Determinants of monozygotic twinning in ART: a systematic review and a meta-analysis

Determinants of monozygotic twinning in ART: a systematic review and a meta-analysis Abstract BACKGROUND The incidence of monozygotic twins (MZT) after ART appears to be higher than the incidence after spontaneous conceptions contradicting the aim of ART to avoid multiple pregnancies because of the associated risks. OBJECTIVE AND RATIONALE The aim was to study the frequency of MZT after IVF and ICSI and how it is influenced by the day of embryo transfer, maternal age, zona pellucida manipulation, controlled ovarian stimulation, stimulation protocol, culture media and embryo quality. SEARCH METHODS Original studies and reviews were identified by searching the PubMed, Embase and Cochrane databases up to March 2017. The inclusion criterion was publications focusing on the five study questions related to MZT in our study. The exclusion criteria were articles that did not include blastocyst transfer, were on non-humans, were not published in peer-reviewed journals, and were based only on case studies. All of the articles were categorized according to the Oxford Centre for Evidence-based Medicine’s ‘Levels of Evidence’, and quality and risk of bias assessment was performed with ‘The Cochrane Collaboration’s Risk of Bias Tools’. A meta-analysis was performed to study the impact of the day of embryo transfer on the MZT rate. OUTCOMES The literature search resulted in a total of 42 articles, including 38 original studies, for analysis. The included original studies reported a MZT rate with blastocyst transfer from zero to 13.2%. Our meta-analysis found a higher frequency of MZT after blastocyst transfer compared with cleavage-stage embryos transfer: odds ratio = 2.18, 95% CI: 1.93–2.48 (fixed effect meta-analysis). A younger maternal age may increase the MZT rate, and recent studies regarding the use of zona pellucida manipulating techniques have disagreed with the previous suspicion of a higher MZT rate after the use of these methods. The extended culture to-blastocyst stage is a potential risk factor for MZT, but it is uncertain whether this phenomenon is due to the extended time, culture media or greater likelihood of younger oocytes to reach the blastocyst stage. An increased frequency of MZT following the GnRH-agonist suppression protocol has been suggested, as well as a decreased frequency of MZT with high gonadotrophin doses, which could reflect an age-related effect. Only limited literature has focused on the role of embryo morphology in the MZT rate, therefore, this issue remains unresolved. WIDER IMPLICATIONS We found blastocyst transfer to be a risk factor for MZT. Hence, the results of this meta-analysis may weaken the previously proposed view that greater experience with blastocyst transfer and improved culture media could decrease the high rate of MZT after blastocyst transfer. To minimize the rate of MZT and the associated complications, the mechanisms underlying blastocyst transfer and MZT pregnancy must be elucidated. monozygotic twins, cleavage-stage, blastocyst, ART, IVF, ICSI, assisted hatching, culture media, meta-analysis Introduction For decades, it has been possible to help infertile women and men have biological children by using ART (Edwards, 2005). ART has been associated with multiple gestations as a result of the transference of two or more embryos, which may develop into dizygotic twins (Sobek et al., 2016). However, monozygotic twins (MZT) arise from one fertilized oocyte that divides into two or even three or four. So although MZT can occur after multiple embryo transfer (ET), only one fertilized oocyte is required (Sobek et al., 2016). Nevertheless, the frequency of MZT after ART varies from 0 to 13.2% compared with a spontaneous conception rate of 0.4% of live births (Milki et al., 2003; Jain et al., 2004; Papanikolaou et al., 2006; Moayeri et al., 2007; Parisaei et al., 2008; Vitthala et al., 2009; Kawachiya et al., 2011; Franasiak et al., 2015). However, in most of the included studies, the rate of MZT after spontaneous conception was calculated based on live births and not early ultrasound. Thus, the MZT rate after spontaneous conception is likely to be considerably higher if it is also based on early ultrasound. The chorionicity of MZT is assumed to depend on the day of embryo division (Fig. 1). If the oocyte divides shortly after fertilization, the foetuses will be dichorionic diamniotic (DC-DA), which occurs in 20–30% of cases after spontaneous conception. If the division occurs at the blastocyst stage, the gestation will most likely become monochorionic diamniotic (MC-DA), which is observed in 70–75% of cases. After 1–2 weeks, division of the embryo will result in monochorionic monoamniotic (MC-MA) twins only in 1–2% of live MZT births (Aston et al., 2008; Sobek et al., 2016). Figure 1 View largeDownload slide A schematic drawing showing the three different types of monozygotic twins and their potential origins. Figure 1 View largeDownload slide A schematic drawing showing the three different types of monozygotic twins and their potential origins. A multiple pregnancy involves numerous risks for the fetuses as well as for the mother, demonstrating some relationship to the chorionicity of the foetuses. Obstetric risks include low birthweight, preterm birth, pre-eclampsia, nausea and vomiting of the mother, vanishing twin syndrome and perinatal mortality (Pinborg et al., 2005; Pinborg, 2005; Aston et al., 2008; Fuchs and Senat, 2016). Furthermore, MZT are at risk for developing twin-to-twin transfusion syndrome (TTTS) (Kontopoulos et al., 2016). TTTS is most likely to occur in monochorionic twins, but exceptions are observed with dichorionic monozygotic and even with dichorionic-dizygotic twins (Kontopoulos et al., 2016). TTTS arises when there is an imbalance in blood flow between two MZT because of vascular anastomoses in the placenta of monochorionic twins, or between the two placentas with dichorionic monozygotic (or rarely seen dizygotic) twins. TTTS occurs in 5–20% of monochorionic twins, and left untreated, the risk of pregnancy loss can reach 95% (Kontopoulos et al., 2016). MC-MA twins also have a risk of umbilical cord entanglement because they share the same amniotic sac. Given the higher risks, the overall goal is to minimize the number of multiple gestations in ART pregnancies. Numerous studies have found a higher clinical pregnancy rate with the transfer of a blastocyst compared with a cleavage-stage embryo (Sharara and Abdo, 2010; Kawachiya et al., 2011; Sotiroska et al., 2015). This allows, in most cases, a practice to transfer only one embryo and reduce the risk of multiple gestations. Previous studies have disagreed upon the incidence of MZT in ART, as well as the underlying mechanisms. Investigations of potential risk factors, such as the age of the oocyte, extended culture and assisted hatching (AH), have provided no clear explanation of the aetiology (Fig. 2) (Knopman et al., 2014). Likewise, the embryo quality may be associated with MZT (Franasiak et al., 2015). Finally, some studies have not found a correlation of MZT with blastocyst transfer compared with cleavage-stage ET (Moayeri et al., 2007; Papanikolaou et al., 2010; Osianlis et al., 2014; Wu et al., 2014). Figure 2 View largeDownload slide Some of the possible mechanisms underlying the increased rate of monozygotic twins in ART. MZT, monozygotic twins; ZP, zona pellucida; AH, assisted hatching. Figure 2 View largeDownload slide Some of the possible mechanisms underlying the increased rate of monozygotic twins in ART. MZT, monozygotic twins; ZP, zona pellucida; AH, assisted hatching. The aim of this systematic review is to examine the risk factors for MZT in ART treatment by addressing the following five questions: – Does the transfer of cleavage-stage versus blastocyst stage embryos (Day 2/3 versus Day 5) influence the frequency of MZT? A meta-analysis is conducted. – Does controlled ovarian stimulation, the ART protocol or the culture media affect the frequency of MZT? – Does maternal age affect the frequency of MZT? – Does the cause of infertility or the fertilization method (IVF versus ICSI) affect the frequency of MZT? – Does the morphology of the embryo affect the frequency of MZT? Methods Sources and literature search We used the PubMed, Embase and Cochrane databases up until March 2017 to identify the articles included in this review. To cover the subject and study questions, the following search strings were used: ‘Monozygotic twins AND blastocyst’, ‘Monozygotic twins AND ART’, ‘Monozygotic twins AND media’, ‘Monozygotic twins AND ovarian stimulation’, ‘Monozygotic twins AND maternal age’, ‘Monozygotic twins AND cause of infertility’, ‘Monozygotic twins AND Day 5’ and ‘Monozygotic twins AND assisted reproductive technology’. We used the same search strings with the exchange of the word ‘Monozygotic twins’ with ‘Monochorionic twins’. The MeSH terms used were as follows: monozygotic twins, monochorionic twins, blastocyst, assisted reproductive technology, ART, media, ovarian stimulation, maternal age, cause of infertility and Day 5. Study selection The different searches identified 2722 scientific articles. We excluded articles in non-English languages. In a single search, it was necessary to apply the criterion of only articles from the last 10 years (‘Monozygotic twins AND maternal age’). After removal of non-English articles and non-accessible full-length articles, a total of 1841 articles were selected for screening. The title and/or abstract of the 1841 articles were screened. Articles were included in the review based on the following criteria: Inclusion criteria: articles focusing on one of the five study-questions listed in the Introduction. Exclusion criteria: articles that did not include blastocyst transfer, were on non-humans, were not published in peer-reviewed journals, and were based only on case studies. A few of the included articles did not directly focus on blastocyst transfer, but were selected because they were included in two major meta-analyses (Chang et al., 2009; Vitthala et al., 2009). Finally, we included 42 articles after screening (Table I). The search results are shown in Fig. 3. Table I Quality of studies, based on The Cochrane Collaboration’s Risk of Bias Tools, included in a systematic review and meta-analysis of determinants of monozygotic twinning in ART. Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Table I Quality of studies, based on The Cochrane Collaboration’s Risk of Bias Tools, included in a systematic review and meta-analysis of determinants of monozygotic twinning in ART. Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Figure 3 View largeDownload slide PRISMA flow chart of the study selection (March 2017). Figure 3 View largeDownload slide PRISMA flow chart of the study selection (March 2017). Four of the 42 articles were reviews and therefore not included in Table I (Hankins and Saade, 2005; Toledo, 2005; Aston et al., 2008; Sobek et al., 2016). Study quality All articles were categorized according to ‘Levels of Evidence’ (Oxford Centre for Evidence-based Medicine guidelines; http://www.cebm.net/oxford-centre-evidence-based-medicine-levels-evidence-march-2009/; 28 March 2018, date last accessed), and quality and bias were assessed with ‘The Cochrane Collaboration’s Risk of Bias Tools’ (http://methods.cochrane.org/sites/methods.cochrane.org.bias/files/public/uploads/Tool%20to%20Assess%20Risk%20of%20Bias%20in%20Cohort%20Studies.pdf; 28 March 2018, date last accessed). In summary, the risk of bias tools consisted of the following questions: whether exposed and non-exposed cohorts are drawn from the same population; confidence in the assessment of exposure; confidence that the outcome of interest was not present at the start of study; exposed and unexposed were matched for all variables associated with the outcome of interest; confidence in the assessment of the presence or absence of prognostic factors; confidence in the assessment of the outcome; whether the cohort follow-up was adequate; whether the co-interventions were similar between groups. Although two of the studies (Papanikolaou et al., 2006; Ren et al., 2013) were categorized as 1b (individual RCT (narrow CI)) by the Levels of Evidence, they were assessed by the same questions regarding quality and bias. All studies showed good quality and a low risk of bias, and hence, none of the studies were categorized as low quality (Table I). Based on a review of the results using The Cochrane Collaboration’s Risk of Bias Tools, the results were categorized as either medium or high quality. Study selection for the meta-analysis Studies published after the latest two meta-analyses by Chang et al. (2009) and Vitthala et al. (2009) (search ending in 2007) were chosen for this meta-analysis. The 6 of the 17 identified studies had to be excluded from the meta-analysis (Verpoest et al., 2009; Knopman et al., 2010; Ren et al., 2013; Luke et al., 2014; Sotiroska et al., 2015; Otsuki et al., 2016). Exclusion criteria were as follows: studies not reporting accurate numbers for MZT following either blastocyst transfer or cleavage-stage transfer, no control group (cleavage-stage embryos), if part of the data had been published in another study, and the lack of a defined method for identifying MZT. The study by Osianlis et al. (2014) addressed the question of concurrent natural conception and showed an inconsistency in their original MZT data because of gender discordance between twins born after single ET (SET). We chose to include Osianlis et al. (2014) in the meta-analysis because all the SET studies showed limitations in identifying MZT twins, as noted in the study by Osianlis et al. (2014). All the included SET studies defined twin pregnancies as monozygotic and thereby overlooked the possibility of dizygotic twins following concurrent natural conception. This is an unavoidable source of error with the definition used in the included SET studies None of the included SET studies descried their patient criteria in more detail with respect to the infertility definition and length of infertility, which might impact the pregnancy outcome and the possibility of concurrent natural conception. Furthermore, public clinics may have more strict criteria for their patient population than private clinics. Statistical analysis The meta-analysis was performed using Review Manager (RevMan) [Computer programme] (Version 5.3, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). Both a fixed effect and a random effects meta-analysis were performed using the Mantel–Haenszel method. The heterogeneity of the studies was tested with a I2-test, and CI of 95% and a P-value < 0.05 were chosen. A sensitivity analysis was conducted for all the studies and showed very little change in the overall results of the meta-analysis. To examine any publication bias, a funnel plot was generated, and the result was acceptable. Results Blastocyst versus cleavage-stage transfer The search of the PubMed, Embase and Cochrane database resulted in a total of 42 articles, including 38 original studies (Table II). The 21 of the 38 original articles examined or discussed the day of transfer as a potential risk factor for MZT. The studies reported a MZT rate after blastocyst transfer in the range from 0 to 22.2% (Sheiner et al., 2001; Jain et al., 2004; Papanikolaou et al., 2006; Kawachiya et al., 2011; Nakasuji et al., 2014; Sotiroska et al., 2015). The reported rates are based on a large variation in the number of analysed pregnancies. Table II Summary of the 38 articles and meta-analyses of the frequency of MZT after ART. Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) AOR, adjusted odds ratio; BL, blastocyst; CS, cleavage-stage embryo; ET, embryo transfer; GS, gestational sac; MZT, monozygotic twins; OR, odds ratio; SET, single embryo transfer; Sig. diff., significant difference. Table II Summary of the 38 articles and meta-analyses of the frequency of MZT after ART. Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) AOR, adjusted odds ratio; BL, blastocyst; CS, cleavage-stage embryo; ET, embryo transfer; GS, gestational sac; MZT, monozygotic twins; OR, odds ratio; SET, single embryo transfer; Sig. diff., significant difference. Overall, 16 articles reported a significantly higher MZT rate after blastocyst transfer (Kawachiya et al., 2011; Knopman et al., 2014; Luke et al., 2014; Mateizel et al., 2016; Nakasuji et al., 2014; Sharara and Abdo, 2010; Sotiroska et al., 2015), and five studies showed insignificant results (Moayeri et al., 2007; Papanikolaou et al., 2010; Sharara and Abdo, 2010; Osianlis et al., 2014; Wu et al., 2014). Two large studies reported a significantly higher rate of MZT after blastocyst transfer compared with cleavage-stage ET, but when adjusted for potential risk factors (patient age, time period, day of embryo biopsy, embryos for cryopreservation, the number of embryos transferred and embryo quality) the statistical significance disappeared (Knopman et al., 2010; Franasiak et al., 2015). Knopman et al. (2014) published a new cohort study, and apart from a trend toward a lower MZT rate over time, they showed a significantly higher MZT rate after blastocyst transfer in an adjusted model (OR = 1.66; 95% CI: 1.04–2.64; P = 0.04) that included the following co-variates: number of oocytes retrieved at age ≥35 years, donor oocytes, ART treatment cycle after 2004, IVF/ICSI and extended culture. Moayeri et al. (2007) found no difference in the rate of MZT after blastocyst or cleavage-stage transfer (Moayeri et al., 2007). However, by comparing the MZT rate after blastocyst transfer in two different time periods Moayeri et al. (2007) observed that the MZT rate was 2.3% in the late time period (1 March 2002 to 31 December 2005), compared with 5.6% in the early time period before March 2002, demonstrating a significantly lower value in the second interval (P = 0.03). In conclusion, based on 19 studies with high quality (Table I), the majority reported a higher rate of MZT after blastocyst transfer compared with cleavage-stage transfer. Meta-analysis Eleven studies published after 2007 were included in the meta-analysis with a total of 161 885 clinical pregnancies. The 10 out of the 11 included studies showed a higher incidence of MZT after blastocyst transfer compared with cleavage-stage transfer, although only seven reported a significant difference. The fixed effect and random effects models for the meta-analyses revealed comparable results. Both analyses showed a significantly higher rate of MZT following blastocyst transfer compared with cleavage-stage transfer (fixed effect meta-analysis: OR = 2.18, 95% CI: 1.93–2.48; random effects meta-analysis: OR = 2.00, 95% CI: 1.48–2.70) (Fig. 4). Figure 4 View largeDownload slide Meta-analysis and forest plot of MZT after cleavage-stage embryo transfer versus blastocyst transfer. Figure 4 View largeDownload slide Meta-analysis and forest plot of MZT after cleavage-stage embryo transfer versus blastocyst transfer. Controlled ovarian stimulation, ART protocol and culture media Different aspects of ART have been investigated to identify a mechanism and provide an explanation for the high rate of MZT in ART. A few studies have focused on the controlled ovarian stimulation and ART protocol that might affect MZT. Papanikolaou et al. (2010) studied the influence of several types of controlled ovarian stimulation, as well as the protocol used for ART, on the MZT rate following transfer of either blastocyst or cleavage-stage embryos. However, no significant difference was observed, which was in agreement with the study by Nakasuji et al. (2014). One large study focused on the effect of controlled ovarian stimulation and the protocol among other variables in a multivariate analysis (Luke et al., 2014). The findings indicated that the GnRH agonist suppression protocol increased the rate (AOR = 1.10; 95% CI: 1.00–1.22; P = 0.04) and that higher FSH doses (≥3000 IU) decreased the rate (AOR = 0.7; 95% CI: 0.62–0.8; P < 0.0001) of MZT (Luke et al., 2014). The culture media may play a role in relation to the higher frequency of MZT in ART. Several studies have discussed different aspects of this topic, but only a few studies have investigated this phenomenon in practice. One study examined the effect of fresh culture media (grouped in Sage, Medicult, Vitrolife and ‘other’), but they found no differences related to the examined culture media (Mateizel et al., 2016). Another study evaluated at two different media (Medicult and Vitrolife) used for both cleavage-stage and blastocyst embryos, but neither found a difference between the two media (Papanikolaou et al., 2010). In summary only four studies of medium-high quality (Table I) have investigated the role of the controlled ovarian stimulation, the ART protocol or the culture media on the MZT rate, suggesting an increased frequency of MZT following the GnRH agonist suppression protocol and a decreased rate with high FSH doses. Maternal age Fourteen studies have examined maternal age, or oocyte age, as a risk factor for MZT. Eight of the studies found that a younger maternal age increased the rate of MZT although only six were statistically significant. Franasiak et al. (2015) observed a significant reduction in the rate of MZT with increasing age (OR = 0.95; 95% CI: 0.92–0.98; P = 0.001). In a study design from 2015, women undergoing ART were divided into groups based on age (<36 years and >36 years) as well as transfer day (Sotiroska et al., 2015). The study did not only yield a higher multiple gestation rate for the group aged <36 years (twins and triplets, varied from one to three embryos transferred), but a higher MZT rate of 0.5% for cleavage-stage embryo and 2.9% for blastocyst as compared with 0% in the group aged >36 years. Another study reported a 4% decrease in the MZT rate for each year the oocyte age increased (Knopman et al., 2014). A study from 2010 addressed whether the oocyte age or maternal age influences the MZT rate (Knopman et al., 2010). First, they described a significant difference in the MZT rate between autologous cycles and donor oocyte cycles (1.7 versus 3.3%, respectively, P = 0.005). Then they examined the same age group (<35 years), but divided the group into autologous cycles and donor oocyte cycles, revealing no significant difference between the two groups. When examining autologous cycles, a significant difference was observed between women age < 35 and >35 years with respect to the rate of MZT (3.1 versus 1.4%, P = 0.001). This finding suggests that oocyte age is the determining factor in investigations of the influence of age on the MZT rate. In contrast, Mateizel et al. (2016), as well as five other studies, found no influence of maternal age on the MZT rate (da Costa et al., 2001; Alikani et al., 2003; Frankfurter et al., 2004; Skiadas et al., 2008; Nakasuji et al., 2014). Based on 14 studies with a high quality (Table I), maternal age seems to influence the rate of MZT, as some of the most recent studies have reported a higher risk of MZT with younger maternal age or with the use of younger oocytes. Cause of infertility and fertilization method Among several potential risk factors, one study has focused on a possible influence of the indication for ART treatment on the MZT rate, with the following indications: male factor, tubal factor, endometriosis, PCOS and/or anovulation, idiopathic and combined (Papanikolaou et al., 2010). None of the listed indications were associated with the MZT rate. In studies of the influence of the method of fertility treatment on the MZT rate, several aspects should be considered. First, the effects of using ZP manipulation methods, including ICSI, PGD or AH, must be considered. Furthermore, the importance of fresh or frozen ET can be analysed. Several studies have observed a higher MZT rate after the use of ZP manipulation methods. Skiadas et al. (2008) discovered a 2.42-fold increase in the likelihood of MC-twins with the use of ICSI compared with IVF. A small study by Tarlatzis et al. (2002) based only on blastocyst transfers reported a 5.9% rate of MZT after ICSI and 0% after conventional IVF, demonstrating a significant difference (P = 0.033), while Hershlag et al. (1999) reported a significantly higher MZT rate with the use of AH. In contrast, in several other studies, the fertilization method did not influence the MZT rate: ICSI versus conventional IVF was studied by Papanikolaou et al. (2010) and Franasiak, et al. (2015); PGD by Verpoest et al. (2009); ICSI, frozen ET and PGD was studied by Knopman et al. (2014); and only AH and ICSI were studied by Knopman et al. (2010), Mateizel et al. (2016), and Milki et al. (2003). Another study examined the influence of the location of the artificial opening with the use of AH, and found no difference in the numbers of MZT in the two studied locations (Ren et al., 2013). However, Mateizel et al. (2016) discovered a higher MZT rate with the use of fresh ET in comparison with frozen ET (P = 0.008). Based on more than 10 studies of high quality (Table I), there is no clear consensus regarding the influence of the infertility cause and methods of fertility treatment on the frequency of MZT. Embryo morphology One group studied embryo quality based on the number of embryos with six- to eight-cells (Franasiak et al., 2015). Women with four or more embryos in the six- to eight-cell stage and women with more than 75% of all embryos in the six- to eight-cell stage (high quality) had a higher risk of MZT with blastocyst transfer (P = 0.05 and P = 0.04, respectively) than women with one to three embryos in the six- to eight-cell stage embryos or in the category in which <75% of all embryos had six- to eight-cells. The study by Otsuki et al. (2016) found the opposite association, namely, an association between a high embryo quality score and a decreased frequency of MC-DA twins. They used an EmbryoScope time-lapse system (Vitrolife) to retrospectively analyse the quality of the embryos in their cohort. The focus was loosening of the inner cell mass (ICM), defined as the presence of more than five cells loosely arranged in the ICM (in comparison to an earlier observation of tightly grouped cells). The study showed that a low-grade ICM resulted in a 1.38% risk of MC-DA twins, in contrast to a 0.38% risk related to a high-grade ICM (grade A) (P = 0.033) (Otsuki et al., 2016). The true influence of embryo morphology on the MZT rate is still controversial and based only on two high-quality studies (Table I). Discussion A higher rate of MZT is consistently observed after the use of ART; however, the underlying mechanisms are not fully understood. Although our meta-analysis showed a significantly higher rate of MZT after blastocyst transfer, the rate still varied considerably between studies. The causes of this phenomenon may be due to heterogeneity between studies and the varying definition of MZT (Table II). Almost all of the studies defined MZT as two foetal poles within one single gestational sac, or if the total number of foetal poles exceeded the number of embryos transferred, often with the addition of ‘cardiac activity’ for both foetuses (Knopman et al., 2010). This definition clearly excludes DC-DA twins and miscarried MZT, but it also eliminates the possibility of a monozygotic multiple gestation, in which the number of foetal poles did not exceed the number of transferred embryos. The study by Mateizel et al. (2016) (MZT: 103/3672 blastocysts) would not have found a significantly higher MZT rate after blastocyst transfer, if foetal heart beats had been used as a criterion for MZT. This possible definition mis-match was confirmed by a meta-analysis performed in 2009 that reported a difference in the MZT rate based on different definitions (Vitthala et al., 2009). To fully evaluate MZT, DNA testing could be a possible solution (Toledo, 2005). Furthermore, the MZT rate in naturally conceived children is reported per delivery, and hence is, per se, lower than and not directly comparable to the MZT rate reported per clinical pregnancy after ART (Alikani et al., 2003). Some heterogeneity among these studies should be expected as the meta-analysis test of heterogeneity showed (I2=77%, Fig. 4). It would be very difficult to eliminate any sort of heterogeneity when combining studies, or simply data, from clinics worldwide because the definition of infertility, access to ART, methods, and materials may vary to large extents. However, this variability is important to keep in mind when examining and discussing this topic. The true incidence of MZT in ART, in general, as well as after the transfer of a blastocyst, can at the moment best be assessed in studies based on SET, taking into consideration the possibility of concurrent natural conceptions (Osianlis et al., 2014). Ideally, a genetic test should be the gold standard, but the high cost often eliminates this opportunity, in addition to difficulties associated with postpartum follow-up (Knopman et al., 2014). Many studies have claimed that the culture medium developed for culturing blastocysts was the cause of the higher MZT rate after blastocyst transfer. Two studies showed a decrease in MZT after blastocyst transfer over time (Knopman et al. (2014): MZT 107/4494 blastocysts, Moayeri et al. (2007): MZT: 9/385 blastocysts). They both speculated whether the decline was a result of better experiences with both blastocyst culture and transfer. Moayeri et al. (2007) changed both the cleavage-stage culture and blastocyst culture media during their study period and considered it as a possible explanation for the shift in the MZT rate over time. The culture medium used for the control group (data originally published in a study from 2003 (Milki et al., 2003): MZT 11/197 blastocysts) showed a greater difference in glucose levels between cleavage-stage and blastocyst media compared with the medium used for the study group. Cleavage-stage embryos in the control group were cultured in a glucose-free medium and blastocysts in a medium with high glucose levels. The cleavage-stage embryos in the study group were cultured in a medium with a low level of glucose, and the blastocyst medium had an intermediate glucose level. Several studies have proposed that a high level of glucose in the culture medium is a factor influencing the MZT rate after blastocyst transfer because it produces more free radicals leading to apoptosis and weakening and disruption of the ICM (Cassuto et al., 2003; Moayeri et al., 2007; Chang et al., 2009; Papanikolaou et al., 2010). Another theory is that the changes in glucose levels in cultures (cleavage-stage versus blastocyst media) rather than the culture medium itself direct the embryo toward a ‘stress reaction’ that leads to splitting of the ICM (Moayeri et al., 2007). One proposition is the development of more favourable culture conditions (Chang et al., 2009). The presence of co-cultures and growth factors may also play a part in the MZT rate. A lower concentration of grow factors might lead to metabolic stress, resulting in apoptosis or weakened of cellular adhesion (Milki et al., 2003; Moayeri et al., 2007). Another theory concerns the exposure of the blastocyst to a lower level of calcium than that found in the endometrium, and animal models show that blastocyst culture weakens ICM intercellular bonds, which leads to embryo splitting (Milki et al., 2003; Papanikolaou et al., 2010). The results of the current meta-analysis may indicate a contradiction to the presumption that better culture conditions and experience with blastocyst transfer would diminish the higher incidence of MZT after blastocyst transfer. Although the incidence might have decreased to some degree over time (from 3.9–5.6% in 2001–2003 to 2.2–2.9% in 2015/2016), the higher rate of MZT following blastocyst transfer has persisted (da Costa et al., 2001; Tarlatzis et al., 2002; Milki et al., 2003; Sotiroska et al., 2015; Mateizel et al., 2016). Ideally, multicentre studies should be performed that allow a comparison of different media to elucidate the overall influence of these factors. It would be optimal to study the specific media independently in controlled trials to understand the influence of the medium composition on the MZT rate, including O2 tension, although such studies may be controversial and raise ethical questions about embryo research. A majority of recent studies has reported a significant influence of maternal age on the MZT rate (Franasiak et al. (2015): MZT 234/9969 ET; Kawachiya et al. (2011): MZT 151/14 956 ET; Knopman et al. (2010): MZT 98/4976 ET; Knopman et al. (2014): MZT 131/6223 ET; Sotiroska et al. (2015): MZT 7/1400 ET). These studies showed an increase in the MZT rate in cases of young maternal age. Although there is no consensus among all the studies included in this review, it can be speculated that the true underlying mechanism might be the oocyte age, or even the quality of the oocyte, rather than maternal age itself as shown in the study by Knopman et al. (2010). For several years, ZP manipulation was thought to play a role in the higher incidence of MZT after ART (Alikani et al. (2003): MZT 81/4305 ET; Hershlag et al. (1999): MZT 8/391 ET; Saito et al. (2000): MZT 9/279 ET; Schieve et al. (2000): MZT 22/11 247 ET; Skiadas et al. (2008): MZT 41/2501 ET; Tarlatzis et al. (2002): MZT 6/181 cycles). It is plausible that hatching through an artificial hole created in the ZP may cause disruption and splitting of the ICM, resulting in herniation of the blastomeres and embryo (Schieve et al., 2000; Tarlatzis et al., 2002; Milki et al., 2003). However, recent studies have questioned the role of ZP manipulation because these studies did not find an increased rate of MZT after ICSI and AH (Franasiak et al. (2015): MZT 234/9969 ET; Knopman et al. (2010): MZT 98/4976 ET; Knopman et al. (2014): MZT 131/6223 ET; Mateizel et al. (2016): MZT 136/6096 ET; Milki et al. (2003): MZT 18/554 ET; Papanikolaou et al. (2010): MZT 13/579 ET). However, one study showed that ZP disruption raised the MZT rate, questioning the complete role of ZP (Alikani et al. (2003): MZT 81/4305 ET). It has been speculated whether ZP disruption per se or the circumstances leading to the technique caused the increased rate of MZT. Nevertheless, this issue requires further research for clarification, potentially also based on the cause of infertility. Another intriguing aspect is the sex of monozygotic children born after the use of ART. Such a study of the sex requires a follow-up after birth, which unfortunately, was not performed in most of the included studies. However, two different studies, one of which was included in the meta-analysis by Chang et al. (2009), found a higher rate of male infants born after the transfer of a blastocyst compared with the transfer of a cleavage-stage embryo (Sotiroska et al., 2015). A possible thought, as shown in animal studies, is that male embryos develop faster than female embryos (Chang et al., 2009). As the most developed embryos are preferred for extended culture and blastocyst transfer, one may assume that more male blastocysts are transferred and thus a higher number of male MZT. Although highly speculative, the large number reported in the two studies might simply be based on a larger number of developed male embryos—and therefore a higher rate of male blastocysts transferred—rather than on the male sex itself. In total, 42 articles, including systemic reviews, meta-analyses and original studies, were included in the current review. The purpose of using many different search strings as well as MeSH terms was to include as many suitable studies on the subject as possible. However, there are several limitations to this search. Many more search strings could have been used to answer the study questions, which naturally gives rise to some limitations. To streamline the current review, the search words, strings and MeSH terms used were focused on the main question: the role of blastocyst transfer in the MZT rate, and studies may have been excluded throughout the search. Not all of the included original studies in the two meta-analyses were included in this review, which might be an example of the limitations of the search (Chang et al., 2009; Vitthala et al., 2009). However, this has mainly involved older studies, whereas the current review has included several new studies. Conclusion In this meta-analysis, we observed a significantly higher rate of MZT following blastocyst transfer than cleavage-stage embryos (fixed effect meta-analysis: OR = 2.18, 95% CI: 1.93–2.48 and random effects meta-analysis: OR = 2.00, 95% CI: 1.48–2.70). We consider the results of this meta-analysis to question the trend theory that more experience with blastocyst transfer and better culture media decrease the high rate of MZT after blastocyst transfer, which has been proposed previously. Several studies, although not all, have also reported a reduced rate of MZT with increased maternal age, although how age influences the frequency of MZT remains to be explored. Difficulties related to how to correctly define and identify MZT also predominate as obstacles that must to be solved. The current review clearly indicates that more research is needed to fully understand the mechanism underlying MZT after ART and to provide useful guidelines for fertility treatment to minimize the rate of MZT and, thereby, the risks associated with multiple monozygotic gestation and potential subsequent complications. Large multi-centre studies based on SET alone are needed to explore the potential risk factors, such as those discussed in the current review. Sufficiently powered studies are important to capture the low incidence of MZT. An ideal study would also include DNA testing as well as embryo time-lapse examination by EmbryoScope to discover potential signs of embryo splitting. Acknowledgements We thank Prof. T. Hviid for creating the schematic drawing presented in Figure 1. Authors’ roles K.V.R.H. performed the search and data collection, wrote the first version of the article, and performed the meta-analysis. H.S.N. conceived the basic idea and supervised the review and data collection. A.P., S.S.M., H.S.N and K.V.R.H. participated in the quality assessment and critically discussed, read and approved the final article. Funding K.V.R.H. has obtained a research scholarship from Rigshospitalet, Copenhagen. Conflicts of interest None declared. References Alikani M , Cekleniak NA , Walters E , Cohen J . Monozygotic twinning following assisted conception: an analysis of 81 consecutive cases . Hum Reprod 2003 ; 18 : 1937 – 1943 . Google Scholar CrossRef Search ADS PubMed Aston KI , Peterson CM , Carrell DT . Monozygotic twinning associated with assisted reproductive technologies: a review . Reproduction 2008 ; 136 : 377 – 386 . Google Scholar CrossRef Search ADS PubMed Behr B , Fisch JD , Racowsky C , Miller K , Pool TB , Milki AA . Blastocyst-ET and monozygotic twinning . J Assist Reprod Genet 2000 ; 17 : 349 – 351 . Google Scholar CrossRef Search ADS PubMed Blickstein I , Verhoeven HC , Keith LG . Zygotic splitting after assisted reproduction . N Engl J Med 1999 ; 340 : 738 – 739 . Google Scholar CrossRef Search ADS PubMed Blickstein I , Jones C , Keith LG . Zygotic-splitting rates after single-embryo transfers in in vitro fertilization . N Engl J Med 2003 ; 348 : 2366 – 2367 . Google Scholar CrossRef Search ADS PubMed Cassuto G , Chavrier M , Menezo Y . Culture conditions and not prolonged culture time are responsible for monozygotic twinning in human in vitro fertilization . Fertil Steril 2003 ; 80 : 462 – 463 . Google Scholar CrossRef Search ADS PubMed Chang HJ , Lee JR , Jee BC , Suh CS , Kim SH . Impact of blastocyst transfer on offspring sex ratio and the monozygotic twinning rate: a systematic review and meta-analysis . Fertil Steril 2009 ; 91 : 2381 – 2390 . Google Scholar CrossRef Search ADS PubMed da Costa ALE , Abdelmassih S , de Oliveira FG , Abdelmassih V , Abdelmassih R , Nagy ZP , Balmaceda JP . Monozygotic twins and transfer at the blastocyst stage after ICSI . Hum Reprod 2001 ; 16 : 333 – 336 . Google Scholar CrossRef Search ADS PubMed Edwards RG . An astonishing journey into reproductive genetics since the 1950’s . Reprod Nutr Dev 2005 ; 45 : 299 – 306 . Google Scholar CrossRef Search ADS PubMed Elizur SE , Levron J , Shrim A , Sivan E , Dor J , Shulman A . Monozygotic twinning is not associated with zona pellucida micromanipulation procedures but increases with high-order multiple pregnancies . Fertil Steril 2004 ; 82 : 500 – 501 . Google Scholar CrossRef Search ADS PubMed Franasiak JM , Dondik Y , Molinaro TA , Hong KH , Forman EJ , Werner MD , Upham KM , Scott RT Jr . Blastocyst transfer is not associated with increased rates of monozygotic twins when controlling for embryo cohort quality . Fertil Steril 2015 ; 103 : 95 – 100 . Google Scholar CrossRef Search ADS PubMed Frankfurter D , Trimarchi J , Hackett R , Meng L , Keefe D . Monozygotic pregnancies from transfers of zona-free blastocysts . Fertil Steril 2004 ; 82 : 483 – 485 . Google Scholar CrossRef Search ADS PubMed Fuchs F , Senat MV . Multiple gestations and preterm birth . Semin Fetal Neonatal Med 2016 ; 21 : 113 – 120 . Google Scholar CrossRef Search ADS PubMed Graham MC , Hoeger KM , Phipps WR . Initial IVF-ET experience with assisted hatching performed 3 days after retrieval followed by day 5 embryo transfer . Fertil Steril 2000 ; 74 : 668 – 671 . Google Scholar CrossRef Search ADS PubMed Hankins GVD , Saade GR . Factors influencing twins and zygosity . Paediatr Perinat Epidemiol 2005 ; 19 : 8 – 9 . Google Scholar CrossRef Search ADS PubMed Hershlag A , Paine T , Cooper GW , Scholl GM , Rawlinson K , Kvapil G . Monozygotic twinning associated with mechanical assisted hatching . Fertil Steril 1999 ; 71 : 144 – 146 . Google Scholar CrossRef Search ADS PubMed Jain JK , Boostanfar R , Slater CC , Francis MM , Paulson RJ . Monozygotic twins and triplets in association with blastocyst transfer . J Assist Reprod Genet 2004 ; 21 : 103 – 107 . Google Scholar CrossRef Search ADS PubMed Kawachiya S , Bodri D , Shimada N , Kato K , Takehara Y , Kato O . Blastocyst culture is associated with an elevated incidence of monozygotic twinning after single embryo transfer . Fertil Steril 2011 ; 95 : 2140 – 2142 . Google Scholar CrossRef Search ADS PubMed Knopman J , Krey LC , Lee J , Fino ME , Novetsky AP , Noyes N . Monozygotic twinning: an eight-year experience at a large IVF center . Fertil Steril 2010 ; 94 : 502 – 510 . Google Scholar CrossRef Search ADS PubMed Knopman JM , Krey LC , Oh C , Lee J , McCaffrey C , Noyes N . What makes them split? Identifying risk factors that lead to monozygotic twins after in vitro fertilization . Fertil Steril 2014 ; 102 : 82 – 89 . Google Scholar CrossRef Search ADS PubMed Kontopoulos E , Chmait RH , Quintero RA . Twin-to-twin transfusion syndrome: definition, staging, and ultrasound assessment . Twin Res Hum Genet 2016 ; 19 : 175 – 183 . Google Scholar CrossRef Search ADS PubMed Luke B , Brown MB , Wantman E , Stern JE . Factors associated with monozygosity in assisted reproductive technology pregnancies and the risk of recurrence using linked cycles . Fertil Steril 2014 ; 101 : 683 – 689 . Google Scholar CrossRef Search ADS PubMed Mateizel I , Van Landuyt L , Tournaye H , Verheyen G . Deliveries of normal healthy babies from embryos originating from oocytes showing the presence of smooth endoplasmic reticulum aggregates . Hum Reprod 2013 ; 28 : 2111 – 2117 . Google Scholar CrossRef Search ADS PubMed Mateizel I , Santos-Ribeiro S , Done E , Van Landuyt L , Van de Velde H , Tournaye H , Verheyen G . Do ARTs affect the incidence of monozygotic twinning? Hum Reprod 2016 ; 31 : 2435 – 2441 . Google Scholar CrossRef Search ADS PubMed Milki AA , Jun SH , Hinckley MD , Behr B , Giudice LC , Westphal LM . Incidence of monozygotic twinning with blastocyst transfer compared to cleavage-stage transfer . Fertil Steril 2003 ; 79 : 503 – 506 . Google Scholar CrossRef Search ADS PubMed Moayeri SE , Behr B , Lathi RB , Westphal LM , Milki AA . Risk of monozygotic twinning with blastocyst transfer decreases over time: an 8-year experience . Fertil Steril 2007 ; 87 : 1028 – 1032 . Google Scholar CrossRef Search ADS PubMed Nakasuji T , Saito H , Araki R , Nakaza A , Nakashima A , Kuwahara A , Ishihara O , Irahara M , Kubota T , Yoshimura Y et al. . The incidence of monozygotic twinning in assisted reproductive technology: analysis based on results from the 2010 Japanese ART national registry . J Assist Reprod Genet 2014 ; 31 : 803 – 807 . Google Scholar CrossRef Search ADS PubMed Osianlis T , Rombauts L , Gabbe M , Motteram C , Vollenhoven V . Incidence and zygosity of twin births following transfers using a single fresh or frozen embryo . Hum Reprod 2014 ; 29 : 1438 – 1443 . Google Scholar CrossRef Search ADS PubMed Otsuki J , Iwasaki T , Katada Y , Sato H , Furuhashi K , Tsuji Y , Matsumoto Y , Shiotani M . Grade and looseness of the inner cell mass may lead to the development of monochorionic diamniotic twins . Fertil Steril 2016 ; 106 : 640 – 644 . Google Scholar CrossRef Search ADS PubMed Papanikolaou EG , Camus M , Kolibianakis EM , Van Landuyt L , Van Steirteghem A , Devroey P . In vitro fertilization with single blastocyst-stage versus single cleavage-stage embryos . N Engl J Med 2006 ; 354 : 1139 – 1146 . Google Scholar CrossRef Search ADS PubMed Papanikolaou EG , Fatemi H , Venetis C , Donoso P , Kolibianakis E , Tournaye H , Tarlatzis B , Devroey P . Monozygotic twinning is not increased after single blastocyst transfer compared with single cleavage-stage embryo transfer . Fertil Steril 2010 ; 93 : 592 – 597 . Google Scholar CrossRef Search ADS PubMed Parisaei M , Shailendra A , Dutta R , Broadbent JA Obstetrics and Gynaecology , 2nd edn . Edinburgh: Mosby Ltd , 2008 . Pinborg A . IVF/ICSI twin pregnancies: risks and prevention . Hum Reprod Update 2005 ; 11 : 575 – 593 . Google Scholar CrossRef Search ADS PubMed Pinborg A , Lidegaard O , la Cour Freiesleben N , Andersen AN . Consequences of vanishing twins in IVF/ICSI pregnancies . Hum Reprod 2005 ; 20 : 2821 – 2829 . Google Scholar CrossRef Search ADS PubMed Ren X , Liu Q , Chen W , Zhu G , Zhang H . Effect of the site of assisted hatching on vitrified-warmed blastocyst transfer cycles: a prospective randomized study . J Assist Reprod Genet 2013 ; 30 : 691 – 697 . Google Scholar CrossRef Search ADS PubMed Saito H , Tsutsumi O , Noda Y , Ibuki Y , Hiroi M . Do assisted reproductive technologies have effects on the demography of monozygotic twinning? Fertil Steril 2000 ; 74 : 178 – 179 . Google Scholar CrossRef Search ADS PubMed Schieve LA , Meikle SF , Peterson HB , Jeng G , Burnett NM , Wilcox LS . Does assisted hatching pose a risk for monozygotic twinning in pregnancies conceived through in vitro fertilization? Fertil Steril 2000 ; 74 : 288 – 294 . Google Scholar CrossRef Search ADS PubMed Sharara FI , Abdo G . Incidence of monozygotic twins in blastocyst and cleavage stage assisted reproductive technology cycles . Fertil Steril 2010 ; 93 : 642 – 645 . Google Scholar CrossRef Search ADS PubMed Sheiner E , Har-Vardi I , Potashnik G . The potential association between blastocyst transfer and monozygotic twinning . Fertil Steril 2001 ; 75 : 217 – 218 . Google Scholar CrossRef Search ADS PubMed Skiadas CC , Jackson KV , Racowsky C . Early compaction on day 3 may be associated with increased implantation potential . Fertil Steril 2006 ; 86 : 1386 – 1391 . Google Scholar CrossRef Search ADS PubMed Skiadas CC , Missmer SA , Benson CB , Gee RE , Racowsky C . Risk factors associated with pregnancies containing a monochorionic pair following assisted reproductive technologies . Hum Reprod 2008 ; 23 : 1366 – 1371 . Google Scholar CrossRef Search ADS PubMed Sobek A , Prochazka M , Klaskova E , Lubusky M , Pilka R . High incidence of monozygotic twinning in infertility treatment . Biomed papers 2016 ; 160 : 358 – 362 . Google Scholar CrossRef Search ADS Sotiroska V , Petanovski Z , Dimitrov G , Hadji-Lega M , Shushleski D , Saltirovski S , Matevski V , Shenbakar S , Panov S , Johansson L . The day of embryo transfer affects delivery rate, birth weights, female-to-male ratio, and monozygotic twin rate . Taiwan J Obstet Gynecol 2015 ; 54 : 716 – 721 . Google Scholar CrossRef Search ADS PubMed Tarlatzis BC , Qublan HS , Sanopoulou T , Zepiridis L , Grimbizis G , Bontis J . Increase in the monozygotic twinning rate after intracytoplasmic sperm injection and blastocyst stage embryo transfer . Fertil Steril 2002 ; 77 : 196 – 198 . Google Scholar CrossRef Search ADS PubMed Toledo MG . Is there increased monozygotic twinning after assisted reproductive technology? Aust N Zeal J Obstet Gynaecol 2005 ; 45 : 360 – 364 . Google Scholar CrossRef Search ADS Van Landuyt L , De Vos A , Joris H , Verheyen G , Devroey P , Van Steirteghem A . Blastocyst formation in in vitro fertilization versus intracytoplasmic sperm injection cycles: influence of the fertilization procedure . Fertil Steril 2005 ; 83 : 1397 – 1403 . Google Scholar CrossRef Search ADS PubMed Vaughan DA , Robin R , Penzias AS , Norwitz ER , Sakkas D . Clustering of monozygotic twinning in IVF . Assist Reprod Technol 2015 ; 33 : 19 – 26 . Google Scholar CrossRef Search ADS Verpoest W , Van Landuyt L , Desmyttere S , Cremers A , Devroey P , Liebaers I . The incidence of monozygotic twinning following PGD is not increased . Hum Reprod 2009 ; 24 : 2945 – 2950 . Google Scholar CrossRef Search ADS PubMed Vitthala S , Gelbaya TA , Brison DR , Fitzgerald CT , Nardo LG . The risk of monozygotic twins after assisted reproductive technology: a systematic review and meta-analysis . Hum Reprod Update 2009 ; 15 : 45 – 55 . Google Scholar CrossRef Search ADS PubMed Wright V , Schieve LA , Vahratian A , Reynolds MA . Monozygotic twinning associated with day 5 embryo transfer in pregnancies conceived after IVF . Hum Reprod 2004 ; 19 : 1831 – 1836 . Google Scholar CrossRef Search ADS PubMed Wu D , Huang SY , Wu HM , Chen CK , Soong YK , Huang HY . Monozygotic twinning after in vitro fertilization/intracytoplasmic sperm injection treatment is not related to advanced maternal age, intracytoplasmic sperm injection, assisted hatching, or blastocyst transfer . Taiwan J Obstet Gynecol 2014 ; 53 : 324 – 329 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Human Reproduction Update Oxford University Press

Determinants of monozygotic twinning in ART: a systematic review and a meta-analysis

Loading next page...
 
/lp/ou_press/determinants-of-monozygotic-twinning-in-art-a-systematic-review-and-a-AV8Kwaksmm
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1355-4786
eISSN
1460-2369
D.O.I.
10.1093/humupd/dmy006
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND The incidence of monozygotic twins (MZT) after ART appears to be higher than the incidence after spontaneous conceptions contradicting the aim of ART to avoid multiple pregnancies because of the associated risks. OBJECTIVE AND RATIONALE The aim was to study the frequency of MZT after IVF and ICSI and how it is influenced by the day of embryo transfer, maternal age, zona pellucida manipulation, controlled ovarian stimulation, stimulation protocol, culture media and embryo quality. SEARCH METHODS Original studies and reviews were identified by searching the PubMed, Embase and Cochrane databases up to March 2017. The inclusion criterion was publications focusing on the five study questions related to MZT in our study. The exclusion criteria were articles that did not include blastocyst transfer, were on non-humans, were not published in peer-reviewed journals, and were based only on case studies. All of the articles were categorized according to the Oxford Centre for Evidence-based Medicine’s ‘Levels of Evidence’, and quality and risk of bias assessment was performed with ‘The Cochrane Collaboration’s Risk of Bias Tools’. A meta-analysis was performed to study the impact of the day of embryo transfer on the MZT rate. OUTCOMES The literature search resulted in a total of 42 articles, including 38 original studies, for analysis. The included original studies reported a MZT rate with blastocyst transfer from zero to 13.2%. Our meta-analysis found a higher frequency of MZT after blastocyst transfer compared with cleavage-stage embryos transfer: odds ratio = 2.18, 95% CI: 1.93–2.48 (fixed effect meta-analysis). A younger maternal age may increase the MZT rate, and recent studies regarding the use of zona pellucida manipulating techniques have disagreed with the previous suspicion of a higher MZT rate after the use of these methods. The extended culture to-blastocyst stage is a potential risk factor for MZT, but it is uncertain whether this phenomenon is due to the extended time, culture media or greater likelihood of younger oocytes to reach the blastocyst stage. An increased frequency of MZT following the GnRH-agonist suppression protocol has been suggested, as well as a decreased frequency of MZT with high gonadotrophin doses, which could reflect an age-related effect. Only limited literature has focused on the role of embryo morphology in the MZT rate, therefore, this issue remains unresolved. WIDER IMPLICATIONS We found blastocyst transfer to be a risk factor for MZT. Hence, the results of this meta-analysis may weaken the previously proposed view that greater experience with blastocyst transfer and improved culture media could decrease the high rate of MZT after blastocyst transfer. To minimize the rate of MZT and the associated complications, the mechanisms underlying blastocyst transfer and MZT pregnancy must be elucidated. monozygotic twins, cleavage-stage, blastocyst, ART, IVF, ICSI, assisted hatching, culture media, meta-analysis Introduction For decades, it has been possible to help infertile women and men have biological children by using ART (Edwards, 2005). ART has been associated with multiple gestations as a result of the transference of two or more embryos, which may develop into dizygotic twins (Sobek et al., 2016). However, monozygotic twins (MZT) arise from one fertilized oocyte that divides into two or even three or four. So although MZT can occur after multiple embryo transfer (ET), only one fertilized oocyte is required (Sobek et al., 2016). Nevertheless, the frequency of MZT after ART varies from 0 to 13.2% compared with a spontaneous conception rate of 0.4% of live births (Milki et al., 2003; Jain et al., 2004; Papanikolaou et al., 2006; Moayeri et al., 2007; Parisaei et al., 2008; Vitthala et al., 2009; Kawachiya et al., 2011; Franasiak et al., 2015). However, in most of the included studies, the rate of MZT after spontaneous conception was calculated based on live births and not early ultrasound. Thus, the MZT rate after spontaneous conception is likely to be considerably higher if it is also based on early ultrasound. The chorionicity of MZT is assumed to depend on the day of embryo division (Fig. 1). If the oocyte divides shortly after fertilization, the foetuses will be dichorionic diamniotic (DC-DA), which occurs in 20–30% of cases after spontaneous conception. If the division occurs at the blastocyst stage, the gestation will most likely become monochorionic diamniotic (MC-DA), which is observed in 70–75% of cases. After 1–2 weeks, division of the embryo will result in monochorionic monoamniotic (MC-MA) twins only in 1–2% of live MZT births (Aston et al., 2008; Sobek et al., 2016). Figure 1 View largeDownload slide A schematic drawing showing the three different types of monozygotic twins and their potential origins. Figure 1 View largeDownload slide A schematic drawing showing the three different types of monozygotic twins and their potential origins. A multiple pregnancy involves numerous risks for the fetuses as well as for the mother, demonstrating some relationship to the chorionicity of the foetuses. Obstetric risks include low birthweight, preterm birth, pre-eclampsia, nausea and vomiting of the mother, vanishing twin syndrome and perinatal mortality (Pinborg et al., 2005; Pinborg, 2005; Aston et al., 2008; Fuchs and Senat, 2016). Furthermore, MZT are at risk for developing twin-to-twin transfusion syndrome (TTTS) (Kontopoulos et al., 2016). TTTS is most likely to occur in monochorionic twins, but exceptions are observed with dichorionic monozygotic and even with dichorionic-dizygotic twins (Kontopoulos et al., 2016). TTTS arises when there is an imbalance in blood flow between two MZT because of vascular anastomoses in the placenta of monochorionic twins, or between the two placentas with dichorionic monozygotic (or rarely seen dizygotic) twins. TTTS occurs in 5–20% of monochorionic twins, and left untreated, the risk of pregnancy loss can reach 95% (Kontopoulos et al., 2016). MC-MA twins also have a risk of umbilical cord entanglement because they share the same amniotic sac. Given the higher risks, the overall goal is to minimize the number of multiple gestations in ART pregnancies. Numerous studies have found a higher clinical pregnancy rate with the transfer of a blastocyst compared with a cleavage-stage embryo (Sharara and Abdo, 2010; Kawachiya et al., 2011; Sotiroska et al., 2015). This allows, in most cases, a practice to transfer only one embryo and reduce the risk of multiple gestations. Previous studies have disagreed upon the incidence of MZT in ART, as well as the underlying mechanisms. Investigations of potential risk factors, such as the age of the oocyte, extended culture and assisted hatching (AH), have provided no clear explanation of the aetiology (Fig. 2) (Knopman et al., 2014). Likewise, the embryo quality may be associated with MZT (Franasiak et al., 2015). Finally, some studies have not found a correlation of MZT with blastocyst transfer compared with cleavage-stage ET (Moayeri et al., 2007; Papanikolaou et al., 2010; Osianlis et al., 2014; Wu et al., 2014). Figure 2 View largeDownload slide Some of the possible mechanisms underlying the increased rate of monozygotic twins in ART. MZT, monozygotic twins; ZP, zona pellucida; AH, assisted hatching. Figure 2 View largeDownload slide Some of the possible mechanisms underlying the increased rate of monozygotic twins in ART. MZT, monozygotic twins; ZP, zona pellucida; AH, assisted hatching. The aim of this systematic review is to examine the risk factors for MZT in ART treatment by addressing the following five questions: – Does the transfer of cleavage-stage versus blastocyst stage embryos (Day 2/3 versus Day 5) influence the frequency of MZT? A meta-analysis is conducted. – Does controlled ovarian stimulation, the ART protocol or the culture media affect the frequency of MZT? – Does maternal age affect the frequency of MZT? – Does the cause of infertility or the fertilization method (IVF versus ICSI) affect the frequency of MZT? – Does the morphology of the embryo affect the frequency of MZT? Methods Sources and literature search We used the PubMed, Embase and Cochrane databases up until March 2017 to identify the articles included in this review. To cover the subject and study questions, the following search strings were used: ‘Monozygotic twins AND blastocyst’, ‘Monozygotic twins AND ART’, ‘Monozygotic twins AND media’, ‘Monozygotic twins AND ovarian stimulation’, ‘Monozygotic twins AND maternal age’, ‘Monozygotic twins AND cause of infertility’, ‘Monozygotic twins AND Day 5’ and ‘Monozygotic twins AND assisted reproductive technology’. We used the same search strings with the exchange of the word ‘Monozygotic twins’ with ‘Monochorionic twins’. The MeSH terms used were as follows: monozygotic twins, monochorionic twins, blastocyst, assisted reproductive technology, ART, media, ovarian stimulation, maternal age, cause of infertility and Day 5. Study selection The different searches identified 2722 scientific articles. We excluded articles in non-English languages. In a single search, it was necessary to apply the criterion of only articles from the last 10 years (‘Monozygotic twins AND maternal age’). After removal of non-English articles and non-accessible full-length articles, a total of 1841 articles were selected for screening. The title and/or abstract of the 1841 articles were screened. Articles were included in the review based on the following criteria: Inclusion criteria: articles focusing on one of the five study-questions listed in the Introduction. Exclusion criteria: articles that did not include blastocyst transfer, were on non-humans, were not published in peer-reviewed journals, and were based only on case studies. A few of the included articles did not directly focus on blastocyst transfer, but were selected because they were included in two major meta-analyses (Chang et al., 2009; Vitthala et al., 2009). Finally, we included 42 articles after screening (Table I). The search results are shown in Fig. 3. Table I Quality of studies, based on The Cochrane Collaboration’s Risk of Bias Tools, included in a systematic review and meta-analysis of determinants of monozygotic twinning in ART. Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Table I Quality of studies, based on The Cochrane Collaboration’s Risk of Bias Tools, included in a systematic review and meta-analysis of determinants of monozygotic twinning in ART. Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Study Study type Levels of evidence The Cochrane Collaboration’s Risk of Bias Tools score Mateizel et al. (2016) Retrospective cohort study 2b High Otsuki et al. (2016) Retrospective cohort study 2b High Sotiroska et al. (2015) Retrospective cohort study 2b High Vaughan (2015) Retrospective cohort study 2b High Franasiak et al. (2015) Retrospective cohort study 2b High Wu et al. (2014) Retrospective cohort study 2b High Knopman et al. (2014) Retrospective cohort study (Nested-case–control) 2b High Osianlis et al. (2014) Retrospective cohort study 2b High Luke et al. (2014) (Database/historical) Cohort study 2b Medium Nakasuji et al. (2014) Retrospective cohort study 2b High Ren et al. (2013) Randomized case control 1b High Kawachiya et al. (2011) Retrospective cohort study 2b High Knopman et al. (2010) Retrospective cohort study 2b High Papanikolaou et al. (2010) Retrospective cohort study 2b High Sharara and Abdo (2010) Retrospective cohort study 2b High Verpoest et al. (2009) Retrospective cohort study 2b High Skiadas et al. (2008) Retrospective cohort study 2b High Moayeri et al. (2007) Retrospective cohort study 2b Medium Papanikolaou et al. (2006) Randomized case control 1b High Jain et al. (2004) Retrospective cohort study 2b Medium Elizur et al. (2004) Retrospective cohort study 2b Medium Frankfurter et al. (2004) Retrospective cohort study 2b Medium Wright et al. (2004) Retrospective cohort study 2b High Milki et al. (2003) Retrospective cohort study 2b High Blickstein et al. (2003) Retrospective cohort study 2b Medium Cassuto et al. (2003) Retrospective cohort study 2b Medium Alikani et al. (2003) Retrospective cohort study 2b High Tarlatzis et al. (2002) Retrospective cohort study 2b High Sheiner et al. (2001) Retrospective cohort study 2b Medium da Costa et al. (2001) Retrospective cohort study 2b High Schieve et al. (2000) Retrospective cohort study 2b High Saito et al. (2000) Retrospective cohort study 2b High Behr et al. (2000) Retrospective cohort study 2b High Graham et al. (2000) Retrospective cohort study 2b Medium Hershlag et al. (1999) Retrospective cohort study 2b Medium Figure 3 View largeDownload slide PRISMA flow chart of the study selection (March 2017). Figure 3 View largeDownload slide PRISMA flow chart of the study selection (March 2017). Four of the 42 articles were reviews and therefore not included in Table I (Hankins and Saade, 2005; Toledo, 2005; Aston et al., 2008; Sobek et al., 2016). Study quality All articles were categorized according to ‘Levels of Evidence’ (Oxford Centre for Evidence-based Medicine guidelines; http://www.cebm.net/oxford-centre-evidence-based-medicine-levels-evidence-march-2009/; 28 March 2018, date last accessed), and quality and bias were assessed with ‘The Cochrane Collaboration’s Risk of Bias Tools’ (http://methods.cochrane.org/sites/methods.cochrane.org.bias/files/public/uploads/Tool%20to%20Assess%20Risk%20of%20Bias%20in%20Cohort%20Studies.pdf; 28 March 2018, date last accessed). In summary, the risk of bias tools consisted of the following questions: whether exposed and non-exposed cohorts are drawn from the same population; confidence in the assessment of exposure; confidence that the outcome of interest was not present at the start of study; exposed and unexposed were matched for all variables associated with the outcome of interest; confidence in the assessment of the presence or absence of prognostic factors; confidence in the assessment of the outcome; whether the cohort follow-up was adequate; whether the co-interventions were similar between groups. Although two of the studies (Papanikolaou et al., 2006; Ren et al., 2013) were categorized as 1b (individual RCT (narrow CI)) by the Levels of Evidence, they were assessed by the same questions regarding quality and bias. All studies showed good quality and a low risk of bias, and hence, none of the studies were categorized as low quality (Table I). Based on a review of the results using The Cochrane Collaboration’s Risk of Bias Tools, the results were categorized as either medium or high quality. Study selection for the meta-analysis Studies published after the latest two meta-analyses by Chang et al. (2009) and Vitthala et al. (2009) (search ending in 2007) were chosen for this meta-analysis. The 6 of the 17 identified studies had to be excluded from the meta-analysis (Verpoest et al., 2009; Knopman et al., 2010; Ren et al., 2013; Luke et al., 2014; Sotiroska et al., 2015; Otsuki et al., 2016). Exclusion criteria were as follows: studies not reporting accurate numbers for MZT following either blastocyst transfer or cleavage-stage transfer, no control group (cleavage-stage embryos), if part of the data had been published in another study, and the lack of a defined method for identifying MZT. The study by Osianlis et al. (2014) addressed the question of concurrent natural conception and showed an inconsistency in their original MZT data because of gender discordance between twins born after single ET (SET). We chose to include Osianlis et al. (2014) in the meta-analysis because all the SET studies showed limitations in identifying MZT twins, as noted in the study by Osianlis et al. (2014). All the included SET studies defined twin pregnancies as monozygotic and thereby overlooked the possibility of dizygotic twins following concurrent natural conception. This is an unavoidable source of error with the definition used in the included SET studies None of the included SET studies descried their patient criteria in more detail with respect to the infertility definition and length of infertility, which might impact the pregnancy outcome and the possibility of concurrent natural conception. Furthermore, public clinics may have more strict criteria for their patient population than private clinics. Statistical analysis The meta-analysis was performed using Review Manager (RevMan) [Computer programme] (Version 5.3, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). Both a fixed effect and a random effects meta-analysis were performed using the Mantel–Haenszel method. The heterogeneity of the studies was tested with a I2-test, and CI of 95% and a P-value < 0.05 were chosen. A sensitivity analysis was conducted for all the studies and showed very little change in the overall results of the meta-analysis. To examine any publication bias, a funnel plot was generated, and the result was acceptable. Results Blastocyst versus cleavage-stage transfer The search of the PubMed, Embase and Cochrane database resulted in a total of 42 articles, including 38 original studies (Table II). The 21 of the 38 original articles examined or discussed the day of transfer as a potential risk factor for MZT. The studies reported a MZT rate after blastocyst transfer in the range from 0 to 22.2% (Sheiner et al., 2001; Jain et al., 2004; Papanikolaou et al., 2006; Kawachiya et al., 2011; Nakasuji et al., 2014; Sotiroska et al., 2015). The reported rates are based on a large variation in the number of analysed pregnancies. Table II Summary of the 38 articles and meta-analyses of the frequency of MZT after ART. Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) AOR, adjusted odds ratio; BL, blastocyst; CS, cleavage-stage embryo; ET, embryo transfer; GS, gestational sac; MZT, monozygotic twins; OR, odds ratio; SET, single embryo transfer; Sig. diff., significant difference. Table II Summary of the 38 articles and meta-analyses of the frequency of MZT after ART. Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) Study, country MZT/total pregnancies Ovarian stimulation protocol Type of ET (cleavage-stage (CS)/Blastocyst (BL) Culture medium MZT (%) Method used for MZT diagnosis Mateizel et al. (2016), Belgium 136/6096 Natural cycles or stimulated Urinary (Menopur, Ferring Pharmaceuticals A/S, Copenhagen, Denmark) or recombinant FSH And GnRH antagonist or agonist Mateizel et al. (2013) Both CS/BL only SET Irvine HTF medium (Irvine Scientific, Dublin, Ireland) or Cook IVF Cleavage and Blastocyst Media (Cook Belgium, Strombeek-Bever, Belgium) or Vitrolife Sequential media G series (Vitrolife, Goteborg, Sweden), or Medicult, EmbryoAssist and BlastAssist Media (Medicult, De Pinte, Belgium) or Sage Quinns Advantage Protein Plus Cleavage and Blastocyst Media (Cooper Surgical, USA) MZT overall: 2.2% BL (early or advanced) versus CS Early: AOR: 2.70 (95% CI 1.36–5.34) Advanced: AOR: 2.05 (95% CI 1.29–3.26) Sig. diff. BL and CS Ultrasound scan: More than one GS was visualized or when the number of foetal poles exceeded the number of GSs Otsuki et al. (2016), Japan 80/3445 Standard GnRH agonist/FSH protocols or antagonist/FSH protocol Only SET BL Universal IVF Medium (Origio a/s) overlaid with mineral oil (Irvine Scientific). After 21–23 h global medium (LifeGlobal) with 10% human serum albumin (HSA; LifeGlobal) 2.32% Ultrasound scan showing two foetal poles Sotiroska et al. (2015), Macedonia 7/1400 <36 years: 7/958 >36 years: 0/442 Long protocol: Down-regulation was initiated with buserelin acetate Injections of recombinant follitropin beta (Puregon, N.V. Oregon, Os, The Netherlands) were initiated on the third day of menstrual cycle Both CS/BL All oocytes and embryos were cultured in preincubated Quinns Advantage sequential media under mineral oil (SAGE, Cooper- Surgical, Trumbull, CT, USA) <36 years: CS: 0.5% BL: 2.9% (P < 0.05) Sig. diff. >36 years: no MZT Not reported Vaughan (2015), USA 95/8598 Ovarian stimulation protocols with gonadotropins and either a GnRH agonist or antagonist Both CS/BL Different types of embryo culture, protein supplements over the stud period from January 2002 to December 2013 CS: 0.7% BL: 2.7% (P < 0.05) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Franasiak et al. (2015), USA 234/9969 Long down-regulation: FSH or recombinant FSH and LH or hMG with GnRH agonist or GnRH antagonist Both CS/BL Quinns Advantage (Cooper Surgical) followed by BlastAssist (Origio) CS: 1.9% BL: 2.8% (P = 0.002) Sig. diff. Multivariate logistic analysis for embryo cohort quality: OR: 0.92 (95% CI 0.47–1.8; P = 0.8) No sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Wu et al. (2014), Taiwan 17/1338 Long protocol (downregulation protocol) using a GnRH agonist (Lupron; Takeda, Osaka, Japan) or the GnRH antagonist (Cetrotide; Serono, Geneva, Switzerland) protocol Both CS/BL Cleavage culture medium equilibrated with 6% CO2 in air until Day 3. If embryos were cultured beyond Day 3, blastocyst culture medium (Cook IVF) MZT overall: 1.3% CS: 1.3% BL: 1.4% No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Knopman et al. (2014), USA 131/6223 GnRH down-regulation and antagonist options. Half GnRH agonist down-regulation + FSH and/or hMG, Other half GnRH-antagonist. Both CS/BL Quinns cleavage medium 18 h to Day 3. Days 3–5: Quinns blastocyst medium MZT overall: 2.1% BL: 2.4% OR: 1.59, (95% CI: 1.01–2.48) Sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Osianlis et al. (2014), Australia 109/4701 Down-regulation was achieved with either a GnRH agonist or a GnRH antagonist Both CS/BL only SET COOK culture system (COOK Medical, Australia) MZT overall: 2.9% CS: 2.0% BL: 2.4% P > 0.05 No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Luke et al. (2014), USA 2829/197 327 Not reported Both CS/BL Not reported MZT overall: 1.4% one embryo: AOR: 1.59; (95% CI 1.27–2.00) two embryos: AOR: 2.33 (95% CI 2.08–256; P < 0.001) Sig. diff. Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred. Nakasuji et al. (2014), Japan 425/30.405 Based on Japanese ART national registry. Ovarian stimulation protocols (no stimulation versus clomiphene citrate, HMG or FSH) Both CS/BL Only SET Not reported CS: 0.2% BL: 0.6% (P < 0.01) Sig. diff. Ultrasound scan: More than one foetal pole observed Ren et al. (2013), China 5/105 Ovarian stimulation was performed using FSH (Gonal-F, Serono, Sweden) with GnRH agonists using either long or short protocol BL Days 1–3: G1 medium (Vitrolife Ltd, Gothenburg, Sweden) until Day 3 Days 4–6: G2 medium (Vitrolife Ltd, Gothenburg, Sweden) under 5% O2, 6% CO2 and 89% N2 BL: 4.8% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Kawachiya et al. (2011), Japan 151/14 969 Clomiphene citrate in combination with small amount of HMG or rFSH. GnRH agonist for ovulation Both CS/BL Cleavage-stage medium (SAGE, Pasadena, CA). Blastocysts: Quinns advantage; SAGE enriched by 10% protein supplement CS: 0.6% BL: 1.2% AOR: 2.04; (95% CI 1.29–4.48) Sig. diff. Detected at first-trimester ultrasound examination Knopman et al. (2010), USA 98/4976 Majority of patients: Hypothalamic down-regulation with GnRH agonist; this was followed by combinations of injectable recombinant FSH and/or hMG In a minority of cases, LH suppression was achieved using GnRH antagonists Both CS/BL Human tubal fluid media (HTF; Irvine Scientific, Irvine, CA) supplemented with 6% plasmanate (Bayer, Elkhart, IN) overlaid with Sage mineral oil (Cooper Surgical, Trumbull, CT) MZT overall: 2% CS: 1.0% BL: 2.6% OR: 2.62 (95% CI 1.47–5.03) Sig. diff. AOR: 2.13 (95% CI 0.67–6.79) No sig. diff. Ultrasound scan: If two foetal poles were observed within a single sac or if the total number of foetal poles exceeded the number of embryos transferred Papanikolaou et al. (2010), Belgium 13/579 rFSH or hMG in combination with GnRH antagonist or GnRH agonist Both CS/BL only SET Medicult (Jyllinge, Denmark) or Vitrolife (Goteborg, Sweden) CS: 2.6% BL: 1.8% (difference +0.8%; 95% CI −1.97 to +3.41) No sig. diff. Ultrasound scan: More than one foetal heartbeat observed Sharara and Abdo (2010), USA 4/300 Three types: Long lyteal GnRHa or GnRH antagonist, or microdose flara GnRHa Both CS/BL P1/cleavage-stage, Irvine Scientific (2003–2004) and Global Medium, Life Global (2004–present (2010)). Blastocyst medium, Irvine Scientific (2003–2004) and Global medium, Life Global (2004–present) (2010) MZT overall: 1.3% CS: 0 BL: 1.57% Ultrasound scan: More than one foetal heartbeat observed Verpoest et al. (2009), Belgium 29/1.565 Female patients underwent ovarian stimulation using urinary or recombinant FSH in combination with GnRH antagonist or agonist Van Landuyt et al. (2005) Only BL Different blastocyst media used: Medicult Blastassist system (Jyllinge, Denmark); Vitrolife G2 and G3 series (Goteborg, Sweden); Cook medical Sydney IVF media (Limerick, Ireland) BL: 1.9% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Skiadas et al. (2008), USA 41/2501 A microdose lupron protocol with lupron 0.05 mg SC twice a day started cycle Day 1 of a period after oral contraceptive (OC) pill lead in and baseline ultrasound testing performed on Day 2 or A GnRH antagonist (GnRH-a) protocol using an OC pill for three weeks, then baseline ultrasound testing cycle Day 2, with GnRH-a initiation at a dose of 0.25 mg/day SC starting stimulation Day 6 Skiadas et al. (2006) Both CS/BL Days 1–3: P1 (Irvine Scientific, Santa Ana, CA) IVF-500 (Scandinavian IVF Science/Vitrolife, Gothenburg, Sweden) G1.2 (Scandinavian) G1.3 (Scandinavian) Days 3–5: Blastocyst Medium S2 following IVF-500; G2.2 (d) G2.3 Monochorionic twins overall: 1.6% CS: 1.4% BL: 4.2% OR: 2.48 (95% CI 1.62–3.80) Sig. diff. Ultrasound scan: The number of chorions was determined by the number of separate gestational sacs. A gestational sac containing two embryos was diagnosed as a monochorionic pair Moayeri et al. (2007), USA 19/1486 FSH/HMG stimulation protocols with GnRH agonist or antagonist Both CS/BL P1/cleavage stage medium, Irvine Scientific for control group (before March 2002) Quinns Advantage cleavage stage medium (Saga) for study group (March 2002–December 2005) Blastocyst: Blastocyst Medium (Irvine Scientific) or Quinns Advantage Blastocyst Medium (Sage) Control group: CS: 1.96% BL: 5.6% (P = 0.02) Sig. diff. Study group: CS: 1.8% BL: 2.3% (P = 0.59) No sig. diff. Control group BL and Study group BL (P = 0.03) Sig diff. between control and study group Ultrasound scan: More than one foetus with heartbeat in same gestations sac Papanikolaou et al. (2006), Belgium 2/99 Multifollicular ovarian stimulation was performed with a GnRH antagonist and rFSH Both CS/BL Only SET Not reported MZT overall: 2.0% CS: 4.9% BL: 0 Detected at first-trimester ultrasound examination Jain et al. (2004), USA 6/85 Leuprolide (Lupron, TAP Pharmaceuticals, Deerfield, IL, USA) downregulation followed by HMG or FSH stimulation. Both CS/ BL P1 media, Irvine Scientific and Blastocyst Media (Irvine Scientific) CS: 2.1% BL: 13.2% (P < 0.05) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Elizur et al. (2004), Israel 10/1066 Approximately 85% of the patients underwent the long protocol of GnRH analogue (GnRH-a) administered either in the midluteal or early follicular phase of the menstrual cycle followed by gonadotrophin administration after verification of complete ovarian suppression. The remaining 15% of the patients were treated with ovarian stimulation using the flare-up protocol (the GnRH-a was used from Day 1 of the menstrual cycle) and stimulation with gonadotrophin was started on Day 3 Both CS/BL (only a few) Not reported MZT overall: 0.9% Ultrasound scan: The number of embryos exceeded the number of gestational sacs Frankfurter et al. (2004), USA 8/111 Not reported BL P1 Medium (Irvine Scientific, Santa Ana, CA) through postretrieval Day 3 at which point they were placed in Blastocyst Medium (Irvine Scientific) until the day of ET BL: 7.2% Ultrasound scan: If two heartbeats were observed within a single sac or if the total number of gestational sacs exceeded the number of embryos transferred Wright et al. (2004) 206/39 198 Not reported Both CS/BL Not reported MZT overall: O.6% CS: 0.4% BL: 1.5% (P < 0.05) Sig. diff. Ultrasound scan: If the number of foetal heart beats exceeded the number of embryos transferred Milki et al. (2003), USA 18/554 GnRH agonist down-regulation followed FSH/HMG and HCG Both CS/BL P1 media, Irvine Scientific Blastocyst: Blastocyst Medium (Irvine Scientific) MZT overall: 3.2% CS: 2% BL: 5.6% (P < 0.03) Sig. diff. Ultrasound scan: More than one foetus with heartbeat in same gestational sac Blickstein et al. (2003), Israel 25/1395 Not reported SET (Stage not reported) Not reported MZT overall: 1.8% SET Cassuto et al. (2003), France 9/1263 The vast majority of the controlled ovarian hyperstimulation cycles were performed using GnRH analogues in long protocols Both CS/BL a) Days 1–2: ISM1 (Medicult, Copenhagen, Denmark) medium Days 3−5: ( for BL) ISM2 (Medicult) b) Day 1: ISM1 (Medicult, Copenhagen, Denmark) medium Days 2–3: ISM2 Days 3–5: M3 (Medicult) Media b) CL: 0.45% BL: 5.5% (P = 0.0098) Sig. diff. Not reported Alikani et al. (2003), United Kingdom 81/4305 (i) Mid-luteal (cycle Day 21) down-regulation with leuprolide acetate (0.5 mg) followed by stimulation with gonadotrophins on Day 3 of the subsequent menses; or (ii) a microdose leuprolide acetate (50 mg twice a day) ̄are protocol utilizing leuprolide on Day 2 of an oral contraceptive withdrawal bleed followed by gonadotrophin stimulation just 1 day later (Day 3 of menses) Both CS/BL Not reported MZT overall: 1.88% Ultrasound scan: The number of yolk sacs, foetal poles or foetal heartbeats exceeded the number of embryos replaced, or two or more yolk sacs, foetal poles or heart beats were observed within a single gestational sac Tarlatzis et al. (2002), Greece 6/181 cycles Not reported Only BL Not reported MZT overall: 3.3% per cycle IVF: 0 ICSI: 5.9% (P = 0.033) Sig. diff. Detected at first-trimester ultrasound examination Sheiner et al. (2001), Israel 3/204 Not reported Both CS/BL Not reported MZT overall: 1.5% BL: (2/9) 22.2% OR: 55 (95% CI 3.3–1778.3; P = 0.005) Sig. diff. Not reported da Costa et al. (2001), Brazil 11/943 Luprolide acetate and recombinant FSH Both CS/BL IVF 50 medium (Scandinavian IVF Science) with mineral oil. BL: after IVF 50 medium, Day 3 in S2 medium (Scandinavian IVF Science) MZT overall: 1.2% CS: 0.7% BL: 3.9% (OR: 5.606; P = 0.00675) Sig. diff. Detected at first-trimester ultrasound examination Schieve et al. (2000), USA 22/11 247 Not reported Stage not reported Not reported MZT overall: 0.2% Ultrasound scan: If the number of foetal heartbeats exceeded the number of embryos transferred Saito et al. (2000), Japan 17/497 cycles Not reported SET Not reported MZT overall: 3.4% SET Behr et al. (2000), USA 10/199 GnRH-a/FSH protocols Only BL a) Days 0–3: P1 media, Irvine Scientific Days 3–5: blastocyst medium (Irvine Scientific) b) Days 0–3: IVF50 (Scandinavian IVF Science) Day 3–5: S2 (Scandinavian IVF Science). BL: 5% Ultrasound scan: More than one foetus with heartbeat in same gestational sac and confirmed at birth Graham et al. (2000), USA 1/53 Combined leuprolide acetate/gonadotrophin regimens Both CS/BL Days 1–3: the HTF/ P1/blastocyst media system (Irvine Scientific, Santa Anna, CA) and later in the IVF/G1/G2 system (Scandinavia IVF Science, Gothenburg, Sweden) Days 3–5: blastocyst or G2 MZT overall: 1.9% Not reported Hershlag et al. (1999), USA 8/391 Step-down protocols with hMG either alone or in combination with FSH Not reported Not reported MZT overall: 2% Ultrasound scan: A single chorionic membrane, with or without a single amnion; and confirmed by pathologic evaluation of the placenta in all live births Blickstein et al. (1999), Israel 4/82 Not reported SET (Stage not reported) Not reported MZT overall: 4.9% SET Chang et al. (2009), South Korea Meta-analysis 9 studies: 40 917 (Meta-analysis) Both CS/BL (Meta-analysis) BL versus CS OR: 3.04; (95% CI 1.54–6.01) Sig. diff. BL and CS (Meta-analysis) Vitthala et al. (2009), UK Meta-analysis 9 studies: 43 876 (Meta-analysis) Both CS/BL (Meta-analysis) MZT overall: 0.9% (95% CI 0.8–0.9) BL: 1.7% (95% CI 1.4–2.0) Sig. diff. (Meta-analysis) AOR, adjusted odds ratio; BL, blastocyst; CS, cleavage-stage embryo; ET, embryo transfer; GS, gestational sac; MZT, monozygotic twins; OR, odds ratio; SET, single embryo transfer; Sig. diff., significant difference. Overall, 16 articles reported a significantly higher MZT rate after blastocyst transfer (Kawachiya et al., 2011; Knopman et al., 2014; Luke et al., 2014; Mateizel et al., 2016; Nakasuji et al., 2014; Sharara and Abdo, 2010; Sotiroska et al., 2015), and five studies showed insignificant results (Moayeri et al., 2007; Papanikolaou et al., 2010; Sharara and Abdo, 2010; Osianlis et al., 2014; Wu et al., 2014). Two large studies reported a significantly higher rate of MZT after blastocyst transfer compared with cleavage-stage ET, but when adjusted for potential risk factors (patient age, time period, day of embryo biopsy, embryos for cryopreservation, the number of embryos transferred and embryo quality) the statistical significance disappeared (Knopman et al., 2010; Franasiak et al., 2015). Knopman et al. (2014) published a new cohort study, and apart from a trend toward a lower MZT rate over time, they showed a significantly higher MZT rate after blastocyst transfer in an adjusted model (OR = 1.66; 95% CI: 1.04–2.64; P = 0.04) that included the following co-variates: number of oocytes retrieved at age ≥35 years, donor oocytes, ART treatment cycle after 2004, IVF/ICSI and extended culture. Moayeri et al. (2007) found no difference in the rate of MZT after blastocyst or cleavage-stage transfer (Moayeri et al., 2007). However, by comparing the MZT rate after blastocyst transfer in two different time periods Moayeri et al. (2007) observed that the MZT rate was 2.3% in the late time period (1 March 2002 to 31 December 2005), compared with 5.6% in the early time period before March 2002, demonstrating a significantly lower value in the second interval (P = 0.03). In conclusion, based on 19 studies with high quality (Table I), the majority reported a higher rate of MZT after blastocyst transfer compared with cleavage-stage transfer. Meta-analysis Eleven studies published after 2007 were included in the meta-analysis with a total of 161 885 clinical pregnancies. The 10 out of the 11 included studies showed a higher incidence of MZT after blastocyst transfer compared with cleavage-stage transfer, although only seven reported a significant difference. The fixed effect and random effects models for the meta-analyses revealed comparable results. Both analyses showed a significantly higher rate of MZT following blastocyst transfer compared with cleavage-stage transfer (fixed effect meta-analysis: OR = 2.18, 95% CI: 1.93–2.48; random effects meta-analysis: OR = 2.00, 95% CI: 1.48–2.70) (Fig. 4). Figure 4 View largeDownload slide Meta-analysis and forest plot of MZT after cleavage-stage embryo transfer versus blastocyst transfer. Figure 4 View largeDownload slide Meta-analysis and forest plot of MZT after cleavage-stage embryo transfer versus blastocyst transfer. Controlled ovarian stimulation, ART protocol and culture media Different aspects of ART have been investigated to identify a mechanism and provide an explanation for the high rate of MZT in ART. A few studies have focused on the controlled ovarian stimulation and ART protocol that might affect MZT. Papanikolaou et al. (2010) studied the influence of several types of controlled ovarian stimulation, as well as the protocol used for ART, on the MZT rate following transfer of either blastocyst or cleavage-stage embryos. However, no significant difference was observed, which was in agreement with the study by Nakasuji et al. (2014). One large study focused on the effect of controlled ovarian stimulation and the protocol among other variables in a multivariate analysis (Luke et al., 2014). The findings indicated that the GnRH agonist suppression protocol increased the rate (AOR = 1.10; 95% CI: 1.00–1.22; P = 0.04) and that higher FSH doses (≥3000 IU) decreased the rate (AOR = 0.7; 95% CI: 0.62–0.8; P < 0.0001) of MZT (Luke et al., 2014). The culture media may play a role in relation to the higher frequency of MZT in ART. Several studies have discussed different aspects of this topic, but only a few studies have investigated this phenomenon in practice. One study examined the effect of fresh culture media (grouped in Sage, Medicult, Vitrolife and ‘other’), but they found no differences related to the examined culture media (Mateizel et al., 2016). Another study evaluated at two different media (Medicult and Vitrolife) used for both cleavage-stage and blastocyst embryos, but neither found a difference between the two media (Papanikolaou et al., 2010). In summary only four studies of medium-high quality (Table I) have investigated the role of the controlled ovarian stimulation, the ART protocol or the culture media on the MZT rate, suggesting an increased frequency of MZT following the GnRH agonist suppression protocol and a decreased rate with high FSH doses. Maternal age Fourteen studies have examined maternal age, or oocyte age, as a risk factor for MZT. Eight of the studies found that a younger maternal age increased the rate of MZT although only six were statistically significant. Franasiak et al. (2015) observed a significant reduction in the rate of MZT with increasing age (OR = 0.95; 95% CI: 0.92–0.98; P = 0.001). In a study design from 2015, women undergoing ART were divided into groups based on age (<36 years and >36 years) as well as transfer day (Sotiroska et al., 2015). The study did not only yield a higher multiple gestation rate for the group aged <36 years (twins and triplets, varied from one to three embryos transferred), but a higher MZT rate of 0.5% for cleavage-stage embryo and 2.9% for blastocyst as compared with 0% in the group aged >36 years. Another study reported a 4% decrease in the MZT rate for each year the oocyte age increased (Knopman et al., 2014). A study from 2010 addressed whether the oocyte age or maternal age influences the MZT rate (Knopman et al., 2010). First, they described a significant difference in the MZT rate between autologous cycles and donor oocyte cycles (1.7 versus 3.3%, respectively, P = 0.005). Then they examined the same age group (<35 years), but divided the group into autologous cycles and donor oocyte cycles, revealing no significant difference between the two groups. When examining autologous cycles, a significant difference was observed between women age < 35 and >35 years with respect to the rate of MZT (3.1 versus 1.4%, P = 0.001). This finding suggests that oocyte age is the determining factor in investigations of the influence of age on the MZT rate. In contrast, Mateizel et al. (2016), as well as five other studies, found no influence of maternal age on the MZT rate (da Costa et al., 2001; Alikani et al., 2003; Frankfurter et al., 2004; Skiadas et al., 2008; Nakasuji et al., 2014). Based on 14 studies with a high quality (Table I), maternal age seems to influence the rate of MZT, as some of the most recent studies have reported a higher risk of MZT with younger maternal age or with the use of younger oocytes. Cause of infertility and fertilization method Among several potential risk factors, one study has focused on a possible influence of the indication for ART treatment on the MZT rate, with the following indications: male factor, tubal factor, endometriosis, PCOS and/or anovulation, idiopathic and combined (Papanikolaou et al., 2010). None of the listed indications were associated with the MZT rate. In studies of the influence of the method of fertility treatment on the MZT rate, several aspects should be considered. First, the effects of using ZP manipulation methods, including ICSI, PGD or AH, must be considered. Furthermore, the importance of fresh or frozen ET can be analysed. Several studies have observed a higher MZT rate after the use of ZP manipulation methods. Skiadas et al. (2008) discovered a 2.42-fold increase in the likelihood of MC-twins with the use of ICSI compared with IVF. A small study by Tarlatzis et al. (2002) based only on blastocyst transfers reported a 5.9% rate of MZT after ICSI and 0% after conventional IVF, demonstrating a significant difference (P = 0.033), while Hershlag et al. (1999) reported a significantly higher MZT rate with the use of AH. In contrast, in several other studies, the fertilization method did not influence the MZT rate: ICSI versus conventional IVF was studied by Papanikolaou et al. (2010) and Franasiak, et al. (2015); PGD by Verpoest et al. (2009); ICSI, frozen ET and PGD was studied by Knopman et al. (2014); and only AH and ICSI were studied by Knopman et al. (2010), Mateizel et al. (2016), and Milki et al. (2003). Another study examined the influence of the location of the artificial opening with the use of AH, and found no difference in the numbers of MZT in the two studied locations (Ren et al., 2013). However, Mateizel et al. (2016) discovered a higher MZT rate with the use of fresh ET in comparison with frozen ET (P = 0.008). Based on more than 10 studies of high quality (Table I), there is no clear consensus regarding the influence of the infertility cause and methods of fertility treatment on the frequency of MZT. Embryo morphology One group studied embryo quality based on the number of embryos with six- to eight-cells (Franasiak et al., 2015). Women with four or more embryos in the six- to eight-cell stage and women with more than 75% of all embryos in the six- to eight-cell stage (high quality) had a higher risk of MZT with blastocyst transfer (P = 0.05 and P = 0.04, respectively) than women with one to three embryos in the six- to eight-cell stage embryos or in the category in which <75% of all embryos had six- to eight-cells. The study by Otsuki et al. (2016) found the opposite association, namely, an association between a high embryo quality score and a decreased frequency of MC-DA twins. They used an EmbryoScope time-lapse system (Vitrolife) to retrospectively analyse the quality of the embryos in their cohort. The focus was loosening of the inner cell mass (ICM), defined as the presence of more than five cells loosely arranged in the ICM (in comparison to an earlier observation of tightly grouped cells). The study showed that a low-grade ICM resulted in a 1.38% risk of MC-DA twins, in contrast to a 0.38% risk related to a high-grade ICM (grade A) (P = 0.033) (Otsuki et al., 2016). The true influence of embryo morphology on the MZT rate is still controversial and based only on two high-quality studies (Table I). Discussion A higher rate of MZT is consistently observed after the use of ART; however, the underlying mechanisms are not fully understood. Although our meta-analysis showed a significantly higher rate of MZT after blastocyst transfer, the rate still varied considerably between studies. The causes of this phenomenon may be due to heterogeneity between studies and the varying definition of MZT (Table II). Almost all of the studies defined MZT as two foetal poles within one single gestational sac, or if the total number of foetal poles exceeded the number of embryos transferred, often with the addition of ‘cardiac activity’ for both foetuses (Knopman et al., 2010). This definition clearly excludes DC-DA twins and miscarried MZT, but it also eliminates the possibility of a monozygotic multiple gestation, in which the number of foetal poles did not exceed the number of transferred embryos. The study by Mateizel et al. (2016) (MZT: 103/3672 blastocysts) would not have found a significantly higher MZT rate after blastocyst transfer, if foetal heart beats had been used as a criterion for MZT. This possible definition mis-match was confirmed by a meta-analysis performed in 2009 that reported a difference in the MZT rate based on different definitions (Vitthala et al., 2009). To fully evaluate MZT, DNA testing could be a possible solution (Toledo, 2005). Furthermore, the MZT rate in naturally conceived children is reported per delivery, and hence is, per se, lower than and not directly comparable to the MZT rate reported per clinical pregnancy after ART (Alikani et al., 2003). Some heterogeneity among these studies should be expected as the meta-analysis test of heterogeneity showed (I2=77%, Fig. 4). It would be very difficult to eliminate any sort of heterogeneity when combining studies, or simply data, from clinics worldwide because the definition of infertility, access to ART, methods, and materials may vary to large extents. However, this variability is important to keep in mind when examining and discussing this topic. The true incidence of MZT in ART, in general, as well as after the transfer of a blastocyst, can at the moment best be assessed in studies based on SET, taking into consideration the possibility of concurrent natural conceptions (Osianlis et al., 2014). Ideally, a genetic test should be the gold standard, but the high cost often eliminates this opportunity, in addition to difficulties associated with postpartum follow-up (Knopman et al., 2014). Many studies have claimed that the culture medium developed for culturing blastocysts was the cause of the higher MZT rate after blastocyst transfer. Two studies showed a decrease in MZT after blastocyst transfer over time (Knopman et al. (2014): MZT 107/4494 blastocysts, Moayeri et al. (2007): MZT: 9/385 blastocysts). They both speculated whether the decline was a result of better experiences with both blastocyst culture and transfer. Moayeri et al. (2007) changed both the cleavage-stage culture and blastocyst culture media during their study period and considered it as a possible explanation for the shift in the MZT rate over time. The culture medium used for the control group (data originally published in a study from 2003 (Milki et al., 2003): MZT 11/197 blastocysts) showed a greater difference in glucose levels between cleavage-stage and blastocyst media compared with the medium used for the study group. Cleavage-stage embryos in the control group were cultured in a glucose-free medium and blastocysts in a medium with high glucose levels. The cleavage-stage embryos in the study group were cultured in a medium with a low level of glucose, and the blastocyst medium had an intermediate glucose level. Several studies have proposed that a high level of glucose in the culture medium is a factor influencing the MZT rate after blastocyst transfer because it produces more free radicals leading to apoptosis and weakening and disruption of the ICM (Cassuto et al., 2003; Moayeri et al., 2007; Chang et al., 2009; Papanikolaou et al., 2010). Another theory is that the changes in glucose levels in cultures (cleavage-stage versus blastocyst media) rather than the culture medium itself direct the embryo toward a ‘stress reaction’ that leads to splitting of the ICM (Moayeri et al., 2007). One proposition is the development of more favourable culture conditions (Chang et al., 2009). The presence of co-cultures and growth factors may also play a part in the MZT rate. A lower concentration of grow factors might lead to metabolic stress, resulting in apoptosis or weakened of cellular adhesion (Milki et al., 2003; Moayeri et al., 2007). Another theory concerns the exposure of the blastocyst to a lower level of calcium than that found in the endometrium, and animal models show that blastocyst culture weakens ICM intercellular bonds, which leads to embryo splitting (Milki et al., 2003; Papanikolaou et al., 2010). The results of the current meta-analysis may indicate a contradiction to the presumption that better culture conditions and experience with blastocyst transfer would diminish the higher incidence of MZT after blastocyst transfer. Although the incidence might have decreased to some degree over time (from 3.9–5.6% in 2001–2003 to 2.2–2.9% in 2015/2016), the higher rate of MZT following blastocyst transfer has persisted (da Costa et al., 2001; Tarlatzis et al., 2002; Milki et al., 2003; Sotiroska et al., 2015; Mateizel et al., 2016). Ideally, multicentre studies should be performed that allow a comparison of different media to elucidate the overall influence of these factors. It would be optimal to study the specific media independently in controlled trials to understand the influence of the medium composition on the MZT rate, including O2 tension, although such studies may be controversial and raise ethical questions about embryo research. A majority of recent studies has reported a significant influence of maternal age on the MZT rate (Franasiak et al. (2015): MZT 234/9969 ET; Kawachiya et al. (2011): MZT 151/14 956 ET; Knopman et al. (2010): MZT 98/4976 ET; Knopman et al. (2014): MZT 131/6223 ET; Sotiroska et al. (2015): MZT 7/1400 ET). These studies showed an increase in the MZT rate in cases of young maternal age. Although there is no consensus among all the studies included in this review, it can be speculated that the true underlying mechanism might be the oocyte age, or even the quality of the oocyte, rather than maternal age itself as shown in the study by Knopman et al. (2010). For several years, ZP manipulation was thought to play a role in the higher incidence of MZT after ART (Alikani et al. (2003): MZT 81/4305 ET; Hershlag et al. (1999): MZT 8/391 ET; Saito et al. (2000): MZT 9/279 ET; Schieve et al. (2000): MZT 22/11 247 ET; Skiadas et al. (2008): MZT 41/2501 ET; Tarlatzis et al. (2002): MZT 6/181 cycles). It is plausible that hatching through an artificial hole created in the ZP may cause disruption and splitting of the ICM, resulting in herniation of the blastomeres and embryo (Schieve et al., 2000; Tarlatzis et al., 2002; Milki et al., 2003). However, recent studies have questioned the role of ZP manipulation because these studies did not find an increased rate of MZT after ICSI and AH (Franasiak et al. (2015): MZT 234/9969 ET; Knopman et al. (2010): MZT 98/4976 ET; Knopman et al. (2014): MZT 131/6223 ET; Mateizel et al. (2016): MZT 136/6096 ET; Milki et al. (2003): MZT 18/554 ET; Papanikolaou et al. (2010): MZT 13/579 ET). However, one study showed that ZP disruption raised the MZT rate, questioning the complete role of ZP (Alikani et al. (2003): MZT 81/4305 ET). It has been speculated whether ZP disruption per se or the circumstances leading to the technique caused the increased rate of MZT. Nevertheless, this issue requires further research for clarification, potentially also based on the cause of infertility. Another intriguing aspect is the sex of monozygotic children born after the use of ART. Such a study of the sex requires a follow-up after birth, which unfortunately, was not performed in most of the included studies. However, two different studies, one of which was included in the meta-analysis by Chang et al. (2009), found a higher rate of male infants born after the transfer of a blastocyst compared with the transfer of a cleavage-stage embryo (Sotiroska et al., 2015). A possible thought, as shown in animal studies, is that male embryos develop faster than female embryos (Chang et al., 2009). As the most developed embryos are preferred for extended culture and blastocyst transfer, one may assume that more male blastocysts are transferred and thus a higher number of male MZT. Although highly speculative, the large number reported in the two studies might simply be based on a larger number of developed male embryos—and therefore a higher rate of male blastocysts transferred—rather than on the male sex itself. In total, 42 articles, including systemic reviews, meta-analyses and original studies, were included in the current review. The purpose of using many different search strings as well as MeSH terms was to include as many suitable studies on the subject as possible. However, there are several limitations to this search. Many more search strings could have been used to answer the study questions, which naturally gives rise to some limitations. To streamline the current review, the search words, strings and MeSH terms used were focused on the main question: the role of blastocyst transfer in the MZT rate, and studies may have been excluded throughout the search. Not all of the included original studies in the two meta-analyses were included in this review, which might be an example of the limitations of the search (Chang et al., 2009; Vitthala et al., 2009). However, this has mainly involved older studies, whereas the current review has included several new studies. Conclusion In this meta-analysis, we observed a significantly higher rate of MZT following blastocyst transfer than cleavage-stage embryos (fixed effect meta-analysis: OR = 2.18, 95% CI: 1.93–2.48 and random effects meta-analysis: OR = 2.00, 95% CI: 1.48–2.70). We consider the results of this meta-analysis to question the trend theory that more experience with blastocyst transfer and better culture media decrease the high rate of MZT after blastocyst transfer, which has been proposed previously. Several studies, although not all, have also reported a reduced rate of MZT with increased maternal age, although how age influences the frequency of MZT remains to be explored. Difficulties related to how to correctly define and identify MZT also predominate as obstacles that must to be solved. The current review clearly indicates that more research is needed to fully understand the mechanism underlying MZT after ART and to provide useful guidelines for fertility treatment to minimize the rate of MZT and, thereby, the risks associated with multiple monozygotic gestation and potential subsequent complications. Large multi-centre studies based on SET alone are needed to explore the potential risk factors, such as those discussed in the current review. Sufficiently powered studies are important to capture the low incidence of MZT. An ideal study would also include DNA testing as well as embryo time-lapse examination by EmbryoScope to discover potential signs of embryo splitting. Acknowledgements We thank Prof. T. Hviid for creating the schematic drawing presented in Figure 1. Authors’ roles K.V.R.H. performed the search and data collection, wrote the first version of the article, and performed the meta-analysis. H.S.N. conceived the basic idea and supervised the review and data collection. A.P., S.S.M., H.S.N and K.V.R.H. participated in the quality assessment and critically discussed, read and approved the final article. Funding K.V.R.H. has obtained a research scholarship from Rigshospitalet, Copenhagen. Conflicts of interest None declared. References Alikani M , Cekleniak NA , Walters E , Cohen J . Monozygotic twinning following assisted conception: an analysis of 81 consecutive cases . Hum Reprod 2003 ; 18 : 1937 – 1943 . Google Scholar CrossRef Search ADS PubMed Aston KI , Peterson CM , Carrell DT . Monozygotic twinning associated with assisted reproductive technologies: a review . Reproduction 2008 ; 136 : 377 – 386 . Google Scholar CrossRef Search ADS PubMed Behr B , Fisch JD , Racowsky C , Miller K , Pool TB , Milki AA . Blastocyst-ET and monozygotic twinning . J Assist Reprod Genet 2000 ; 17 : 349 – 351 . Google Scholar CrossRef Search ADS PubMed Blickstein I , Verhoeven HC , Keith LG . Zygotic splitting after assisted reproduction . N Engl J Med 1999 ; 340 : 738 – 739 . Google Scholar CrossRef Search ADS PubMed Blickstein I , Jones C , Keith LG . Zygotic-splitting rates after single-embryo transfers in in vitro fertilization . N Engl J Med 2003 ; 348 : 2366 – 2367 . Google Scholar CrossRef Search ADS PubMed Cassuto G , Chavrier M , Menezo Y . Culture conditions and not prolonged culture time are responsible for monozygotic twinning in human in vitro fertilization . Fertil Steril 2003 ; 80 : 462 – 463 . Google Scholar CrossRef Search ADS PubMed Chang HJ , Lee JR , Jee BC , Suh CS , Kim SH . Impact of blastocyst transfer on offspring sex ratio and the monozygotic twinning rate: a systematic review and meta-analysis . Fertil Steril 2009 ; 91 : 2381 – 2390 . Google Scholar CrossRef Search ADS PubMed da Costa ALE , Abdelmassih S , de Oliveira FG , Abdelmassih V , Abdelmassih R , Nagy ZP , Balmaceda JP . Monozygotic twins and transfer at the blastocyst stage after ICSI . Hum Reprod 2001 ; 16 : 333 – 336 . Google Scholar CrossRef Search ADS PubMed Edwards RG . An astonishing journey into reproductive genetics since the 1950’s . Reprod Nutr Dev 2005 ; 45 : 299 – 306 . Google Scholar CrossRef Search ADS PubMed Elizur SE , Levron J , Shrim A , Sivan E , Dor J , Shulman A . Monozygotic twinning is not associated with zona pellucida micromanipulation procedures but increases with high-order multiple pregnancies . Fertil Steril 2004 ; 82 : 500 – 501 . Google Scholar CrossRef Search ADS PubMed Franasiak JM , Dondik Y , Molinaro TA , Hong KH , Forman EJ , Werner MD , Upham KM , Scott RT Jr . Blastocyst transfer is not associated with increased rates of monozygotic twins when controlling for embryo cohort quality . Fertil Steril 2015 ; 103 : 95 – 100 . Google Scholar CrossRef Search ADS PubMed Frankfurter D , Trimarchi J , Hackett R , Meng L , Keefe D . Monozygotic pregnancies from transfers of zona-free blastocysts . Fertil Steril 2004 ; 82 : 483 – 485 . Google Scholar CrossRef Search ADS PubMed Fuchs F , Senat MV . Multiple gestations and preterm birth . Semin Fetal Neonatal Med 2016 ; 21 : 113 – 120 . Google Scholar CrossRef Search ADS PubMed Graham MC , Hoeger KM , Phipps WR . Initial IVF-ET experience with assisted hatching performed 3 days after retrieval followed by day 5 embryo transfer . Fertil Steril 2000 ; 74 : 668 – 671 . Google Scholar CrossRef Search ADS PubMed Hankins GVD , Saade GR . Factors influencing twins and zygosity . Paediatr Perinat Epidemiol 2005 ; 19 : 8 – 9 . Google Scholar CrossRef Search ADS PubMed Hershlag A , Paine T , Cooper GW , Scholl GM , Rawlinson K , Kvapil G . Monozygotic twinning associated with mechanical assisted hatching . Fertil Steril 1999 ; 71 : 144 – 146 . Google Scholar CrossRef Search ADS PubMed Jain JK , Boostanfar R , Slater CC , Francis MM , Paulson RJ . Monozygotic twins and triplets in association with blastocyst transfer . J Assist Reprod Genet 2004 ; 21 : 103 – 107 . Google Scholar CrossRef Search ADS PubMed Kawachiya S , Bodri D , Shimada N , Kato K , Takehara Y , Kato O . Blastocyst culture is associated with an elevated incidence of monozygotic twinning after single embryo transfer . Fertil Steril 2011 ; 95 : 2140 – 2142 . Google Scholar CrossRef Search ADS PubMed Knopman J , Krey LC , Lee J , Fino ME , Novetsky AP , Noyes N . Monozygotic twinning: an eight-year experience at a large IVF center . Fertil Steril 2010 ; 94 : 502 – 510 . Google Scholar CrossRef Search ADS PubMed Knopman JM , Krey LC , Oh C , Lee J , McCaffrey C , Noyes N . What makes them split? Identifying risk factors that lead to monozygotic twins after in vitro fertilization . Fertil Steril 2014 ; 102 : 82 – 89 . Google Scholar CrossRef Search ADS PubMed Kontopoulos E , Chmait RH , Quintero RA . Twin-to-twin transfusion syndrome: definition, staging, and ultrasound assessment . Twin Res Hum Genet 2016 ; 19 : 175 – 183 . Google Scholar CrossRef Search ADS PubMed Luke B , Brown MB , Wantman E , Stern JE . Factors associated with monozygosity in assisted reproductive technology pregnancies and the risk of recurrence using linked cycles . Fertil Steril 2014 ; 101 : 683 – 689 . Google Scholar CrossRef Search ADS PubMed Mateizel I , Van Landuyt L , Tournaye H , Verheyen G . Deliveries of normal healthy babies from embryos originating from oocytes showing the presence of smooth endoplasmic reticulum aggregates . Hum Reprod 2013 ; 28 : 2111 – 2117 . Google Scholar CrossRef Search ADS PubMed Mateizel I , Santos-Ribeiro S , Done E , Van Landuyt L , Van de Velde H , Tournaye H , Verheyen G . Do ARTs affect the incidence of monozygotic twinning? Hum Reprod 2016 ; 31 : 2435 – 2441 . Google Scholar CrossRef Search ADS PubMed Milki AA , Jun SH , Hinckley MD , Behr B , Giudice LC , Westphal LM . Incidence of monozygotic twinning with blastocyst transfer compared to cleavage-stage transfer . Fertil Steril 2003 ; 79 : 503 – 506 . Google Scholar CrossRef Search ADS PubMed Moayeri SE , Behr B , Lathi RB , Westphal LM , Milki AA . Risk of monozygotic twinning with blastocyst transfer decreases over time: an 8-year experience . Fertil Steril 2007 ; 87 : 1028 – 1032 . Google Scholar CrossRef Search ADS PubMed Nakasuji T , Saito H , Araki R , Nakaza A , Nakashima A , Kuwahara A , Ishihara O , Irahara M , Kubota T , Yoshimura Y et al. . The incidence of monozygotic twinning in assisted reproductive technology: analysis based on results from the 2010 Japanese ART national registry . J Assist Reprod Genet 2014 ; 31 : 803 – 807 . Google Scholar CrossRef Search ADS PubMed Osianlis T , Rombauts L , Gabbe M , Motteram C , Vollenhoven V . Incidence and zygosity of twin births following transfers using a single fresh or frozen embryo . Hum Reprod 2014 ; 29 : 1438 – 1443 . Google Scholar CrossRef Search ADS PubMed Otsuki J , Iwasaki T , Katada Y , Sato H , Furuhashi K , Tsuji Y , Matsumoto Y , Shiotani M . Grade and looseness of the inner cell mass may lead to the development of monochorionic diamniotic twins . Fertil Steril 2016 ; 106 : 640 – 644 . Google Scholar CrossRef Search ADS PubMed Papanikolaou EG , Camus M , Kolibianakis EM , Van Landuyt L , Van Steirteghem A , Devroey P . In vitro fertilization with single blastocyst-stage versus single cleavage-stage embryos . N Engl J Med 2006 ; 354 : 1139 – 1146 . Google Scholar CrossRef Search ADS PubMed Papanikolaou EG , Fatemi H , Venetis C , Donoso P , Kolibianakis E , Tournaye H , Tarlatzis B , Devroey P . Monozygotic twinning is not increased after single blastocyst transfer compared with single cleavage-stage embryo transfer . Fertil Steril 2010 ; 93 : 592 – 597 . Google Scholar CrossRef Search ADS PubMed Parisaei M , Shailendra A , Dutta R , Broadbent JA Obstetrics and Gynaecology , 2nd edn . Edinburgh: Mosby Ltd , 2008 . Pinborg A . IVF/ICSI twin pregnancies: risks and prevention . Hum Reprod Update 2005 ; 11 : 575 – 593 . Google Scholar CrossRef Search ADS PubMed Pinborg A , Lidegaard O , la Cour Freiesleben N , Andersen AN . Consequences of vanishing twins in IVF/ICSI pregnancies . Hum Reprod 2005 ; 20 : 2821 – 2829 . Google Scholar CrossRef Search ADS PubMed Ren X , Liu Q , Chen W , Zhu G , Zhang H . Effect of the site of assisted hatching on vitrified-warmed blastocyst transfer cycles: a prospective randomized study . J Assist Reprod Genet 2013 ; 30 : 691 – 697 . Google Scholar CrossRef Search ADS PubMed Saito H , Tsutsumi O , Noda Y , Ibuki Y , Hiroi M . Do assisted reproductive technologies have effects on the demography of monozygotic twinning? Fertil Steril 2000 ; 74 : 178 – 179 . Google Scholar CrossRef Search ADS PubMed Schieve LA , Meikle SF , Peterson HB , Jeng G , Burnett NM , Wilcox LS . Does assisted hatching pose a risk for monozygotic twinning in pregnancies conceived through in vitro fertilization? Fertil Steril 2000 ; 74 : 288 – 294 . Google Scholar CrossRef Search ADS PubMed Sharara FI , Abdo G . Incidence of monozygotic twins in blastocyst and cleavage stage assisted reproductive technology cycles . Fertil Steril 2010 ; 93 : 642 – 645 . Google Scholar CrossRef Search ADS PubMed Sheiner E , Har-Vardi I , Potashnik G . The potential association between blastocyst transfer and monozygotic twinning . Fertil Steril 2001 ; 75 : 217 – 218 . Google Scholar CrossRef Search ADS PubMed Skiadas CC , Jackson KV , Racowsky C . Early compaction on day 3 may be associated with increased implantation potential . Fertil Steril 2006 ; 86 : 1386 – 1391 . Google Scholar CrossRef Search ADS PubMed Skiadas CC , Missmer SA , Benson CB , Gee RE , Racowsky C . Risk factors associated with pregnancies containing a monochorionic pair following assisted reproductive technologies . Hum Reprod 2008 ; 23 : 1366 – 1371 . Google Scholar CrossRef Search ADS PubMed Sobek A , Prochazka M , Klaskova E , Lubusky M , Pilka R . High incidence of monozygotic twinning in infertility treatment . Biomed papers 2016 ; 160 : 358 – 362 . Google Scholar CrossRef Search ADS Sotiroska V , Petanovski Z , Dimitrov G , Hadji-Lega M , Shushleski D , Saltirovski S , Matevski V , Shenbakar S , Panov S , Johansson L . The day of embryo transfer affects delivery rate, birth weights, female-to-male ratio, and monozygotic twin rate . Taiwan J Obstet Gynecol 2015 ; 54 : 716 – 721 . Google Scholar CrossRef Search ADS PubMed Tarlatzis BC , Qublan HS , Sanopoulou T , Zepiridis L , Grimbizis G , Bontis J . Increase in the monozygotic twinning rate after intracytoplasmic sperm injection and blastocyst stage embryo transfer . Fertil Steril 2002 ; 77 : 196 – 198 . Google Scholar CrossRef Search ADS PubMed Toledo MG . Is there increased monozygotic twinning after assisted reproductive technology? Aust N Zeal J Obstet Gynaecol 2005 ; 45 : 360 – 364 . Google Scholar CrossRef Search ADS Van Landuyt L , De Vos A , Joris H , Verheyen G , Devroey P , Van Steirteghem A . Blastocyst formation in in vitro fertilization versus intracytoplasmic sperm injection cycles: influence of the fertilization procedure . Fertil Steril 2005 ; 83 : 1397 – 1403 . Google Scholar CrossRef Search ADS PubMed Vaughan DA , Robin R , Penzias AS , Norwitz ER , Sakkas D . Clustering of monozygotic twinning in IVF . Assist Reprod Technol 2015 ; 33 : 19 – 26 . Google Scholar CrossRef Search ADS Verpoest W , Van Landuyt L , Desmyttere S , Cremers A , Devroey P , Liebaers I . The incidence of monozygotic twinning following PGD is not increased . Hum Reprod 2009 ; 24 : 2945 – 2950 . Google Scholar CrossRef Search ADS PubMed Vitthala S , Gelbaya TA , Brison DR , Fitzgerald CT , Nardo LG . The risk of monozygotic twins after assisted reproductive technology: a systematic review and meta-analysis . Hum Reprod Update 2009 ; 15 : 45 – 55 . Google Scholar CrossRef Search ADS PubMed Wright V , Schieve LA , Vahratian A , Reynolds MA . Monozygotic twinning associated with day 5 embryo transfer in pregnancies conceived after IVF . Hum Reprod 2004 ; 19 : 1831 – 1836 . Google Scholar CrossRef Search ADS PubMed Wu D , Huang SY , Wu HM , Chen CK , Soong YK , Huang HY . Monozygotic twinning after in vitro fertilization/intracytoplasmic sperm injection treatment is not related to advanced maternal age, intracytoplasmic sperm injection, assisted hatching, or blastocyst transfer . Taiwan J Obstet Gynecol 2014 ; 53 : 324 – 329 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Human Reproduction UpdateOxford University Press

Published: Mar 12, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off