Blastocyst culture selects for euploid embryos: comparison of blastomere and trophectoderm biopsies

Blastocyst culture selects for euploidembryos: comparison of blastomere andtrophectoderm biopsies Alexis Adler *, Hsaio-Ling Lee, David H McCulloh, Esmeralda Ampeloquio,Melicia Clarke-Williams, Brooke Hodes Wertz, James Grifo New York University Fertility Center, New York University Langone School of Medicine, 660 1st Ave., 5th Floor, New York,NY 10016, United States* Corresponding author. E-mail address: (A Adler).
Alexis Adler is supervisor of embryology at New York University. Her clinical experience in reproductivemedicine began in 1987 at Cornell, moving to Saint Barnabas before joining NYU in 1995. She consults atprogrammes throughout the USA, China, Taiwan, Israel, Korea, South Africa, and Belgium. Her research inclinical embryology has resulted in abstracts, papers and national and international lectures. Ms Adler receivedher BSc from Barnard College, Columbia University. She has chaired the New York Embryologist Society andASRM's laboratory affiliated society. She also is a member of ACE, Alpha and ESHRE.
genetic diagnosis and screening improves the chances of achieving a viable pregnancy, not only free of undesired single-gene defects but also aneuploidy. In addition, improvements in vitrification provide an efficient means of preserv-ing embryos (blastocysts). By combining trophectoderm biopsy with recent improvements in vitrification methods, only thoseembryos that have proved themselves viable and potentially more competent are tested. Using array comparative genomic hybrid-ization (aCGH) to assess all 24 chromosomes, aneuploidy rates were compared between day-3 blastomere biopsy and day-5 troph-ectoderm biopsy. Of those 1603 embryos, 31% were euploid, 62% were aneuploid and 7% not analysable. A significantly largerproportion of embryos were euploid on day-5 biopsy (42%) compared with day-3 biopsy (24%, P < 0.0001). The number of euploidembryos per patient was not significantly different. Combining extended culture, trophectoderm biopsy and aneuploidy assessmentby aCGH and subsequent vitrification can provide a more efficient means of achieving euploid pregnancies in IVF. RBM ª 2014, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved.
KEYWORDS: aneuploidy, array comparative genomic hybridization, blastomere biopsy, preimplantation genetic screening, trophectodermbiopsy, whole-genome amplification resulting embryos/fetuses. There are aneuploidies that arepaternal in origin or are post meiotic, occurring after fertil- A limiting issue in IVF is the age-dependent high incidence of ization but the majority has been shown originating from aneuploidy with an abnormal number of chromosomes in the meiotic nondisjunction in the oocyte ().
1472-6483/ª 2014, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved.
A Adler et al.
This has been shown to increase with advancing maternal patients regarding the possible benefits of performing biop- age ). These untested, sies only on blastocysts. Recruiting patients to participate in abnormal embryos, whether transferred or conceived via this study with cryopreservation and no fresh transfer as described below was a challenge. Each case was evaluated (or result in a fetus with an abnor- and options discussed with patients. Some patients were mal chromosome compliment often leading to termination.
interested in the variation in treatment options, while oth- order to screen embryos for euploid status, tradition- ers were more reticent. Others became aware of the new ally day-3 cleavage-stage blastomere biopsy is performed procedure and requested trophectoderm biopsy. Patients with chromosome analysis of the most common aneuploi- were enlisted for the study between 1 January 2010 and 1 dies diagnosed by fluorescence in-situ hybridization (FISH).
September 2012. All patients desiring to undergo blasto- The cells' DNA can also be amplified using PCR to test for mere biopsy with preimplantation genetic screening for known single-gene disorders and can provide results in time aneuploidy determination with or without gender selection for a day-5 embryo transfer were considered for the study. Patients with preimplanta- Higher pregnancy rates are obtained in some patient groups tion genetic diagnosis for single-gene disorders or using using day-3 blastomere biopsy with fresh transfer on day 4 donor oocytes were excluded from this study. Patients who presented prior to onset of trophectoderm biopsy were the advancement in the detection of the full chro- provided only the day-3 biopsy option.
mosome complement available with array comparative Several patients undergoing trophectoderm biopsy had genomic hybridization (aCGH), embryos can be more fully suffered repeated miscarriages and were diagnosed with screened to avoid transferring embryos that do not contain Asherman's syndrome or other uterine issues and subse- a normal karyotype ). In addition, with quently successfully used gestational carriers to carry their improvement in media as well as culture conditions, blasto- euploid embryos to term. These patients were excluded cyst formation can reach an average of 40–60% ( from the study.
Blastocyst culture has been shown toimprove pregnancy and implantation rates by allowing Ovarian stimulation selection to occur in the Petri dish. The extended culturegives an embryo time to better express its potential and All underwent ovarian stimulation involving: (i) luteal-phase has become an invaluable tool for embryo selection.
down-regulation with a gonadtrophin-releasing hormone Observing development in a cohort of embryos and basing (GnRH) analogue; (ii) follicular-phase flare using GnRH ana- selection of embryos for transfer on blastocyst morphology logue; and (iii) antagonist-controlled stimulations in which alone gives better implantation and birth rates GnRH antagonist was administered when lead follicles However, there are some embryos that arrest at the achieved diameters of roughly 12 mm. All cycles included cleavage stage: they simply stop growing. These embryos daily injections of gonadotrophins, either recombinant FSH prove that they have no further potential, and most have (Gonal F, EMD Serono, Rockland, MA, USA; or Follistim, degraded DNA and/or are aneuploid ( Merck, Whitehouse Station, NJ, USA) or a combination of work hoped to improve success by combining two recombinant FSH and human menopausal gonadotrophin selection tools: selecting euploid embryos (genetic selec- (Menopur, Merck). Injection of human chorionic gonadotro- tion) that had achieved the blastocyst stage (developmental phin (HCG; 10,000 IU urinary HCG or two vials of recombinant HCG (Ovidrel; EMD Serono) or leuprolide ace- performing trophectoderm biopsy and 24-chromosome pre- tate (Lupron, Tap Pharmaceuticals, Lake Forrest, IL, USA) implantation genetic screening using aCGH onlyon embryos plus 1000 IU urinary HCG was used to trigger the final mat- reaching the blastocyst stage ( uration, when at least two follicles had reached 18 mm, ). We compared the euploidy rates of day-3 although individualized criteria are often used when a blastomere biopsy to trophectoderm biopsy at the blastocyst patient has had prior cycles. The oocyte retrieval was stage to determine if extended embryo culture also selects scheduled 36 h after the trigger.
for euploid status and then compared pregnancy and implan-tation rates between the two different days of biopsy.
Day-3 biopsy and genetic screening Cleavage-stage blastomere biopsy was performed on day-3embryos that had at least 4 cells. A hole was created in the zona pellucida using acidified Tyrode's solution (Sigma,St Louis, MO, USA). Calcium/magnesium-free medium presented to New York University Fertility Center (Global, Guilford, CT, USA) was used to loosen the cell–cell with interest in preimplantation genetic screening since it adhesion allowing for less traumatic suction removal of the had been performed here for years prior to this study. The cell from the embryo. The biopsied cell was then loaded use of aCGH as the method for genetic screening began in into a PCR tube (Eppendorf) labelled with the corresponding 2010 as the centre was convinced that screening all chromo- embryo number and sent to the reference laboratory somes was an improvement over screening a limited number (Reprogenetics, Livingston, NJ, USA) on dry ice for analysis of chromosomes using FISH probes. Recruiting patients for a using whole-genome amplification. Embryos were then new unproven technique (trophectoderm biopsy) was diffi- placed back in culture until aCGH results were obtained in cult. Laboratory personnel and physicians counselled time for a day-5 embryo transfer. The best-quality euploid Comparing euploid rates of blastomere versus trophectoderm biopsies embryo(s) were then transferred using a Sureview Wallace catheter (Smith's Medical, UK) with ultrasound guidance.
Progesterone supplementation was provided to patients total of 305 patients underwent ovarian stimulation: 131 with fresh transfers of biopsied embryos by injection of pro- in anticipation of blastomere biopsy and 174 in anticipation gesterone in oil or by vaginal administration (Crinone, of trophectoderm biopsy. There were no significant differ- Endometrin or suppositiories) beginning on the day follow- ences in patient age, maximum day-2 or -3 FSH or inci- ing oocyte retrieval.
dences of most of the diagnoses or stimulation protocolsbetween the day-3 and trophectoderm biopsy groups Trophectoderm biopsy and genetic screening Patients in the trophectoderm biopsy group hada higher incidence of uterine factor infertility. Oocyte For blastocyst biopsy, a narrow channel was created in the retrieval was cancelled because of poor ovarian response zona pellucida using a laser (Cronus, Research Instruments, in 9.9% (13/131) of patients in the day-3 biopsy group and UK) on day 3. On days 5 and 6 and on rare occasions day 7, 4.6% (8/174) of patients in the trophectoderm biopsy group.
embryos were assessed and any fully differentiated total of 284 patients underwent oocyte retrievals: 118 good-quality blastocyst was biopsied. Using suction, the for day-3 biopsy and 166 for trophectoderm biopsy. Patient trophectoderm cells extruding from the expanded blasto- treatment parameters were not significantly different, cyst were gently pulled and then a few pulses of the laser including day-2 FSH and oestradiol concentrations, total were used at cell junctions in order to safely remove a gonadotrophin, FSH and/or human menopausal gonadotro- few cells without disrupting the inner cell mass. The cells phin administered and days of gonadotrophin administration were washed and loaded into PCR tubes labelled with corre- sponding embryo number and sent to the reference labora- total, 1603 embryos were analysed (936 biopsied on tory on dry ice for 24-chromosome analysis. Following day 3, 667 biopsied at the blastocyst stage) over the trophectoderm biopsy, blastocysts were vitrified using Sage 2.5-year period. Maternal age at biopsy was (mean ± stan- (Trumbull, CT, USA) or Irvine (Santa Ana, CA, USA) media dard deviation) 38.1 ± 4.3 years (range 22–47 years) and and vitrification kits and CryoLocks (BioTech, Cummin, was not significantly different between the two groups.
GA, USA) according to the manufacturer's instructions.
Trophectoderm biopsies were performed on day 5 (59.6%), All biopsies were sent to a reference genetics laboratory day 6 (38.7%) and day 7 (1.7%) just prior to vitrification.
for preimplantation genetic screening using whole-genome was no difference between patient parameters in amplification and aCGH as described previously ( the two biopsy groups when considering the numbers of oocytes retrieved, patients with intracytoplasmic sperm After the aCGH results were obtained, the patient's prep- injection, mature oocytes, 2PN zygotes, embryos cultured aration for transfer of vitrified–warmed embryos began with beyond day 3 (extended culture) or cells in day-3 embryos oral oestradiol tablets (Estrace; Warner Chilcott, Rockaway, ). There were significant differences between the NJ, USA) beginning on day 2 of the next menstrual cycle.
number of embryos biopsied per patient (8.8 day-3 versus Once the uterine lining was at least 7 mm with a ring pattern, 4.7 trophectoderm biopsy; P < 0.0001). There were also sig- injections of progesterone in oil (Watson Pharma, Parsippa- nificant differences in the number of blastocysts forming on ny, NJ, USA) or suppositories (prepared by local pharmacies) days 5 or 6 (4.1 day-3 versus 5.3 trophectoderm biopsy; were begun. On day 6 of progesterone, the desired number P = 0.031) and the number of blastocysts that achieved (generally one and sometimes two) euploid embryos were stage 2 or greater (3.4 day-3 versus 4.7 trophectoderm warmed using Sage or Irvine warming kits and then trans- biopsy; P = 0.012; ).
ferred to the patient's uterus under ultrasound guidance.
Pregnancy results were obtained on day 28 of the cycle and confirmed on day 35 with ultrasound on day 42–49 forpresence of intrauterine sac and fetal heart activity.
those 1603 embryos that were biopsied, 31.3% (n = 501) Specific outcome measures included rates of euploidy, were euploid, 62.1% were aneuploid (n = 995) and 6.7% implantation and clinical pregnancy. Euploidy rate was cal- (n = 107) had no diagnosis. There were 6.5 ± 4.6 embryos culated using the number of euploid embryos divided by (range 1–35) biopsied and 2.0 ± 3.0 (range 1–17) were the total number of embryos that were biopsied. Implanta- found to be normal. The proportion of biopsied embryos tion rate was calculated as the number of intrauterine ges- that were euploid was significantly larger following trophec- tational sacs visualized on ultrasound divided by the total toderm biopsy compared with day-3 biopsy (42% versus 24% number of embryos transferred. A clinical pregnancy was expressed per biopsied embryo, P < 0.0001; 32% versus 23% defined as the presence of an intrauterine gestational per biopsied patient, P = 0.011; The number of sac(s) with fetal cardiac activity as documented by ultra- euploid embryos per patient was not significantly different sound. Variables in the study groups were compared by between the day-3 and trophectoderm biopsy groups (2.1 chi-squared analyses and Fisher's exact test or Student's versus 2.0).
t-test, as appropriate. Statistical analysis was performed patients (29%) who had embryos biopsied on using Microsoft Excel, SPSS software and/or GraphPad day 3 had no euploid embryos. Twelve additional patients online. Patients' acceptance of informed consent was indi- with at least one euploid embryo had no euploid blastocysts cated by their signatures in accordance with Institutional at the time of transfer (day 5) or the following day (day 6).
Review Board protocol (NYU IRB no. 512-03373, approved (These patients underwent transfer of euploid morulae or 10 October 2012).
euploid cleavage-stage embryos without blastocysts on A Adler et al.
Characteristics of patients who underwent ovarian stimulation.
Blastomere biopsy Trophectoderm biopsy Maximum FSH (IU/l) Infertility diagnosis Diminished ovarian reserve Ovulatory dysfunction (PCO) Luteal down-regulation Cycle cancellation Values are mean ± SD or % (n).
PCO = polycystic ovaries.
aP = 0.003.
Cycle and embryological data of patients who underwent oocyte retrieval.
Blastomere biopsy Trophectoderm biopsy Day-2 FSH (mIU/ml) Day-2 oestradiol (pg/ml) Gonadotrophin (total IU) Days of gonadotrophin Oestradiol (pg/ml) on trigger day Patients with ICSI Embryos in extended culture Stage-2 blastocysts Patients with biopsy Values are mean ± SD or % (n).
HMG = human menopausal gonadotrophin; ICSI = intracytoplasmic sperm injection.
aP = 0.031.
bP = 0.012.
day 5.) Therefore, there were 43 patients (41%) in the blas- had no euploid blastocysts. The incidence of patients with tomere biopsy group with no euploid blastocysts on day 5 or no euploid blastocysts was not significantly different when 6. Fifty-four patients who had trophectoderm biopsies (38%) expressed per biopsied patient (41%, 43/106 versus 38%, Comparing euploid rates of blastomere versus trophectoderm biopsies Results of patients who underwent embryo biopsy.
Blastomere biopsy Aneuploid embryos Embryos with no determination Embryos biopsied/biopsy patient Euploid embryos/biopsy patient Euploid embryos/biopsied embryo (%) Euploid embryos/cycle Cycles with no euploid embryos Cycles with no euploid blastocysts Values are mean ± SD (total numbers in parentheses) or % (n).
NS = not statistically significant.
54/141), per retrieval (47%, 55/118 versus 48%, 79/166) or significantly different (t-test, P = 0.0001), comparisons of per cycle initiated (52% 68/131 versus 50%, 87/174).
clinical pregnancy rates may not be appropriate.
the euploid embryos that were transferred, 31 babies Transfer and pregnancy results were born in the blastomere biopsy group (25% of thosetransferred) and 45 babies were born in the trophectoderm There were 75 embryo transfers in the blastomere biopsy biopsy group (58% of those transferred). There were three group and 68 in the trophectoderm biopsy group, for which miscarriages and one stillbirth in the blastomere biopsy 125 embryos (1.7 ± 0.6 embryos per transfer) and 78 group (11.4%) versus six miscarriages in the trophectoderm (1.2 ± 0.4 embryos per transfer) were transferred. There biopsy group (15.4%). There were no misdiagnoses found was a significantly higher implantation rate in the trophec- in any of the babies born or in tissue recovered from the ter- toderm biopsy group compared with the embryo biopsy minations that occurred due to nonviable pregnancies.
group (60% versus 41%, respectively, P = 0.004), but the clin- the blastomere biopsy group, there were 12 fresh ical pregnancy rate was not significantly different (47% ver- transfers that included only euploid nonblastocysts (moru- sus 57%, respectively) (). However, since the lae and cleavage-stage embryos). Only one pregnancy numbers of embryos transferred in the two groups were resulted from transfer of two morulae (implantation rate14%, 2/14). In contrast, when only euploid embryos weretransferred and the transferred embryos included at least one blastocyst, the implantation rate was 44% (49/111)and was significantly different from the implantation rate for nonblastocysts (Fisher's exact test, P = 0.043). Theimplantation rate of embryos biopsied on day 3 when blas- tocysts were transferred (44%) was not significantly differ-ent implantation rate of embryos undergoing trophectoderm biopsy, vitrification, rewarming and transfer (58%, 45/78).
observations provide evidence to support three important ideas: (i) extended culture enables selection of embryos with a greater chance of implantation and live delivery; (ii) selection of embryos that achieve the blasto- Implantation rates and clinical pregnancy rates per cyst stage enriches for euploid embryos; and (iii) extended transfer for embryo and trophectoderm biopsies. Implantation culture does not lead to a significant loss of euploid embryos rates were significantly different between embryo and troph- capable of implantation.
ectoderm biopsy (P = 0.004) when considering all patients with present observations indicate that among euploid embryo transfers. Clinical pregnancy rates were not signifi- embryos biopsied on day 3, those that become blastocysts cantly different between the two groups. CPR = clinical preg- by the time of transfer are more likely to implant than those IR = implantation that do not become blastocysts by the time of transfer. It is TEB = trophectoderm biopsy.
clear that transferred euploid blastocysts selected on day 5 A Adler et al.
have greater potential for implantation than euploid Advances in cryobiology, specifically vitrification, have embryos that had not reached the blastocyst stage at trans- allowed for cryopreservation and very high post-warming fer. This suggests that there is a qualitative component of survival of the blastocysts post biopsy. Trophectoderm the embryo that is associated with ability to implant but biopsy would not be universally feasible without reliable that is unassociated with embryo ploidy. This qualitative cryopreservation techniques. The ability to vitrify blasto- component may simply be the embryo's potential to implant cysts with reliable, excellent warming results, allows time or a more complex embryo–uterus synchrony issue. This for analysis ) and minimal loss of precious, confirms Munne's results on discarded embryos and mor- euploid embryos. Also there is evidence that an unstimu- phology along with also confirming day-5 embryo transfer lated cycle provides a potentially superior uterine environ- yielding higher pregnancy and implantation rates ( ). Extended culture transfer (). This allowed for transfer of enables selection of embryos with a greater chance of warmed embryos in a programmed warming cycle. Transfer implantation and live delivery.
of vitrified–warmed euploid blastocysts was certainly not data demonstrate that embryos biopsied at the blas- inferior to transfer of fresh euploid blastocysts in the cur- tocyst stage (trophectoderm biopsy) have a higher euploidy rent study centre. In addition, the implantation rate of vit- rate than embryos biopsied as cleavage-stage embryos rified, euploid blastocysts was significantly greater than (embryo biopsy). Extending culture and performing biopsy implantation rate of fresh transferred embryos selected at the blastocyst stage also results in lower aneuploidy rates solely based on euploidy. It has been suggested that the as compared with day-3 biopsy. The selection process that uterine environment after exogenous oestrogen and proges- occurs in the Petri dish eliminates the need to biopsy those terone may be less prone to detrimental effects thought to embryos not competent to advance beyond the cleavage occur during ovarian stimulation. The success with this stage and supports the concept that selection is occurring approach demonstrates that single warmed euploid embryo in the laboratory. Biopsying at the blastocyst stage identi- transfer is an extremely effective method for achieving fies only those embryos with the best chance for being chro- euploid pregnancy.
mosomally normal and developmentally competent. This is Numbers of embryos per patient undergoing extended potentially the reason that extended culture provides higher culture (beyond day 3) and the number of cells in embryos pregnancy, implantation and live birth rates. Selection of (on day 3) were not significantly different. In contrast, from embryos that achieve the blastocyst stage enriches for this similar number of embryos with similar quality, a signif- euploid embryos.
icantly lower number of blastocysts and blastocysts (stage 2 distinction should be made between euploid embryos or greater) arose in the blastomere biopsy group compared and euploid blastocysts in the blastomere biopsy group with the trophectoderm biopsy group. The number of where some of the biopsied embryos found to be euploid euploid embryos averaged 1.9 ± 2.3 per retrieval following failed to progress to the blastocyst stage by the time of blastomere biopsy. This number declined to 1.2 ± 1.6 transfer (day 5) or cryopreservation (day 6). This attrition euploid blastocysts on day 5 following blastomere biopsy accounts for the difference between numbers of euploid due to failure of biopsied embryos to progress to the blasto- embryos and euploid blastocysts in the blastomere biopsy cyst stage. The performance of day-3 biopsy is one remark- group. In the trophectoderm biopsy group, only blastocysts ably different feature that stands out when comparing these were biopsied so there was no difference between the num- two groups of embryos. This suggests that day-3 biopsy itself ber of euploid embryos and the number of euploid blasto- is responsible for poorer development to the blastocyst cysts. The enrichment in euploid embryos obtained by stage and supports the contention of selection of blastocysts for biopsy was achieved with no sig- that blastomere biopsy on day 3 is detrimental to nificant loss of euploid embryos since neither the number of embryo development, although this study cannot exclude euploid embryos per patient nor the number of euploid blas- the possibility of simple, inherent poor quality of some of tocysts per patient was significantly different when compar- the day-3 euploid embryos. The number of euploid blasto- ing the two biopsy groups. Excluding the embryos that failed cysts averaged 1.7 ± 2.8 following trophectoderm biopsy.
to progress to become blastocysts from testing had no The lack of significance between the numbers of euploid impact on the patient other than to reduce the number of blastocysts with embryo versus trophectoderm biopsy sug- embryos biopsied. Extended culture does not lead to a sig- gests that there was no significant loss of euploid embryos nificant loss of euploid embryos capable of implantation.
or blastocysts in association with extended culture without advantages of trophectoderm biopsy over blasto- biopsy. This lack of significance gives confidence in delaying mere biopsy are not only selection, but also with trophecto- biopsy to the blastocyst stage.
derm biopsy more cells can be removed, compared with one The limitations of this study are that it is retrospective blastomere at day 3 (). This permits and that the study was performed longitudinally. The intent a more robust amplification and aneuploidy assessment.
of this paper was to describe this study centre's experience Although there are more cells removed for testing, a lesser with implementing trophectoderm biopsy and some notable percentage of the embryo is removed and the inner mass observations about the success of trophectoderm biopsy rel- cells destined for fetal development are not compromised, ative to blastomere biopsy. Despite its longitudinal nature, presumably avoiding damage to the tissue that becomes the the patients in the two study groups were quite similar as fetus. Along with the advanced development at the blasto- indicated in which show that the patients' cyst stage, there is evidence of less mosaicism found com- ages, ovarian reserves, diagnoses, treatment parameters pared with that of day-3 embryos ( Comparing euploid rates of blastomere versus trophectoderm biopsies similarity between study groups demonstrates that compar- isons of outcomes are meaningful.
In conclusion, by extending embryo culture to the blasto- cyst stage, a selection process occurs that enriches for euploid embryos. This further elucidates why higher preg- nancy, implantation and live birth rates can be achieved with blastocyst transfer in unscreened embryos compared with day-3 transfer. This is accomplished by eliminating those embryos not competent enough to develop to the blastocyst stage in addition to enrichment for euploidy.
However, trophectoderm biopsy adds a further tool by being able to screen out blastocysts (viable embryos) that are aneuploid, thereby improving chances of implantation by a normal embryo and delivering a healthy, chromosomally normal baby. Many morphologically normal blastocysts are aneuploid and selecting only euploid embryos lowers mis- carriage rates ). Combining extended culture, trophectoderm biopsy and aneuploidy assessment by aCGH and subsequent vitrification can pro- vide a highly efficient means of achieving singleton euploid pregnancies in IVF.
Thanks to Tsai-Ling (Cindy) Lee, Anne Goldstein-Tufaro, Yael Kramer, Fran Hooper and Imelda Weill. In addition, the authors thank all of the members of the NYU Fertility Centre as well as the Reprogenetics team.
Declaration: The authors report no financial or commercial conflicts of interest.
Received 9 January 2013; refereed 26 November 2013; accepted 27 November 2013.

Source: http://www.irvinesci.sitesetup.net/uploads/technical-documentations/Adler_A_2014_-_Blastocyst_culture_selects_for_euploid_embryos.pdf