was first reported to be effective for ovulation induction and appeared to avoid the unfavourable effects of clomiphene citrate and gonadotrophin (Mitwally and Casper, 2001). In recent years many randomized controlled trials (RCT) have been conducted to assess the efficiency and safety of letrozole in PCOS, but conclusions have not been completely consistent. Furthermore, there has been no meta-analysis to provide definite evidence as to whether letrozole is superior to clomiphene citrate. Therefore this meta-analysis was performed to systematically evaluate the difference in letrozole and clomiphene citrate for PCOS on the basis of the available evidence. Materials and methods Search strategy This study systematically searched the Cochrane Central Register of Controlled Trials (CENTRAL, Issue 2, 2010), Pub- Med (January 1978 June 2010), Embase (January 1978 June 2010), CBMdisc (January 1970 June 2010) and CNKI (January 1979 June 2010) to identify studies without language restriction. The following keywords were used: letrozole, aromatase inhibitors, clomiphene citrate, polycystic ovary syndrome, hyperandrogenism, anovulation, ovulation induction and randomized controlled trial. References of the retrieved articles were also screened manually. Full texts were obtained by contacting the author when it could not be obtained online. Study selection Citations selected from this initial search were subsequently screened for eligibility using the following inclusion criteria: (i) an RCT; (ii) infertile patients with PCOS, diagnosed as fulfilling the Rotterdam 2003 criteria (Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2004), by two of the following three features: oligo- or anovulation, clinical and/or biochemical signs of hyperandrogenism and polycystic ovaries; (iii) letrozole versus clomiphene citrate for ovulation induction; and (iv) the outcome included at least one of the following: ovulation rate, pregnancy rate, endometrial thickness at human chorionic gonadotrophin (HCG), mature follicles, abortion rate, multiple pregnancy rate and OHSS rate. Studies were excluded if they included patients with another known cause of infertility or were unable to obtain adequate details of study methodology or results from the article or study investigators. Quality assessment Two reviewers independently assessed each included study for the six specific domains (sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting and other issues ) based on the recommended approach in the Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.2 (Higgins and Green, 2009). Doubts were solved by consensus after re-review of the publications. Data collection The data, which were collected independently by the two reviewers, included first author, publication year, country, baseline characteristics of the patients (number, age, race, body mass index), the number of withdrawals and dropouts and the reasons, and clinical outcomes. Clinical outcomes The clinical outcomes were determined as follows. For ovulation, blood samples for the measurement of progesterone concentrations were obtained on cycle days 21 23 and 3 ng/ml was considered as a marker for ovulation. For pregnancy, serum HCG was determined 2 weeks after HCG injection in the absence of menstruation for diagnosis of biochemical pregnancy and ultrasound was performed 4 weeks after a positive pregnancy test to confirm clinical pregnancy by fetal cardiac activity and gestational sac. Endometrial thickness was measured on the same day as intramuscular HCG, at the dose of 5000 IU or 10,000 IU to trigger ovulation when at least one mature follicle was developed. Also recorded were the number of mature follicles (follicles with diameter 18 mm), multiple pregnancies (number of gestational sacs 2) and abortions (pregnancy loss after clinical pregnancy). Statistical analysis The data analysis was performed using the fixed-effects model or random-effects model according to heterogeneity with Review Manager version 5.0 (The Nordic Cochrane Centre, The Cochrane Collaboration, 2008, Copenhagen). Dichotomous outcomes were summarized by calculating the relative risk (RR) and continuous outcomes were analysed by calculating the standardized mean difference (SMD) for each study, both with 95% confidence intervals (CI). Heterogeneity between studies was explored by the chi-squared test, with P 0.10 indicating significant heterogeneity. If any heterogeneity existed, the following techniques were employed to explain it: (i) random-effects model; (ii) subgroup analysis; and (iii) sensitivity analysis performed by excluding the trials which potentially biased the results. Funnel plots were used to test publication bias. Sample size calculation for a RCT was performed using the program G*power 3.1 (Faul et al., 2007). Results Of 1530 articles screened, 27 RCT were identified as potentially eligible for inclusion. Only six RCT (Atay et al., 2006; Aygen et al., 2007; Badawy et al., 2009; Bayar et al., 2006; Dehbashi, 2009; Zeinalzadeh et al., 2010) fulfilled the criteria in the meta-analysis. A total of 21 RCT were excluded for

Letrozole versus clomiphene citrate in PCOS 93 Potentially relevant trials identified and screened for retrieval (n = 1530) Trials retrieved for more detailed evaluation (n = 27) Potentially appropriate trials to be included in meta-analysis (n = 17) Trials included in meta-analysis (n = 6) Not randomized controlled trials (n = 1503) Not comparing letrozole with CC (n = 10) Improper randomization method (n = 7) Diagnosis did not fulfill the inclusion criteria for PCOS (n = 3) CC used before recruitment into the study (n = 1) Figure 1 Process of study selection of the randomized controlled trials included in this study. CC = clomiphene citrate; PCOS = polycystic ovary syndrome. the following reasons: the patients did not fulfill the diagnostic criteria of PCOS (three); trials did not compare letrozole with clomiphene citrate (10); use of improper randomization methods (seven); and clomiphene citrate used before recruitment into the study (one). Figure 1 shows the search process. Of the six eligible studies, five were performed in Asia and one in Africa. Funnel plots showed no suggestion of asymmetry due to publication bias. A total of 841 participants (1617 cycles) were enrolled, of which 414 (813 cycles) belonged to the letrozole group and 427 (804 cycles) to the clomiphene citrate group. In all studies, both drugs were administered during days 3 7 of a menstrual cycle. Follicular diameter was determined by calculating the mean of two perpendicular diameters measured at the largest plane of the follicle. Intramuscular HCG at the dose of 5000 10,000 IU was administered to trigger ovulation when at least one mature follicle had developed followed by timed intercourse. The baseline of all trials was comparable. The characteristics of the included studies are listed in Table 1. The process of randomization was adequate in four studies (Badawy et al., 2009; Bayar et al., 2006; Dehbashi, 2009; Zeinalzadeh et al., 2010) and was unclear in two studies (Atay et al., 2006; Aygen et al., 2007). Allocation sequence concealment was carried out and reported in only two studies (Bayar et al., 2006; Dehbashi, 2009) and was unclear in the remainder. The researchers and participants were clearly masked to the intervention in two studies (Bayar et al., 2006; Dehbashi, 2009), while in most of the remaining cases masking was unclear. There were dropouts in only one study (Bayar et al., 2006), but the reason was not reported. Results of quality assessment can be seen in Figure 2. The results of all meta-analyses are detailed in Table 2. Ovulation rate per cycle Information on ovulation rate per cycle was available in all RCT. The pooled analysis showed a significant heterogeneity (chi-squared = 16.90; P = 0.005). With no subgroups described in the studies, sensitivity analysis was performed. According to the sensitivity analysis, two studies (Badawy et al., 2009; Bayar et al., 2006) were excluded leaving a total of 360 cycles (174 in the letrozole group and 186 in the clomiphene citrate group). Analysis of the included studies showed a significant difference in ovulation rate per cycle between the two groups Table 1 Characteristics of included studies. Publication Country Interventions Patients (n) Cycles (n) Outcome measures Quality assessment Atay et al. (2006) Turkey 2.5 mg LE 51 51 Ovulation rate, pregnancy rate, multiple B 100 mg CC 55 55 pregnancy, endometrial thickness, mature follicles Aygen et al. (2007) Turkey 2.5 mg LE 5 23 Ovulation rate, pregnancy rate, B 100 mg CC 5 24 endometrial thickness Bayar et al. (2006) Turkey 2.5 mg LE 40 99 Ovulation rate, pregnancy rate, abortion B 100 mg CC 40 95 rate, endometrial thickness Badawy et al. (2009) Egypt 5 mg LE 218 540 Ovulation rate, pregnancy rate, B 100 mg CC 220 523 endometrial thickness, mature follicles, abortion rate, multiple pregnancy, OHSS Dehbashi (2009) Iran 5 mg LE 50 50 Ovulation rate, pregnancy rate, multiple A 100 mg CC 50 50 pregnancy, endometrial thickness, mature follicles, abortion rate Zeinalzadeh et al. (2010) Iran 5 mg LE 50 50 Ovulation rate, pregnancy rate, OHSS B 100 mg CC 57 57 All treatments were for 5 days. CC = clomiphene citrate; LE = letrozole; OHSS = ovarian hyperstimulation syndrome. A The result of every quality evaluation item is yes. B At least one result of the quality evaluation item is unclear and none is no.

94 D He, F Jiang Adequate sequence generation Atay 2006 Aygen 2007 Badawy 2009 Allocation concealment Blinding Incomplete outcome data addressed Free of selective reporting Free of other bias Badawy et al., 2009; Bayar et al., 2006; Dehbashi, 2009). A greater endometrial thickness was detected in the letrozole group compared with the clomiphene citrate group in the study by Atay et al. (2006) (8.4 ± 0.18 mm versus 5.2 ± 0.12 mm; P < 0.05) and by Aygen et al. (2007) (10.4 ± 1.4 mm versus 6.8 ± 0.5 mm; P < 0.05). However, the result in the study by Badawy et al. (2009) was the opposite, the endometrial thickness being greater in the women of the clomiphene citrate group (8.1 ± 0.2 mm versus 9.2 ± 0.7 mm; P = 0.021). In the other two studies (Bayar et al., 2006; Dehbashi, 2009), no significant difference was found between the two groups. Pooled analysis was not performed due to the significant heterogeneity (chi-squared = 274.87; P < 0.00001). Mature follicles per cycle (RR 1.29; 95% CI 1.12 to 1.49; P = 0.0006; heterogeneity chi-squared = 5.00). However, when the two excluded studies were included in the sensitivity analysis, the difference was not significant (RR 1.01; 95% CI 0.94 to 1.07). Pregnancy rate Bayar 2006 Dehbashi 2009 Zeinalzadeh 2010 Figure 2 Risk of bias item for each included study. = Yes; = Unclear. All RCT reported pregnancy rate. Of the 841 enrolled (414 in the letrozole group and 427 in the clomiphene citrate group), there was no heterogeneity among the studies (chi-squared = 6.51). The pooled analysis with fixed-effects model showed no significant difference in pregnancy rate between the two groups (RR 0.97; 95% CI 0.79 to 1.18) (Figure 3). Endometrial thickness at HCG Five studies reported the endometrial thickness on the day of HCG administration (Atay et al., 2006; Aygen et al., 2007; Mature follicles per cycle were reported in three studies (Atay et al., 2006; Badawy et al., 2009; Dehbashi, 2009). Significant heterogeneity (chi-squared = 40.22; P < 0.00001) among the studies was found. After excluding one study (Dehbashi, 2009) by sensitivity analysis, a total of 1169 cycles (591 in the letrozole group and 578 in the clomiphene citrate group) were included in the analysis. The pooled analysis with the fixed-effects model demonstrated fewer mature follicles in the letrozole group than in the clomiphene citrate group (SMD 1.41; 95% CI 1.54 to 1.28; P < 0.00001; heterogeneity chi-squared = 0.06). Multiple pregnancy rate There were three studies (Atay et al., 2006; Badawy et al., 2009; Dehbashi, 2009) that noted multiple pregnancy rate on 644 patients (319 in the letrozole group and 325 in the clomiphene citrate group). No heterogeneity (chi-squared = 0.54) was found. The pooled analysis with the fixed-effects model showed no significant difference in the multiple pregnancy rate between the two groups (RR 0.34; 95% CI 0.07 to 1.72). Abortion rate Three studies (Badawy et al., 2009; Bayar et al., 2006; Dehbashi, 2009) reported the abortion rate on 618 patients Table 2 Outcome Methods and results of the meta-analysis. Numbers of participant (cycles) Statistical method RR/SMD 95% CI P-Value Heterogeneity (chi-squared) Ovulation 323 (360) Odds ratio (M H, fixed, 95% CI) 1.29 1.12 to 1.49 0.0006 5.00 Pregnancy 841 (1617) Odds ratio (M H, fixed, 95% CI) 0.97 0.79 to 1.18 NS 6.51 Mature follicles 544 (1169) SMD (I V, fixed, 95% CI) 1.41 1.54 to 1.28 <0.00001 0.06 Multiple pregnancy 644 (1269) Odds ratio (M H, fixed, 95% CI) 0.34 0.07 to 1.72 NS 0.54 Abortion 618 (1357) Odds ratio (M H, fixed, 95% CI) 1.38 0.48 to 3.96 NS 0.22 All chi-squared tests were NS. CI = confidence interval; M H = Mantel Haenszel; I V = Inverse-Variance; NS = not statistically significant; RR = relative risk; SMD = standardized mean difference.

Letrozole versus clomiphene citrate in PCOS 95 Study or Subgroup Atay 2006 Aygen 2007 Badawy 2009 Bayar 2006 Dehbashi 2009 Zeinalzadeh 2010 LE CC Risk Ratio Risk Ratio Events 5 2 82 9 13 10 Total 51 5 218 40 50 50 Events 11 1 94 7 7 8 Total 55 5 220 40 50 57 Weight 8.4% 0.8% 73.9% 5.5% 5.5% 5.9% M-H, Fixed, 95% CI M-H, Fixed, 95% CI 0.49 [0.18, 1.31] 2.00 [0.26, 15.62] 0.88 [0.70, 1.11] 1.29 [0.53, 3.12] 1.86 [0.81, 4.26] 1.43 [0.61, 3.33] Total (95% CI) 414 427 Total events 121 128 Heterogeneity: Chi² = 6.51, df = 5 (P = 0.26); I² = 23% Test for overall effect: Z = 0.34 (P = 0.73) 100.0% 0.97 [0.79, 1.18] 0.01 0.1 1 10 100 Favours experimental Favours control Figure 3 Forest plot of pregnancy rate associated with comparison of letrozole (LE) with clomiphene citrate (CC). CI=confidence intervals; M H=Mantel-Haenszal. (308 in the letrozole group and 310 in the clomiphene citrate group). No heterogeneity (chi-squared = 0.22) was found. The pooled analysis with the fixed-effects model showed no significant difference in the abortion rate between the two groups (RR 1.38; 95% CI 0.48 to 3.96). Sample size calculation This study performed a sample size calculation using the program G*Power 3.1, using an alpha of 0.05 with 80% power (b = 0.2) and a large effect size, d = 0.8 (t-test) or w = 0.5 (chi-squared test). Based on the results, the smallest sample size required for a proper RCT was n = 52 for continuous outcomes and n = 32 for dichotomous outcomes. Most of the studies included in this meta-analysis had the eligible sample size (except for Aygen et al., 2007). Discussion Ovulation induction is consistently regarded as an essential treatment procedure for PCOS patients. Although clomiphene citrate is still the traditional agent used for inducing ovulation in this condition, clomiphene resistance, which refers to persistence of anovulation after standard clomiphene citrate therapy, occurs in 20 25% of patients (Hughes et al., 2000). Aromatase inhibitors are agents that suppress the biosynthesis of oestrogen and, therefore, reduce the negative feedback effect on the hypothalamic pituitary system. This results in increased secretion of FSH that can lead to follicle selection and maturation. The third-generation compound, letrozole, has been recently used for ovulation induction in PCOS women resistant to clomiphene citrate or with inadequate endometrial thickness during clomiphene citrate treatment. Many clinical studies have indicated that letrozole treatment gives better ovulation and pregnancy rates in comparison to clomiphene citrate in patients with PCOS (Begum et al., 2009; Ganesh et al., 2009; Wang et al., 2008). However, a systematic review of the efficacy and safety of letrozole is important. This meta-analysis showed a higher ovulation rate per cycle in the letrozole group compared with the clomiphene citrate group. Two RCT were excluded due to significant heterogeneity before pooled analysis: one (Bayar et al., 2006) had dropouts of 10% and intention-to-treat analysis was not conducted and the other (Badawy et al., 2009) had unclear allocation concealment and blinding (the author was contacted to seek clarification but no response was received). But the weighting coefficients of these two studies were larger than the others because of the bigger sample size and the result of the meta-analysis was not consistent when tested by sensitivity analysis. Consequently, the current evidence was not sufficient to support either letrozole or clomiphene citrate as being superior in terms of ovulation rate. Because letrozole has no oestrogen-like effects, mono-ovulation was expected to occur in most cases, as a result of which a lower rate of multiple pregnancies and OHSS may be expected. In this meta-analysis, fewer mature follicles per cycle were observed in the letrozole group, yet the multiple pregnancy rate was similar in the two groups, which maybe related to the shortage of included studies, as well as the OHSS rate. On the other hand, Begum et al. (2009) assumed that, because of the much shorter life span (2 days) and absence of anti-oestrogenic effects in the late follicular phase, oestradiol concentration increased and consequently the endometrium grew faster, and thus a greater endometrial thickness on the day of HCG administration, in favour of nidation, and a higher pregnancy rate was to be expected theoretically (Zhang et al., 2007). In the current study, a greater endometrial thickness was detected in the letrozole group compared with the clomiphene citrate group in two studies (Atay et al., 2006; Aygen et al., 2007). However, the result in Badawy et al. (2009) was the opposite: the endometrial thickness was greater in the clomiphene citrate group. In addition, the sample size of the trial by Badawy et al. was much bigger than the former two. Finally, the efficacy of an agent for inducing ovulation should be evaluated by the pregnancy rate. All studies reported similar pregnancy rates in two groups and these data were supported by the meta-analysis. It is thought that this is probably due to the lower number of mature follicles per cycle found in the letrozole group. There was no significant difference in the abortion rate in the two groups in the meta-analysis. Whether letrozole has harmful or teratogenic effects on the fetus was not reported in any of the included studies. An animal study

96 D He, F Jiang found that letrozole has toxic effects on prenatal development in rats (Tiboni et al., 2008). But so far, no embryonic or fetal malformations have been reported in humans. A study published in 2006 examined 514 babies born to mothers who had used letrozole to conceive and compared them with 397 babies born to mothers who had conceived using clomiphene citrate (Tulandi et al., 2006). There were no increased rates of major and minor malformations in the letrozole group compared with the clomiphene citrate group. Additionally, the number of cardiac anomalies in the letrozole group was slightly lower than in the clomiphene citrate group. As a result, letrozole can be seen as a safe agent for ovulation induction like clomiphene citrate. This study provides, as far as is known for the first time, quantitative estimates of the efficiency of letrozole in ovulation induction for PCOS. The methodology used in this meta-analysis was rigorous because all eligible studies were prospective RCT; the quality level of all the studies is A or B (Table 1). The baseline characteristics of the patients in the RCT were largely comparable, which suggested that the patient population was representative. The inclusion criteria for all PCOS patients enrolled in the RCT fulfilled the Rotterdam 2003 criteria, thus certain bias in the individual trials was controlled. This meta-analysis has several limitations. First, most trials were conducted in Asia. Second, the dose of letrozole administered was not completely the same in all trials, which could influence the results of the meta-analysis. Third, it is possible that studies with negative results, which showed no trend in favour of either intervention, may remain unpublished, leading to publication bias. Finally, one study (Aygen et al., 2007) was really small. Based on sample size calculation, the current study recommends that at least 52 participants are included for a proper RCT to compare both drugs. In conclusion, although there are several limitations in this meta-analysis, it suggests that letrozole may be as effective as clomiphene citrate for ovulation induction in patients with PCOS. Properly designed randomized trials are essential to confirm whether letrozole can be introduced as a first-line treatment with PCOS. References Atay, V.C., Muhcu, M., Cam, M., Karateke, A., 2006. Comparison of letrozole and clomiphene citrate in women with polycystic ovaries undergoing ovarian stimulation. J. Int. Med. Res. 34, 73 76. Aygen, E.M., Guzel, Z., Ozgun, T., Atakul, T., Sahin, Y., 2007. The use of letrozole for ovulation induction in infertile women with polycystic ovarian syndrome. Erciyes Tip Derg. 29, 195 200. Badawy, A., Abdel Aal, I., Abulatta, M., 2009. Clomiphene citrate or letrozole for ovulation induction in women with polycystic ovarian syndrome: a prospective randomized trial. Fertil. Steril. 92, 849 852. Bayar, U., Basaran, M., Kiran, S., Coskun, A., Gezer, S., 2006. Use of an aromatase inhibitor in patients with polycystic ovary syndrome: a prospective randomized trial. Fertil. Steril., 1447 1451. Begum, M.R., Ferdous, J., Begum, A., Quadir, E., 2009. Comparison of efficacy of aromatase inhibitor and clomiphene citrate in induction of ovulation in polycystic ovarian syndrome. Fertil. Steril., 853 857. Dehbashi, S.K., 2009. Comparison of the effects of letrozole and clomiphene citrate on ovulation and pregnancy rate in patients with polycystic ovary syndrome. Iran J. Med. Sci. 34, 23 28. Faul, F., Erdfelder, E., Lang, A.-G., Buchner, A., 2007. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav. Res. Methods 39, 175 191. Ganesh, A., Goswami, S.K., Chattopadhyay, R., Chaudhury, K., Chakravarty, B., 2009. Comparison of letrozole with continuous gonadotropins and clomiphene gonadotropin combination for ovulation induction in 1387 PCOS women after clomiphene citrate failure: a randomized prospective clinical trial. J. Assist. Reprod. Genet. 26, 19 24. Higgins, J.P.T., Green, S. (Eds.), Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.2. The Cochrane Collaboration, 2009. Available from: <www.cochrane-handbook.org> (updated September 2009). Hughes, E., Collins, J., Vandekerckhove, P., 2000. Clomiphene citrate for ovulation induction in women with oligo-amenorrhoea. Cochrane Database Syst. Rev., CD000056. Mitwally, F.M., Casper, R.F., 2001. Use of an aromatase inhibitor for induction of ovulation in patients with an inadequate response to clomiphene citrate. Fertil. Steril. 75, 305 309. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2004. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil. Steril. 81, 19 25. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2008. Consensus on infertility treatment related to polycystic ovary syndrome. Hum. Reprod. 23, 462 477. Tiboni, G.M., Marotta, F., Rossi, C., Giampietro, F., 2008. Effects of the aromatase inhibitor letrozole on in utero development in rats. Hum. Reprod. 23, 1719 1723. Tulandi, T.M.J., Al-Fadhli, R., Kabli, N., Forman, R., Hitkari, J., Librach, C., Greenblatt, E., Casper, R.F., 2006. Congenital malformations among 911 newborns conceived after infertility treatment with letrozole or clomiphene citrate. Fertil. Steril. 85, 1761 1765 (Epub 2006 May 2). Wang, L., Lu, S., Cao, Z., Li, L., 2008. Effects of letrozole and clomiphene citrate on ovulation induction in polycystic ovarian syndrome. Acad. J. Xian Jiaotong Univ. 29, 699 701. Zeinalzadeh, M., Basirat, Z., Esmailpour, M., 2010. Efficacy of letrozole in ovulation induction compared to that of clomiphene citrate in patients with polycystic ovarian syndrome. J. Reprod. Med. 55, 36 40. Zhang, H.W., Ll, L., Yu, L.Q., 2007. Effects of letrozole versus clomiphene citrate on endometrial receptivity in mouse. Shengzhi Yu Biyun 27, 251 254. Declaration: The authors report no financial or commercial conflicts of interest. Received 14 December 2010; refereed 22 March 2011; accepted 24 March 2011.