The type of culture medium and the duration of in vitro culture do not influence birthweight of ART singletons

Transcription

1 Human Reproduction, Vol.30, No.1 pp , 2015 Advanced Access publication on November 17, 2014 doi: /humrep/deu286 ORIGINAL ARTICLE Embryology The type of culture medium and the duration of in vitro culture do not influence birthweight of ART singletons A.De Vos 1, *, R. Janssens 1,H.VandeVelde 1, P. Haentjens 2,M.Bonduelle 3, H. Tournaye 1, and G. Verheyen 1 1 Centre for Reproductive Medicine, Universitair Ziekenhuis Brussel, Laarbeeklaan 101, B-1090 Brussels, Belgium 2 Centre for Outcomes Research and Laboratory for Experimental Surgery, Universitair Ziekenhuis Brussel, Laarbeeklaan 101, B-1090 Brussels, Belgium 3 Centre for Medical Genetics, Universitair Ziekenhuis Brussel, Laarbeeklaan 101, B-1090 Brussels, Belgium *Correspondence address. Tel: ; Fax: ; anick.devos@uzbrussel.be Submitted on June 4, 2014; resubmitted on October 2, 2014; accepted on October 9, 2014 study question: Does the type of in vitro culture medium or the duration of in vitro culture influence singleton birthweight after IVF/ICSI treatment? summaryanswer: In a comparison of two culture media, neither the medium nor the duration of culture (Day 3 versus Day 5 blastocyst transfer) had any effect on mean singleton birthweight. what is known already: Previous studies indicated that in vitro culture of human embryos may affect birthweight of live born singletons. Both the type of culture medium and the duration of culture may be implicated. However, these studies are small and report conflicting results. study design, size, duration: A large retrospective analysis was performed including all singleton live births after transferring fresh Day 3 or Day 5 embryos. IVF and ICSI cycles performed between April 2004 and December 2009 at a tertiary care centre were included for analysis. participants/materials, setting, methods: A total of 2098 singleton live births resulting from singleton pregnancies were included for analysis. Two different sequential embryo culture media were concurrently used in an alternating way: Medicult (n ¼ 1388) and Vitrolife (n ¼ 710). Maternal age, maternal and paternal BMI, maternal parity, maternal smoking, main cause of infertility, cycle rank, stimulation protocol, method of fertilization (IVF or ICSI), time in culture and number of embryos transferred were taken into account. Embryo transfers were performed either on Day 3 (n ¼ 1234) or on Day 5 (n ¼ 864). Singleton birthweight was the primary outcome parameter. Gestational age and gender of the newborn were accounted for in the multiple regression analysis. main results and the role of chance: No significant differences in mean singleton birthweight were observed between the two culture media: Medicult 3222 g (+15 SE) and Vitrolife 3251 g (+21 SE) (P ¼ 0.264). The mean singleton birthweight was not different between Day 3 embryo transfers ( g) and Day 5 blastocyst transfers ( g; P ¼ 0.209). Multiple regression analysis controlling for potential maternal, paternal, treatment and newborn confounders confirmed the non-significant differences in mean singleton birthweight between the two culture media. Likewise, the adjusted mean singleton birthweight was not different according to the duration of in vitro culture (P ¼ 0.521). limitations, reasons for caution: The conclusions are limited by its retrospective design; however, the two different sequential culture systems were used in an alternating way during the same time period. Pregnancy-associated factors possibly influencing birthweight (such as diabetes, hypertension, pre-eclampsia) were not included in the analysis. wider implications of the findings: This large retrospective study does not support earlier concerns that both the type of culture medium and the duration of embryo culture influence singleton birthweight. However, a continuous surveillance of human embryo culture procedures (medium type, culture duration and other culture conditions) should remain a priority within assisted reproduction technology. study funding/competing interest(s): None. Key words: IVF/ICSI / embryo culture / birthweight / perinatal outcome / human & The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please journals.permissions@oup.com

2 Culture medium and duration do not affect birthweight 21 Introduction The widespread use of assisted reproduction technology (ART) has raised concerns about its safety especially towards the children born. The introduction of single embryo transfer has reduced the incidence of multiple births and has overcome to a large extent the problems associated with multiple pregnancies. However, compared with spontaneous conceptions, ART singletons are also reported to have an increased risk of preterm birth, low birthweight and being small for gestational age (SGA) (Pandey et al., 2012; Pinborg et al., 2012). Several factors are proposed to explain the poorer outcome in ART singletons. Subfertility itself is one major risk factor for adverse perinatal outcome in ART singletons (Pinborg et al., 2012) but the influence of ovarian hyperstimulation and embryo culture on the perinatal outcome should not be neglected. As far as in vitro embryo culture is concerned, both the type of culture medium used as well asthe duration of culture is reported to be implicated. Recent studies have indicated that in vitro culture may affect birthweight of live born singletons (Dumoulin et al., 2010; Nelissen et al., 2012). In retrospect, several groups from all over the world have analysed their birthweight data of singleton deliveries after fresh embryo transfer with respect to the embryo culture medium used (Eaton et al., 2012; Vergouw et al., 2012; Carrasco et al., 2013; Lin et al., 2013). Most of these studies compared different culture media to those initially reported; however, they did not confirm any significant association between the embryo culture medium and birthweight (Eaton et al., 2012; Vergouw et al., 2012; Lin et al., 2013). Moreover, a retrospective as well as a prospective comparison with the same culture media as studied by Dumoulin et al. (2010) did not confirm any relationship between the medium used for in vitro culture and the birthweight of singletons born after IVF/ICSI (Carrasco et al., 2013). The length of culture has also been associated with an increased risk of preterm birth (Källén et al., 2010; Kalra et al., 2012; Daret al., 2013). Only Källén et al. (2010) reported an increased risk for low birthweight. In line with Kalra et al. (2012) and Daret al. (2013), reporting an increasedriskof preterm birth without concomitant increased risk for low birthweight, Mäkinen et al. (2012) observed an increase in large for gestational age (LGA) babies according to the length of embryo culture. In contrast, others have not reported any significant effect of the time embryos spent in culture on birthweight (Wang et al., 2009), on the rate of very preterm birth or SGA babies (Sazonova et al., 2011) or on any other perinatal outcome (Fernando et al., 2012). The majority of studies comparing different culture media with respect to the resulting singleton birthweight, and especially the first reports, involved limited numbers of children born (Dumoulin et al., 2010; Eaton et al., 2012; Nelissen et al., 2012; Vergouwet al., 2012). In contrast, most studies comparing Day 3 and Day 5 transfers included large numbers of children born. These studies are based on national registers and therefore maybe biased due to the use of different culture media in different clinics, due to the selection criteria used for Day 5 transfer (Forman et al., 2013) or due to other inter-clinic variations. Two singlecentre studies reported on only 69 (Mäkinen et al., 2012) and 96 (Zhu et al., 2014) blastocyst transfers, with conflicting results. In view of the existing controversies raised by these different studies, additional data would be more than welcome to assess whether the type of culture medium used and the length of embryo culture may have any significant influence on the perinatal outcome of singleton babies born after fresh embryo transfer. Therefore, in the present study, we analysed in a large single-centre dataset, the use of two different sequential embryo culture media in terms of singleton birthweight. The influence of culture duration, being 3 days or 5 days, was also evaluated. Materials and Methods Study design We analysed retrospectively all singleton live births obtained after fresh embryo transfer performed in our centre between April 2004 and December A total of 2098 singleton live births resulting from only singleton pregnancies were included. As such, vanishing twins were excluded from the analysis. We restricted the dataset to cycles with a female age 40 years at oocyte retrieval. Infants born from donated oocytes were also excluded from the study. The cycles in which donor semen was used were included. Patients who had PGD were excluded. Live singleton birthweights were analysed together with maternal age, maternal and paternal BMI, parity of the mother, maternal smoking, main cause of infertility, rank of cycle, stimulation protocol, method of fertilization (IVF or ICSI), type of sequential culture medium, time spent in culture and number of embryos transferred. Embryo transfer was performed on Day 3 or Day 5 after oocyte retrieval. The final dataset contained 2098 singleton live births: 357 obtained after IVF and 1741 obtained after ICSI. Of these, 1234 resulted from a Day 3 embryo transfer while 864 resulted from Day 5 transfer. Two different culture media were used: Medicult (n ¼ 1388) and Vitrolife (n ¼ 710). Other variables analysed included gestational age and gender of the newborn. The present retrospective study was approved by the local ethical committee of the UZ Brussel (Commissie Medische Ethiek O.G. 016) on 16 October Procedures Ovarian stimulation was performed using either a long GnRH-agonist protocol (Suprefact, Aventis Pharma, Frankfurt, Germany) in combination with hmg (Menopur, Ferring Pharmaceuticals A/S) (Papanikolaou et al., 2005) or in combination with recombinant FSH (Puregon, MSD Organon, Oss, The Netherlands or Gonal-F, Merck-Serono, Overijse, Belgium) (n ¼ 931). Alternatively, using a GnRH-antagonist protocol, Orgalutran (MSD Organon, Oss, The Netherlands) or Cetrotide (Merck-Serono, Overijse, Belgium) was administered in order to inhibit a premature LH surge (Kolibianakis et al., 2004)(n ¼ 1138). In a limited number of cycles (n ¼ 29), no agonist or antagonist was used (17 of them were natural cycles, 12 were Menopur stimulated). Final oocyte maturation was achieved by 5000 or IU hcg (Pregnyl, MSD Organon, Oss, The Netherlands) when at least three follicles with a minimum diameter of 17 mm were present at ultrasound. Oocyte retrieval was carried out by vaginal ultrasound-guided puncture 36 h after hcg administration. Intravaginally administered progesterone (Utrogestan, Besins, Brussels, Belgium) was used for luteal phase supplementation. Two different sequential culture systems were used in the same time period, each including fertilization medium, cleavage medium and blastocyst medium. Universal IVF Medium, EmbryoAssist TM and BlastAssist w (Medicult, De Pinte, Belgium) were used between April 2004 and April Vitrolife sequential media GIII Series were used from October 2004, followed from September 2008 onwards by G5 Series TM until December 2009 (Vitrolife, Göteborg, Sweden). Compared with the GIII series, the G5 Series TM additionally included lipoic acid and methionine in the cleavage medium. Gentamicin was used as antibiotic instead of Penicillin G in the GIII Series. Medicult and Vitrolife were used in parallel based on alternation (consecutive treatment cycles were strictly alternated between the two media types on the day of oocyte retrieval). Three intermittent periods involved the exclusive use of Medicult medium, explaining the higher number of offspring in that

3 22 De Vos et al. group. In the Medicult group 1388 children were born, in the Vitrolife group 710 children were born. Both media were pre-supplemented with human serum albumin by the manufacturer and used according to the manufacturer s instructions. A minority of all cycles included were IVF treatments (n ¼ 357, 17.0%), whereas the others were ICSI treatments (n ¼ 1741, 83.0%). Oocytes for IVF were placed in fertilization medium immediately after oocyte retrieval. Insemination was performed with 5000 to progressively motile spermatozoa in 25 ml microdroplets containing one or two oocyte complexes. The next morning, oocytes were denuded and transferred into 25 ml pre-equilibrated droplets of cleavage medium. ICSI oocytes were placed in fertilization medium after oocyte retrieval and into cleavage medium immediately after microinjection. Sperm origin was mainly the ejaculate (91.1%, similar among the two media groups), whereas a minority represented surgically retrieved sperm (8.9%). The embryos were cultured individually in microdroplets under oil (Irvine Scientific, Wicklow, Ireland or Ovoil, Vitrolife, Göteborg, Sweden) at 378C, under 5% O 2 and 6% CO 2. Fertilization was checked h after insemination or ICSI. Embryo development was evaluated daily. Embryos were transferred to blastocyst medium on Day 3. Embryo transfer was performed on either Day 3 or Day 5 after oocyte retrieval. Blastocyst transfer was performed mainly in women under 36 years of age in their first or second trial of IVF (Papanikolaou et al., 2006; Van Landuyt et al., 2006). Data collection and statistical analysis Only singleton live births resulting from singleton pregnancies after transfer of fresh embryos were included in our analysis. Gestational age was calculated from the day of oocyte retrieval, defined as Day 14 of the menstrual cycle. Preterm birth and very preterm birth were defined as delivery before 37 and 32 completed weeks of gestation, respectively. Low birthweight, very low birthweight and high birthweight were defined as birthweight,2500,,1500 and.4500 g. Smoking habits, parental weight and height were documented at the start of the treatment cycle. Continuous data are presented as mean (+SD) for baseline characteristics and as mean (+SE) for outcome variables. Categorical data are presented as number of events and percentages. Between-group differences were assessed using Student s t-test or Chisquare test for continuous or categorical data, respectively. Next, a multiple linear regression analysis was performed to examine the independent effect of culture medium and culture duration on birthweight while controlling for potential confounders. The confounders included: maternal age, maternal and paternal BMI, maternal parity, maternal smoking, main infertility cause, cycle rank, stimulation protocol (agonist, antagonist, or other), fertilization method (IVF or ICSI), number of embryos transferred, gestational age and gender of the newborn. A P-value of,0.05 was considered statistically significant. For all analyses we used IBM SPSS Statistics Version 22. Results Parental and treatment characteristics according to culture medium used The main maternal, paternal and treatment characteristics according to culture medium used are shown in Table I. No differences in maternal and paternal BMI, maternal parity or maternal smoking behaviour were observed between the two culture media. Other parameters for which a significant difference between the two culture media was observed were included as potential confounders in the multiple linear regression analysis. These were maternal age, stimulation protocol, day of embryo transfer and number of embryos transferred. Perinatal outcome according to culture medium used Table II shows the parameters of delivery and children born according to the culture medium used. No differences were observed among the two Table I Main maternal, paternal and treatment characteristics according to culture medium used. Medicult Vitrolife P-value (n ) (n 5 710) Maternal age (years)* 31.9 (4.2) 32.4 (4.3) Maternal BMI (kg/m 2 )* 23.6 (4.5) 23.3 (4.1) Paternal BMI (kg/m 2 )* 25.6 (3.5) 25.9 (3.7) Maternal parity Nulliparous 846 (71.9) 424 (73.9) Parous 1 child 283 (24.1) 123 (21.4) Parous 2 children 47 ( 4.0) 27 ( 4.7) Maternal smoking Non-smoker 1252 (93.2) 651 (94.3) Smoker 91 ( 6.8) 39 ( 5.7) Main infertility cause Female 280 (20.2) 138 (19.4) Male 724 (52.2) 366 (51.6) Mixed 171 (12.3) 99 (13.9) Unexplained 171 (12.3) 86 (12.1) Other 42 (3.0) 21 (3.0) Cycle rank First 757 (54.5) 411 (57.9) Second 336 (24.2) 156 (22.0) Higher order 295 (21.3) 143 (20.1) Stimulation protocol Agonist 600 (43.2) 331 (46.6) Antagonist 773 (55.7) 365 (51.4) Fertilization method IVF 250 (18.0) 107 (15.1) ICSI 1138 (82.0) 603 (84.9) Embryo transfer day Day (56.3) 452 (63.7) Day (43.7) 258 (36.3) Number transferred* 1.46 (0.61) 1.56 (0.69) SET 831 (59.9) 385 (54.2) DET 486 (35.0) 263 (37.1) Higher order 71 ( 5.1) 62 ( 8.7) Please note some missing datafor maternal parity and maternal smoking, not summing to 1388 and 710 cases, respectively, while 15 and 14 cases, respectively, involved a different stimulation protocol, as described in Materials and Methods. Continuous variables were compared by Student s t-test; categorical variables were compared by Chi-square test. SET, single embryo transfer; DET, double embryo transfer. *Mean (SD) or number (%).

4 Culture medium and duration do not affect birthweight 23 Table II Pregnancy and infant characteristics according to culture medium used. Medicult Vitrolife P-value (n ) (n 5 710) Gestational age (weeks)* 38.7 (0.1) 38.9 (0.1) 0.121,32 (very preterm) 22 (1.6) 12 (1.7) 32 to 37 (preterm) 127 (9.1) 52 (7.3) (89.3) 646 (91.0) Newborn gender Male 685 (49.4) 355 (50.0) Female 703 (50.6) 355 (50.0) Mean birthweight (g)* 3222 (15) 3251 (21) 0.264,1500 g (very low) 18 (1.3) 12 (1.7),2500 g (low) 81 (5.8) 39 (5.5) g (high) 7 (0.5) 6 (0.8) Continuous variables (gestational age and birthweight) were compared by Student s t-test; categorical variables (newborn gender) were compared by the Chi-square test. *Mean (SE) or number (%). Table III Pregnancy and infant characteristics according to the two different versions of Vitrolife medium used. Vitrolife Vitrolife P-value GIII Series G5 Series (n 5 309) (n 5 401) Gestational age (weeks)* 38.8 (0.1) 38.9 (0.1) Newborn gender Male 152 (49.2) 203 (50.6) Female 157 (50.8) 198 (49.4) Mean birthweight (g)* 3270 (33) 3237 (28) Continuous variables (gestational age and birthweight) were compared by Student s t-test; categorical variables (newborn gender) were compared by the Chi-square test. *Mean (SE) or number (%). media groups in gestational age, gender of the newborn or mean singleton birthweight. Likewise, when comparing the two Vitrolife media used, no differences arose in these three parameters (Table III). The long GnRH-agonist protocol resulted in a mean singleton birthweight of 3202 g (+18) while the GnRH-antagonist protocol resulted in a mean singleton birthweight of 3257 g (+16). Table IV Maternal, paternal and treatment characteristics according to culture duration. Day 3 Day 5 P-value (n ) (n 5 864) Maternal age (years)* 33.0 (4.2) 30.7 (3.9),0.001 Maternal BMI (kg/m 2 )* 23.7 (4.5) 23.2 (4.2) Paternal BMI (kg/m 2 )* 25.8 (3.6) 25.5 (3.5) Maternal parity Nulliparous 760 (61.6) 510 (59.0) Parous 1 child 221 (17.9) 185 (21.4) Parous 2 children 50 (4.1) 24 (2.8) Maternal smoking Non-smoker 1129 (91.5) 774 (89.6) Smoker 68 (5.6) 62 (7.2) Main infertility cause Female 261 (21.2) 157 (18.2) Male 641 (51.9) 449 (52.0) Mixed 160 (13.0) 110 (12.7) Unexplained 136 (11.0) 121 (14.0) Other 36 (2.9) 27 (3.1) Cycle rank First 625 (50.6) 543 (62.8) Second 299 (24.2) 193 (22.3),0.001 Higher order 310 (25.1) 128 (14.8) Stimulation protocol Agonist 617 (50.0) 314 (36.3),0.001 Antagonist 593 (48.1) 545 (63.1) Fertilization method IVF 212 (17.2) 145 (16.8) ICSI 1022 (82.8) 719 (83.2) Culture medium Medicult 782 (63.4) 606 (70.1) Vitrolife 452 (36.6) 258 (29.9) Number transferred* 1.64 (0.69) 1.27 (0.48),0.001 SET 576 (46.7) 640 (74.1),0.001 DET 537 (43.5) 212 (24.5) Higher order 121 (9.8) 12 (1.4) Please note some missing datafor maternal parity and maternal smoking, not summing to, respectively, 1234 and 864 cases, and 24 and 5 cases, respectively, involved a different stimulation protocol, as described in Materials and Methods. Continuous variables were compared by Student s t-tests; categorical variables were compared by Chi-square test. *Mean (SD) or number (%). Parental and treatment characteristics according to duration of culture Patients receiving day 3 or Day 5 transfer were different patient populations (Table IV). Day 5 transfer patients were significantly younger ( versus years, P, 0.001), more often attempting their first cycle (P, 0.001) and mainly receiving a single embryo at transfer (P, 0.001). These and additional differences in maternal parity, cycle rank, stimulation protocol, type of culture medium used and number of embryos transferred were included in the multiple regression analysis as potential confounding parameters. Perinatal outcome according to culture duration The mean singleton birthweight (+SE) after 5 days of embryo culture was 3250 g (+19) (Table V). This was not different from the mean singleton birthweight (+SE) when embryo culture was limited to 3 days

5 24 De Vos et al. (3219 g (+16), P ¼ 0.209). Gestational age and newborn gender distribution were not different between 3 days or 5 days of embryo culture. When stratifying according to maternal parity, no differences were observed in gestational age, gender distribution or mean singleton birthweight (Table VI). Mean singleton birthweight according to culture medium used and stratified for culture duration Table VII shows the mean singleton birthweights according to the culture medium used but stratified for the duration of culture. No significant differences were observed. Likewise, when separating the Vitrolife group Table V Pregnancy and infant characteristics according to culture duration. Day 3 Day 5 P-value (n ) (n 5 864) Gestational age (weeks)* 38.8 (0.1) 38.8 (0.1) 0.821,32 (very preterm) 17 (1.4) 17 (2.0) 32 to 37 (preterm) 117 (9.5) 62 (7.2) (89.1) 785 (90.9) Newborn gender Male 612 (49.6) 428 (49.5) Female 622 (50.4) 436 (50.5) Mean birthweight (g)* 3219 (16) 3250 (19) 0.209,1500 g (very low) 19 (1.5) 11 (1.3),2500 g (low) 73 (5.9) 47 (5.4) g (high) 6 (0.5) 7 (0.8) Continuous variables (gestational age and birthweight) were compared by Student s t-tests; categorical variables (newborn gender) werecomparedby the Chi-squaretest. *Mean (SE) or number (%). into the two versions GIII and G5, no differences in birthweight were observed between Day 3 and Day 5 transfer (data not shown). Multiple linear regression analysis Multiple linear regression was used to determine the relationship between embryo culture medium or embryo culture duration and singleton birthweight (Table VIII). The data from this analysis indicated that there were no significant differences between the two culture media after correction for the confounding factors. Also, neither culture duration nor the stimulation protocol significantly affected the singleton birthweight. Singleton birthweight after embryo culture was associated with maternal parity, maternal smoking, gestational age and newborn gender. Discussion Our retrospective analysis shows that the use of two different commercially available sequential embryo culture media does not significantly influence the mean birthweight in live born singletons born after fresh embryo (Day 3) or blastocyst (Day 5) transfer. Table VII Mean singleton birthweights according to culture medium used and stratified according to day of transfer. Medicult Vitrolife P-value (n ) (n 5 710) Embryo transfer day Day 3 (n ¼ 1234) 3203 (20) 3247 (25) Day 5 (n ¼ 864) 3247 (22) 3259 (39) P-value Mean (SE). Mean singleton birthweights were compared by Student s t-tests. Table VI Pregnancy and infant characteristics according to maternal parity for Day 3 and Day 5 transfers. Day 3 Day 5 P-value... Gestational age (weeks)* Nulliparous 38.8 (0.1) 38.8 (0.1) Parous 1 child 38.8 (0.1) 38.8 (0.1) Parous 2 children 38.2 (0.4) 38.4 (0.3) Newborn gender (male/female) Nulliparous 381/379 (50.1/49.9) 247/263 (48.4/51.6) Parous 1 child 106/115 (48.0/52.0) 90/95 (48.6/51.4) Parous 2 children 33/17 (66.0/34.0) 14/10 (58.3/41.7) Mean birthweight (g)* Nulliparous 3188 (20) 3208 (25) Parous 1 child 3303 (35) 3353 (40) Parous 2 children 3265 (97) 3408 (89) Continuous variables (gestational age and birthweight) were compared by Student s t-tests; categorical variables (newborn gender) were compared by the Chi-square test. *Mean (SE) or number (%).

6 Culture medium and duration do not affect birthweight 25 Table VIII Results of multiple linear regression analysis among live born singletons. Birthweight Beta P-value (SE) Medicult (reference) Vitrolife 22.0 (12.6) Day 5 (versus Day 3) 8.4 (13.1) Maternal age (per year) 22.3 (3.3) Maternal BMI (per kg/m 2 ) 5.4 (3.0) Paternal BMI (per kg/m 2 ) 20.2 (3.5) Maternal smoking (53.4) Nulliparous (reference) Parous 1 child (31.8) 0.086,0.001 Parous 2 children (65.5) 0.083,0.001 Female infertility (reference) Male infertility (35.1) Mixed infertility 43.9 (44.0) Unexplained infertility (44.6) First cycle (reference) Second cycle 1.9 (38.3) Higher order cycle 21.3 (33.3) Agonist (reference) Antagonist (33.8) ICSI (versus IVF) 38.1 (37.5) SET (reference) DET 12.4 (79.4) Higher order 19.6 (76.7) Gestational age (per week) (6.5) 0.672,0.001 Newborn gender, female (24.2) ,0.001 (versus male) Beta is the regression coefficient. Our data are in line with accumulating reports showing that the type of embryo culture medium does not influence mean birthweight (Eaton et al., 2012; Vergouw et al., 2012; Carrasco et al., 2013; Lin et al., 2013). Different types of culture media, both single-step and sequential, have been compared in these studies (human tubal fluid or HTF, Global, Cook K-SICM, Medicult Universal IVF and ISM1, Sage Quinn s advantage, Vitrolife G1.3, G1.5 and G-1 PLUS). These different culture media were mostly used in consecutive time periods. Our comparisons are closer to those reported by Lin et al. (2013) and Carrasco et al. (2013) in terms of medium type used. According to Eskild et al. (2013), however, a significantly lower birthweight was obtained using Medicult ISM1 when compared with Medicult Universal or Vitrolife G-1 PLUS, even after a limited culture period of 1 or 2 days. All the above-mentioned studies have looked into embryo culture for 2 or 3 days without evaluating the effect of extended culture to the blastocyst stage on birthweight. So far, eight recent studies have compared blastocyst transfer with cleavage stage transfer concerning the effect of culture duration on singleton birthweight. An increased risk for preterm delivery was observed after blastocyst culture (Källén et al., 2010; Kalraet al., 2012; Daret al., 2013), together with an increased risk for low birthweight (Källén et al., 2010). The latter study might, however, be biased by the inclusion of frozen embryo transfers. The duration of the embryo culture was found to be a significant factor in determining the birthweight, as shown by an increased proportion of LGA babies resulting from blastocyst culture (Mäkinen et al., 2012). However, the absolute mean birthweights did not differ between Day 3 and Day 5 transfers (Mäkinen et al., 2012). In contrast, recent findings by Zhu et al. (2014) show that the mean absolute birthweight and gestational age- and gender-adjusted birthweight (Z-scores) of singletons born after blastocyst transfer were significantly higher than singletons born after Day 3 transfer. Both studies, however, are limited by a small sample size in the blastocyst transfer group (respectively only 69 and 96 blastocyst transfers were considered). Other authors did not confirm these differences in perinatal outcome in relation to the duration of embryo culture (Wang et al., 2009; Sazonova et al., 2011; Fernando et al., 2012). When reporting birthweights it is important to adjust the absolute birthweights at least for gestational age and newborn gender. Additionally, some other confounding factors should be considered. Parental characteristics, such as maternal age, maternal and paternal BMI, maternal parity, maternal smoking should be included. Differences in parental phenotype of the Cook and the Vitrolife study group might indeed have influenced the outcome in the studies performed by Dumoulin et al. (2010) and Nelissen et al. (2012). Some cycle characteristics (duration and cause of infertility, rank of cycle, stimulation protocol) and laboratory parameters (IVF or ICSI, number of embryos replaced, day of transfer) can influence birthweight as well (Pinborg et al., 2012) and were corrected for in the present study. The inclusion or exclusion of vanishing twins mayalter the outcome (Pinborg et al., 2007; Luke et al., 2009). Vanishing twins were excluded from the present analysis, although not all published studies on singleton birthweight have consistently done so. Most studies evaluating the type of culture medium have reported mean birthweights and have used multiple linear regression analysis including the appropriate confounding parameters. However, the studies evaluating the length of culture have mainly used risk estimations for defined categories such as preterm and very preterm birth, low and very low birthweight, SGA and LGA. Categorization has limited power to detect small associations between exposures and fetal weight (Land, 2006). In addition to linear regression analysis including the appropriate confounding factors, the use of Z-scores has been advised for perinatal outcome reporting (Land, 2006). Adjusted mean birthweights are then related to a national reference. Less than half of the aforementioned studies have included this Z-score (Dumoulin et al., 2010; Nelissen et al., 2012; Mäkinen et al., 2012; Vergouw et al., 2012; Carrasco et al., 2013; Zhu et al., 2014). We have not calculated the Z-scores because a substantial part of our cohort represents foreign patients (13.6%), for which a Belgian standard would be irrelevant. Additionally, the aim of this study was to compare the two media groups, or between Day 3 and Day 5 transfers, without reference to a national standard. In our cohort, low birthweight was observed in only 6% of the singleton births. None of the above-mentioned studies reported frequencies higher than 10%. Preterm birth was observed in 7 9% of our singleton deliveries, which is in agreement with other reports (Wang et al., 2009; Dumoulin et al., 2010; Källén et al., 2010; Fernando et al., 2012; Nelissen et al., 2012; Lin et al., 2013). The lowest frequencies of preterm birth (1 2%) were reported by a Swedish group (Sazonova et al., 2011), whereas rates of preterm birth were 17.2 and 18.6% in

7 26 De Vos et al. Canadian (Dar et al., 2013) and American (Kalra et al., 2012) ART register databases, respectively. So far, a final answer on whether different types of culture media are affecting singleton birthweight differentially is still lacking. If such an influence would truly exist, it is unclear to what components it would be ascribed. Basically, culture media are bicarbonate buffers containing glucose, pyruvate and amino acids. Media may differ in their protein supplement. Other differences are the addition of taurine, hyaluronan, vitamins and/or growth factors. Whereas the components of commercially available media are often disclosed, their individual concentrations remain unknown. However, most recently, commendable efforts were made to address this gap in available information, by a systematic analysis of the composition of culture media from seven suppliers (Morbeck et al., 2014). Of utmost importance is to provide the embryos with the required nutrients, mimicking physiological conditions as close as possible, and to offer the embryos protection in a critical phase of preimplantation development. Media formulations have been adjusted over time, based on optimized blastocyst formation or increasing knowledge about the oviductal/uterine environment. Studies in animal models are predominant, far more numerous than studies using human embryos. Additionally, focus has been put on success rates (blastocyst formation, clinical implantation), whereas the safety issue needs to be closely followed as well, preferably in large cohorts. It seems justified to have compared Medicult and Vitrolife media because they do differ in several components. Medicult is supplemented with synthetic serum replacement throughout fertilization medium, cleavage medium and blastocyst medium. Vitrolife fertilization medium contains fructose, lactate, non-essential amino acids and EDTA, whereas Medicult fertilization medium does not. Medicult cleavage medium contains HEPES, inositol as an additional carbohydrate and eight different vitamins. Vitrolife cleavage medium instead contains methionine, hyaluronan, lipoic acid and EDTA. Both blastocyst media are quite similar in their amino acid composition except for arginine, which is present in Vitrolife but not in Medicult blastocyst medium. Both blastocyst media contain four vitamins in common, whereas Medicult blastocyst medium also contains D-biotin, folic acid and niacinamide. Medicult blastocyst medium contains inositol as well as ethanolamine. Vitrolife blastocyst medium is supplemented with hyaluronan. If the culture medium type or the culture duration has any effect on neonatal singleton birthweight or gestational length, it is believed to be exerted, at least in part, via epigenetic changes. Many imprinted genes are crucial for fetal growth and development and may thus determine birthweight (Young, 2001, Behr and Wang, 2004). Low birthweight has been associated with increased risk for chronic disease expressed later in life, such as cardiovascular disease, hypertension and type 2 diabetes (Barker, 2004). These associations are thought to be the consequences of developmental plasticity in response to different environmental conditions during development. Indeed, culture environment may be suboptimal, compromising imprint maintenance during preimplantation development (Denomme and Mann, 2012). Aberrant imprinting in cultured embryos has been reported mainly in mouse (Mann et al., 2004; Fauque et al., 2007; Market-Velker et al. 2010) but also in human embryos (Chen et al., 2010). The interplay between culture environment and epigenetic regulation is far from understood (McEwen et al., 2013). Especially in the human, limited availability of research material and the lack of in vivo developed control embryos contribute to this ignorance. Assessment is further complicated by confounding factors, such as subfertility, ovarian stimulation and the effect of gamete manipulation, to bring about fertilization, including IVF and ICSI. The present large single-centre retrospective study does not support earlier concerns that the type of culture medium may influence singleton birthweight. Similarly, no evidence was obtained to support the hypothesis that embryo culture duration may affect singleton birthweight. Our retrospective findings are far from a final answer to this highly debated issue, but at least they may contribute to this debate. Subtle epigenetic changes without overt influence on neonatal birthweight can however not be excluded. A continuous surveillance of human embryo culture procedures (medium type, culture duration and other culture conditions) in terms of efficacy but also safety should remain a priority in the field of ART. Acknowledgements The authors wish to thank the laboratory, clinical and paramedical team of the Centre for Reproductive Medicine. Authors roles A.D.V. and G.V. conceived the manuscript, A.D.V. acquired the data and all authors contributed to the analysis and interpretation of the data. A.D.V. wrote the manuscript, which was critically revised by all authors. All the authors approved the submitted version. Funding None. Conflict of interest None declared. References Barker DJ. The developmental origins of adult disease. J Am Coll Nutr 2004; 23(6 Suppl):588S 595S. Behr B, Wang H. Effects of culture conditions on IVF outcome. Eur J Obstet Gynecol Reprod Biol 2004;115(Suppl 1):S72 S76. Carrasco B, Boada M, Rodriguez I, Coroleu B, Barri PN, Veiga A. Does culture medium influence offspring birth weight? Fertil Steril 2013; 100: Chen SL, Shi XY, Zheng HY, Wu FR, Luo C. Aberrant DNA methylation of imprinted H19 gene in human preimplantation embryos. Fertil Steril 2010; 94: Dar S, Librach CL, Gunby J, Bissonnette F, Cowan L; IVF Directors Group of Canadian Fertility and Andrology Society. Increased risk of preterm birth in singleton pregnancies after blastocyst versus Day 3 embryo transfer: Canadian ART Register (CARTR) analysis. Hum Reprod 2013; 28: Denomme MM, Mann MR. Genomic imprints as a model for the analysis of epigenetic stability during assisted reproductive technologies. Reproduction 2012;144: Dumoulin JC, Land JA, Van Montfoort AP, Nelissen EC, Coonen E, Derhaag JG, Schreurs IL, Dunselman GA, Kester AD, Geraedts JP et al. Effect of in vitro culture of human embryos on birthweight of newborns. Hum Reprod 2010;25:

8 Culture medium and duration do not affect birthweight 27 Eaton JL, Lieberman ES, Stearns C, Chinchilla M, Racowsky C. Embryo culture media and neonatal birthweight following IVF. Hum Reprod 2012; 27: Eskild A, Monkerud L, Tanbo T. Birthweight and placental weight; do changes in culture media used for IVF matter? Comparisons with spontaneous pregnancies in the corresponding time periods. Hum Reprod 2013; 28: Fauque P, Jouannet P, Lesaffre C, Ripoche MA, Dandolo L, Vaiman D, Jammes H. Assisted reproductive technology affects developmental kinetics, H19 imprinting control region methylation and H10 gene expression in individual mouse embryos. BMC Dev Biol 2007;7:116. Fernando D, Halliday JL, Breheny S, Healy DL. Outcomes of singleton births after blastocyst versus nonblastocyst transfer in assisted reproductive technology. Fertil Steril 2012;97: Forman EJ, Werner MD, Scott RT. Extended culture and the risk of preterm delivery in singletons: confounding by indication? Hum Reprod 2013; 28: Källén B, Finnström O, Lindam A, Nilsson E, Nygren KG, Olausson PO. Blastocyst versus cleavage stage transfer in in vitro fertilization: differences in neonatal outcome? Fertil Steril 2010;94: Kalra SK, Ratcliffe SJ, Barnhart KT, Coutifaris C. Extended embryo culture and an increased risk of preterm delivery. Obstet Gynecol 2012;120: Kolibianakis EM, Zikopoulos K, Verpoest W, Camus M, Joris H, Van Steirteghem AC, Devroey P. Should we advise patients undergoing IVF to start a cycle leading to a day 3 or day 5 transfer? Hum Reprod 2004; 19: Land JA. How should we report on perinatal outcome? Hum Reprod 2006; 21: Lin S, Li M, Lian Y, Chen L, Liu P. No effect of embryo culture media on birthweight and length of newborns. Hum Reprod 2013;28: Luke B, Brown MB, Grainger DA, Stern JE, Klein N, Cedars MI. The effect of early fetal losses on singleton assisted-conception pregnancy outcomes. Fertil Steril 2009;91: Mäkinen S, Söderström-Anttila V, Vainio J, Suikkari AM, Tuuri T. Does long in vitro culture promote large for gestational age babies. Hum Reprod 2012; 28: Mann MR, Lee SS, Doherty AS, Verona RI, Nolen LD, Schultz RM, Bartolomei MS. Selective loss of imprinting in the placenta following preimplantation development in culture. Development 2004; 131: Market-Velker BA, Fernandes AD, Mann MR. Side-by-side comparison of five commercial media systems in a mouse model: suboptimal in vitro culture interferes with imprint maintenance. Biol Reprod 2010; 83: McEwen KR, Leitch HG, Amouroux R, Hajkova P. The impact of culture on epigenetic properties of pluripotent stem cells and pre-implantation embryos. Biochem Soc Trans 2013;41: Morbeck DE, Krisher RL, Herrick JR, Baumann NA, Matern D, Moyer T. Composition of commercial media used for human embryo culture. Fertil Steril 2014;102: Nelissen EC, Van Montfoort AP, Coonen E, Derhaag JG, Geraedts JP, Smits LJ, Land JA, Evers JL, Dumoulin JC. Further evidence that culture media affect perinatal outcome: findings after transfer of fresh and cryopreserved embryos. Hum Reprod 2012;27: Pandey S, Shetty A, Hamilton M, Bhattacharya S, Maheshwari A. Obstetric and perinatal outcomes in singleton pregnancies resulting from IVF/ICSI: a systematic review and meta-analysis. Hum Reprod Update 2012; 18: Papanikolaou EG, D haeseleer E, Verheyen G, Van de Velde H, Camus M, Van Steirteghem A, Devroey P, Tournaye H. Live birth rate is significantly higher after blastocyst transfer than after cleavage-stage embryo transfer when at least four embryos are available on day 3 of embryo culture. A randomized prospective study. Hum Reprod 2005;11: Papanikolaou EG, Camus M, Kolibianakis EM, Van Landuyt L, Van Steirteghem A, Devroey P. In vitro fertilization with single blastocyststage versus single cleavage-stage embryos. N Engl J Med 2006;354: Pinborg A, Lidegaard O, Freiesleben NL, Andersen AN. Vanishing twins: a predictor of small-for-gestational age in IVF singletons. Hum Reprod 2007;22: Pinborg A, Wennerholm UB, Romundstad LB, Loft A, Aittomaki K, Söderström-Anttila V, Nygren KG, Hazekamp J, Bergh C. Why do singletons conceived after assisted reproduction technology have adverse perinatal outcome? Systematic review and meta-analysis. Hum Reprod Update 2012;19: Sazonova A, Källen K, Thurin-Kjellberg A, Wennerhom U, Bergh C. Factors affecting obstetric outcome of singletons born after IVF. Hum Reprod 2011; 26: Van Landuyt L, Verheyen G, Tournaye H, Camus M, Devroey P, Van Steirteghem A. New Belgian embryo transfer policy leads to sharp decrease in multiple pregnancy rate. Reprod Biomed Online 2006; 13: Vergouw CG, Kostelijk EH, Doejaaren E, Hompes PG, Lambalk CB, Schats R. The influence of the type of embryo culture medium on neonatal birthweight after single embryo transfer in IVF. Hum Reprod 2012; 27: Wang YA, Sullivan EA, Healy DL, Black DA. Perinatal outcomes after assisted reproductive technology treatment in Australia and New Zealand: single versus double embryo transfer. Med J Aust 2009;190: Young LE. Imprinting of genes and the Barker hypothesis. Twin Res 2001; 4: Zhu J, Lin S, Li M, Chen L, Lian Y, Liu P, Qiao J. Effect of in vitro culture period on birthweight of singleton newborns. Hum Reprod 2014; 29: