Advertisement

Co-treatment with letrozole during ovarian stimulation for IVF/ICSI – a systematic review and meta-analysis

Published:December 21, 2021DOI:https://doi.org/10.1016/j.rbmo.2021.12.006

      ABSTRACT

      Letrozole reduces serum oestradiol by inhibiting the aromatase enzyme and has growing clinical indications in fertility. The available evidence of letrozole's role in ovarian stimulation (OS) for IVF/ICSI was assessed concerning clinical outcomes. Medline, Cochrane, and ClinicalTrials.gov databases were systematically searched up until August 2021, including 31 studies (n=16 RCTs; n=15 observational studies). Live birth rate (LBR) in poor responders significantly increased by 7% (95% CI:[1%; 13%], P=0.03) with letrozole co-treatment. Concomitantly, the gonadotrophin consumption were significantly reduced, without decreasing the number of retrieved oocytes. In normal responders, number of oocytes increased with 1.8 oocytes (95% CI:[0.35;3.27], P=0.01) with letrozole co-treatment. However, no significant effect on LBR, clinical pregnancy rate (CPR), or OHSS rate was demonstrated. Only two studies reported on high responders and revealed no effect on LBR or CPR. Overall, the endometrium thickness, miscarriage rate, and cancellation rate were unaffected by letrozole co-treatment. None of the included studies reported on neonatal outcomes. The quality of evidence was high or moderate in the RCTs and low in the observational studies. In conclusion, poor responders may benefit from co-treatment with letrozole during OS for IVF, while letrozole for normal and high responders requires further investigation with larger, high-quality studies.

      Abbreviations:

      AFC (Antral follicle count), AMH (Anti-Müllerian hormone), CI (Confidence interval), CPR (Clinical pregnancy rate), FSH (Follicle-stimulating hormone), GnRH (Gonadotrophin-releasing hormone), hMG (Human menopausal gonadotrophin), IVF (In vitro fertilization), LBR (Live birth rate), LH (Luteinizing hormone), MD (Mean difference), OPR (Ongoing pregnancy rate), OS (Ovarian stimulation), PCOS (Polycystic ovary syndrome), POR (Poor ovarian responders), RCT (Randomized controlled trial), RD (Risk difference)
      INTRODUCTION
      Letrozole reduces circulating oestrogens by reversibly inhibiting the aromatization of androgens to oestrogens in the granulosa cells in the ovaries, without affecting oestrogen receptors in the peripheral tissue (Bossche et al., 1994; Buzdar et al., 2002).
      These pharmacological actions led to its early adoption as a chemotherapeutic agent in oestradiol receptor-positive breast tumours and for the past 20 years, it has assumed an increasing role in the treatment of subfertility. In 2001, Casper and Mitwally first proposed letrozole as an ovulation induction agent with the rationale that by suppressing oestradiol, it will diminish the negative feedback on the hypothalamic-pituitary axis, and enhance endogenous follicle stimulating hormone (FSH) release (Mitwally and Casper, 2001). A few years later, however, its emerging clinical use and further investigation of its applications were largely abandoned following the presentation of data linking letrozole fertility treatment to locomotor disabilities and congenital cardiovascular anomalies in offspring (Biljan et al., 2005). However, this study was later discredited because of several methodological flaws and was never published in a peer-reviewed journal. The concerns have since been alleviated by larger, well conducted studies showing no such association (Pundir et al., 2020). The role of letrozole has further developed and is now recommended by ESHRE and the international guideline consensus group as the first-line ovulation induction agent for anovulatory infertility, including patients with polycystic ovary syndrome (PCOS), despite its use is still awaiting approval by local medical authorities (Teede et al., 2018; Gadalla et al., 2020).
      The efficient anti-oestrogenic action has made letrozole an established co-treatment in ovarian stimulation (OS) for fertility preservation in breast cancer patients, where the supra-physiological oestrogen levels are suppressed to avoid stimulating tumour growth (Reddy and Oktay, 2012). In recent years, the interest in co-treatment with letrozole during OS for in vitro fertilization (IVF) has grown in fertility patients, aiming to normalize the disrupted endocrine milieu created by the multiple follicle development (Macklon et al., 2006). The supra-physiological levels of oestrogen in IVF have previously been shown to be suppressed to physiological levels with letrozole co-treatment (Oktay et al., 2006), this reduces the negative feedback at the hypothalamic-pituitary axis and increase endogenous gonadotrophin production. It has also been demonstrated that co-treatment with letrozole increases the precursor androgens that may serve to upregulate FSH receptors. These two factors present a rationale for the use of co-treatment with letrozole in poor ovarian responders (POR) (Weil et al., 1999; Nielsen et al., 2011; Lossl et al., 2020).
      Moreover, it has been hypothesized that restricting supra-physiological levels of oestrogen may reduce the risk of ovarian hyper-stimulation syndrome (OHSS). Published data are conflicting regarding co-treatment with letrozole during OS in the follicular phase, but a recent systematic review shows a reduction in moderate-severe OHSS in high responders, when using letrozole in the luteal phase (Zhao et al., 2020).
      Suppressing oestrogen levels during stimulation may affect endometrial receptivity by reducing the embryo-endometrial asynchrony thereby affecting implantation and miscarriage rates (van der Gaast et al., 2002; Bourgain and Devroey, 2003; Ma et al., 2003; Horcajadas et al., 2008; Boomsma et al., 2010; Sunkara et al., 2014). The prevention of supra-physiological oestrogen may also contribute to an earlier recovery of the pituitary suppression in the luteal phase, thus accounting for the need to provide luteal hormonal support in IVF cycles (Beckers et al., 2003). Conversely, disrupting the endocrine milieu could have detrimental effects on reproductive functions and clinical outcomes, and most published studies have been underpowered to detect these.
      The evidence supporting the role of letrozole, as an ovulation induction agent, has recently been assessed in a Cochrane systematic review (Franik et al., 2018). However, no systematic review and meta-analysis has investigated the efficacy of co-treatment with letrozole in IVF/ICSI regardless of the expected response. This systematic assessment aims to synthesize the evidence of co-treatment with letrozole on reproductive outcomes during OS for fresh IVF/ICSI cycles.
      MATERIALS AND METHODS
      This systematic review and meta-analysis was conducted and reported according to the PRISMA guidelines and were registered at PROSPERO, no: CRD42020194333 before the literature search was initiated (Shamseer et al., 2015). The study protocol is accessible at https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020194333.
      Search strategy
      A systematic search was conducted with help from a research librarian in PubMed/MEDLINE database with the string (aromatase inhibitor OR letrozole) AND (IVF OR infertility OR subfertility OR fertility), and at the Cochrane Library with: (letrozole). No limits or restrictions were applied to the searches and all published articles were screened for eligibility. To avoid publication bias, published protocols were searched in ClinicalTrials.gov with the string (condition: IVF OR ICSI OR ovarian stimulation; other terms: letrozole) with no limits. All authors with presumably ended unpublished trials were invited to share data if contact information was provided, and relevant studies were included. A thorough reference check was made in all included studies using the ‘snowballing method’ (Greenhalgh and Peacock, 2005). In all relevant systematic reviews, eligible studies from the reference list were included. The last systematic search was performed on August 18th, 2021.
      Selection of studies
      The search results were imported into Covidenceࣨ and duplicates removed. Covidenceࣨ is an online software program designed to manage studies in a systematic review with screening and data extraction. Title, summaries, and abstracts were screened for eligibility. NSB and MDH screened all summaries and abstracts independently, and relevant studies were included for full-text screening. Any disagreements were resolved by discussion, or by consulting additional reviewer(s).
      The full-text screening was performed independently by NSB and MDH. Any doubts or discrepancies were resolved in the study group with at least one other author.
      Criteria for inclusion: Co-treatment with letrozole in the follicular phase with gonadotrophins started simultaneously with letrozole, during gonadotrophin OS, or sequentially with gonadotrophins. Stimulation with gonadotrophins should be initiated 24 hours after letrozole had been stopped at the latest. Studies with random start OS and/or luteal phase OS were excluded. Studies investigating fertility preservation were excluded, as the primary outcomes for this study would not be available. To ensure comparable high-quality outcomes of the review, only randomized controlled trials, case-control studies, cohort study and observational studies with matched controls were included. Cases, case-series, abstracts without full text, reviews, opinions, and “letter to the editor” were excluded. Systematic reviews were screened for references, but not included for data extraction. Studies in other languages than English were excluded.
      All included studies were printed in two hard copies for individual data extraction, assessment of bias, and quality of evidence. The selection process is reported in the PRISMA flow chart (Figure 1). Further explanation of the excluded full-text studies is presented in Supplementary Table 1.
      Data extraction and outcome measures
      Following were extracted independently by two authors. Continuous outcomes were measured in mean and standard deviation and dichotomous outcomes in proportions. Primary outcomes: LBR, defined as LBR > gestational age (GA) 22 weeks; OPR, defined as ultrasound scan > GA 12 with foetal heartbeat; CPR, defined as ultrasound scan > GA 5-7 with foetal heartbeat or gestational sac; and miscarriage rate, defined as pregnancy loss > GA 5-22 weeks. Secondary outcomes: the number of oocytes retrieved, total dose of FSH, length of OS in days, endometrial thickness at the day of oocyte trigger, cancellation rate pr. started cycle, and neonatal outcome.
      If any doubts or questions regarding the data presented, or if data were presented as median instead of mean, the corresponding author of the relevant article was contacted by e-mail to obtain mean and standard deviation of continuous outcomes or proportion of dichotomous outcomes. If contact information was not valid, a search on Google, a relevant university/hospital or Facebook was conducted until a valid email address was found. If contact was not established or the right data was not available, the study was excluded from the meta-analysis with the specific outcomes.
      Assessment of bias
      The risk of bias of the included randomized controlled trials (RCTs) was assessed and reported according to “Cochrane collaboration's tool for assessing the risk of bias in randomized trials”, as recommended by the Cochrane Handbook for Systematic Reviews of interventions (Cumpston et al., 2019). The assessed bias consisted of (1) selection bias -random sequence generation, (2) -allocation concealment, (3) performance bias -blinding of participants and personnel, (4) detection bias -blinding of outcome assessors, (5) attrition bias -incomplete outcome data, (6) reporting bias -selective outcome reporting, (7) other potential sources of bias, and (8) overall bias. Bias was classified as low, high, or unclear risk of bias.
      The assessment of the risk of bias for observational studies was conducted using the Robins-I tool as low, moderate, serious, critical, or no information (Sterne et al., 2016). The following domains were applied: (1) confounding bias, (2) selection bias, (3) intervention bias, (4) deviations from intended interventions (5) information bias -missing data, (6) outcomes (7) selection of reported results, and (8) overall risk of bias.
      To compare bias in RCTs and observational studies, critical and serious were grouped as “serious”, moderate or no information as “unclear” and low as “low”. The overall risk of bias was assessed on the primary outcome. A standard operating procedure was described using the Robin-I tool before assessing the bias to ensure consistent assessment. All bias assessments were conducted independently by two authors and in case of discrepancy, it was discussed until agreement. The risk-of -bias assessments are presented next to the Forest plots and in Supplementary Table 2.
      Assessing the overall quality of EVIDENCE
      The overall quality of evidence was assessed using the GRADE criteria and was conducted independently by two authors and discussed until agreement as well (Schünemann et al., 2014). Four levels of quality were assessed: high, moderate, low, or very low quality. RCTs were graded as high quality but downgraded to a moderate level of evidence if high risk of bias was found. Observational studies were graded as a low grade of quality and downgraded to very low if containing ≥2 areas with high risk of bias and upgraded to moderate quality if no high risk of bias was found. If any disagreement the quality assessment was discussed until agreement. The assessments of overall quality of evidence are presented in Supplementary Table 2.
      Data synthesis and statistics
      Dichotomous variables were analysed using the Mantel-Haenszel method, reporting the risk differences (RD) with 95% confidence interval (CI). Continuous variables were analysed with the Inverse variance method and expressed as mean difference (MD) with a 95% CI. Heterogeneity across studies was evaluated using Forest plots and I2 statistics (Higgins, 2003). The random-effect model was applied, because of expected high heterogeneity within the studies. Funnel plots for each outcome were used to detect publication bias.
      Subgroup analyses on expected poor, normal and high responders were carried out for CPR, number of oocytes retrieved, total FSH consumption and days of OS. Poor, normal, and high responders were categorized according to the classification used in the specific studies.
      Sensitivity analyses were carried out to increase the homogeneity of the studies and decrease the bias in the studies included. Only high-quality studies (overall low risk to moderate risk of bias) and studies comparing similar stimulation protocols were included. They were carried out for the following outcomes: CPR, number of oocytes retrieved, and FSH consumption.
      Statistical analyses were conducted out using Review Manager, Version 5.4. (The Cochrane Collaboration, 2020).
      RESULTS
      The literature search yielded 1,126 studies after removing duplicates, of which 1,046 were excluded based on summary or abstract. A total of 80 studies were included for full-text screening of which 49 studies were excluded. The reasons for exclusion of the full-text articles are reported in Supplementary Table 1. This systematic review and meta-analysis included 31 studies (Figure 1: PRISMA Flow Diagram), 16 RCTs and 15 retrospective or prospective observational studies with appropriate control groups.
      Most of the studies included in the meta-analysis addressed the use of letrozole in POR (n=20) (Goswami et al., 2004; Garcia-Velasco et al., 2005; Schoolcraft et al., 2008; Ozmen et al., 2009; Davar et al., 2010; Elassar et al., 2011; Lee et al., 2011, 2014; Mohsen and El Din, 2013; Cenksoy et al., 2014; Lazer et al., 2014; Yucel et al., 2014; Nabati et al., 2015; Bastu et al., 2016; Ebrahimi et al., 2017; Yu et al., 2018; Moini et al., 2019; Liu et al., 2020; Shapira et al., 2020; Khojah et al., 2021). Ten studies investigated expected normal responders (Verpoest et al., 2006; Mukherjee et al., 2012; Yasa et al., 2013; Haas et al., 2017, 2018; Siristatidis et al., 2017; Eftekhar and Saeed, 2020; Kim et al., 2020; Shapira et al., 2020; Bülow et al., 2021), while only two studies included high responders (D'Amato et al., 2018; Yang et al., 2019). The number of patients in the studies ranged from 12 to 534. In one study (Shapira et al. 2020), the same patients (n=12) participated as control group and within six months in a subsequent cycle in the intervention group. Data were published from 13 different countries from 2004 to 2021, and 22 of the studies were published within the last 10 years. The characteristics of the included studies are presented in Table 1.
      Table 1Included studies
      Publication, Country of originStudy typeAuthors contactedPopulation, n(Letrozole+ Control)Inclusion criteriaOvarian stimulation protocolI = InterventionC= ControlsFertilizationTransfer day after OPU(Number of embryos)Measured OutcomesFundingGRADE
      Bastu et al. (2016), USARCTNoPoor responders

      64 (33+31)

      2/3 Bologna criteria

      Age: 18 -42 years, Cycle: 25-34 days, BMI: 19.3-28.9

      No metabolic/ endocrine disorders, normal hormones HSG/ hysteroscopy
      I: CD2/3 letrozole 5 mg/day (5 days)

      + 75 IU FSH + 75 IU hMG

      C: 150 IU FSH + 150 IU hMG
      ICSI

      Day 3

      (1/2)
      CPR, OR, FSH, Endo,

      NAModerate ⊕⊕⊕⊝
      Bulow et al.

      (2021), Denmark
      RCTYes, data included

      Normal

      Responders

      159 (80+79)
      AMH 8-32 pmol/l

      Regular menstrual cycle

      BMI<35

      Not PCOS
      I: CD2/3 letrozole 5 mg/day (5 days)

      + 150 U FSH

      C: Same protocol without letrozole
      IVF/ICSI

      Day2/3/5

      (1)
      CPR, OHSSEuropean Union, Interreg V ÖKS and Ferring Pharmaceuticals and Roche DiagnosticsHigh ⊕⊕⊕⊕
      Cenksoy et al. (2014), TurkeyRetro-spective cohortYesPoor responders

      153 (70+83)
      BMI ≤ 30

      AMH<1.1ng/mL and or

      Oocytter ≤3

      Primary infertility
      I: CD3 letrozole 2,5 mg/day (5 days)+

      450 IU FSH

      C: Pre-treatment 7 days OCP and Agonist + 450 IU FSH
      ICSI

      Day 3/5

      (1/2)
      CPR, OR, FSH, Endo,

      NAVery low ⊕⊝⊝⊝
      D'Amato et al. (2018), ItalyRetro-spective

      cohort
      Yes,

      no response
      High responders

      125 (80+45)
      Non-obese PCOS and mild protocol with letrozole and hMG OR standard IVF with hMG AND received metformin 1200-2000 mg/day in the preceding 3-6 monthsI: CD2 letrozole 2,5 mg/day (5 days) +

      150 IU hMG

      C: Same protocol without letrozole
      IVF/ICSI

      Day 3/5

      (NA)

      CPR,

      NAVery low ⊕⊝⊝⊝
      Davar et al. (2010), IranRCTNoPoor responders

      94 (45+49)
      ≥1 failed IVF cycle with ≤3 follicles ≤16 mm and/or E2 on trigger day ≤500 pg/mL.Pre-treatment 21 days of OCP

      I: Letrozole 5 mg/day (5 days)

      300-450 IU FSH or hMG

      C: Agonist + 300-450 IU FSH
      IVF/ICSI

      Day 2/3

      (NA)

      CPR, OR, FSH, Endo

      NAModerate ⊕⊕⊕⊝
      Ebrahimi et al. (2017), IranRCTNoPoor responders

      70 (35+35)
      2/3 Bologna criteria.

      Poor ovarian response in previous cycle: ≥ 1 previous failed IVF/ICSI cycle with conventional long-agonist protocol and < 4 mature oocytes.
      I: CD3 letrozole 2,5 mg/day (5 days) +

      225 IU FSH adjusted

      C: Same protocol without letrozole
      ICSI

      Day 2/3

      (NA)

      CPR, OR, FSH, Endo

      NAHigh ⊕⊕⊕⊕
      Ecemis et al. (2015), TurkeyRetro-spective cohortNoNormal

      responders

      320 (105+215)
      Age 18 – 40 years

      Cycle day 3 FSH < 15 IU/L

      I: CD 2 letrozole 5 mg (5 days) +

      150-225 IU FSH

      C: Same protocol without letrozole
      ICSI

      Day 3/5

      (NA)

      CPR, OR, FSH, EndoNAVery low ⊕⊝⊝⊝
      Eftekhar et al. (2020), IranRCTYes,

      no response
      Normal

      responders

      100 (50+50)
      Age 18-40 years

      AFC >7 and AMH 1.1-3.5 ng/mL

      I: CD2 letrozole 5 mg until trigger day +150 IU FSH

      C: Same protocol without letrozole
      IVF/ICSI

      Day 2/3

      (NA)
      EndoNALow ⊕⊕⊝⊝
      Elassar et al. 2011, USARetro-spective cohortNoPoor responders

      99 (47+52)
      Age <42 years

      ≥ 2 previous IVF with FSH ≥300 IU that yielded <5 oocytes or

      a previous cycle cancellation due to ≤3 follicles ≥15 mm in diameter, after 10 days of stimulation. No prior ovarian surgery or exposure to chemotherapy or radiotherapy.
      I: CD2 letrozole 5 mg (5 days) + CD5 450 IU FSH and 150 IU hMG

      C: Pretreatment with E2 and agonist. + CD2, antagonist 450 IU FSH and 150 IU hMG
      IVF/ICSI

      Day 3

      (NA)

      CPR, OPR, Misc., OR,FSH,

      Days
      NAVery low ⊕⊝⊝⊝
      Garcia-Velasco et al. (2005), SpainPro-

      spective cohort
      NoPoor responders

      147 (71+76)
      One previous cancelled

      IVF attempt (long protocol) with

      ≤4 follicle at OPU and/or

      E2 ≤500pg/mL and FSH <12 IU/L
      I: CD NA letrozole 2,5 mg +

      225 IU FSH +150 hMG adjusted

      C: Same protocol without letrozole
      NA

      Day 3

      (NA)

      CPR, Misc, OR, FSH, EndoNAVery low ⊕⊝⊝⊝
      Goswami et al. (2004), IndiaRCTNoPoor responders

      38 (13+25)
      Age ≥ 35 years

      1-3 previous failed IVF due to poor ovarian response
      I: CD 3 letrozole 2,5 mg/day (5 days)

      75 IU FSH only on CD 3 and 8.

      C: Agonist + 300 IU FSH adjusted
      IVF

      Day 2

      (NA)
      CPR, OR, FSH, EndoNAHigh ⊕⊕⊕⊕
      Haas et al. (2018), CanadaPro-

      spective cohort
      NoNormal responders

      26 (13+13)
      Not specified

      Not PCOS or fulfilling Bologna criteria
      I: CD3 letrozole 5 mg/day until trigger + FSH and from day of leading follicle ≥ 13 mm/or E2 levels exceeded 1200 pmol/L supplemented with hMG or LH.

      C: Same protocol without letrozole
      NA

      Day NA

      (NA)

      OPR, OHSS, OR, FSHNAVery low ⊕⊝⊝⊝
      Haas et al. (2017), CanadaRetro-

      spective cohort
      NoNormal responders

      174 (87+87)
      Age ≤ 42 years

      AFC ≥ 4
      I:CD3 letrozole 5 mg/day until trigger +

      FSH flexible

      C: FSH ± 75 IU as matched intervention
      NA

      Day NA

      (NA)

      OPR, OR, FSHNAVery low ⊕⊝⊝⊝
      Khojah et al. (2021), CanadaRetro-spective cohortNoPoor responders

      309 (62 + 247)
      Age <42 years

      Regular menstruation cycle

      and 1-2 previous IVF cycles with 0-2 mature follicles
      I: CD3 letrozole 20 mg/day until trigger + hMG 150 IU flexible up to 300 IU

      C: FSH 300-450 IU
      IVF/ICSI

      Day 3

      (1/2)
      CRP, FSH, Days, EndoNAVery low ⊕⊝⊝⊝
      Kim et al. (2020), KoreaRetro-spective cohortYes,

      no response
      Normal responders

      64 (38 + 26)
      Endometriosis diagnosed by biopsy at endometrioma surgery OR endometrioma seen on ultrasound

      Baseline mean AMH 18 pmol/L
      I: CD3 letrozole 5 mg/day until trigger +

      FSH flexible or FSH and LH

      C: Same protocol without letrozole
      NA

      Day 3

      (NA)
      CRP, FSH, EndoSNUBH research fundVery low ⊕⊝⊝⊝
      Lazer et al. (2014), CanadaRetro-spective cohortNoPoor responders

      141 (70+71)
      Bologna criteria and

      AMH ≤8 pmol/L and/or

      AFC ≤5

      Only 1st IVF cycle
      I: CD2 letrozole 2,5 mg/day (5 days) +

      150 IU hMG for 3 days and thereafter 150-225 IU hMG

      C: Same protocol without letrozole
      NA

      Day 3

      (NA)
      LBR, CPR, OR, FSHNAVery low ⊕⊝⊝⊝
      Lee et al. (2014), KoreaRetro-spective cohortNoPoor responders

      103 (46+57)
      Age 20-39 years

      Bologna criteria and

      Menstrual cycle: 24 – 35 days

      BMI: 18-25 kg/m2

      I: CD2 letrozole 2,5 mg (5 days) +

      225 IU rhFSH

      C: Same protocol without letrozole
      IVF/ICSI

      Day 3

      (1-4)

      LBR, CPR,OPR, Misc, OR, FSHNAModerate ⊕⊕⊕⊝
      Lee et al. (2011), ChinaRCTYes,

      data received on CPR and number of oocytes
      Poor responders

      53 (26+27)
      <4 oocytes in a previous failed IVF cycles OR AFC < 5.I: CD2 letrozole 2,5 mg (5 days) + CD7 225 IU hMG, day after letrozole was stopped.

      C: CD3 225 IU hMG
      IVF

      Day 2

      (1/2)
      LBT, OPR,Misc, OR, FSH

      Hong Kong

      OG

      Trust fund
      High ⊕⊕⊕⊕
      Liu et al. (2020), ChinaRCTYes,

      data received on FSH
      Poor responders

      191 (97 + 94)
      Age ≥40 years and/or other risk factor for POR

      Previous cycle with ≤6 oocytes

      AFC<7 or AMH<1.1ng/ml

      Bologna criteria
      I:CD3 letrozole 5 mg (5 days) + CD4 and 6: 150 IU FSH

      CD8 150 FSH until trigger

      C: Agonist protocol

      CD start: NA, 300 IU FSH until trigger
      IVF/ICSI

      Day 3

      (1/2)

      LBR, CPR, OPR, Misc, OR, FSH

      NAModerate ⊕⊕⊕⊝
      Mohsen et al. (2013), EgyptRCTNoPoor responders

      60 (30+30)
      Regular cycle, BMI <30 AND

      ≥1 cycle with ≤4 oocytes and 300FSH
      I: CD2 letrozole 5 mg (5 days) + CD7 150 IU hMG

      C: Agonist + 300 IU hMG
      ICSI

      Day 2/3

      (NA)
      CPR, OR, FSH, Endo

      NAModerate ⊕⊕⊕⊝
      Moini et al. (2019), IranRCTNoPoor responders

      160 (80+80)
      2/3 Bologna criteria:

      ≤ 3 oocytes retrieved; ≥ 40 years/other risk factor for POR

      AFC < 5-7 follicles/AMH < 0.5-1.1 ng/mL

      BMI < 30 kg/m2
      I: CD2/3 letrozole 5 mg (5 days) + CD2/3 150 IU FSH and 150 IU hMG adjusted

      C: Same protocol without letrozole
      ICSI

      Day 2/3

      (NA)

      CPR, OR, FSH, Endo

      NAHigh ⊕⊕⊕⊕
      Mukherjee et al. (2012), IndiaRCTNoNormal responders

      94 (42+52)
      Age 25-35 years

      Partner with azoospermia needing TESA

      Normal endocrinology

      thyroid, prolaktin, gonadotropins.
      I: CD3 letrozole 5 mg (5 days) +

      CD5 75 IU FSH

      C:CD2 150-225 IU FSH
      ICSI

      Day 2

      (2)
      CPR, Misc, OHSS, OR, FSH, EndoNAModerate ⊕⊕⊕⊝
      Nabati et al. (2015), IranRCTNoPoor responders

      123 (62+61)
      Age >40 years

      Previously cycle ≤3 oocytes

      Cycle day 3: FSH 15-110mIU/m and E2 <1500pg/m
      I: CD3 letrozole 10 mg (5 days) + CD3 450 IU FSH

      C: Pre-treatment: OCP

      3 days after OCP start agonist+ 5 days after OCP: 450 IU FSH, adjusted
      ICSI

      Day 1

      (1-3)
      CPR, OR, FSH, Endo

      Mazandaran University

      of medical sciences
      Low ⊕⊕⊝⊝
      Özmen et al. (2009), TurkeyRCTNoPoor responders

      70 (35+35)
      Previous cycle <4 oocytes or

      Previous cycle cancellation due to low E2 on CD6 or trigger day.
      I:CD3 letrozole 5 mg (5 days) +

      CD5 450 IU FSH

      C: Same protocol without letrozole
      ICSI

      Day 3

      (NA)
      CPR, OR, FSH, Endo

      NAModerate ⊕⊕⊕⊝
      Schoolcraft et al. (2008), USAPro-

      spective cohort
      Yes,

      no response
      Poor responders

      534 (179+355)
      Age >41 years or

      CD2 FSH > 10 IU/mL or

      AFC < 6 or

      Previous cycle cancellation

      Previous poor response to COH (peak E2 < 500 pg/mL and/or < 6 oocytes retrieved)
      I: CD3 letrozole 2,5 mg (5 days) + CD3 300 IU FSH + 150 IU hMG

      C: Pre-treatment: 14-21 days of OCP, 3 days after OCP start agonist

      5 days after OCP, FSH 300 IU and

      hMG 150 IU per day.
      IVF/ICSI

      Day 3/5

      (NA)

      OR, FSH

      NAVery low ⊕⊝⊝⊝
      Shapira et al. (2020), IsraelRetro- spective

      cohort
      NoPoor and suboptimal responders

      24 (24+24)
      Treated within a one-year period

      Poor (1-3 oocytes) or suboptimal (4-9 oocytes) response to treatment

      Subsequent cycle same patients with letrozole.
      I: CD2-3 letrozole 5 mg until trigger + 300-450 rFSH

      C: Same protocol without letrozole
      IVF

      NA

      (NA)
      FSH, OR, DaysNo fundingVery low ⊕⊝⊝⊝
      Verpoest et al. (2006),

      Belgium
      RCTNoNormal responder

      20 (10+10)
      Age < 39 years

      Subfertility > 1 year

      Regular menstrual cycle

      Basal FSH <15 IU/L

      1 or 2nd IVF/ICS

      Not WHO ovulation disorder
      I:CD2 letrozole 2,5 mg (5 days) + CD2 150 IU FSH

      C: Same protocol without letrozole
      NA

      Day 3/5

      (NA)

      CPR, OR, FSHNAModerate ⊕⊕⊕⊝
      Yang et al. (2019),

      China
      RCTYes,

      no response
      High responders

      130 (65+65)
      Age 21-35 years

      AFC 15-23

      1rst or 2nd IVF cycle

      BMI 18-28 kg/m2
      I: Agonist, CD NA letrozole 2,5 mg

      100-225 IU FSH

      C: Same protocol without letrozole
      ICSI

      Day 3

      (1/2)

      LBR, CPR, Misc

      NoModerate ⊕⊕⊕⊝
      Yasa et al. (2013), TurkeyPro-

      spective

      cohort
      NoNormal responders

      50 (25+25)
      Age ≤ 40 years

      1rst IVF cycle

      BMI ≤3 0, ≥ 2 years of infertility.

      Male factor, unexplained, ovarian factor or tuboperitoneal factor infertility
      I: CD2 letrozole 2,5 mg (5 days) +

      CD 3 150-225 IU, flexible

      C: Same protocol without letrozole
      IVF

      Day 3

      (1-3)

      CPR, OPR, OHSS, OR, FSH, Endo

      NAVery low ⊕⊝⊝⊝
      Yu et al. (2018), ChinaRCTNoPoor responders

      112 (52+60)
      Age ≤42

      AFC ≤ 8 and AMH ≤ 1.5ng/ml and

      FSH ≥15 and

      BMI < 23 and

      Tuba factor or male factor
      I: CD3 letrozole 5 mg (5 days) +

      hMG 75 IU from CD8

      C: hMG 150 IU, adjusted
      IVF/ICSI

      Day 3

      (1-3)

      LBR, CPR, OR, FSH, days, EndoNatural science foundation of China, Zhejiang Provincial Natural science foundation,

      Wenzhou Science and Technology Bureau Foundation
      Low ⊕⊕⊝⊝
      Yucel et al. (2014), TurkeyRetro-spective cohortYes,

      no response
      Poor responders

      121 (42+79)
      2/3 Bologna criteria

      ≥ 40 year

      ≤ 3 oocytes retrieved /cycle cancellation

      AFC < 6 / AMH < 0.5-1.1 ng/mL
      I: Pretreatment E2 for 2 weeks added P for 2 week.

      4 days pause start letrozole 2.5mg (5 days)

      2 days after letrozole + FSH 200IU& hMG 150 IU

      C: OCP for 21 days, 2days pause start similar stimulation and agonist the following day
      ICSI

      Day 3

      (NA)

      OR, FSH

      NAVery low ⊕⊝⊝⊝
      E2: Estradiol
      LBR: Live birth rate/cycle
      LH: Luteinizing hormone
      CD: Cycle day in the menstrual cycle
      CPR: Clinical pregnancy rate/cycle
      FSH: Follicle-stimulating hormone
      OPR: On-going pregnancy rate/cycle
      Misc: Miscarriage rate/cycle
      NA: Not applicable
      OHSS: Ovarian hyperstimulation syndrome/cycle
      OCP: Oral contraceptives pills
      OR: Number of oocytes retrieved
      P: Progesterone
      Days: Days of stimulation with gonadotropins
      Endo: Endometrium thickness (mm) at trigger day
      Letrozole and gonadotropin doses are per day unless other is stated.
      Antagonist protocols are uses unless stated otherwise
      Expected responses to OS were defined in the included studies by the authors as poor, normal, or high responders. POR was categorized according to the Bologna criteria (≥2 of 3 of the following: (1) advanced maternal age (≥40 years) or any other risk factor for POR, (2) a previous POR (≤3 oocytes with a conventional stimulation protocol), (3) an abnormal ovarian reserve test (antral follicle count (AFC) < 5–7 follicles or Anti-Müllerian hormone (AMH) 3.6-7.9 pmol/L (Ferraretti et al., 2011). Expected normal responders to OS is lacking a precise definition, but ovarian reserve tests with either AMH or AFC are recognized as the best predictors of ovarian response (Broer et al., 2013a, 2013b). Hence, in one study with no statement of expected response (Mukherjee et al., 2012), the following criteria for expected normal response were used: regular menstrual cycle between 21-35 days, AFC between 7-15, or FSH < 15 in the early follicular phase. Participants from one study (Shapira et al., 2020), with a previous response of 4-9 oocytes in a conventional stimulation protocol was categorized as expected normal responders. The definition of high responders in the studies was AFC >15 or PCOS according to the Rotterdam revised 2003 criteria (2 of 3, (1) oligo- and/or anovulation, (2) clinical and/or biochemical signs of hyperandrogenism, (3) polycystic ovaries, and exclusion of other aetiologies) (Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group, 2004)(Table 1).
      Risk of bias assessment and quality of evidence
      Among the 16 included RCTs, four studies had an overall low risk of bias, ten studies had a moderate and two studies had an overall high risk of bias. The quality of evidence of the RCTs was graded as high, moderate, and low in the five, eight, three studies respectively. Of the 15 observational studies, only one had a moderate overall risk of bias and 14 studies had an overall serious risk or critical of bias.
      The 15 observational studies generally had a high risk of bias and very low quality of evidence, only one study qualified as having a moderate risk of bias and moderate quality of evidence. The risk of bias assessment and grading of the quality of evidence can be viewed in detail next to the Forest plots (Figure 2.12.10) with the Risk of bias legend (Figure 2.0) and in (Supplementary Table 2). To reveal any potential risk of publication-bias a funnel plot was created for each outcome. No high risk of publication bias was revealed (Supplementary Figure 2.1-2.10: Funnel plots).
      Figure 20
      Figure 2.0Forest plots
      Legend for the Risk of Bias in the Forest plots
      Systematic reviews screened
      Systematic reviews of co-treatment with letrozole in IVF/ICSI, but with different protocols, scopes and restrictions were identified: (Requena et al., 2008; Sunkara et al., 2011; Bosdou et al., 2012; Song et al., 2014; Jeve and Bhandari, 2016; Bechtejew et al., 2017; Kamath et al., 2017; Zhang et al., 2020; Zhao et al., 2020; Montoya-Botero et al., 2021; Qin, 2021). All references included in the previous systematic reviews were screened for missing studies for possible inclusion. The screening did not identify any further studies.
      Live birth rate
      Analyses of six studies that reported LBR, including 656 women, showed no significant impact of letrozole co-treatment on LBR, with an RD of 0.04 (95%CI: [-0.05, 0.12]; I2=46%, P=0.39). The overall quality of evidence was low to high according to GRADE (Supplementary Table 2).
      Subgroup analysis with five studies (n=526) of POR reported LBR, revealed a significant effect favouring letrozole with an RD of 0.07 (95%CI: [0.01, 0.13]; I2=0%, P=0.03; low to high quality of evidence). Only one study reported LBR in high responders (n = 130 women), and it showed no significant effect of letrozole (RD = -0.14, 95% CI: [-0.30, 0.03] P=0.1; moderate quality of evidence). No studies of normal responders reported LBR (Figure 2.1).
      On-going pregnancy rate
      Seven studies, including 622 women, were included in this meta-analysis which showed no overall effect of letrozole with a RD of 0.03 (95%CI: [-0.03, 0.09]; I2=0%, P=0.32). The overall quality of evidence was low to high according to GRADE (Supplementary Table 2).
      The subgroup analysis of POR (4 studies; n=372 women, RD 0.05 (95%CI: [-0.03, 0.13]; I2=0%, P=0.23) as well as normal responders (3 studies; n=250 women, RD 0.01 (95%CI: [-0.07, 0.09]; I2=0%, P=0.82), revealed no significant effect of letrozole co-treatment on OPR. No studies with high responders reported OPR.
      Clinical pregnancy rate
      Twenty-five studies, including 2,800 women, provided data on CPR. There was no significant effect of letrozole with a pooled estimate with a RD of 0.02 (95%CI: [-0.02, 0.06]; I2=43%, P=0.39). A relatively high degree of heterogeneity was observed among the studies (Figure 2.3a). The overall quality of evidence was low to high according to GRADE (Supplementary Table 2).
      The subgroup analysis of POR (17 studies; n=1,913 women), showed a RD of 0.03 (95% CI: [-0.01, 0.08]; I2=35%, P=0.14, low to high level of quality) with consistently high heterogeneity. In the normal-responders (6 studies; n=657 women, RD -0.05 (95% CI: [-0.13, 0.03]; I2=14%, P=0.21; low to high level of quality) and high responders (2 studies; n=230 women, RD of 0.02 (95% CI: [-0.03, 0.35]; I2=84%, P=0.89; low to moderate level of quality).
      A sensitivity analysis of CPR, aimed at increasing the homogeneity of the studies, was performed. It revealed an increased RD for POR favouring letrozole of 0.08 (95% CI: [0.01, 0.15]; I2=0%, P=0.03) (Figure 2.3b). Only two studies in the normal and one study in the high responder group fulfilled the criteria as a study of high quality of evidence. Neither analysis showed any effect of letrozole.
      Miscarriage rate
      Seven studies reported miscarriage rates and included 743 women (Figure 2.4). No difference was noted in the overall effect (RD=0.00, 95% CI: [-0.05, 0.05]; I2=40%, P=0.93). The quality of evidence varied from low to high.
      The subgroup analysis of POR (5 studies; n=519 women, RD =-0.01, 95% CI: [-0.07, 0.08]; I2=58%, P=0.85; low to high quality of evidence), normal responders (one study; n=94 women; moderate quality of evidence), and high responders (one study; n=130 women; moderate quality of evidence) showed similar miscarriages rates in all subgroups.
      Number of oocytes retrieved
      A total of 26 studies with 3,091 women displayed data regarding the number of oocytes retrieved for the meta-analysis (Figure 2.5). No difference was observed in the overall effect with a MD of 0.09 (95% CI: [-0.41, 0.60]; I2=89%, P=0.72). A high degree of heterogeneity was present among the studies, whilst the overall quality of evidence was low to high according to GRADE (Supplementary Table 2).
      The subgroup analysis for POR, including 2,287 women, had a MD of -0.27 (95% CI: [-0.82, 0.28]; I2=91%, P=0.33; low to high quality of evidence). In the normal responders (8 studies; n=804 women; low to moderate quality of evidence) there was a significant difference favouring the letrozole group with a MD of 1.81 (95% CI: [0.35, 3.27]; I2=69%, P=0.01). No included studies of high responders reported number of oocytes retrieved.
      The sensitivity analysis of number of oocytes retrieved revealed no effect of letrozole for POR with a MD of 0.16 (95% CI: [-0.59, 0.92]; I2=79%, P=0.67) and continued to show a high degree of heterogeneity (Supplementary Figure 2.1). Only one study in the normal responder group fulfilled the criteria for studies of high quality of evidence, which did not show any effect of letrozole.
      Total FSH consumption
      A total of 27 studies, including 3,364 women, which showed an overall significant reduction of the total FSH consumption with a MD of 1,067 IU in total FSH consumption after co-treatment with letrozole (95% CI: [-1556, -577]; I2=100%, P<0.0001) (Figure 2.6). However, a high degree of heterogeneity was observed amongst the studies, whilst the overall quality of evidence was low to high according to GRADE (Supplementary Table 2).
      The significant effect of letrozole was also present in the POR (20 studies; 2,596 women), where the total FSH consumption was decreased, MD of 1383 IU (95% CI: [-2128, -638]; I2=100%, P=0.0003). No significant difference was observed in normal responders (8 studies; n=768 women; MD = -252 IU (95% CI: [-768, 264]; I2=96%, P=0.34); low to moderate quality of evidence). No included studies provided data on the total FSH consumption for high responders.
      In the sensitivity analysis of the total FSH consumption, five studies of POR, including 256 women, were included. A decrease of total FSH consumption favouring letrozole was confirmed with a MD of -507 (95% CI: [-803, -210]; I2=91%, P<0.001) (Supplementary Figure 2.2). Only one study in the normal responder group fulfilled the criteria for studies of high quality of evidence, which did not show any effect of letrozole.
      Days of stimulation with FSH
      A total of 21 studies, including 2,882 women, showed an overall significant effect favouring the letrozole with a MD of -0.9 days (95% CI: [-1.2, -0.5]; I2=89%, P<0.0001) (Figure 2.7).
      A high degree of heterogeneity was observed amongst the studies, whilst the overall quality of evidence was low to high according to GRADE (Supplementary Table 2).
      The significant decrease in days of stimulation with FSH was also present in the subgroup analysis of POR (16 studies; 2,228 women; low to high quality of evidence) which revealed a reduction of 1.1 days in the letrozole group (95% CI: [-1.6, -0.7]; I2=90%, P<0.00001). No difference was observed for normal responders (6 studies; n=654 women, MD=-0.1 days, 95% CI: [-0.8; 0.6]; I2=75%, P=0.80); all very low quality of evidence). No studies with high responders reported days of stimulation.
      OHSS rate
      Three studies, including 303 women, all normal responders, were included in this meta-analysis (Figure 2.8). There was a non-significant tendency towards a reduction of OHSS risk favouring letrozole with an RD of -0.06 (95% CI: [-0.14, 0.03]; I2=68%, P=0.19). The quality of evidence ranged from low to high according to GRADE (Supplementary Table 2).
      Endometrium thickness
      Fifteen studies, including 1,852 women, were included in this meta-analysis. An overall significant thinner endometrium thickness was revealed in the letrozole group with a MD of -0.54 mm (95% CI: -0.99, -0.08, I2=94%, P=0.02) (Figure 2.9). A high degree of heterogeneity was observed amongst the studies, whilst the overall quality of evidence was low to high according to GRADE (Supplementary Table 2).
      Figure 29
      Figure 2.9Forest plots
      Endometrial Thickness at trigger day
      The subgroup analysis of POR (10 studies; n=1,224 women; MD= -0.52 mm (95% CI: -1.18, 0.14, I2=94%, P=0.12); low to high quality of evidence) showed similar endometrial thickness in both groups. Similar findings were obtained in the subgroup of normal responders (five studies; n=628 women; low to moderate quality of evidence) (MD= -0.58 mm, 95% CI: -1.36, -0.20, I2=91%, P=0.14).
      Cancellation rate
      Twelve studies reported on this outcome, including 1,322 women, and no effect of letrozole on the cancellation rate was demonstrated. There was an overall RD of 0.03 (95% CI: -0.09, 0.15, I2=84%, P=0.64) (Figure 2.10). The overall quality of evidence was low to moderate according to GRADE (Supplementary Table 2).
      The subgroup analysis showed similar results for the POR with a RD of 0.03 (95% CI: -0.10, 0.16, I2=86%, P=0.63), and only one study of normal responders provided data on cancellation rate. No studies of high responders reported cancellation rate.
      Neonatal outcomes
      None of the included articles reported any neonatal outcomes.
      DISCUSSION
      This systematic review and meta-analysis is to our knowledge the first study to assess the impact of letrozole co-treatment in OS for IVF/ICSI in expected normal responders and the largest study to assess the impact of POR and high responders. A total of 3,952 women were included in this analysis.
      Impact of letrozole co-treatment in poor responders.
      Most published studies of co-treatment with letrozole during OS for IVF/ICSI assessed the number of retrieved oocytes and pregnancy rates in POR. In this patient group, the pooled estimate of LBR was significantly increased by 7% (15% to 22%), with letrozole co-treatment. No significant difference was present for CPR. However, the included studies were very heterogeneous regarding both primary outcome, choice of protocol (antagonist/agonist), a different starting dose of FSH, and trigger criteria. In the sensitivity analyses, only comprising studies using the same protocol with similar gonadotrophin starting dose and excluding studies with serious risk of bias, the pooled estimate for CPR was significantly higher by 8% (15% to 23%) with adjuvant letrozole treatment. The definition of CPR differed among the included studies, with only one study in the high-quality group (Lee et al., 2014) accepted CPR as the presence of a gestational sac. Leaving it out of the meta-analysis maintained the significant increase in CPR. Only a few studies reported (OPR), showing an overall increase of 5%, though it did not reach statistical significance, whereas miscarriage rates were similar regardless of letrozole treatment.
      The first RCTs investigation of letrozole as co-treatment in OS (Goswami et al., 2004) focused on lowering the FSH consumption to reduce the cost. This meta-analysis confirms the finding with a significantly lower MD of 1398 IU FSH consumption, and one day less of stimulation with the same average number of oocytes retrieved when using letrozole.
      The effect of letrozole on CPR may be explained by the impact on the endometrium, as letrozole normalizes oestradiol levels during stimulation, which could optimize the window of implantation by reducing the embryo-endometrium asynchrony. The included data did reveal a statistical difference in endometrial thickness at trigger day. Yet, a difference of only 0.13 mm indicates that the beneficiary effect lies in the function of the endometrium rather than crude thickness. It has been hypothesized that adding letrozole to stimulation protocols could reduce the number of cancelled cycles. However, this is not supported by the current literature.
      The meta-analysis for PORs showed an increased CPR, fewer days of stimulation, and reduced the total FSH consumption with letrozole co-treatment during OS.
      Impact of letrozole co-treatment in normal responders
      The literature of studies investigating letrozole co-treatment among normal responders is scarce. No studies report on LBR, and only three studies of low-quality evidence on OPR. CPR is reported in six studies in low to high quality, and the meta-analyses for both outcomes show no difference between the letrozole and placebo groups.
      The concept of adding letrozole to OS could increase the number of oocytes retrieved by two different mechanisms. Letrozole changes the balance between the sex-steroids by inhibiting the conversion of androgens to oestrogens. Intraovarian androgens therefore rise, which is believed to increase the FSH receptors and promoting initial recruitment of follicles (Lossl et al., 2020). Oestrogen levels are reduced, thereby reducing negative feedback on the pituitary-hypothalamic axis and enhancing endogenous gonadotrophin production. This may impact the recruitment of follicles both through enhanced serum levels of gonadotrophins and because of the endogenous acidic isoform of FSH, which is lacking in the recombinant exogenous FSH (Andersen, 2017). The number of oocytes was increased by 1.8 oocytes per cycle in the letrozole group with similar FSH consumption. With the promising high pregnancy rates from frozen-thawed embryos, it would be interesting to investigate, whether the blastocyst formation would be higher as well, which may lead to a higher cumulative pregnancy rate per cycle.
      OHSS is an iatrogenic and occasional fatal event and is associated with high oestradiol levels, why letrozole may reduce the incidence (Aboulghar, 2003; Haas et al., 2018). Interestingly, a trend towards fewer women experienced OHSS among letrozole users in expected normal responders in the three studies, which were all using the same antagonist protocol and 10,000 IU hCG triggering final oocyte maturation. However, the dose of letrozole varied from 2.5-5 mg, and the number of stimulation days with letrozole varied from five days to all days of stimulation. The missing data regarding this outcome is explained by the freeze-all strategy to avoid fresh embryo transfer in patients at risk for OHSS. As OHSS is a rare event in OS a very high number of patients is required to reveal a significant outcome and the pooled estimate with 303 women is still insufficient. A more profound impact of letrozole on OHSS may be seen if continuing letrozole in the luteal phase. A recent systematic review demonstrated a decrease of moderate and severe OHSS in high-risk women, but no significant effect on early or mild OHSS (Zhao et al., 2020).
      Impact of letrozole treatment in high responders.
      Only two studies of letrozole in high responders with a total of 255 women were included and found opposing effects of letrozole on CPR. The RCT, with the highest quality of evidence, reported a tendency towards a decrease in CPR and LBR in the letrozole group. The miscarriage rate was similar in both groups. The studies did not include data on FSH consumption, stimulation length, number of oocytes retrieved, or most importantly OHSS and cancellation risk. The hypothesis of using letrozole in high responders to reduce exogenous FSH and oestradiol levels, thereby decreasing the risk of OHSS, as well as optimizing the endometrium for implantation in the fresh cycle remains unanswered by the existing literature.
      Endometrium thickness
      In contrast to clomiphene citrate, letrozole does not block oestrogen receptors in the reproductive tract, why it is not believed to have any direct impact on the endometrium. Although co-treatment with letrozole resembles oestrogens levels in the natural cycle, there have been concerns about creating a suboptimal endometrium for implantation. The endometrial thickness on ultrasound scans remains a crude but acknowledged indicator of receptivity in IVF as it is correlated to the pregnancy rate and LBR (Kasius et al., 2014) To increase the validity of the analysis three studies were excluded, as the day of measuring endometrial thickness was not reported (Goswami et al., 2004; Garcia-Velasco et al., 2005; Davar et al., 2010). The pooled analysis of endometrial thickness at trigger day showed a significant, but not clinically relevant, reduction of only 0.5 mm in the letrozole group compared with controls. This difference was not present in the subgroup analyses. Furthermore, mean values in all studies lie within the earlier reported optimal range at trigger day (Simeonov et al., 2020), hence, we found no signs of a negative impact of letrozole on the endometrium. It has been proposed that by suppressing oestradiol levels during OS, letrozole possibly has a beneficial effect on the endometrial receptivity, but data supporting this is lacking at the present.
      Cancellation rate
      A large proportion of the included studies reported cancellation rates, however with different and often ill-defined criteria, making a valid meta-analysis difficult for this outcome. However, the crude data suggest that the cancellation rate is unaffected by co-treatment with letrozole. The hypothesis supporting better early follicle recruitment (Lossl et al., 2020), which may reduce the cancellation rate, could not be supported or rejected by this review. Further, larger studies of good quality are needed.
      Neonatal outcomes
      A gap in the literature concerning neonatal outcomes after letrozole co-treatment in IVF/ICSI was identified, since none of the included studies reported on this. Any potential impact of letrozole on off-spring has to be extrapolated from studies of ovulation induction. The extensive reporting after this treatment has been summarized in a recent review, which confirms no increase in the rate of congenital malformations or pregnancy loss (Pundir et al., 2020). These data are reassuring, though the use of co-treatment with letrozole in OS for IVF/ICSI differs with regards to dose and duration, hence attention on the neonatal outcomes in the future is still warranted.
      Efficacy of letrozole
      Letrozole has previously been demonstrated to reduce serum oestradiol levels in a dose-depending manner (He et al., 2014). In this study, significant suppression of oestradiol at trigger day was demonstrated in seven of 11 studies using 2.5 mg letrozole per day, as well as in the majority of studies (11/13) using 5 mg letrozole per day, and also in the two studies using 10 mg or 20 mg letrozole per day. In the studies (n=6)(Garcia-Velasco et al., 2005; Verpoest et al., 2006; Mohsen IA. and El-Din RE., 2013; Yucel O. et al., 2014; Ebrahimi M. et al., 2017; Moini A. et al., 2019), without significantly reduced oestradiol levels, there was no association to oestradiol level, number of oocytes or expected ovarian response. Three of the included studies did not report the oestradiol levels at trigger day(Lee et al., 2011; Bastu et al., 2016). A stimulation protocol with a minimum of 5 mg letrozole per day for at least 5 days seems appropriate to ensure a significant suppression of oestradiol in most of the women. However, larger high-quality studies are needed to explore the appropriate dose of co-treatment with letrozole without detrimental effect on the clinical outcome.
      Strengths and limitations
      The systematic review and meta-analyses were conducted according to the PRISMA statement, thereby securing a high methodological quality. Since the review was not restricted to RCTs, a total of 3,952 women were included in the different meta-analyses. This is to our knowledge the largest systematic review on the reproductive outcome of co-treatment with letrozole for IVF. The quality of the evidence was evaluated consistently with GRADE, and to ensure high validity of the meta-analyses, a sensitivity analysis using only studies with high quality of evidence was conducted. This study has some limitations that need to be addressed. The heterogeneity was overall high and included different study types, different patient groups, different stimulation protocols with different doses, and the timing of FSH and letrozole. Heterogeneity was expected and managed by predefined subgroup analyses, sensitivity analyses where appropriate, and by using the random effect model. Selection bias could potentially have been introduced because of the English or Scandinavian language restrictions, though there were very few studies not fulfilling these criteria. Publication bias was assessed through a systematic approach. The responsible researchers of all unpublished studies registered on clinicaltrials.gov were contacted if contact details were public. Not all responded or were in the process of publishing. Funnel plots were made for all outcomes and none revealed any suspicion of publication bias (Supplementary Figure 1). Nevertheless, this present systematic review and meta-analysis summarizes the current published literature on co-treatment with letrozole for IVF and should be valuable to the clinician.
      Future research
      More high-quality studies are needed to assess the impact of co-treatment with letrozole for expected normal and high responders, as more evidence is needed before reaching any clinical recommendations. Furthermore, a dose-response study of co-treatment with letrozole would be of great value to know if letrozole can be increased and if it is more effective in higher doses without detrimental effects on the endometrium or the oocytes. The OHSS risk may be reduces by continuing letrozole co-treatment in the luteal phase but require large studies to assess this rare event. To further elaborate on the effect of letrozole on embryos and their quality, studies assessing cumulative LBR from one cycle, including following frozen embryo transfers would provide interesting insight. The LBR was increased in POR, but the true clinical effect should be confirmed in an RCT with LBR as the primary outcome in similar protocols in POR. The substantial heterogeneity amongst the studies warrants an individual participant data meta-analysis. The identified gap in the literature concerning neonatal outcomes after co-treatment with letrozole in OS should be closed with a large register study. The risk of OHSS was only reported in three studies of normal responders and no studies of high responders, hence a prospective study of high and normal responders may render information of this rare event.
      CONCLUSIONS
      This present systematic review and meta-analyses indicate that POR may benefit from co-treatment with letrozole during ovarian stimulation for IVF/ICSI, as this subgroup had significantly higher LBR and CPR, a similar number of oocytes retrieved, whilst the total FSH consumption and days of stimulation were significantly decreased. The effect of letrozole suppressing oestradiol levels at trigger day was most consistently achieved using an antagonist protocol with 5 mg letrozole a day for a minimum of 5 days. Normal responders gained almost two oocytes more per cycle in the letrozole group but had no other significant effect of letrozole in any other outcomes. Only very few studies of high responders were included in this study and they showed no significant effect of letrozole on LBR, CPR, miscarriage rate, and there were no data on other outcomes. Based on the current literature there is no evidence for recommending co-treatment with letrozole in expected normal or high responders to improve the clinical outcome. However, more well-designed good-quality studies are needed.
      A table with description of the excluded full text studies, and a detailed table of the bias and quality assessment are presented in the supplementary. Furthermore, funnel plots for all outcomes, and sensitivity analysis with high quality studies regarding number of oocytes retrieved and FSH consumption are reported in the supplementary.
      ACKNOWLEDGEMENTS
      The author group is grateful for the corresponding authors, who responded with missing data for the analyses, and for the shared experience in developing the search strategy from information specialist Anders Larsen, UCSF, Medical Research Centre, University Hospital, Copenhagen. We also acknowledge the constructive guidance regarding statistical assistance from associate professor Julie Lyng Forman, Section on Biostatistics, Copenhagen University.
      AUTHOR'S ROLES
      All authors participated in designing the study. NM and SOS conceived the study. NB and MDH drafted the protocol, which was revised and approved by all authors. NB and MDH searched databases, screened all summaries, abstracts, and full-text articles. NB and MDH extracted most data with contribution from AP and ALM. NB and MDH assessed bias and quality of evidence, consulting AP for advice. NB carried out the meta-analyses and MDH contacted authors of relevant studies from clinicaltrials.gov and authors of included studies, where additional data or clarification were needed. NB and MDH designed all tables and figures, which were approved by all authors. NB and MDH drafted the manuscript and all authors revised the manuscript and approved the final version.
      DATA
      Templates, collection forms, and extracted data used in the analyses can be requested through the corresponding author.
      FUNDING
      MDH was supported by an unrestricted research grant from Gedeon Richter. NB's salary was covered by the Fertility Clinic, Rigshospitalet, and the Gynaecological and obstetrics department, Herlev Hospital, both a part of the University hospitals, Capital Region of Denmark.
      CONFLICT OF INTEREST
      The authors had no conflicts of interest to declare in relation to this work.
      R*EFERENCES
      Aboulghar M. Prediction of ovarian hyperstimulation syndrome (OHSS). Estradiol level has an important role in the prediction of OHSS. Hum Reprod 2003;18:1140–1141.
      Andersen CY. Inhibin-B secretion and FSH isoform distribution may play an integral part of follicular selection in the natural menstrual cycle. Molecular Human Reproduction 2017;23:16–24.
      Bastu E, Buyru F, Ozsurmeli M, Demiral I, Dogan M, Yeh J. A randomized, single-blind, prospective trial comparing three different gonadotropin doses with or without addition of letrozole during ovulation stimulation in patients with poor ovarian response. European Journal of Obstetrics and Gynecology 2016;203:30–34.
      Bechtejew TN, Nadai MN, Nastri CO, Martins WP. Clomiphene citrate and letrozole to reduce follicle-stimulating hormone consumption during ovarian stimulation: systematic review and meta-analysis. Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 2017;50:315–323.
      Beckers N, Macklon NS, Eijkemans MJ, Ludwig M, Felberbaum RE, Diedrich K, Bustion S, Loumaye E, Fauser BCJM. Nonsupplemented luteal phase characteristics after the administration of recombinant human chorionic gonadotropin, recombinant luteinizing hormone, or gonadotropin-releasing hormone (GnRH) agonist to induce final oocyte maturation in in vitro fertilizatio. The Journal of clinical endocrinology and metabolism 2003;88:4186–4192.
      Biljan MM, Hemmings R, Brassard N. The Outcome of 150 Babies Following the Treatment With Letrozole or Letrozole and Gonadotropins. Fertility and sterility 2005;84:S95.
      Boomsma CM, Kavelaars A, Eijkemans MJC, Fauser BCJM, Heijnen CJ, Macklon NS. Ovarian stimulation for in vitro fertilization alters the intrauterine cytokine, chemokine, and growth factor milieu encountered by the embryo. Fertil Steril 2010;94:1764–1768.
      Bosdou JK, Venetis CA, Kolibianakis EM, Toulis KA, Goulis DG, Zepiridis L, Tarlatzis BC. The use of androgens or androgen-modulating agents in poor responders undergoing in vitro fertilization: a systematic review and meta-analysis. Human reproduction update 2012;18:127–145.
      Bossche HV, Moereels H, Koymans LMH. Aromatase inhibitors ? mechanisms for non-steroidal inhibitors. Breast Cancer Res Tr 1994;30:43–55.
      Bourgain C, Devroey P. The endometrium in stimulated cycles for IVF. Hum Reprod Update 2003;9:515–522.
      Broer SL, Disseldorp J van, Broeze KA, Dolleman M, Opmeer BC, Bossuyt P, Eijkemans MJC, Mol B-WJ, Broekmans FJM, on behalf of the IMPORT study group, et al. Added value of ovarian reserve testing on patient characteristics in the prediction of ovarian response and ongoing pregnancy: an individual patient data approach. Human Reproduction Update 2013a;19:26–36.
      Broer SL, Dólleman M, Disseldorp J van, Broeze KA, Opmeer BC, Bossuyt PMM, Eijkemans MJC, Mol BW, Broekmans FJM, Broer SL, et al. Prediction of an excessive response in in vitro fertilization from patient characteristics and ovarian reserve tests and comparison in subgroups: an individual patient data meta-analysis. Fertility and Sterility 2013b;100:420-429.e7.
      Bülow NS, Skouby SO, Warzecha AK, Udengaard H, Andersen CY, Holt MD, Grøndahl ML, Nyboe Andersen A, Sopa N, Mikkelsen ALE, et al. Impact of letrozole co-treatment during ovarian stimulation with gonadotrophins for IVF: a multicentre, randomized, double-blinded placebo-controlled trial. Hum Reprod 2021;deab249.
      Buzdar AU, Robertson JFR, Eiermann W, Nabholtz J-M. An overview of the pharmacology and pharmacokinetics of the newer generation aromatase inhibitors anastrozole, letrozole, and exemestane. Cancer 2002;95:2006–2016.
      Cenksoy PO, Ficicioglu C, Kizilkale O, Suhha Bostanci M, Bakacak M, Yesiladali M, Kaspar C. The comparision of effect of microdose GnRH-a flare-up, GnRH antagonist/aromatase inhibitor letrozole and GnRH antagonist/clomiphene citrate protocols on IVF outcomes in poor responder patients. Gynecol Endocrinol 2014;30:485–489.
      Cumpston M, Li T, Page MJ, Chandler J, Welch VA, Higgins JP, Thomas J. Updated guidance for trusted systematic reviews: a new edition of the Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Database Syst Rev 2019;10:ED000142.
      D'Amato G, Caringella AM, Stanziano A, Cantatore C, Palini S, Caroppo E. Mild ovarian stimulation with letrozole plus fixed dose human menopausal gonadotropin prior to IVF/ICSI for infertile non-obese women with polycystic ovarian syndrome being pre-treated with metformin: a pilot study. Reprod Biol Endocrinol 2018;16:1–8.
      Davar R, Oskouian H, Ahmadi S, Firouzabadi RD. GnRH antagonist/letrozole versus microdose GnRH agonist flare protocol in poor responders undergoing in vitro fertilization. Taiwan J Obstet Gynecol 2010;49:297–301.
      Ebrahimi, Akbari-Asbagh F., Ghalandar-Attar M. Letrozole+ GnRH antagonist stimulation protocol in poor ovarian responders undergoing intracytoplasmic sperm injection cycles: An RCT. International journal of reproductive biomedicine (Yazd, Iran) 2017;15:101–108.
      Eftekhar M, Saeed L. Effect of adding letrozole to gonadotropin on in vitro fertilization outcomes: An RCT. Int J Reprod Biomed 2020;18:287–294.
      Elassar A, Engmann L, Nulsen J, Benadiva C. Letrozole and gonadotropins versus luteal estradiol and gonadotropin-releasing hormone antagonist protocol in women with a prior low response to ovarian stimulation. Fertil Steril 2011;95:2330–2334.
      Ferraretti AP, La Marca A, Fauser BCJM, Tarlatzis B, Nargund G, Gianaroli L, ESHRE working group on Poor Ovarian Response Definition. ESHRE consensus on the definition of “poor response” to ovarian stimulation for in vitro fertilization: the Bologna criteria. Hum Reprod 2011;26:1616–1624.
      Franik S, Eltrop SM, Kremer JA, Kiesel L, Farquhar C. Aromatase inhibitors (letrozole) for subfertile women with polycystic ovary syndrome. In Cochrane Gynaecology and Fertility Group, editor. Cochrane Database of Systematic Reviews 2018;113:218.
      Gaast MH van der, Beckers NGM, Beier-Hellwig K, Beier HM, Macklon NS, Fauser BCJM. Ovarian stimulation for IVF and endometrial receptivity–the missing link. Reprod Biomed Online 2002;5 Suppl 1:36–43.
      Gadalla MA, Norman RJ, Tay CT, Hiam DS, Melder A, Pundir J, Thangaratinam S, Teede HJ, Mol BWJ, Moran LJ. Medical and Surgical Treatment of Reproductive Outcomes in Polycystic Ovary Syndrome: An Overview of Systematic Reviews. Int J fertil Steril [Internet] 2020;13:.
      Garcia-Velasco JA, Moreno L, Pacheco A, Guillén A, Duque L, Requena A, Pellicer A. The aromatase inhibitor letrozole increases the concentration of intraovarian androgens and improves in vitro fertilization outcome in low responder patients: A pilot study. Fertility and sterility 2005;84:82–87.
      Goswami SK, Das T, Chattopadhyay R, Sawhney V, Kumar J, Chaudhury K, Chakravarty BN, Kabir SN. A randomized single-blind controlled trial of letrozole as a low-cost IVF protocol in women with poor ovarian response: a preliminary report. Hum Reprod 2004;19:2031–2035.
      Greenhalgh T, Peacock R. Effectiveness and efficiency of search methods in systematic reviews of complex evidence: Audit of primary sources. British Medical Journal 2005;331:1064–1065.
      Haas J, Bassil R, Gonen N, Meriano J, Jurisicova A, Casper RF. The VEGF and PEDF levels in the follicular fluid of patients co- treated with LETROZOLE and gonadotropins during the stimulation cycle. Reprod Biol Endocrinol 2018;16:1–5.
      Haas J, Bassil R, Meriano J, Samara N, Barzilay E, Gonen N, Casper RF. Does daily co-administration of letrozole and gonadotropins during ovarian stimulation improve IVF outcome? 2017;1–5.
      He Q, Liang L, Zhang C, Li H, Ge Z, Wang L, Cui S. Effects of different doses of letrozole on the incidence of early-onset ovarian hyperstimulation syndrome after oocyte retrieval. Syst Biol Reprod Med 2014;60:355–360.
      Higgins JPT. Measuring inconsistency in meta-analyses. BMJ 2003;327:557–560.
      Horcajadas JA, Mínguez P, Dopazo J, Esteban FJ, Domínguez F, Giudice LC, Pellicer A, Simón C. Controlled ovarian stimulation induces a functional genomic delay of the endometrium with potential clinical implications. J Clin Endocrinol Metab 2008;93:4500–4510.
      Jeve Y, Bhandari H. Effective treatment protocol for poor ovarian response: A systematic review and meta-analysis. Journal of Human Reproductive Sciences 2016;9:70.
      Kamath MS, Maheshwari A, Bhattacharya S, Lor KY, Gibreel A. Oral medications including clomiphene citrate or aromatase inhibitors with gonadotropins for controlled ovarian stimulation in women undergoing in vitro fertilisation. In Group CG and F, editor. Cochrane Database of Systematic Reviews 2017;26:593.
      Kasius A, Smit JG, Torrance HL, Eijkemans MJC, Mol BW, Opmeer BC, Broekmans FJM. Endometrial thickness and pregnancy rates after IVF: a systematic review and meta-analysis. Human Reproduction Update 2014;20:530–541.
      Khojah M, Khayat S, Dahan MH. Comparison of in vitro fertilization cycles stimulated with 20 mg letrozole daily versus high‐dose gonadotropins in Rotterdam Consensus ultra‐poor responders: A proof of concept. Int J Gynecol Obstet 2021;ijgo.13626.
      Kim SJ, Choo CW, Kim SK, Lee JR, Jee BC, Suh CS, Lee WD, Kim SH. The effects of letrozole on women with endometriosis undergoing ovarian stimulation for in vitro fertilization. Gynecol Endocrinol 2020;36:257–260.
      Lazer T, Dar S, Shlush E, Al Kudmani BS, Quach K, Sojecki A, Glass K, Sharma P, Baratz A, Librach CL. Comparison of IVF Outcomes between Minimal Stimulation and High-Dose Stimulation for Patients with Poor Ovarian Reserve. Int J Reprod Med 2014;2014:581451.
      Lee K-H, Kim C-H, Suk H-J, Lee Y-J, Kwon S-K, Kim S-H, Chae H-D, Kang B-M. The effect of aromatase inhibitor letrozole incorporated in gonadotrophin-releasing hormone antagonist multiple dose protocol in poor responders undergoing in vitro fertilization. Obstet Gynecol Sci 2014;57:216–222.
      Lee VCY, Chan CCW, Ng EHY, Yeung WSB, Ho PC. Sequential use of letrozole and gonadotrophin in women with poor ovarian reserve: a randomized controlled trial. Reprod Biomed Online 2011;23:380–388.
      Liu X, Li T, Wang B, Xiao X, Liang X, Huang R. Mild stimulation protocol vs conventional controlled ovarian stimulation protocol in poor ovarian response patients: a prospective randomized controlled trial. Arch Gynecol Obstet 2020;301:1331–1339.
      Lossl K, Freiesleben N la C, Wissing ML, Petersen KB, Holt MD, Mamsen LS, Anderson RA, Andersen CY. Biological and clinical rationale for androgen priming in ovarian stimulation. Frontiers in Endocrinology 2020;11:.
      Ma W, Song H, Das SK, Paria BC, Dey SK. Estrogen is a critical determinant that specifies the duration of the window of uterine receptivity for implantation. Proc Natl Acad Sci USA 2003;100:2963–2968.
      Macklon NS, Stouffer RL, Giudice LC, Fauser BCJM. The science behind 25 years of ovarian stimulation for in vitro fertilization. Endocr Rev 2006;27:170–207.
      Mitwally MF, Casper RF. Use of an aromatase inhibitor for induction of ovulation in patients with an inadequate response to clomiphene citrate. FNS 2001;75:305–309.
      Mohsen IA, El Din RE. Minimal stimulation protocol using letrozole versus microdose flare up GnRH agonist protocol in women with poor ovarian response undergoing ICSI. Gynecol Endocrinol 2013;29:105–108.
      Moini A, Lavasani Z, Kashani L, Mojtahedi MF, Yamini N. Letrozole as co-treatment agent in ovarian stimulation antagonist protocol in poor responders: A double-blind randomized clinical trial. Int J Reprod Biomed 2019;17:653–660.
      Montoya-Botero P, Drakopoulos P, González-Foruria I, Polyzos NP. Fresh and cumulative live birth rates in mild versus conventional stimulation for IVF cycles in poor ovarian responders: a systematic review and meta-analysis. Human Reproduction Open 2021;2021:hoaa066.
      Mukherjee S, Sharma S, Chakravarty BN. Letrozole in a low-cost in vitro fertilization protocol in intracytoplasmic sperm injection cycles for male factor infertility: A randomized controlled trial. J Hum Reprod Sci 2012;5:170–174.
      Nabati A, Peivandi S, Khalilian A, Mirzaeirad S, Hashemi SA. Comparison of GnRh Agonist Microdose Flare Up and GnRh Antagonist/Letrozole in Treatment of Poor Responder Patients in Intra Cytoplaspic Sperm Injection: Randomized Clinical Trial. Glob J Health Sci 2015;8:166–171.
      Nielsen ME, Rasmussen IA, Kristensen SG, Christensen ST, Mollgard K, Wreford Andersen E, Byskov AG, Yding Andersen C. In human granulosa cells from small antral follicles, androgen receptor mRNA and androgen levels in follicular fluid correlate with FSH receptor mRNA. Molecular Human Reproduction 2011;17:63–70.
      Oktay K, Hourvitz A, Sahin G, Oktem O, Safro B, Cil A, Bang H. Letrozole reduces estrogen and gonadotropin exposure in women with breast cancer undergoing ovarian stimulation before chemotherapy. The Journal of clinical endocrinology and metabolism 2006;91:3885–3890.
      Ozmen B, Sönmezer M, Atabekoglu CS, Olmus H. Use of aromatase inhibitors in poor-responder patients receiving GnRH antagonist protocols. Reprod Biomed Online 2009;19:478–485.
      Pundir J, Achilli C, Bhide P, Sabatini L, Legro RS, Rombauts L, Teede H, Coomarasamy A, Zamora J, Thangaratinam S. Risk of foetal harm with letrozole use in fertility treatment: a systematic review and meta-analysis. Human reproduction update 2020;
      Qin Y. Effects of using letrozole in combination with the GnRH antagonist protocol for patients with poor ovarian response: A meta-analysis. J Gynecol Obstet Hum Reprod 2021;50:102139.
      Reddy J, Oktay K. Ovarian stimulation and fertility preservation with the use of aromatase inhibitors in women with breast cancer. Fertility and sterility 2012;98:1363–1369.
      Requena A, Herrero J, Landeras J, Navarro E, Neyro JL, Salvador C, Tur R, Callejo J, Checa MA, Farré M, et al. Use of letrozole in assisted reproduction: a systematic review and meta-analysis. Human reproduction update 2008;14:571–582.
      Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004;19:41–47.
      Schoolcraft WB, Surrey ES, Minjarez DA, Stevens JM, Gardner DK. Management of poor responders: can outcomes be improved with a novel gonadotropin-releasing hormone antagonist/letrozole protocol? Fertility and sterility 2008;89:151-6 AN-.
      Schünemann H, Brożek J, Guyatt G, Oxman A. GRADE handbook for grading quality of evidence and strength of recommendations. Updated October 2013. The GRADE Working Group, 2013 [Internet]. 2014;Available from: https://gdt.gradepro.org/app/handbook/handbook.html.
      Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart LA, the PRISMA-P Group. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ 2015;349:g7647–g7647.
      Shapira M, Orvieto R, Lebovitz O, Nahum R, Aizer A, Segev-Zahav A, Haas J. Does daily co administration of gonadotropins and letrozole during the ovarian stimulation improve IVF outcome for poor and sub optimal responders? J Ovarian Res 2020;13:66.
      Simeonov M, Sapir O, Lande Y, Ben-Haroush A, Oron G, Shlush E, Altman E, Wertheimer A, Shochat T, Shufaro Y. The entire range of trigger-day endometrial thickness in fresh IVF cycles is independently correlated with live birth rate. Reproductive BioMedicine Online 2020;41:239–247.
      Siristatidis C, Dafopoulos K, Vrantza T, Salamalekis G, Basios G, Vogiatzi P, Pergialiotis V, Papantoniou N. Mild versus conventional antagonist ovarian stimulation protocols in expected normal responders undergoing IVF/ICSI: a case–control study. Gynecol Endocrinol 2017;33:553–556.
      Song Y, Li Z, Wu X, Wang X, Xiao J, Wang B. Effectiveness of the antagonist/letrozole protocol for treating poor responders undergoing in vitro fertilization/intracytoplasmic sperm injection: A systematic review and meta-analysis. Gynecological Endocrinology 2014;30:330–334.
      Sterne JA, Hernán MA, Reeves BC, Savović J, Berkman ND, Viswanathan M, Henry D, Altman DG, Ansari MT, Boutron I, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ 2016;i4919.
      Sunkara SK, Khalaf Y, Maheshwari A, Seed P, Coomarasamy A. Association between response to ovarian stimulation and miscarriage following IVF: an analysis of 124 351 IVF pregnancies. Human Reproduction 2014;29:1218–1224.
      Sunkara SK, Pundir J, Khalaf Y. Effect of androgen supplementation or modulation on ovarian stimulation outcome in poor responders: a meta-analysis. Reproductive biomedicine online 2011;22:545–555. Reproductive Healthcare Ltd.
      Teede H, Misso M, Costello M, Dokras A, Laven J, Moran L, Piltonen T, Norman R. International evidence-based guideline for the assessment and management of polycystic ovary syndrome 2018. 2018;1–201.
      Verpoest WMJA, Kolibianakis E, Papanikolaou E, Smitz J, Van Steirteghem A, Devroey P. Aromatase inhibitors in ovarian stimulation for IVF/ICSI: a pilot study. Reprod Biomed Online 2006;13:166-72 AN-.
      Weil S, Vendola K, Zhou J, Bondy CAB. Androgen and follicle-stimulating hormone interactions in primate ovarian follicle development. The Journal of clinical endocrinology and metabolism 1999;84:2951–2956.
      Yang X, Lin G, Lu G, Gong F. Letrozole supplementation during controlled ovarian stimulation in expected high responders: a pilot randomized controlled study. Reprod Biol Endocrinol 2019;17:170.
      Yasa C, Bastu E, Dural O, Celik E, Ergun B. Evaluation of low-dose letrozole addition to ovulation induction in IVF. Clin Exp Obstet Gynecol 2013;40:98–100.
      Yu R, Jin H, Huang X, Lin J, Wang P. Comparison of modified agonist, mild-stimulation and antagonist protocols for in vitro fertilization in patients with diminished ovarian reserve. J Int Med Res 2018;46:2327–2337.
      Yucel O, Ekin M, Cengiz H, Zebitay AG, Yalcinkaya S, Karahuseyinoglu S. Comparison of estradiol and progesterone priming/antagonist/letrozole and microdose flare-up protocols for poor responders undergoing intracytoplasmic sperm injection. Gynecological endocrinology : the official journal of the International Society of Gynecological Endocrinology 2014;30:653–656.
      Zhang Y, Zhang C, Shu J, Guo J, Chang HM, Leung PCK, Sheng JZ, Huang H. Adjuvant treatment strategies in ovarian stimulation for poor responders undergoing IVF: A systematic review and network meta-analysis. Human Reproduction Update 2020;26:247–263.
      Zhao J, Xu B, Huang X, Yan Y, Li Y. Whether Letrozole could reduce the incidence of early ovary hyperstimulation syndrome after assisted reproductive technology? A systematic review and meta-analysis. Reprod Health 2020;17:181.

      Appendix. Supplementary materials

      Biography

      Nathalie Søderhamn Bülow, MD commenced working at the Fertility Clinic in Copenhagen University hospital, Rigshospitalet in October 2019 after completing the specialist training in Gynaecology and Obstetrics. Her Ph.D. project is investigating the clinical use of letrozole in IVF.
      KEY MESSAGE
      Co-treatment with letrozole during ovarian stimulation for IVF/ICSI may benefit poor responders by increasing live birth rate. There is currently insufficient evidence to support a recommendation of co-treatment with letrozole in normal or high responders, as more good quality studies are required.