Faculty Bibliography

Study question: What is the overall value of mitochondrial DNA (mtDNA) assessment in a clinical setting, and how can it help improve IVF outcomes? Summary answer: Retrospective analysis demonstrates that mtDNA quantification has a high negative predictive value and can be used as an independent biomarker in determining embryo viability. What is known already: Chromosomal abnormalities are a major cause of implantation failure. Recently our group has validated the relationship between mtDNA quantification and IVF outcomes in a non-selection study. In particular we have been able to establish a threshold above which the probability of implantation is greatly reduced. While further randomized testing is being conducted, we took this opportunity to retrospectively analyze the overall value of the established mtDNA threshold in previously completed clinical cases. Study design, size, duration: Retrospective study in which mtDNA was assessed in a total of 572 euploid blastocysts obtained from 328 couples (average maternal age 34.95 +/- 0.27 years) undergoing preimplantation genetic screening (PGS). Outcome data was collected from 6 different IVF centers for routine follow up. Implantation outcomes were then utilized to determine the validity of the established threshold. Participants/materials, setting, methods: DNA from blastocyst biopsy samples was amplified (Sureplex, Illumina, USA) and then subjected to aneuploidy analysis using next generation sequencing (NGS, Veriseq protocol, Illumina, USA). Only those embryos classified as chromosomally normal had their mtDNA levels assessed using MitoGrade (Reprogenetics). mtDNA levels were then compared to the pregnancy outcomes to confirm implantation predictions. All embryos were single embryo transfers (SET). Main results and the role of chance: Nearly 14% (80/572) of all blastocysts analyzed contained mtDNA levels above the established threshold and were predicted to have lower chances of implantation. To date, 246 euploid embryos were replaced in SET with a pregnancy rate of 62.3% (153/246). Retrospective assessment of mtDNA levels revealed 216 embryos to contain normal mtDNA levels. Therefore the pregnancy rate post mtDNA quantification was 71% (153/216) in the normal level mtDNA group. Furthermore of the 30 embryos with elevated mtDNA levels, only one led to a successful pregnancy. Therefore the negative predictive value of mtDNA quantification was 96.7% (29/30). This highly significant (p < 0.001) difference between implanting and non-implanting embryos validates the clinical applicability of mtDNA quantification. More importantly, the mtDNA threshold retained its validity across six different IVF centers and was unaffected by maternal age. Limitations, reasons for caution: The study was retrospective and the number of cycles was not enough to show a difference between no selection (62% pregnancy rate) and selection against MitoGrade elevated embryos (71% pregnancy rate). It was however large enough to show a significant difference in implantation rates between elevated and normal MitoGrade embryos. Wider implications of the findings: This study demonstrates the validity of MitoGrade as an independent variable in predicting embryonic implantation potential of euploid embryos. Further research involving the biological significance of mtDNA levels and implantation rates would be invaluable

Study question: Does undetected aneuploidy or mosaicism contribute to pregnancy loss after transfer of euploid embryos by array comparative genomic hybridization (aCGH)? Summary answer: NGS detects more cases of mosaicism and triploidy than aCGH, and mosaicism rates are significantly higher among pregnancies resulting in miscarriage than live birth. What is known already: Array CGH is widely used for pre-implantation genetic screening (PGS). NGS is capable of detecting more cases of mosaicism and triploidy (69XXY), which may assist in reducing the incidence of spontaneous abortion and increase ongoing pregnancy rates. Study design, size, duration: Retrospective study of 183 patients undergoing PGS by aCGH between 8/2012 and 5/2015 at New York University Fertility Center and Oregon Reproductive Medicine. Participants/materials, setting, methods: Saved amplified DNA samples from the 183 blastocyst trophectoderm (TE) biopsies previously diagnosed as euploid by aCGH were re-analyzed using the miSeq NGS platform (Illumina, USA) and VeriSeq NGS technology (Illumina, USA). 44 embryos resulting in a biochemical pregnancy, 62 resulting in miscarriage, and 77 resulting in live birth were available for re-analysis. Main results and the role of chance: 25% (11/44) of embryos resulting in biochemical pregnancies were mosaic, and one embryo was found to be triploid (69, XXY) by NGS. 33.9% (21/62) and 3.2% (2/62) of embryos resulting in miscarriage were mosaic and triploid by NGS, respectively. In contrast, the mosaicism rate among embryos resulting in live birth was only 13% (10/77), which was significantly lower than the rate of mosaicism among miscarriages (p = 0.0062, RR 1.78 with 95% CI 1.23-2.5). Limitations, reasons for caution: This study was limited by its retrospective design. Up to 10% of DNA samples that were undergoing re-analysis were excluded due to degraded DNA, although the frequency was similar in all three groups. Wider implications of the findings: Undetected mosaicism may increase the risk of first trimester pregnancy loss. NGS is more sensitive at picking up mosaicism and triploidy than aCGH. Mosaic embryos can be considered for transfer after genetic counseling and informed consent, but they have a higher miscarriage rate as well as unknown post-natal genetic effects

Study question: Can quantification of mitochondrial DNA (mtDNA) in trophectoderm biopsy specimens provide information concerning embryo viability, potentially enhancing embryo selection and improving IVF treatment outcomes? Summary answer: This study demonstrates that mtDNA levels are highly predictive of embryo potential. Euploid embryos of good morphology, but with high mtDNA levels fail to implant. What is known already: Better methods of embryo selection are highly desirable in order to improve IVF treatment efficiency. Even the transfer of chromosomally normal embryos of high morphological grade cannot guarantee that a pregnancy will follow. Recently, the quantity of mtDNA in embryonic cells has been proposed as a new biomarker of viability-higher levels of mtDNA associated with reduced implantation potential. However, to date no prospective blinded studies have been undertaken to confirm this possibility. The current investigation involves the first evaluation of the predictive power of mtDNA quantification in a prospective, blinded, non-selection setting. Study design, size, duration: mtDNA was quantified in 280 blastocysts, previously biopsied and shown to be chromosomally normal using preimplantation genetic screening (PGS). These were generated by 143 couples (average female age 37.2 years). All patients underwent IVF in a single clinic. The study took place in a blinded, non-selection manner-i.e., mtDNA quantity was not known at the time of single embryo transfer. The fate of the embryos transferred was subsequently compared to the mtDNA levels measured. Participants/materials, setting, methods: Embryos were biopsied at the blastocyst stage. The trophectoderm samples obtained were subjected to whole genome amplification followed by comprehensive chromosome analysis using a next generation sequencing strategy (NGS). The same biopsy specimens were also tested using quantitative PCR, allowing highly accurate mtDNA quantification. After embryo transfer, the code used for blinding was broken and analysis undertaken to reveal whether the amount of mtDNA had any association with blastocyst implantation. Main results and the role of chance: mtDNA analysis of the 280 blastocysts revealed that 15 (5.4%) contained unusually high levels of mtDNA. At the time of writing, outcome data was available for 111 of the blastocysts, transferred after PGS results had been obtained but before mtDNA levels were known. All transfers involved a single chromosomally normal blastocyst of good morphology. Of these, 78 (70%) led to ongoing pregnancies, and all (100%) had mtDNA levels considered to be normal/low. The remaining 33 (30%) blastocysts failed to implant. Among these non-viable embryos there were 7 (21%) with unusually high levels of mtDNA. This meant that the ongoing pregnancy rate for morphologically good, euploid blastocysts, with normal/low levels of mtDNA was 76% (78/103). In contrast, the ongoing pregnancy rate for the same type of embryos, but with elevated mtDNA levels, was 0/7 (0%). This difference was highly statistically significant (P < 0.0001). Limitations, reasons for caution: This study provides strong evidence that mtDNA quantification can serve as a valuable tool to improve the identification of viable blastocysts. However, to determine the true extent of any clinical benefits a randomised clinical trial will be necessary. Research is needed to improve understanding of the biology of mtDNA expansion. Wider implications of the findings: This is the first study to evaluate the clinical impact of increased mtDNA in a prospective blinded manner. Results confirm that embryos with elevated mtDNA rarely implant, supporting its use as a viability biomarker. 76% of euploid embryos with normal/low mtDNA implanted vs. 70% for the cohort as a whole

PURPOSE: The aim of this study was to report the results of IVF with trophectoderm biopsy and preimplantation genetic screening (PGS) following delayed intracytoplasmic sperm injection (ICSI). METHODS: Patients undergoing IVF with PGS and delayed ICSI were included in the study. Indications for delayed ICSI included absent or poor fertilization via standard insemination or more than 50 % immature oocytes, noted post-cumulus stripping for standard ICSI procedure. Delayed ICSI was performed the day after retrieval due to absent or poor fertilization. The immature oocytes were kept in extended culture, and if demonstrated maturity, ICSI was performed. Primary outcome included fertilization rate and blastocyst stage formation, defined by the number of blastocysts for biopsy. Secondary outcome included aneuploidy rate and pregnancy outcomes following single thawed euploid embryo transfers (STEET). RESULTS: Sixteen patients with delayed ICSI were included in the study. Twelve were due to poor fertilization and four secondary to immature oocytes. A total of 219 oocytes were retrieved; ten were frozen upon patient request, 168 had standard insemination, and 13 had routine ICSI on the day of retrieval. A total of 129 oocytes underwent delayed ICSI. Sixty-three (49 %) fertilized, 19 (14.7 %) reached blastocysts for biopsy; fivw of which were chromosomally normal (26.3 %). Three patients underwent STEET of a delayed ICSI embryo; all three resulted in live births, including one embryo biopsied on day 8 of development. CONCLUSION: Fertilization failure or an excessive proportion of immature oocytes in an IVF cycle, necessitating delayed ICSI, showed equivalent fertilization and blast formation rates. With the implementation of trophectoderm biopsy and PGS, these embryos can lead to healthy live born babies.

OBJECTIVE: To assess if SART donor oocyte pregnancy rates are consistent with high percentage of aneuploidy and mosaicism in donor oocyte cycles using Preimplantation Genetic Screening (PGS) with Next Generation Sequencing (NGS). DESIGN: Retrospective study and mathematical analysis. MATERIALS AND METHODS: All trophectoderm biopsy specimens from donor oocyte cycles received by a genetics laboratory were queried for NGS results. The rate of euploidy, aneuploidy, and mosaicism was calculated. A separate analysis was done to find the ongoing pregnancy rate (OPR) from patients' first donor single thawed euploid embryo transfers (STEET) from NGS at a single university fertility center from 2/2015 to 3/2016. Lastly, the 2014 SART National Summary data on non-tested donor embryo transfers and their live birth rates was used to calculate the expected live births from untested donor embryo transfers based on euploidy and mosaicism rates from NGS donor data, applying the the OPR from donor STEET data. The calculated expected live birth rate was then compared to the actual live birth rate of untested donor embryo transfer cycles. RESULTS: 268 donor cycles from 38 IVF centers showed a euploid rate of 48.7%+/-23.8, aneuploid rate of 22.9%+/-20.4 and mosaic rate of 28.4%+/- 21.9 (Table 1). Thirty two patients were included in the donor NGS STEET cycle analysis. The average age at transfer was 42.7 +/-3.9 years with an average donor age of 26.5 +/-2.9 years, for an OPR of 62.2% (n=20/32). The 2014 SART National Summary shows 6070 non-PGS donor egg transfers, averaging 1.7 embryos per transfer, for an estimate of 10319 embryos transferred in total. The reported live birth rate of 53.3% (36.9% singletons, 16.2% twins and 0.2% triplets) gives a cumulative 4242 babies delivered (41.1% per embryo). When the euploid rate of 48.7% is applied, 5025 embryos transferred are estimated to have been euploid. The 62.2% OPR of donor STEETs applied to this gives an expected 3126 singleton deliveries (95% CI 2927-3246). The mosaic rate of 28.4% plus data showing an overall approximate 20%1,2 OPR of mosaic embryos adds 586 (95% CI 421- 524) singleton births. The total expected babies born is 3712 (95% CI 3505-3877) with a live birth rate of 35.9%(95% CI 34.0-37.6) from donor egg transfers. CONCLUSIONS: High rates of aneuploidy and mosaicism in donor oocyte cycles are consistent with the pregnancy rates of untested donor embryo transfers. The use of PGS with NGS can prevent transfer of aneuploid embryos in donor cycles

OBJECTIVE: The recent recognition of increased genetic admixture and advances in technology advances have caused a shift to broader acceptance of pan-ethnic expanded carrier screening (ECS) panels. However, ECS has not been widely accepted across Europe. The resistance is partially due to a belief that the US is genetically diverse from years of admixture while the European population is genetically homogeneous, making pan-ethnic screening unnecessary. We sought to determine if this difference in admixture was present. DESIGN: Retrospective. MATERIALS AND METHODS: Genomic data from an ECS panel was analyzed for 7544 participants, 6297 from US clinics and 1247 from European clinics. For all participants, genetic ancestral origin was predicted by a statistical model based on 672 SNPs validated using samples from the 1000 Genomes Project, and admixture proportions were calculated for 6 ancestral populations (European, Oceania Native, Native American, East Asian, Sub-Saharan African, and South Asian). A comparison of the resulting predicted admixture proportions was made between patients in Europe against patients in the US. Consent to use de-identified genomic data was obtained for all participants. RESULTS: For comparison, the European and American patients were further subdivided into four groups, based upon which ethnicity they self-reported on test requisition forms: European, Mediterranean, Latin American, or African. Across all four comparison groups, our results showed similar average admixture proportions. For example, European patients who self-identified as Mediterranean were genetically predicted to be an average of 86% European, and 5% Sub-Saharan African. American patients who self-identified as Mediterranean were genetically predicted to be an average of 81% European and also 5% Sub-Saharan African. European patients who self-identified as African were predicted to be an average of 78% African and 16% European, which was strikingly similar to the predictions for American patients who self-identified as African (78% African and 12% European). This similarity in admixture proportions was also seen between European and American patients who self-identified as European and Latin American. CONCLUSIONS: These results demonstrate that on a genetic level, both European and American patients demonstrate equal degrees of admixture. The European patient population may not be as genetically homogenous as previously believed. Pan-ethnic carrier screening panels could prove to be effective at identifying carriers in a European population, who would otherwise be missed

OBJECTIVE: hr-NGS detects mosaicism in trophectoderm (TE) biopsies when 10%-90% of cells are abnormal. Mosaic TE biopsies may come from embryos that result from euploid, aneuploid or mosaic pregnancies. Mosaic biopsies have higher risk of miscarrying, lower implantation potential but some may go to term. Mosaic embryos should be classified in between euploid and aneuploid in their priority for transfer. This study aims to determine the rate of euploidy, mosaicism and other abnormalities according to maternal age in a large cohort of embryos for purpose of patient counseling. DESIGN: Analysis of PGS procedures involving TE biopsy and hr-NGS performed by two large genetic reference laboratories serving over 250 fertility clinics. MATERIALS AND METHODS: Two laboratories analyzed 4277 and 2303 PGS procedures using TE biopsy by whole genome amplification method (SurePlex), assay for hr-NGS (VeriSeq PGS assay, Illumina), sequencer (MiSeq, Illumina) and software (BlueFuse Multi analysis, Illumina). EMRs (eIVF, practiceHwy and lab's own) used to query the data. Embryos with 1 or 2 aneuploid chromosomes or 1 aneuploid chromosome and 1 mosaic chromosome were called aneuploid. Embryos with 3 aneuploid chromosomes or two aneuploid and 1 or more mosaic chromosomes were complex abnormal. Embryos with a mixture of normal cells and abnormal were called mosaic. RESULTS: Rate of euploidy, mosaicism and complete abnormalities (aneuploidy, complex or polyploidy) was 41% and 47%, 20% and 14%, and 39% and 39%, respectively. The combined results for both datasets shown (Table1). CONCLUSIONS: Being a postmeiotic abnormality, mosaicism does not increase with advancing maternal age (the apparent decrease in the table is due to some mosaics that are also aneuploid being classified as aneuploid). Differences between the two labs were attributed to difference in scoring criteria where the threshold to call an embryo normal or euploid was set (10-90% vs 17-83% or 1/10-9/10 cells vs 1/6-5/6). Regardless, the amount of mosaic embryos, which have lower potential than euploid embryos, is considerable, 11-23% depending on maternal age. (Table Presented)

OBJECTIVE: Recent reports have suggested that the quantity of mtDNA in embryonic cells may serve as an indicator of embryo viability, higher levels being associated with reduced implantation potential. The current investigation represents the first evaluation of the predictive potential of mtDNA quantification in a prospective, blinded, non-selection setting. DESIGN: Prospective blinded clinical study. MATERIALS AND METHODS: mtDNA was assessed in trophectoderm (TE) samples biopsied from 337 blastocysts that had been shown to be chromosomally normal. These were generated by 195 couples (average female age 36.7 years). All patients underwent IVF in a single clinic. Cytogenetic assessment occurred via next generation sequencing, whereas mtDNA quantification utilized quantitative PCR. The study took place in a blinded, nonselection manner - i.e. mtDNA quantity was not known at the time of single embryo transfer. The fate of the embryos transferred was subsequently compared to the mtDNA levels measured. RESULTS: mtDNA assessment showed that 25/336 (7.4%) of embryos contained elevated mtDNA levels. At the time of writing, 159 of the blastocysts have been transferred. All transfers involved a single chromosomally normal blastocyst of good morphology. mtDNA amounts were not known at the time of transfer. 109 (69%) of these led to ongoing pregnancies, and all (100%) had mtDNA levels in the normal range. The remaining 50 (31%) blastocysts failed to implant. 8 (16%) of these non-viable embryos were found to have elevated quantities of mtDNA. This meant that the ongoing pregnancy rate for morphologically good, euploid blastocysts, with elevated mtDNA levels, was 0/8 (0%), a highly significant difference compared to the 72% pregnancy rate observed for similar embryos with normal levels of mtDNA (P<0.0001). CONCLUSIONS: This is the first study to evaluate the clinical impact of increased mtDNA in a prospective blinded manner. Results confirm that embryos with elevated mtDNA rarely if ever implant, providing support for its use as a viability biomarker. 72% of euploid embryos with normal quantities of mtDNA implanted vs. 69% for the cohort as a whole. No embryos with elevated mtDNA levels implanted