Intention to treat (ITT) analysis identifying patient groups who benefit from preimplantation genetic screening (PGS)

P-454 Wednesday, November 1, 2017 ROUTINE TRANSFER OF MULTIPLE EMBRYOS IS NO LONGER NECESSARY TO ACHIEVE HIGH RATES OF HEALTHY SINGLETON BIRTHS PER TRANSFER IN WOMEN UP TO 42 YEARS OLD. M. C. Schiewe,a F. Garner,b R. E. Anderson,c N. L. Nugent,a J. B. Whitney,a B. S. Shapiro.b a ART Lab, Ovation Fertility, Newport Beach, CA; bOvation Fertility, Fertility Center of Las Vegas, Las Vegas, NV; cOvation Fertility, Southern California Center for Reproductive Medicine (SCCRM), Newport Beach, CA. OBJECTIVE: Based on current literature, our center implemented a 100% freeze-all strategy, with almost all cycles (87.5%) using pre-implantation genetic screening (PGS) and 80-90% of our transfers being single blastocysts. Here, we evaluate the effectiveness of a clinical strategy of routine single euploid embryo transfer (SEET) to achieve acceptable live birth rates and to try to maximize the rate of healthy singleton live births. DESIGN: The reults reported below are those reported by the US Centers for Disease Control and Prevention (CDC) for 2014. The outcome measures were live birth rate per transfer and the rate of term singleton births of normal birthweight (‘‘Healthy singletons’’). MATERIALS AND METHODS: Routine laboratory procedures involved ICSI, group culture in single-step media with protein supplement, tri-gas incubation, blastocyst vitrification following trophectoderm biopsy/PGS. Embryo thaw and transvaginal ultrasound-guided transfer followed within 1-2 months. RESULTS: See Table below summarizing transfers of frozen-thawed embryos. In 2014, we performed no fresh embryo transfers. An average of 1.11.2 thawed embryos were transferred, so that 80-90% of our transfers were single embryos. Despite the routine transfer of single embryos, the live birth rates per transfer were at least 55% in each age group, and the rates of healthy singleton live birth per transfer were at least 48%. These rates were consistently 1.5x the national averages, validating acceptability. CONCLUSIONS: The strategy of routine single euploid blastocyst transfer following a freeze-all cycle was successful at producing high rates of live birth and healthy singleton term births, confirming no need for the transfer of multiple embryos in these age groups. These findings support the recently revised ASRM/SART guidance to transfer a maximum of one embryo following PGS. Based on these encouraging results in 2014, SCCRM’s use of PGS was increased to 99% in 2015, and the mean number of transferred embryos was further reduced to 1.0. Overall, our goal of ‘‘one embryo-one baby’’ has effectively been realized.

Number of transfers Mean number transferred Live birth per transfer Healthy singleton term birth/ET

P-456 Wednesday, November 1, 2017 INTENTION TO TREAT (ITT) ANALYSIS IDENTIFYING PATIENT GROUPS WHO BENEFIT FROM PREIMPLANTATION GENETIC SCREENING (PGS). L. A. Murphy,a E. A. Seidler,b N. Resetkova,c A. Penzias,d K. L. Thornton,e D. Sakkas.e aBoston IVF/BIDMC, Harvard Medical School, Boston, MA; bReproductive Endocrinology & Infertility, BIDMC, Harvard Medical School, Boston, MA; cBoston IVF / Beth Israel Deaconess Medical Center, Boston, MA;

Data Set

Age <35

Age 35-37

Age 38-40

Age 41-42

National Avg OF Clinic National Avg OF Clinic National Avg OF Clinic National Avg OF Clinic

24,740 78 1.6 1.1 46.6% 73.1% 30.7% 56.4%

12,626 71 1.5 1.2 44.0% 66.2% 30.6% 56.3%

9,343 60 1.6 1.2 38.3% 55.0% 27.1% 48.3%

3,460 23 1.7 1.1 32.1% 78.3% 23.3% 69.6%

P-455 Wednesday, November 1, 2017 PREIMPLANTATION GENETIC SCREENING (PGS) GIVES SIMILAR PROCREATIVE ADVANTAGE IN DONOR CYCLE (DC) AND NON-DONOR CYCLE (NDC). S. W. Hong, J. Seo, J. J. Berger. CHA Fertility Center, Los Angeles, CA. OBJECTIVE: The advantages of PGS, higher pregnancy and lower miscarriage rates, have been confirmed in NDC. However, there is limited data available regarding similar effects of PGS in DC. The objective of this study was to demonstrate if PGS would have similar improvements in pregnancy rates and miscarriage rates when used with DC IVF. DESIGN: Retrospective study. MATERIALS AND METHODS: From Sep-2014 through Feb-2017 a total of 456 frozen embryo transfers (FETs) were performed at our facility. The numbers of NDC with and without PGS, and DC with and without PGS, were 165, 127, 57, and 107 respectively. All mature oocytes were inseminated with ICSI. Normally fertilized oocytes were cultured until day 5/6 post retrieval.For PGS, laser was used for biopsies, samples were loaded into PCR tubes. Vitrification technique was previously described. Embryos were equilibrated in 37
C cryoprotectant solution having 1.5 mol/L ethylene glycol (EG) for 5min followed

FERTILITY & STERILITYÒ

by vitrification solution having 5.5 mol/L EG + 1.0 mol/L sucrose for 20sec. The embryo was then placed on the gold electron microscope grid (EMG) and plunged into liquid nitrogen. For warming, the EMG was transferred to dishes sequentially containing 1 ml of warming solutions with varying amounts of sucrose for 2.5 min intervals at 37
C. Transfer embryos were warmed at least 3hrs prior. Non-biopsied embryos had assisted hatching. RESULTS: A total of 456 FETs were performed. Groups 1 (n¼165) and 2 (n¼127) represented NDC with and without PGS respectively. Groups 3 (n¼57) and 4 (n¼107) represented DC with and without PGS respectively. A total of 923 blastocysts were warmed and 886 embryos (96.0%) survived. A total of 782 blastocysts were transferred, a mean of 1.70.5 embryos were transferred per FET. Of the 456 FETs, 322 resulted in pregnancy (70.6%), of those 43 pregnancies ended in miscarriage (13.4%). The pregnancy rate of each group (Group 1-4) was 76.4% (126/165), 66.9% (85/127), 77.2% (44/57), and 62.6% (67/107) respectively. For each group, the miscarriage rate was 7.9% (10/126), 14.1% (16/85), 6.8% (3/44), and 20.9% (12/67) as of March 2017. Miscarriage rates without PGS (Group 2 and Group 4) were significantly higher (P>0.05) than with PGS (Group 1 and Group 3) in both NDC and DC. In the NDC, the miscarriage rate without PGS (Group 2) was about 2 times higher than with PGS (Group 1). In the DC, the miscarriage rate without PGS (Group 4) was about 3 times higher than with PGS (Group 3). Pregnancy rates in the PGS groups were higher than without PGS groups, for both DC and NDC, but this was not statistically significant. Thus far, the number of deliveries per group was 42, 34, 29, and 37 respectively, while the number of ongoing per group was 75, 35, 18 and 16 as of March 2017. CONCLUSIONS: In this study, PGS led to a significantly lower miscarriage rate for both donor and non-donor FETs. These findings are comparable to other studies in non-donor cycles and supports the use of PGS even when using young, donor eggs. If these improved pregnancy rates lead to higher live birth rates have yet to be demonstrated, but thus far the results are promising.

d

Boston IVF / Harvard Medical School, Waltham, MA; eBoston IVF, Waltham, MA.

OBJECTIVE: To evaluate the effectiveness of PGS in an intention to treat setting across multiple age groups compared to IVF without PGS. DESIGN: Retrospective cohort study. MATERIALS AND METHODS: Electronic medical records were searched for all first cycle patients who indicated an intention to pursue PGS (PGS-ITT) from December 2014 to September 2016, with frozen transfers from these intended cycles assessed through December 2016 (N¼300). The cohort of patients was matched by age and oocyte yield with controls (no intention to use PGS) from the same time period. A second group of patients intending PGS for their 2nd and 3rd cycles (N¼175) were also evaluated and matched with controls. Outcome measures are shown in Table 1. Ongoing pregnancy is defined as pregnancy in the second trimester and through delivery. RESULTS: For the 300 patients intending to utilize PGS on their first IVF cycle attempt, PGS did not make a difference in overall ongoing pregnancy rates compared to matched controls (36.6% PGS-ITT vs. 35.4% Control) (Table 1). Patients aged >38 in the PGS-ITT group achieved significantly better pregnancy rates per embryo transfer than controls (62.5% PGS-ITT vs. 31.3% Control) but similar pregnancy rates per oocyte retrieval (35.2%

e281

Analysis of ongoing pregnancy rates per patient and retrieval for PGS-ITT and matched Controls.

Control vs. PGS-ITT

Number of Retrievals

Number of Fresh Transfers

1st cycle Control <38 165 136 PGS-ITT<38 158 15 Control >38 135 110 142 8 PGS-ITT>38 2nd/3rd Cycle Control<38 74 62 PGS-ITT<38 83 9 119 106 Control>38 124 6 PGS-ITT>38 Same letters a:a, b:b, c:c denotes P<0.01

Number of Frozen Transfers

Number of Cycles with Embryos Biopsied

Number of Cycles with Euploid Embryos

89 137 56 72

0 140 3 118

56 76 50 68

0 64 0 86

PGS-ITT vs. 38.5% control). A similar outcome was seen in those undergoing second and third cycles (Table 1). The difference in pregnancy rate per transfer was most apparent in the >38 age group (44.6% PGS-ITT vs. 16.7% Control). Of significance, those intending PGS were more likely to have a single embryo transfer than controls (mean ¼1.0 PGS-ITT vs. 1.4 Control: P<0.001). CONCLUSIONS: PGS does not improve pregnancy rates when comparing ongoing pregnancy per patient. It is not beneficial in patients attempting their first IVF cycle under 38 years of age. In patients >38, fewer transfers involving fewer embryos are needed to achieve pregnancy in those utilizing PGS compared to age/oocyte matched controls. The greatest benefit of PGS is seen in those >38 years who are undergoing repeat IVF cycles. P-457 Wednesday, November 1, 2017 COPY NUMBER VARIANT CALLING ON A 177 GENE EXPANDED CARRIER SCREENING PANEL REVEALS IMPACT OF HBB DELETIONS. K. A. Beauchamp,a P. Grauman,a G. J. Hogan,a K. R. Haas,a G. M. Gould,a K. K. Wong,a G. A. Lazarin,a E. Evans,a D. Muzzey.b aCounsyl, South San Francisco, CA; bCounsyl Inc., South San Francisco, CA. OBJECTIVE: Expanded carrier screening (ECS) identifies couples whose future children are at increased risk of Mendelian conditions. Historically, ECS has been performed with limited or no copy number variant (CNV) calling, often restricted to a handful of founder deletions. The lack of broad CNV calling may reduce the detection rate of ECS. DESIGN: We performed panel-wide CNV calling on an ECS cohort of 65,732 patients tested between Nov. 2016 and April 2017 to determine its impact on detecting at risk couples. MATERIALS AND METHODS: Here we report CNV statistics for a 177gene ECS panel on a cohort of 65,732 anonymized patient samples. We performed CNV deletion and duplication calling on 161 autosomal genes and 10 genes on chromosome X (calls for six genes, such as SMN1, are treated as special cases and excluded from this analysis). Copy number calling was performed using a Hidden Markov Model on next generation sequencing depth data. RESULTS: Although the genes with the most observed deletions (CLN3, GALC, CTNS) contain known founder mutations, 62% of called CNVs were located outside of the six genes for which we called deletions in our previous 112-gene panel (CLN3, CTNS, GALC, HEXA, MCOLN1, and NEB). Beyond the previously studied six gene subset, we also observe a large number of deletions in HBB (50 observations) and FANCA (58 observations); many of these variants are predicted to be pathogenic and significantly impact the population disease risk burden. Furthermore, HBB deletions are particularly common in Southeast Asians and African Americans. CONCLUSIONS: @Using a 65,732 patient cohort, we retrospectively evaluated the impact of CNVs on a 177 gene ECS panel. Most of the observed CNVs were outside the six genes for which we called founder deletions in previous work, indicating that broader use of CNV calling will improve detection rates in ECS. The presented findings have motivated the inclusion of panel-wide deletion calling in our clinical ECS panel, where novel deletions undergo real-time variant interpretation to assess clinical impact. Supported by: All authors are employees of Counsyl, a commercial molecular diagnostics laboratory.

e282

ASRM Abstracts

Number of Ongoing/Deliveries per Transfer

Number of Ongoing/Deliveries per Retrieval

0 128 2 81

106/225 (47.1%) 68/152 (44.7%) 52/166 (31.3%)a 50/80 (62.5%)a

106/165 (64.2%)b 68/158 (43.1%)b 52/135 (38.5%) 50/142 (35.2%)

0 50 0 42

36/118 (30.5%) 31/85 (36.5%) 26/156 (16.7%)c 33/74 (44.6%)c

36/74 (48.6%) 31/83 (37.3%) 26/119 (21.8%) 33/124 (26.6%)

P-458 Wednesday, November 1, 2017 EXPLORING THE CHROMOSOMAL CONCORDANCE BETWEEN TROPHECTODERM AND INNER CELL MASS REVEALS A 6% ‘BIOLOGICAL FALSE NEGATIVE’ RATE DURING PREIMPLANTATION GENETIC SCREENING. M. Viotti,a A. R. Victor,a A. Brake,a J. Tyndall,a A. Murphy,a L. Lepkowsky,a A. Lal,a D. K. Griffin,b C. Zouves,a F. L. Barnes.a aZouves Fertility Center, Foster City, CA; bSchool of Biosciences, Canterbury, United Kingdom. OBJECTIVE: Preimplantation genetic screening (PGS) at the blastocyst stage is based on the assumption that the placenta-forming trophectoderm (TE), the source of cell biopsy used for downstream examination, is an accurate representative of the inner cell mass (ICM), the group of cells that gives rise to the fetus. Previous studies exploring the rates of concordance between TE and ICM have relied on dated technologies or have suffered from small sample sizes. Here, we use NextGen Sequencing (NGS) to explore this question in a large group of embryos. DESIGN: 50 blastocyst embryos deemed aneuploid by original TE biopsy were tested for the karyotypic profile of the ICM. MATERIALS AND METHODS: Blastocysts from freeze-all cycles were subjected to routine PGS by NGS. Embryos with full aneuploidies (complete losses or gains) were thawed and re-biopsied to isolate cells of the ICM and a second TE sample. PGS by NGS was subsequently performed on each embryo’s ICM biopsy, and in case of discordance also on the second TE sample. Three analysts blindly interpreted all resulting karyotype profiles independently. RESULTS: From a population of 50 blastocysts originally designated as aneuploid, three cases (6%) had euploid ICMs. For these three instances, the second TE biopsy is representative of the ICM. A further five embryos (10%) had ICMs that contained the aneuploidy detected in the initial TE biopsy, but had additional minor karyotypic discrepancies with the original profile. For these five embryos the second TE biopsy was concordant with the original profile, suggesting that the entire TE had a different karyotypic makeup than the ICM. The remaining 42 embryos (84%) showed perfect karyotypic concordance between ICM and TE. CONCLUSIONS: In 6% of blastocysts, an initial aneuploid diagnosis is false negative, and the ICM is in effect euploid. This would help explain the reported cases in which a transferred aneuploid embryo leads to a normal and healthy pregnancy. Interestingly, in the discordant embryos the second TE biopsy often represents the ICM, meaning that a TE re-biopsy of aneuploid blastocysts might identify euploid embryos for transfer. P-459 Wednesday, November 1, 2017 PGS DOES NOT IMPROVE PREGNANCY OUTCOMES IN IVF CYCLES USING VITRIFIED DONOR M. J. Hill,a J. Doyle,b OOCYTES. N. Doyle,a W. Caswell,c J. Lim,b M. J. Tucker,b M. O. Stratton,c J. Graham,b A. DeCherney,a K. Devine,b H. L. Hayes,c M. Levy.b aNational Institute of Health, Bethesda, MD; bShady Grove Fertility, Rockville, MD; c Donor Egg Bank USA, Rockville, MD. OBJECTIVE: Theoretically oocyte cryopreservation has the potential to disrupt the meiotic spindle leading to abnormal segregation of chromosomes and increased aneuploidy in embryos. Therefore, PGS might have benefit in

Vol. 108, No. 3, Supplement, September 2017