Faculty Bibliography

OBJECTIVE: With increased availability of genetic testing, particularly expanded carrier screening (ECS) and hereditary cancer (HC) testing, the scope of conditions for which PGT-M is performed is expanding. Our aim was to report on indications for PGT-M from the past decade in our large academic practice. MATERIALS AND METHODS: All PGT-M cases occurring between January 2010 and April 2021 were reviewed.
RESULT(S): A total of 331 patients were identified for which PGT-M was performed for 124 different genes over 582 cycles. Eighteen patients tested for two genes and one patient tested for three genes; therefore, there were a total of 351 unique PGT-M cases. Of the 124 genes tested, 82 (66.1%) were of childhood onset while 16 (12.9%) were of adult onset, and the remaining 26 (21.0%) were of variable onset. Over the entire study period, 70/351 patients (19.9%) tested for 16 genes related to HC syndromes; between 2010-2017, HC-related PGT-M accounted for 12.6% (20/159) of our total PGT-M volume, and since 2018, it rose to 26.0% (50/192). Overall, BRCA1 was the most common gene tested in our practice, and hereditary breast and ovarian cancer syndrome (BRCA1 and BRCA2) accounted for 15.1% (53/351) of our total PGT-M patient population. 181/351 patients (51.6%) tested for 49 genes that are commonly found on ECS, with cystic fibrosis (CFTR) being the most common (34/351) followed by fragile X (FMR1; 32/351); these represented the second and third most common genes tested in our practice (9.7% and 9.1% respectively). Of all patients doing PGT-M for ECS-related conditions, 46.4% (84/181) tested for 41 genes that are not detected by traditional or ethnicity-based carrier screening, with the most common being GJB2-related nonsyndromic hearing loss (the fourth most common condition tested in our practice, representing 6.6% of our total PGT-M volume), followed by 21-hydroxylase deficient congenital adrenal hyperplasia (CYP21A2) and familial Mediterranean fever (MEFV). There were eight patients (2.3%) who either were or could have been identified on our current 283-disease ECS panel but would have been missed on our prior 176-disease ECS panel. Eight patients did PGT-M for HLA matching, and three did non-disclosure PGT-M (two for Huntington's disease/HTT and one for CADASIL/NOTCH3). 80/124 genes tested (64.5%) were unique to a single patient.
CONCLUSION(S): PGT-M is performed for a wide range of genetic conditions, and nearly two-thirds of genes tested in our clinic were unique to a single patient. While most conditions tested are childhood-onset, BRCA1 is the most common gene tested by our patient population, and the proportion of patients testing for HC syndromes has doubled over the past three years. More than half of patients pursued PGT-M for conditions detectable through ECS but not through traditional carrier screening; however, increasing the ECS panel size by more than 100 conditions has only had a minor effect on PGT-M uptake. IMPACT STATEMENT: This large dataset from a single IVF clinic highlights the impact of HC testing and ECS on PGT-M utilization over the past decade

OBJECTIVE: To compare telomere length (TL) and telomerase gene expression in human euploid and aneuploid blastocysts generated from IVF treatment. MATERIALS AND METHODS: TL and telomerase gene expression were measured in cryopreserved aneuploid (N=115) and euploid (N=4) human blastocysts donated by 26 patients who consented research under approval of IRB study #16-00154. Blastocysts were classified according to number of aneuploid chromosomes (A1-one segmental error, A2-one whole chromosome error, A3-two chromosomal errors and A4- >= 3 chromosomal errors). Genomic DNA and messenger RNA were separated simultaneously from individual blastocysts after thawing in vitrification-warming media. Telomerase reverse transcriptase (TERT) and telomerase RNA component (TERC) mRNA levels were determined by RT-qPCR with GAPDH as internal control, and TL was measured by qPCR with 5s rDNA as internal control. Relative gene expression and TL were calculated by DELTADELTA^Ct method, and GraphPad Prism 8 software was used for statistical analysis.
RESULT(S): TL and telomerase gene expression were not normally distributed, so nonparametric tests were used to compare the medians among groups (Table 1). Median TL, TERTand TERC levels didn't differ by number of chromosome errors nor between aneuploid and euploid groups. Intriguingly, TL, TERT and TERC levels in aneuploid blastocysts tended to be greater compared to euploid blastocysts. TL in blastocysts correlated with telomerase TERT expression (R² =0.054, P = 0.011), but not TERC expression (R² =0.0002, P = 0.865).
CONCLUSION(S): To our knowledge, this is the largest study to measure telomere length and telomerase gene expression in human blastocysts. Our data indicated that telomeres are lengthened and telomerase is activated in aneuploid embryos at blastocyst stage. Moreover, telomere length and telomerase gene TERT in human blastocysts correlate regardless of ploidy status. Like cancer cells, TERT is highly expressed in aneuploid blastocysts. IMPACT STATEMENT: Robust TERT expression and telomere maintenance in aneuploid human blastocysts may explain why extended in vitro culture alone is insufficient to cull out aneuploidy embryos during IVF (Table Presented)

OBJECTIVE: It is well known that the embryo aneuploidy rate increases with women's age [1], but the effect of age on complex chromosomal abnormality (CCA) is less clear. Here, we addressed the relationship between maternal age and CCA with a retrospective cohort study. MATERIALS AND METHODS: We reviewed results of preimplantation genetic testing (PGT) by aCGH or NGS of embryo biopsies performed in an academic IVF unit between 2010 and 2019. We excluded PGT results from single gene disorder and egg donation cycles. CCA was defined as>=3 chromosome abnormalities (whole, partial and/or mosaic). Maternal age was categorized according to SART age groups: <35, 35-37, 38-40, 41-42, and >42 years. Statistical analyses were conducted using GraphPad Prism 8.
RESULT(S): 27,423 embryos were biopsied from 3,501 women aged 23 to 48 years. 4,740 embryos (16%) has CCA. Consistent with prior study [2], the most frequent chromosomes involved in CCA were 22, 16, 21 and 15, with incidences of 30.6%, 29.1%, 26.1% and 25.8% respectively. The number of chromosomal errors (from 3 to 42) involved in CCA did not correlate with maternal age (Spearman r = -0.0149, P = 0.3352). However, the rate of complex abnormal embryos tended to increase with advancing maternal age (9.7%, 11.2%, 10.9%, 24.8% and 43.6% in women aged < 35, 35-37, 38-40, 41-42, and > 42 years, respectively). Women over 40 years old had significantly higher rates of CCA compared to those under 40 years (Chisquare test, P < 0.0001). Surprisingly, the relationship between maternal age and CCA followed a U-shaped curve, decreasing from the 25 to 30 year old group (Pearson r = -0.831, P = 0.04) to the 30 to 35 year old group (Pearson r = 0.093, P = 0.861), then increased markedly in the 35 to 48 year old group (Pearson r = 0.921, P < 0.0001).
CONCLUSION(S):We found that CCA embryos share common features of aneuploidy, such as association with maternal age and preferential involvement of shorter chromosomes i.e. 22, 16, 21 and 15. Unexpectedly, our data showed that the relationship between CCA and maternal age assumes a U shape with increased rates at very young and very old ages. Both meiotic and mitotic errors contribute to chromosomal abnormality, and the contribution of each to CCA merits further investigation. IMPACT STATEMENT: The complex relationship between maternal age and embryo aneuploidy, which approximates a U-shape, may inform optimal timing of elective oocyte freezing and oocyte donation

OBJECTIVE: The use of a GnRH trigger in COH cycles has increased due to an improved safety profile but not all patients have adequate response1.We sought to investigate the utility of using serum GN levels to predict response to GnRH trigger. MATERIALS AND METHODS: We performed a retrospective cohort study of all GnRH-antagonist COH cycles at an urban university affiliated fertility center from 2017-2020. Cycles that utilized GnRH-agonist (GnRH-a) alone or in combination with human chorionic GN (hCG) for trigger were included. Patient and cycle characteristics were collected from the electronic medical record, including day 2 baseline follicle stimulating hormone (B-FSH) and earliest in-cycle luteinizing hormone (EIC LH). An optimal response to GnRH-a trigger was defined as a LH R40 mIU/mL on the morning after trigger. Descriptive statistics (median +/- range for continuous variables; frequencies and percentages for categorical variables) were calculated by GnRH-a response. Statistical analyses were performed on SAS (v9.4) and included the chi-square test, Fisher's exact test, or Mann Whitney U test, as appropriate with a p<0.05 considered significant.
RESULT(S): A total of 3,865 COH antagonist cycles were included. Ninetyone percent of patients had an optimal response to GnRH-a trigger. Optimal responders had higher B-FSH levels than those with poor response (6.52 mIU/ml vs 4.36 mIU/ml, p<0.001). Similarly, the EIC LH was higher for optimal responders (4.66 mIU/ml vs 2.16 mIU/ml, p<0.001). Optimal response had a positive association with older age (p<0.00001), lower BMI (p<0.0001), less days of stimulation (p<0.001), lower starting serum estradiol (p<0.0007), and lower total gonadotropin dose (p<0.001). Optimal response was also associated with B-FSH >5 mIU/ml (p<0.0001), EIC LH >1 (p<0.0001), and Clomiphene citrate use (p< 0.009). Asian race was associated with poor response (p<0.006). There was no difference in oocyte maturity rate (p=0.6) or fertilization rate (p=0.5) for optimal or poor response. Cutoffs for B-FSH (>5 mIU/mL) and EIC LH ( >1 mIU/mL) were chosen to be reasonable clinical cutoffs to create a tool or aid to predict patient response to GnRH-a trigger. The incidence of patients with B-FSH >5 IU/ ml who had a poor response was 4.9% compared to 16.0% in patients with B-FSH <5 (p<0.0001). Twenty-four percent of patients with an EIC LH <1 had a poor response, compared to 4% of patients with EIC LH >1 (p<0.0001). The combination of B-FSH >5 IU/ml and EIC LH >1 IU/ml had a 71% sensitivity and 96% PPV in predicting an optimal response. When individually compared to a B-FSH >5 mIU/ml, an EIC LH>1 mIU/ ml had a higher sensitivity (91% vs 76%) and higher PPV (96% vs 95%) in predicting optimal response.
CONCLUSION(S): A B-FSH>5 and EIC LH>1 may be an appropriate threshold and helpful guide for physicians when determining trigger medicine for GnRH-antagonist COH cycles. Further studies are needed to understand predictors of poor response above these thresholds. IMPACT STATEMENT: In an era of personalized medicine, cycle and patient characteristics, such as GN levels, may improve cycle outcomes and provide further individualized care

OBJECTIVE: LBR models used in patient counseling often fail to account for oocyte cryopreservation and donation. Our aim was to evaluate the effect of oocyte cryopreservation and donation on predicted LBRs in STEETs. MATERIALS AND METHODS: All STEETs from 2014-2019 at a large academic center were identified. STEETs were excluded if embryos were CP before the blastocyst stage, embryos underwent >1 biopsy or >1 cryopreservation, blastocyst day or Gardner grade was unknown, age at oocyte cryopreservation or birth outcome was unknown, or reciprocal IVF or gestational carrier was used. The primary outcome was LBR. Statistical analyses included stepwise multiple logistic regressions. Significant parameters were chosen using the Akaike Information Criterion (AIC).
RESULT(S): 3421 STEET cycles were included; the female gametes used for embryo creation included 2951 fresh autologous oocytes (86.3%), 193 CP autologous oocytes (5.6%), 258 fresh donor oocytes (7.5%) and 19 CP donor oocytes (0.6%). See table for actual LBRs for select blastocyst days and Gardner grades. The following AIC regression equation was obtained: ln(OR [live birth / no live birth]) = +0.017 x age at CP (years) -0.027 x age at transfer (years) -0.36 x blastocyst day (coded as day 5 = 0, day 6 = 1, day 7 = 2) -0.39 x blastocyst stage (coded as: not filling zona = 0, filling zona or expanding = 1, hatched = 2) -0.44 x inner cell mass (ICM) grade (coded as A = 0, B = 1, C = 2) -0.62 x trophectoderm (TE) grade (coded as a = 0, beta = 1, c = 2) +2.45. Inclusion of a random variable for neither CP oocytes nor donor oocytes resulted in an e-fold improvement of the likelihood (did not meet the AIC criterion).
CONCLUSION(S): As previously established, blastocyst day and Gardner grade are predictors of LBR in STEET cycles. When adjusting for age at cryopreservation, age at transfer, blastocyst day and stage, and ICM and TE grades, cycles with CP or donor oocytes had no impact on the fit of the regression. Thus, embryos of equal blastocyst day and Gardner grade have similar LBRs when comparing fresh autologous, CP and donor oocytes. IMPACT STATEMENT: To the best of our knowledge, this is the first model that predicts LBRs in STEETs while accounting for whether oocytes were cryopreserved or donated. Our model suggests STEET LBRs are similar regardless of whether the female gamete is a fresh autologous oocyte, a cryopreserved oocyte or a donor oocyte

OBJECTIVE: There has been significant uncertainty surrounding the COVID-19 pandemic and its effect on human reproduction which resulted in a temporary suspension of ART treatments in early stages of the pandemic. The ACE2 receptor used by the virus to infect pulmonary cells is also found in reproductive organs and has fueled speculation as to whether the disease can be sexually transmitted and whether it can cause infertility. Non-viral issues (e.g., pandemic related psychological stress, alternate methods of communication and interaction, and new clinic procedures) may also worsen outcomes. We sought to determine whether clinical outcomes following the frozen embryo transfer (FET) of a euploid embryo were different during the COVID-19 pandemic in 2020 when compared to prior to the pandemic in 2019. MATERIALS AND METHODS: Patients who tested negative for COVID-19 and underwent FET of a single euploid embryo at NYU Fertility Center in NYC over January 2020 through September 2020 were separated by treatment month and compared with patients from the corresponding month in 2019. Patient's age at cycle start and age at freeze were compared using Student's T-Test. Potential cycle outcomes included intrauterine pregnancy (IUG), biochemical pregnancy (Biochem), and no pregnancy, and outcomes were compared between the two years using contingency Chi Square.
RESULT(S): 1,044 patients were compared over the corresponding months. 558 transfers from 2019 and 486 patients from 2020, with no patients in April of 2020. There were no differences noted in patient's age at cycle start, or age at cryopreservation, between any of the months across the two years. Analysis of outcomes following FET further revealed no statistically significant differences between any of the months over the two years, X² = 14.64, p > 0.05. Post hoc analyses comparing the combined months of March, April and May, or the combined 9-month periods, were also not statistically significant (X² = 0.042, p > 0.05; X² = 1.68, p > 0.05; respectively).
CONCLUSION(S): In patients who tested negative for COVID-19, there were no differences in treatment outcomes following FET's when comparing patients treated during the COVID pandemic with those who were treated prior to the pandemic. IMPACT STATEMENT: Providers and patients can be reassured that with proper testing and sanitizing techniques FET outcomes remained unaffected by the pandemic. (Table Presented)

OBJECTIVE: OC is widely used for fertility preservation. Many models predict the live birth (LB) rate of OC, but real-world data is lacking. We reviewed our LBs from OC to develop an OC counseling tool based on real outcomes. MATERIALS AND METHODS: We reviewed patients (pts) who thawed autologous oocytes (AOs) at our academic fertility center from 2004-2020. We included pts who: 1) had a LB or ongoing pregnancy (OP) >12 weeks at last contact, or 2) consumed all AOs and resultant embryos. Pts were excluded if they transferred AOs or resultant embryos to another center or if OC was performed for a medical reason, as research, due to no sperm or a natural disaster, combined with embryos or for use with a gestational carrier. We calculated OP + LB rate (LBR) based on number of AOs and metaphase II oocytes (M2s) thawed. Data were stratified by age (<38y vs. >=38y). For pts who underwent OC at <38y and >=38y, a weighted age was calculated (for each OC cycle, #AOs thawed was multiplied by age at OC; the sum of these numbers was then divided by total #AOs thawed). Statistics included Fisher's exact test (p<0.05 significant).
RESULT(S): We included 462 pts (median age at 1st OC 38.5y). Weighted ages were used for 21 pts (5%). Our pts underwent 650 OCs (90% our center, 9% elsewhere, 1% both), 512 thaws and 385 embryo transfers. OC involved vitrification for 72%, slow freezing for 4% and both for 24% of pts. A total of 7050 AOs and 6178 M2s were thawed. 38% of pts (n=176) have >=1 LB or OP from AO thaw. See table for outcomes. Pts who thawed 0-10 AOs had a lower LBR than pts who thawed 11-20, 21-30, or >30 AOs (p<=0.03). Pts who thawed 0-10 M2s had a lower LBR than pts who thawed 11-20 or 21- 30 M2s (p<=0.02). LBR was not significantly different between pts who thawed 11-20, 21-30, or >30 AOs or M2s.
CONCLUSION(S): Pts who thawed 0-10 AOs had a lower LBR (27%) than pts who thawed >10 AOs (LBR >= 43%), and pts who thawed 0-10 M2s had a lower LBR (30%) than pts who thawed > 10 M2s (LBR >= 42%), but LBR was not different with >10 thawed AOs. IMPACT STATEMENT: Our real-world OC outcomes are not consistent with LBRs in published models. These results provide more realistic expectations about OC outcomes and may help pts decide how many AOs to freeze

OBJECTIVE: The causes of spontaneous abortion (SAB) following euploid embryo transfer (EET) remain poorly understood. Here we describe the frequency of aneuploidy in products of conception (POC) and endometrial dysfunction in women who miscarried after EET. MATERIALS AND METHODS: Between 1/2018 - 8/2020, 255 dilation and curettage (D&C) procedures were performed at a large academic IVF center for SAB following EET. Retrospective chart review was performed to identify D&Cs followed with genetic analysis of POCs. Information collected from the medical record included assessments of endometrial dysfunction based on Endometrial Receptivity Assay (ERA), CD138 for chronic endometritis (CE), and/or BCL6 for endometriosis. Exclusion criteria included an abnormal endometrial cavity on imaging. Demographic factors, clinical parameters and IVF/FET outcomes were reviewed. Additionally, retrospective chart review was performed of all ERAs completed at our institution from 12/2018-9/2020.
RESULT(S): Genetic analysis of 67 POCs after D&C following EET were identified. Fifty-nine POCs (88%) were euploid by SNP microarray. Eight (12%) of the POCs displayed genetic abnormalities: 3 trisomies, 2 partial duplications, 2 mosaic trisomies and 1 triploidy of paternal origin. Of the 51 patients who had endometrial biopsy (EMB), 28 (55%) had normal results. Twenty-three (45%) had abnormal results: 18 with CE, 2 with elevated BCL6 and 3 with pre-receptive ERA. The proportion of SABs unexplained by endometrial dysfunction or genetically abnormal POCs was 38% (26). A total of 44 patients underwent repeat EET. Eleven live births (LB) occurred, six after correction of endometrial dysfunction. Eight patients currently have ongoing pregnancy, 2 after treatment for CE. Three patients experienced repeat SAB, 1 following correction of pre-receptive ERA, and 1 after CE treatment. Four patients had implantation failure, 3 following normal EMB and 1 after treatment of CE. Two patients conceived spontaneously and delivered, 1 after treatment for CE, the other after a normal EMB. Upon review of all ERAs, 82 single EET following ERA guidance were identified. Fifty-nine percent (n=48) resulted in ongoing pregnancy or LB. There was no significant difference in ERA result or post ERA transfer outcome based on ethnicity (p= 0.7, p=0.4) or BMI (p= 0.8, 0.9), respectively. There was also no difference in post ERA transfer outcome based on blastocyst age (day 5 or 6) (p=0.5)
CONCLUSION(S): Aneuploidy and/or endometrial factor can contribute to SAB following EET. Aneuploid POCs could have arisen de novo and/or have passed undetected by trophectoderm biopsy and NGS. Our results are consistent with the 1-2% false negative rate reported for PGT-A. Further studies are needed to characterize the sub-chromosomal genetic variations associated with euploid embryo SABs, as well as endometrial function testing. IMPACT STATEMENT: The etiology behind failed EET may involve more discrete entities such as sub-chromosomal abnormalities in addition to aneuploidy and endometrial dysfunction