Faculty Bibliography

Objective: Morphologic grading of embryos has been an ART standard for nearly 4 decades. More recently, PGT-A has improved embryo selection. Data conflicts regarding whether morphological evaluation improves outcomes of euploid embryo transfers [1, 2, 3]. Our objective was to determine whether morphology is predictive of pregnancy outcomes among single thawed euploid embryo transfers (STEETs).
Design(s): Retrospective cohort study.
Material(s) and Method(s): We reviewed all STEETs at a university-based ART center from 2014-2018. STEETs were excluded if oocytes were cryopreserved, embryos were created at another facility, embryos were frozen or biopsied twice, PGT-M or -SR was used, or an oocyte donor or gestational carrier was used. Only the first STEET during the study period that did not meet any exclusion criterion from each patient was included. Embryo morphology was graded according to Gardner [4]. Outcomes included implantation rate of all transfers and live birth (LB) rate, excluding 154 ongoing pregnancies and 28 pregnancies with unknown birth outcomes. Statistical analysis included chi-square, one-way ANOVA, and 2 multivariable log-binomial regression models to determine the association of predictors (age, expansion, ICM, TE) with implantation and LB.
Result(s): We reviewed 1323 STEETs (mean age 37y; range 24-46y). Overall, implantation was 69% and LB (n=1141) was 55%. ICM and TE were bivariately associated with both implantation and LB (p<0.01), but age and expansion were not. ICM significantly predicted implantation, but TE did not. Both ICM and TE independently predicted LB (see Table for adjusted predicted probabilities of implantation and LB based on ICM and TE grades at mean levels of all covariates in the models).
Conclusion(s): Ploidy status is not the sole determinant of embryo competence. ICM and TE are strong predictors of LB and can improve selection among euploid embryos. Poor ICM is the greatest negative morphological predictor of implantation and LB. Our model can serve as a counseling tool for patients banking embryos. [Figure presented] References: 1) Chen, M., et al. (2015). "Can Comprehensive Chromosome Screening Technology Improve IVF/ICSI Outcomes? A Meta-Analysis." PLoS One10(10): e0140779. 2) Capalbo, A., et al. (2014). "Correlation between standard blastocyst morphology, euploidy and implantation: an observational study in two centers involving 956 screened blastocysts." Hum Reprod29(6): 1173-1181. 3) Irani, M., et al. (2017). "Morphologic grading of euploid blastocysts influences implantation and ongoing pregnancy rates." Fertil Steril107(3): 664-670. 4) Gardner DK, Lane M, Stevens J, Schlenker T, Schoolcraft WB. Blastocyst score affects implantation and pregnancy outcome: towards a single blastocyst transfer. Fertility and Sterility (2000) 73 (6):1155-1158.
Copyright

Objective: PGT-A on TE biopsies (TE Bx/NGS) provides a method of selecting blastocysts with excellent prognosis for establishing clinical pregnancies, minimizing miscarriages and improving live birth rates per ART procedure. However some practitioners distrust the reliability of TE Bx/NGS because mosaicism is seen in normal placentae (derived from the TE) and small numbers of TE cells biopsied may not represent the fetus (derived from the inner cell mass). We examined rebiopsy specimens from the TE and the ICM to determine the reliability of the clinical biopsy to characterize the blastocyst. In particular, we determined concordance between the clinical biopsies and rebiopsy specimens, focusing on 1) chromosomal concordance for disomic results, aneuploid results or mosaic results in the clinical biopsy and 2) clinical concordance (whether biopsy of the ICM was concordant with the initial biopsy's result of "Euploid" vs "NOT euploid") Design: Rebiopsy of blastocysts with known results of clinical PGT-A
Material(s) and Method(s): Results of initial, clinical TE biopsies were obtained from Cooper Genomics. Vitrified blastocysts from patients consenting to research were selected for groups that had no aneuploidy (N = 10), aneuploidy with 1 or 2 aneuploid chromosomes (N = 4) or 1 aneuploid chromosome and 1 mosaic result (N =18). Blastocysts were rewarmed and cells from the TE and ICM were biopsied separately, obtaining as many rebiopsy specimens as possible. Biopsy specimens were subjected to WGA and NGS in our university's core laboratory. NGS results for rebiopsies were compared with results of the clinical biopsy. Rebiopsy chromosomes were considered concordant when the same chromosomal diagnosis was observed Results: Chromosomal concordances were 97.0%, 74.3%, and 13.7% per chromosome, respectively, for disomic (Di), aneuploid (An) and mosaic (Mo) chromosomes in the clinical biopsy. Discordant chromosome results were predominantly mosaic results (2.6%) for Di, mosaic or complementary results (21.6%) for An, or were not seen or non-mosaic aneuploid results (74.5%) for Mo observed for the same chromosome that was seen in the clinical biopsy. These minor discordances can be considered concordant since they mainly confirm the initial results. Counting them as concordant leads to concordances 99.6% for Di, 95.9% for An, and 88.2% for Mo per chromosome. Rebiopsies of inner cell mass were clinically concordant for 100% of the blastocysts (biopsy result of ICM agreed with the clinical result of "euploid" or "not euploid").
Conclusion(s): Despite small number of biopsied cells (required to avoid damage to the blastocyst) and mosaicism (demonstrated by rebiopsy specimens) the excellent chromosomal concordance for rebiopsy specimens (99.6% and 95.9%) and clinical concordance for ICM biopsies (100%) indicate that TE biopsy/NGS provides excellent accuracy in its assessment of ploidy. Within this non-randomly selected subset of blastocysts, mosaics detected in the clinical biopsy outnumbered mosaics detected only by rebiopsy 2.25:1 (18:8).
Copyright

Objective: It is now well-established that embryos diagnosed as chromosomally mosaic can result in healthy live births. Our aim was to report on our clinic's outcomes associated with mosaic embryo transfer (MET), determine which parameters predict MET success, and compare MET outcomes to single thawed euploid embryo transfer (STEET).
Design(s): Retrospective cohort study.
Material(s) and Method(s): All STEET cycles after in vitro fertilization, trophectoderm biopsy, and preimplantation genetic testing for aneuploidy (PGT-A) by next-generation sequencing were identified as controls. Cases included all MET cycles. Statistical analysis included chi-square, with p<0.05 considered significant.
Result(s): A total of 645 PGT-A frozen embryo transfer cycles occurred during the selection period. STEET occurred in 569 cycles (mean age = 35.8), and MET occurred in 70 cycles (mean age = 39.6) with 76 embryos. 47 embryos were diagnosed as segmental mosaic (SM) and 29 embryos were diagnosed as whole chromosome mosaic (WCM; including monosomies and trisomies). 28/47 (59.6%) SM embryos and 10/29 (38.5%) WCM embryos implanted, compared to 408/569 (71.7%) euploid embryos. The ongoing pregnancy/live birth rate was significantly higher in SM embryos (22/47; 46.8%) compared to WCM embryos (5/29; 19.2%; p<0.01). STEET resulted in significantly more ongoing pregnancies (358/569; 62.9%) than both SM (p=0.03) and WCM (p<0.01) embryos. For SM embryos, there was no significant difference in outcomes when stratified by the percentage of mosaicism in the biopsy; however, WCM embryos with 20-40% mosaicism in the biopsy had significantly higher ongoing pregnancy rates compared to those with 41-80% mosaicism (p=0.03). 8/8 WCM embryos with more than 1 chromosome involved failed to implant. There were no significant differences when comparing embryos with a deletion vs. a duplication or a monosomy vs. a trisomy. Maternal age at testing was not associated with differences in pregnancy outcomes in the MET or STEET groups. 15/27 fetuses conceived by MET were reported to be tested by amniocentesis (including 9 with chromosomal microarray), and none were found to have evidence of mosaicism involving the chromosome in question.
Conclusion(s): In the absence of euploid embryos, MET may be considered. While embryos diagnosed as SM have higher reproductive potential compared to those diagnosed as WCM, both categories have inferior outcomes when compared to STEET. While neonatal outcomes are reassuring, additional studies are needed to explore long-term outcomes from babies born following MET.
Copyright

Objective: The use of PGT-A and vitrification to select euploid embryos for transfer has led to improved live birth success in IVF; however, some euploid embryos fail to progress following implantation. Our objective was to compare parameters from 1) the retrieval cycle (IVF) in which blastocysts were biopsied and vitrified, 2) the frozen embryo cycle (FETu) during which the uterus is prepared for transfer, 3) the embryo transfer (FETt), and 4) the embryology (Lab) records all consolidated to determine what best predicts pregnancy loss following establishment of pregnancy by euploid embryos.
Design(s): Multivariate analysis of 45 parameters from IVF, FETu, FETt and Lab and their association with loss of pregnancies after a positive pregnancy test (+hCG).
Material(s) and Method(s): Data were collected from our electronic records for patients with transfers of thawed single euploid embryos diagnosed as euploid by NGS during the IVF cycle. Parameters from IVF (17), FETu (5), FETt (4), and Lab (19) were considered. All cases of STEET using euploid embryos tested with Next Generation Sequencing (908) were considered for analysis. Transfers without +hCG (204) and clinical pregnancies without final outcomes (205) at the time of analysis were excluded. Those 499 remaining cases with a positive pregnancy test (hCG > 5 mIU/mL) and all the required fields. 144 cases failed to progress (75 biochemical pregnancies and 69 SAbs). Stepwise multiple logistic regression (152 combinations of parameters) was performed using the Akaike Information Criterion (AIC) to select parameters associated with loss of pregnancy precluding live birth following +hCG. +hCGs were considered implantations since 1) patients believe they are pregnant when they have a +hCG result and 2) no interfering hCG was administered to these patients.
Result(s): Parameters associated with increased pregnancy loss after positive hCG (in descending magnitude of standard partial regression coefficient) were: more expansion of the trophectoderm prior to biopsy (IVF); more serum estrogen on the day prior to progesterone administration (FETu) and more difficulty of the embryo transfer procedure (FETt ). Age at retrieval, embryo grades, as well as many other parameters were not associated with pregnancy loss.
Conclusion(s): Parameters from 3 categories were associated with loss of pregnancy as biochemical pregnancies or spontaneous abortions. Of these, some are under our control: serum estradiol levels on the day prior to progesterone administration and possibly the difficulty of the transfer and the expansion of the blastocyst prior to biopsy. However, it is possible that these parameters may be aliases for other features such as rate of blastocyst development, patient weight, and/or uterine contractions or presence of blood in the cervix or uterus. No Lab parameters were associated with pregnancy loss. Also notable was the lack of association between embryo grades and pregnancy loss.
Copyright

Objective: Use of a combination trigger (low dose hCG with GnRH-a) is commonly used as part of controlled ovarian hyperstimulation in good responders as it lowers the risk of OHSS compared to hCG alone as well as triggering oocyte maturation. We sought to determine the effects of a second dose of GnRH-a 12 hours after the first, as part of the combination trigger on oocyte and blastocyst quality and risk of OHSS.
Design(s): Retrospective cohort study of all cycles utilizing a combination trigger between 1/1/2017 and 12/31/2018 at our center.
Material(s) and Method(s): Primary outcomes were the number and maturity of oocytes retrieved and incidence of OHSS, as defined by the ASRM practice committee. A subgroup analysis of all cycles utilizing pre-implantation genetic testing for aneuploidy (PGT-A) was performed for the secondary outcomes of number of 2PN, percent of 2PN biopsied as blastocysts and number euploid. Statistical analysis included student t-test, a chi-squared test with and without adjustments for age.
Result(s): 1892 cycles were included in the study, 1394 utilizing a single GnRH-a (G1) and 498 utilizing a double GnRH-a (G2) as part of the combination trigger. Demographics were not different between groups (Table 1). Cycles utilizing G2 had more oocytes retrieved (20.5 v 19.4, p < 0.03), however maturity was not different between groups. The incidence of OHSS was not different between groups, even when controlled for by age and use of cabergoline. Groups were then made by comparing low (<3500pg/mL) v high (>3500pg/mL) E2. The incidence of OHSS was higher in the high E2 group but lower with G2 (11.0% v 14.7%, p = 0.14). A subgroup analysis of cycles utilizing PGT-A was then performed comparing G1 (n=703) vs G2 (n=463). The number of 2PN zygotes was higher in G2 (12.3 v 11.2, p < 0.01) and there was a trend towards more %2pn of all eggs in G2 (60.0% v 59.8% p=0.05). The % of blastocysts biopsied from M2 was not different between groups (45.5% v 46.2%, p = 0.75), however, the total number of euploid embryos was higher with G2 compared to G1 (2.6 v 2.2, p < 0.02).
Conclusion(s): Use of G2 results in more oocytes retrieved and more euploid embryos. While there was no statistically decreased incidence of OHSS in G2 vs G1 overall, when stratified into low vs. high estradiol, there may be a clinically important reduction in OHSS for patients receiving G2 over G1. [Figure presented] Reference: None.
Copyright

Objective: To review outcomes of all STEET procedures based on PGT-A platform used to determine Ploidy status.
Design(s): Retrospective review of all STEET procedures over an 8 year period at a single center.
Material(s) and Method(s): More than 3200 STEET procedures performed over an 8 year period (2011 to 2018) at a single center were reviewed based on the PGT-A platform (NGS, aCGH or SNP) utilized. Our main outcome measures were: Implantation Rate (IR), Clinical Preg rate (FH) and Live Birth (LB) rate. Only embryos reported as euploid were included in the analysis- embryos reported as mosaic or those not yielding a result were omitted. Statistical significance was determined using contingency X² with 1 degree of freedom.
Result(s): TABLE 1. Comparison of STEET outcomes depending on PGT-A Platform [Figure presented] STEET following PGT-A via NGS resulted in a significantly higher IR compared to aCGH /SNP combined (70.1% vs 62.4%). Similarly, ongoing Pregnancy rates and LB rates were significantly improved when NGS was utilized vs aCGH or SNP. SAB rates were not significantly different between platforms but all methods reduced SAB rates compared to age matched controls without PGS (18% ) (Ref 1).
Conclusion(s): STEET results in high IR, high clinical pregnancy rates and high LB rates across all age groups. However, with advances in PGT-A platforms we can continue to improve outcomes and increase safety of ART by maximizing the potential of every ET procedure. With continuing development of PGT-A platforms and interpretation methods used to determine ploidy we can further improve outcomes and safety by transferring a single embryo with the highest implantation potential every time. Reference: Harton et al (2013) Fert & Ster Vol. 100 (6) 1695-1703.
Copyright

Objective: In the era of personalized medicine and the simultaneously increasing use of frozen embryo transfer (FET), assays of the endometrium's receptivity prior to transfer has gained popularity, especially among patients. However, the optimal timing for single thawed euploid embryo transfers (STEET) in a programmed FET has yet to be determined¹. We sought to examine the outcomes of euploid FETs by length of progesterone (P4) exposure at our clinic.
Design(s): Prospective cohort study of all programmed FETs of single euploid embryos between 6/1/2018 and 12/18/2018 at our center.
Material(s) and Method(s): All patients undergoing FET in the inclusion time period were asked to write down the exact time of P4 initiation and then report the start time on the day of P4 serum level check (2 days later) prior to embryo transfer. All FETs were then reviewed. Programmed FET cycles were defined as treatment of oral estradiol daily followed by either 50-75mg intramuscular P4 in oil or vaginal P4 suppository with transfer of a euploid embryo (tested by either array comparative genomic hybridization or Next Generation Sequencing). Programmed FETs with untested, mosaic or double embryo transfers, as well as natural cycles, were excluded. Cycles where exact timing data could not be determined were also excluded. Statistical analysis included ANOVA, receiver operating characteristic curve, and Spearman's Rho correlation where appropriate, with p<0.05 considered significant.
Result(s): 253 programmed STEET cycles met criteria and were included in the analysis. The average patient age at time of ET was 38.0+/-4.5 years. 3 cycles utilized an assay for endometrial receptivity for adjusted timing. The mean duration of P4 exposure was 112.8+/-2.9 hours with a range from 98.25-124.5 hours for all unadjusted cycles. Overall, 166 women had an ongoing pregnancy (OP), 25 had a spontaneous loss (SAB), 12 a biochemical pregnancy (BP) and 50 a negative pregnancy (NP) test, for a 65.6% ongoing pregnancy rate. There was no significant difference in P4 duration between outcome groups (112.8+/-3.1 OP, 112.4+/-4.4 SAB, 111.6+/-1.7 BP, 113.9+/-5.7 NP, F 1.76, df 3, p = 0.156). An ROC curve assessing the ability of P4 duration to predict ongoing pregnancy (OP) had an area under the curve of 0.467 (p=0.38). Furthermore, there was no correlation in the day 28 serum hcg value based on the duration of P4 exposure (rs = 0.058, p = 0.399)
Conclusion(s): Duration of P4 is critical to the success of a programmed FET. At our center, duration of P4 was not associated with FET outcomes. 65.6% of cycles resulted in ongoing pregnancy with our center's standard instructions, which may vary from other centers who have equivalent euploid embryo implantation rates. With growing popularity for individualized testing, these results provide evidence for patient counseling of the high likelihood of desired outcome (ongoing pregnancy) without need for personalized testing. These results also support the need for further prospective or randomized controlled study of optimal FET cycle across clinics and protocols. References: 1. Mackens S, Santos-Ribeiro S, van de Vijver A, Racca A, Van Landuyt L, Tournaye H, Blockeel C. Frozen embryo transfer: a review on the optimal endometrial preparation and timing. Hum Reprod. 2017; 32(11):2234-2242.
Copyright

Objective: To determine the incidence of euploidy in Blastocysts derived from 0PN and 1PN compared with 2PN embryos.
Design(s): Single center retrospective review of PGT-A cases over a 4 year period (2015-2018) where a biopsy and ploidy determination was performed on blastocysts (blasts) derived from zygotes where pronuclei (PNs) were either not evident (0 PN) or only 1 pronucleus (1 PN) was evident at the time of fertilization check.
Material(s) and Method(s): At NYULMC fertilization checks are conducted ~18 hours post insemination or ICSI. The number of PN in each egg is recorded and zygotes are cultured individually. Cases where <=50% of the mature eggs exhibit 2PN are routinely rechecked later on Day 1. In cases for PGT-A, all viable inseminated eggs excluding those with >=3 PN remain in culture to Day 6/7. Good quality blastocysts with a distinct Inner cell mass and cohesive trophectoderm are considered for PGT-A regardless of whether they were 0PN, 1PN or 2PN at fertilization check. PGT-A results are shown in Table 1 along with PGT-A sex of blasts derived from each group.
Result(s): [Figure presented]
Conclusion(s): Prior to utilization of PGT-A and/or timelapse zygotes not exhibiting 2PN at fertilization check were routinely discarded. However, it is now obvious that a percentage of these, albeit small, are fertilized normally and are euploid. Though they account for only a small percentage these may be the only euploid blasts available. Implantation rates and LB rates following transfer of these blasts are similar to those for 2PN blastocysts. Of interest, ratios of XX:XY blasts derived from 1PN and 0PN zygotes were ~2:1 while those from 2PN zygotes were ~1:1. It should be noted that NGS cannot detect pure haploidy (23, XO) or triploidy (69, XXX thereby possibly misdiagnosing these as euploid although our IR and LB results indicate otherwise.
Copyright

Objective: Clinical guidelines on the optimal duration of controlled ovarian stimulation and ideal follicle size were developed for fresh embryo transfer cycles. Whether these apply to freeze all cycles remain unclear. We evaluated the impact of lead follicle size and duration of stimulation on the probability of euploid embryos in women undergoing IVF/PGT-A Design: Cross-sectional study
Material(s) and Method(s): Data from 721 patients undergoing at least two cycles of COS for IVF with preimplantation genetic testing for aneuploidy (PGT-A) via Next Generation Sequencing (NGS) (1859 cycles). Mixed-effect logistic regression, which can account for correlations among repeated outcomes within sample patients, was used to evaluate the association between independent variables and probability of achieving euploid embryos. We first conducted a mixed-effect logistic regression in a univariate manner. All variables then were evaluated in a multivariate model to control for confounding effects. Significant variables to p<0.05 were retained in the final model. p-values <0.05 were considered significant. Statistical analyses were performed using "nlme" and "lme4" package from R project. Results are reported as odds ratios (OR) with 95% confidence intervals (CI).
Result(s): Increasing sum (1.034 [1.022 1.046]/p<0.001) and mean diameter (1.129 [1.046 1.219] p=0.002) of the 5 largest follicles increased the probability of forming euploid embryos. Increasing days of stimulation showed a non-significant trend toward lower chance of forming euploid embryos (0.976 [0.923 1.031]/p=0.382) (Table 1).
Conclusion(s): Allowing the lead follicles to exceed 18mm increases the total number of euploid embryos formed per cycle, presumably by enabling retrieval of additional mature oocytes. Evidence of a detrimental effect of excessive follicle size was not evident in our study, though the number of cycles with follicles exceeding 24 mm was limited. The non-significant trend toward decreased euploid embryos following prolonged stimulation may reflect the effects of poor responders. [Figure presented]
Copyright

Objective: Current data suggests that use of oral antibiotics and corticosteroids (AC) prior to embryo transfer (ET) does not improve ET outcomes. We hypothesized that patients with a history of chronic endometritis (CE) may be an exception to this finding. The objective was to investigate the utility of AC prior to single thawed euploid embryo transfer (STEET) in patients with CE.
Design(s): Retrospective cohort study.
Material(s) and Method(s): Patients who underwent STEET at an academic medical center from 1/2000 to 4/2019 were identified. Cycles prior to 1/2018 received 100 mg doxycycline bid and methylprednisolone 16 mg daily (Pre) prior to ET, and cycles performed after this date did not (Post). Cycles were evaluated for performance of endometrial biopsy (EMB) for CE, with CE defined as presence of plasma cell marker CD-138 (not by hematoxylin and eosin stain alone). Patients positive for CE were treated with 2-3 weeks antibiotics prior to ET cycle start. Outcomes were recorded as not pregnant (NP), biochemical pregnancy (BP), ectopic (E), or intrauterine pregnancy (IUP). Rates of IUP were compared to NP+E and BP. Chi-squared test was used for analysis (p<0.05).
Result(s): 2774 STEET cycles were included. There were 1870 Pre and 904 Post. 462 cycles had an EMB for CE performed. Of these, 238 were positive for CE and 224 were negative. Rates of IUP versus NP, BP, and E combined were not significantly different between all Pre (n=1247, 67%) and all Post (n=628, 68%) with X²(2, N=2765) =2.24 P>.05. Similarly, rates of IUP were not significantly different in patients who had not had an EMB (Untested), or in patients who were tested for CE and found to be negative. In patients with a history of CE there was a small but significant increase in IUP in Post (n=61, 54%) compared to Pre (n=46, 37%) X²(2, N=238) =9.31 P<0.05.
Conclusion(s): Overall, treatment with AC was not associated with higher IUP rates. The use of AC did not improve outcomes in patients with a history of CE, and unexpectedly resulted in lower IUP rates. [Figure presented]
Copyright