- Email: [email protected]
Next generation sequencing for preimplantation genetic screening improves pregnancy outcomes compared with array comparative genomic hybridization in single thawed euploid embryo transfer cycles
ORIGINAL ARTICLES: ASSISTED REPRODUCTION
Next generation sequencing for preimplantation genetic screening improves pregnancy outcomes compared with array comparative genomic hybridization in single thawed euploid embryo transfer cycles , Ph.D.,b Yael Kramer, M.S.,a Jenna Friedenthal, M.D.,a Susan M. Maxwell, M.D.,a Santiago Munne David H. McCulloh, Ph.D.,a Caroline McCaffrey, Ph.D.,a and James A. Grifo, M.D., Ph.D.a a
New York University Langone Fertility Center, New York, New York; and b Cooper Genomics, Livingston, New Jersey
Objective: To evaluate whether the use of next generation sequencing (NGS) for preimplantation genetic screening (PGS) in single thawed euploid embryo transfer (STEET) cycles improves pregnancy outcomes compared with array comparative genomic hybridization (aCGH). Design: Retrospective cohort study. Setting: Single university-based fertility center. Patient(s): A total of 916 STEET cycles from January 2014 to December 2016 were identified. Cases included 548 STEET cycles using NGS for PGS and controls included 368 STEET cycles using aCGH for PGS. Intervention(s): Patients having a STEET after undergoing IVF and PGS with either NGS or aCGH. Main Outcome Measure(s): Primary outcomes were implantation rate, ongoing pregnancy/live birth rate (OP/LBR), biochemical pregnancy rate (PR), and spontaneous abortion (SAB) rate. Result(s): The implantation rate was significantly higher in the NGS group compared with the aCGH group (71.6% vs. 64.6%). The OP/ LBR was also significantly higher in the NGS group (62% vs. 54.4%), and there were significantly more biochemical pregnancies in the aCGH group compared with the NGS group (15.1% vs. 8.7%). After adjustment for confounding variables with a multiple logistic regression analysis, OP/LBR remained significantly higher in the NGS group. The SAB rate was not significantly different in the NGS group compared with the aCGH group (12.4% vs. 12.7%). Conclusion(s): Preimplantation genetic screening using NGS significantly improves pregnancy outcomes versus PGS using aCGH in STEET cycles. Next-generation sequencing has the ability to identify and screen for embryos with reduced viability such as mosaic embryos and those with partial aneuploidies or triploidy. Pregnancy outcomes with NGS may be improved due to the exclusion of these abnormal embryos. (Fertil SterilÒ 2018;109:627–32. Ó2017 by American Society for Reproductive Medicine.) Key Words: Next generation sequencing, array comparative genomic hybridization, preimplantation genetic screening, mosaicism Discuss: You can discuss this article with its authors and other readers at https://www.fertstertdialog.com/users/16110-fertilityand-sterility/posts/28723-25156
Received October 15, 2017; revised November 30, 2017; accepted December 15, 2017; published online March 28, 2018. J.F has nothing to disclose. S.M.M. has nothing to disclose. S.M. is an employee of CooperGenomics. Y.K. has nothing to disclose. D.H.M. reports compensation from ReproART: Georgian-American Center for Reproductive Medicine, Tbilisi, Georgia, Biogenetics Corporation, Mountainside, NJ, and Sperm and Embryo Bank of New York, New York, NY, outside of the submitted work. C.M. has nothing to disclose. J.A.G. has nothing to disclose. Reprint requests: Jenna Friedenthal, M.D., Department of Obstetrics and Gynecology, New York University Langone Fertility Center, New York, New York 10016 (E-mail: [email protected]). Fertility and Sterility® Vol. 109, No. 4, April 2018 0015-0282/$36.00 Copyright ©2017 American Society for Reproductive Medicine, Published by Elsevier Inc. https://doi.org/10.1016/j.fertnstert.2017.12.017 VOL. 109 NO. 4 / APRIL 2018
627
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
C
hromosomal abnormalities are the most frequent cause of first trimester pregnancy losses, and account for >50% of spontaneous abortions (1–3). In patients with advanced maternal age or recurrent pregnancy loss, IVF with preimplantation genetic screening (PGS) was developed as a means to identify and exclude chromosomally abnormal embryos, thereby increasing implantation rates and decreasing spontaneous abortion (SAB) rates. Various genetic platforms are available for comprehensive chromosomal screening for PGS. The most used platforms are array comparative genomic hybridization (aCGH) (4–6), single nucleotide polymorphism array (7, 8), quantitative polymerase chain reaction (PCR) (9–11), and next generation sequencing (NGS) (12–14). Next generation sequencing is the newest platform for PGS, which performs high throughput and high resolution sequencing by synthesis. Table 1 summarizes the detection capability of the different platforms according to the published literature. All platforms can assess aneuploidy of full chromosomes with low error rates, but not all platforms can reliably detect unbalanced translocations, segmental aneuploidies, polyploidy, or mosaicism. Mosaicism is defined as the presence of two or more populations of cells, each with different genotypes, within the same embryo and results from mitotic errors occurring after fertilization. Next generation sequencing has gained in popularity due to its ability to identify unbalanced translocations, segmental aneuploidies, some triploidies (20), and lower levels of mosaicism than other techniques (17, 21). This may be clinically important because mosaic embryos seem to have impaired viability and a decreased potential to result in a live birth. Maxwell et al. (20) re-examined embryos that were diagnosed as euploid by aCGH and resulted in miscarriage. Using NGS, they found that 31.6% of these miscarriages were from mosaic embryos and 5.2% were from triploid embryos. Array CGH was unable to detect these chromosomal abnormalities in whole genome amplified DNA from the same trophectoderm biopsy samples. In a multicenter study, Munne et al. (17) found that mosaic embryos diagnosed with NGS had significantly lower implantation rates than euploid embryos. Complex mosaic embryos and mosaic embryos with 40%– 80% abnormal cells also had lower ongoing implantation rates than other types of mosaic embryos. It is estimated that 20%–30% of blastocysts undergoing PGS with NGS are diagnosed as mosaic.
The purpose of this study was to determine whether the use of NGS, with its ability to exclude more mosaic embryos, improves pregnancy outcomes compared with aCGH when performing single thawed euploid ETs (STEET). We hypothesized that the clinical implementation of NGS would increase implantation and ongoing pregnancy rates (PRs) and decrease SAB rates among patients undergoing IVF with PGS.
MATERIALS AND METHODS This study was approved by the Institutional Review Board at New York University School of Medicine (NYU IRB#1300389). This is a retrospective cohort study of all STEET cycles from January 2014 to December 2016 at a single large university-based fertility center. Cases included all STEET cycles using NGS for PGS and controls included all STEET cycles using aCGH for PGS. The primary outcomes were implantation rate, SAB rate, biochemical PR, and ongoing PR/live birth rate. The implantation rate was calculated as the number of gestational sacs visualized on transvaginal ultrasound divided by the total number of embryos transferred. The SAB rate was defined as a pregnancy failure after a previously documented gestational sac on transvaginal ultrasound divided by the total number of clinical pregnancies. Biochemical pregnancies were defined as a positive hCG level on cycle days 28–30 R5 mIU/mL followed by declining hCG levels before the development of a gestational sac on transvaginal ultrasound. The biochemical PR was calculated as the number of biochemical pregnancies per ET with a subsequent positive hCG. The ongoing PR/live birth rate was defined as the number of ongoing pregnancies after the presence of a fetal pole with fetal heart tones and/or live births divided by the total number of embryos transferred. Monozygotic twins resulting from the transfer of one embryo were counted as one implantation and one ongoing pregnancy or live birth. At our center, all physicians used aCGH in 2014. Starting in January of 2015, NGS became available for PGS; therefore, two-thirds of our physicians elected to use NGS exclusively for their PGS cycles. One-third of our physicians continued to use aCGH for all of their PGS cycles in 2015. As of 2016, all physicians in the practice used NGS for PGS. We performed a subanalysis comparing only PGS cycles in 2015 in an attempt to control for changes in laboratory practices over time.
TABLE 1 A comparison of current preimplantation genetic screening platforms for comprehensive chromosomal screening. Characteristics Total independent data signalsa (reads per sample) Resolution in million megabytes Misdiagnosis of aneuploidies (4, 9, 12, 13, 15) Unbalanced translocations (16) Partial aneuploidies Polyploidy Percent mosaicism detectable (17, 18, 19)
qPCR
aCGH
SNP array
High resolution NGS
96 20 1% No No No No
2,700 6 2% Yes Yes No 40%–60%
32,000 6 2% Yes Yes Yes No
700,000 3 0 Yes Yes Yes 20%–80%
Note: aCGH ¼ array comparative genomic hybridization; NGS ¼ next generation sequencing; qPCR ¼ quantitative polymerase chain reaction; SNP ¼ single nucleotide polymorphism. a Number of reads per run number of samples per run percent of reads lost ¼ number of reads per sample. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
628
VOL. 109 NO. 4 / APRIL 2018
Fertility and Sterility® Demographic and baseline clinical data were collected on all patients including age, gravidity, parity, day 2 E2 and FSH levels, and infertility diagnosis. Exclusion criteria included patients who underwent double ET, mosaic ET, transfer of an embryo with degraded DNA after PGS biopsy, oocyte thaws, embryo thaws for biopsy, or those with incomplete outcome data. In addition, in an attempt to control for variation in analytic processes between laboratories, exclusion criteria also excluded any embryos not analyzed by CooperGenomics. All patients underwent controlled ovarian hyperstimulation (COH) using recombinant FSH or a combination of recombinant FSH with hMGs. Luteinizing hormone suppression was performed using either a GnRH antagonist starting around cycle day 8 or a GnRH agonist in a long or short protocol. Ovulation was triggered with either 10,000 IU of hCG or a combination of 40 U of a GnRH agonist and 1,000 IU of hCG when lead follicles reached 18 mm. Oocytes were retrieved 35 hours later and fertilized using standard insemination when possible. Intracytoplasmic sperm injection (ICSI) was performed for semen total motile counts <2 million or for patients with a history of poor fertilization or sperm acquired by testicular biopsy. Embryos were cultured to the blastocyst stage and underwent assisted hatching followed by trophectoderm biopsy on days 5 or 6. The trophectoderm biopsy specimens were sent to CooperGenomics, where they were analyzed by either aCGH or NGS per previously published protocols (4, 20). Statistical analyses of categorical data were performed using Fisher's exact test, and continuous variables were analyzed using Student's t-test. P values < .05 were considered significant. For the primary outcomes, relative risk with 95% confidence intervals were reported. Multiple logistic regression analysis was performed on the primary outcomes to adjust for confounding variables. Adjusted percentages with 95% confidence intervals were reported. Statistical significance was determined if the 95% confi-
dence intervals did not overlap the mean of the comparison group.
RESULTS A total of 916 STEET cycles were performed at our center from January 2014 to December 2016. Of these, 368 STEET cycles had PGS of their embryos using aCGH and 548 had PGS with NGS. Twenty-seven transfer cycles were excluded due to incomplete demographic or baseline clinical data: 12 from the NGS cohort and 15 from the aCGH cohort. Twelve additional transfer cycles were excluded from the NGS group (5 for mosaic ET and 2 for PGS analysis at a laboratory other than CooperGenomics), leaving 524 transfer cycles available for analysis. In the aCGH group, 353 transfer cycles were available for analysis. Of the five NGS mosaic ETs excluded, two did not implant and the remaining three resulted in SABs. Demographic and baseline clinical data are provided in Table 2. There was no difference in age, gravidity, parity, day 2 E2 and FSH values, or infertility diagnosis between groups. The average age of patients in the study was 36.2 years for the aCGH group and 35.6 years for the NGS group (P¼ .06). Patients with PGS using NGS had significantly more embryos remaining in storage after ET than those using PGS with aCGH (3.2 vs. 2.2; P¼ .0001). However, when excluding mosaic embryos, patients using NGS for PGS had significantly fewer embryos remaining in storage than those who used aCGH (1.7 vs. 2.2; P¼ .0066). The results of the primary outcome measures before and after adjustment can be found in Table 3. The implantation rate was significantly higher in the NGS group compared with the aCGH group (71.6% vs. 64.6%; P¼ .03), but after adjusting for day 2 FSH, gravidity, and uterine factor infertility, there was not a significant difference. The aCGH cohort had a significantly higher rate of biochemical pregnancies (15.1% vs. 8.7%; P¼ .013), but again after adjustment for day 2
TABLE 2 Demographic and baseline clinical data of patients undergoing single thawed euploid ET. Characteristics Age (y) Gravity Parity Day 2 E2 Day 2 FSH No. of embryos remaining No. of embryos remaining (excluding mosaic embryos) Infertility diagnosis Diminished ovarian reserve Endometriosis Male factor PCOS Recurrent pregnancy loss Tubal disease Uterine Unexplained
aCGH (n [ 353)
NGS (n [ 529)
P value
36.24 4.79 1.38 1.76 0.31 0.58 41.75 14.06 6.80 2.64 2.23 3.02 2.23 3.02
35.64 4.58 1.23 1.56 0.29 0.56 43.44 12.81 6.68 2.55 3.22 3.28 1.74 2.25
.06 .19 .67 .07 .51 .0001 .0066
17.8% (63/353) 3.7% (13/353) 10.2% (36/353) 11.3% (40/353) 9.1% (32/353) 3.7% (13/353) 3.1% (11/353) 41.1% (145/371)
15.3% (81/529) 4.5% (24/529) 11% (58/529) 14.6% (77/529) 6.2% (33/529) 4.7% (25/529) 2.3% (12/529) 41.2% (218/529)
.35 .61 .74 .19 .12 .41 .52 1.00
Note: aCGH ¼ array comparative genomic hybridization; NGS ¼ next generation sequencing; PCOS ¼ polycystic ovary syndrome. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
VOL. 109 NO. 4 / APRIL 2018
629
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
TABLE 3 Pregnancy outcomes for patients undergoing single thawed euploid ET with preimplantation genetic screening using either array comparative genomic hybridization or next generation sequencing from 2014 to 2016 and adjusted values after multiple logistic regression. Characteristics IR (%) BPR (%) SABR (%) OPR/LBR (%)
aCGH (n [ 353)
NGS (n [ 529)
P value
RR with 95% CI
Adjusted aCGH with 95% CI
Adjusted NGS with 95% CI
64.6% (228/353) 15.1% (41/272) 12.7% (29/228) 54.4% (192/353)
71.6% (379/529) 8.7% (36/416) 12.4% (47/329) 62% (328/529)
.032 .013 .90 .026
0.9 (0.82–0.99) 1.7 (1.44–2.65) 0.89 (0.58–1.37) 0.88 (0.78–0.99)
64.7 (9.6–10.8)a 14.2 (8.9–17.7)b 12 (3.8–5.0)c 53.8 (5.9–6.0)d
71.8 (7.9–9.4)a 8.3 (5.4–12.4)b 11.7 (3.1–3.8)c 62.1 (4.8–4.9)d
Adjusted significant No No No Yes
Note: aCGH ¼ array comparative genomic hybridization; BPR ¼ biochemical pregnancy rate; CI ¼ confidence interval; IR ¼ implantation rate; OPR/LBR ¼ ongoing pregnancy rate/live birth rate; NGS ¼ next generation sequencing; SABR ¼ spontaneous abortion rate. a Adjusted for FSH, gravidity, uterine factor infertility. b Adjusted for day 2 E2 and FSH, male factor infertility. c Adjusted for parity and infertility diagnosis of endometriosis. d Adjusted for gravidity and infertility diagnosis of endometriosis. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
E2 and FSH and male factor infertility, there was not a significant difference in the biochemical PR. An infertility diagnosis of endometriosis and a lower parity were associated with a higher miscarriage rate. Before and after adjustment for these confounding variables, the SAB rate was not significantly different between groups. The ongoing PR/live birth rate was significantly higher in the NGS group (62% vs. 54.4%; P¼ .026), and it remained significantly different after adjustment for gravidity and the infertility diagnosis of endometriosis. Results of the subanalysis of STEET cycles performed in 2015 only are shown in Supplemental Table 1(available online). The PGS using NGS had a higher implantation rate (69% vs. 61.3%; P¼ .15) and ongoing PR/live birth rate (59.8% vs. 50.1%; P¼ .087) than PGS using aCGH, but these
were not statistically significant. The biochemical PR was significantly higher when aCGH was used compared with NGS (20.6% vs. 8.6%; P¼ .006). The SAB rate was not significantly different between NGS and aCGH cycles (11% vs. 12.6%; P¼ .84). Figure 1 illustrates the percentage of STEET cycles that reached the different stages of early pregnancy development in the aCGH and NGS groups. The percentage of STEET cycles with a positive hCG was similar between patients using aCGH and NGS (77.1% vs. 78.6%). However, in the aCGH group, 11.6% of patients had a biochemical pregnancy compared with 6.8% of those in the NGS group. In the NGS group, 65.4% of embryos transferred led to a pregnancy with fetal heart tones visualized, compared with 60.3% of embryos transferred in the aCGH group. Sixty-two percent of
FIGURE 1 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0
aCGH NGS
PosiƟve HCG
GestaƟonal Sac
Fetal Heartbeat
Ongoing Pregnancy/Live Birth
Percentage of ET cycles that reached the different stages of early pregnancy development in the array comparative genomic hybridization and next generation sequencing groups. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
630
VOL. 109 NO. 4 / APRIL 2018
Fertility and Sterility® pregnancies were ongoing or resulted in a live birth in the NGS group compared with 54.39% in the aCGH group.
DISCUSSION The clinical implementation of NGS for PGS resulted in significantly improved ongoing PR/live birth rates compared with aCGH. Most of the pregnancy losses in the aCGH group were biochemical pregnancies before the formation of a gestational sac. The number of miscarriages between groups was not significantly different. The improved PRs seen may be attributed to several advances provided by NGS. First, NGS protocols may offer enhanced detection of segmental aneuploidies. Some NGS protocols have been shown to detect partial chromosomal gains and losses as small as 1.8–3 Mb (22, 23). In addition, NGS has greater potential to detect mosaicism compared with aCGH to due to its increased dynamic range (21, 24). Various NGS protocols have been compared with aCGH for the detection of embryo aneuploidy. The NGS results have been found to be concordant with aCGH 100% of the time (12, 13, 23), but when NGS detected mosaicism some results were classified by aCGH as normal and some abnormal (25). Few studies, however, have compared pregnancy outcomes with the clinical implementation of NGS compared with aCGH for PGS. Yang et al. (6) performed a pilot randomized control study comparing NGS and aCGH for PGS. They similarly found that NGS was concordant with aCGH 100% of the time for 24-chromosome diagnosis and more precisely detected segmental changes when compared with aCGH. Implantation, clinical pregnancy, and ongoing PRs were higher with NGS, but this was not statistically significant, possibly due to a small sample size. Most patients undergoing PGS in this study underwent a double euploid ET. The outcomes, therefore, do not reflect the more common practice of single ET when using comprehensive chromosomal screening. Our results show that pregnancy outcomes are improved with PGS with NGS compared with aCGH for patients undergoing single euploid ETs. Coates et al. (26) performed a retrospective study comparing birth outcomes in patients using PGS versus no PGS in the donor egg population and found that PGS significantly improved live birth rates with double ETs, but birth rates were not significantly different when single ET was performed. In their study, they used aCGH and NGS for PGS, and they did not perform a head-to-head comparison of the two modalities. Based on our findings of improved ongoing PR/ live birth rates with NGS compared with aCGH, studies comparing PGS versus no PGS in the donor egg population should be repeated using NGS for PGS and single ET. It is important to consider the potential consequences of exclusion of mosaic embryos from transfer. Mosaic embryos seem to have reduced viability. Fragouli et al. (27) studied 43 embryos transferred with known diploid-aneuploid mosaicism. Of these, 62% did not implant, 12% led to SAB, and 26% led to ongoing pregnancies. The pregnancy outcomes for mosaic embryos in this cohort were significantly decreased compared with euploid embryos. Our patients VOL. 109 NO. 4 / APRIL 2018
undergoing PGS with NGS had significantly fewer embryos available for transfer due to the detection and exclusion of mosaic embryos. By excluding mosaic embryos from the NGS group, the SAB rates were not significantly different between patients using NGS or aCGH for PGS. There was, however, a decrease in the rate of biochemical pregnancies with NGS. This provides a potential explanation for the results seen in our study. NGS may have the ability to identify and screen out embryos with reduced implantation potential. Because the embryos excluded by NGS are lost before implantation, they are removed from the pool before they are able to miscarry, which could explain the minimal difference seen in SAB rates. Miscarriages after euploid ET and implantation may, therefore, be due to other factors unrelated to embryo ploidy, such as uterine factors or spontaneous errors occurring during fetal development. Despite their reduced pregnancy potential, mosaic embryos have been reported to result in live births. Greco et al. (18) published a case series in which six live births were reported in a group of patients undergoing single mosaic ET. In addition, Maxwell et al. (20) found in their reanalysis of trophectoderm biopsies with NGS that 15.8% of live births had resulted from mosaic embryos. In addition, a recent multicenter analysis by Munne et al. (17) found that 41% of mosaic embryos transferred in their study led to ongoing pregnancies. This suggests that some mosaic embryos have the capacity to produce live births, and it may be acceptable to transfer certain types of mosaic embryos after extensive genetic counseling (24). In recent studies, Munne et al. (17) and Fragouli et al. (27) further analyzed which mosaic embryos are more likely to implant. As described, Fragouli et al. (27) found that, among diploid-aneuploid mosaic embryos transferred, most did not implant. In addition, Munne et al. (17) determined that complex mosaic embryos have significantly lower implantation rates (10%) than either aneuploidy mosaic, double aneuploidy mosaic, or segmental mosaics (50%, 45%, and 41%, respectively). This study was performed at a single institution with PGS results obtained by a single genetics laboratory. Therefore, all embryos undergoing PGS were subjected to the same PGS platform protocols with uniform analysis of the PGS results. Another strength of this study was its large sample size. One limitation is that aCGH and NGS were used at different points in time at our facility (aCGH being the predominant form of PGS in 2014, with NGS used exclusively as of 2016). To attempt to control for this, we performed a subanalysis of pregnancy outcomes using STEET cycles performed in 2015 when both aCGH and NGS were used for PGS. The 2015 subanalysis showed a trend toward improved outcomes with NGS. However, these results were underpowered to detect a significant difference with a power of 39.5%. The greatest limitation of this study was its retrospective design with the potential for confounding variables. Multiple logistic regression analysis was performed to minimize these effects. Randomized control trials should be performed to compare pregnancy outcomes between NGS and aCGH for PGS. Ideally, these studies should be expanded to multiple centers, which would allow for a greater sample size with increased generalizability of the results. 631
ORIGINAL ARTICLE: ASSISTED REPRODUCTION In conclusion, in this retrospective cohort study, we found that the use of NGS for PGS is associated with a significant improvement in ongoing PR/live birth rate when compared with aCGH for PGS. This may be due to the improved ability of NGS to detect mosaicism, triploidy, and partial deletions/duplications. Mosaic embryos may have reduced implantation potential. Therefore, the exclusion of mosaic embryos with NGS may explain these results. Larger, randomized, prospective studies are needed to confirm these findings.
12.
13.
14.
15.
REFERENCES 1. 2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
632
Petracchi F, Colaci DS, Igarzabal L, Gadow E. Cytogenetic analysis of first trimester pregnancy loss. Int J Gynaecol Obstet 2009;104:243–4. Warburton D, Fraser FC. Spontaneous abortion risks in man: data from reproductive histories collected in a medical genetics unit. Am J Hum Genet 1964;16:1–25. Hassold T, Chen N, Funkhouser J, Jooss T, Manuel B, Matsuura J, et al. A cytogenetic study of 1000 spontaneous abortions. Ann Hum Genet 1980; 44(Pt 2):151–78. Gutierrez-Mateo C, Colls P, Sanchez-Garcia J, Escudero T, Prates R, Ketterson K, et al. Validation of microarray comparative genomic hybridization for comprehensive chromosome analysis of embryos. Fertil Steril 2011; 95:953–8. Capalbo A, Wright G, Elliott T, Ubaldi FM, Rienzi L, Nagy ZP. FISH reanalysis of inner cell mass and trophectoderm samples of previously array-CGH screened blastocysts shows high accuracy of diagnosis and no major diagnostic impact of mosaicism at the blastocyst stage. Hum Reprod 2013;28: 2298–307. Yang Z, Lin J, Zhang J, Fong WI, Li P, Zhao R, et al. Randomized comparison of next-generation sequencing and array comparative genomic hybridization for preimplantation genetic screening: a pilot study. BMC Med Genom 2015;8:30. Treff NR, Su J, Tao X, Levy B, Scott RT Jr. Accurate single cell 24 chromosome aneuploidy screening using whole genome amplification and single nucleotide polymorphism microarrays. Fertil Steril 2010;94:2017–21. Johnson DS, Gemelos G, Baner J, Ryan A, Cinnioglu C, Banjevic M, et al. Preclinical validation of a microarray method for full molecular karyotyping of blastomeres in a 24-h protocol. Hum Reprod (Oxf, Engl) 2010;25:1066–75. Treff NR, Tao X, Ferry KM, Su J, Taylor D, Scott RT Jr. Development and validation of an accurate quantitative real-time polymerase chain reactionbased assay for human blastocyst comprehensive chromosomal aneuploidy screening. Fertil Steril 2012;97:819–24. Scott RT Jr, Upham KM, Forman EJ, Hong KH, Scott KL, Taylor D, et al. Blastocyst biopsy with comprehensive chromosome screening and fresh embryo transfer significantly increases in vitro fertilization implantation and delivery rates: a randomized controlled trial. Fertil Steril 2013;100:697–703. Forman EJ, Hong KH, Ferry KM, Tao X, Taylor D, Levy B, et al. In vitro fertilization with single euploid blastocyst transfer: a randomized controlled trial. Fertil Steril 2013;100:100–7.e1.
16.
17.
18. 19.
20.
21.
22.
23. 24. 25.
26.
27.
Kung A, Munne S, Bankowski B, Coates A, Wells D. Validation of nextgeneration sequencing for comprehensive chromosome screening of embryos. Reprod Biomed Online 2015;31:760–9. Wells D, Kaur K, Grifo J, Glassner M, Taylor JC, Fragouli E, et al. Clinical utilisation of a rapid low-pass whole genome sequencing technique for the diagnosis of aneuploidy in human embryos prior to implantation. J Med Genet 2014;51:553–62. Fiorentino F, Biricik A, Bono S, Spizzichino L, Cotroneo E, Cottone G, et al. Development and validation of a next-generation sequencing-based protocol for 24-chromosome aneuploidy screening of embryos. Fertil Steril 2014; 101:1375–82. Scott RT Jr, Ferry K, Su J, Tao X, Scott K, Treff NR. Comprehensive chromosome screening is highly predictive of the reproductive potential of human embryos: a prospective, blinded, nonselection study. Fertil Steril 2012;97: 870–5. Yeboah E, Munne S, Escudero T, Wagner Coughlin C, Surrey M, Ghadir S, et al. Chromosome abnormalities found in >60,000 blastocysts tested via array CGH. Fertil Steril 2015;104:e275. Munne S, Blazek J, Large M, Martinez-Ortiz PA, Nisson H, Liu E, et al. Detailed investigation into the cytogenetic constitution and pregnancy outcome of replacing mosaic blastocysts detected with the use of high-resolution next-generation sequencing. Fertil Steril 2017;108:62– 71.e8. Greco E, Minasi MG, Fiorentino F. Healthy babies after intrauterine transfer of mosaic aneuploid blastocysts. N Engl J Med 2015;373:2089–90. Tormasi S, Capaldi R, Gouw F, Munne S, Wells D, Fragouli E. The use of next generation sequencing to determine the developmental potential of mosaic blastocysts. Fertil Steril 2016;105:e41. Maxwell SM, Colls P, Hodes-Wertz B, McCulloh DH, McCaffrey C, Wells D, et al. Why do euploid embryos miscarry? A case-control study comparing the rate of aneuploidy within presumed euploid embryos that resulted in miscarriage or live birth using next-generation sequencing. Fertil Steril 2016;106: 1414–9.e5. Munne S, Wells D. Detection of mosaicism at blastocyst stage with the use of high-resolution next-generation sequencing. Fertil Steril 2017;107:1085– 91. Zheng H, Jin H, Liu L, Liu J, Wang WH. Application of next-generation sequencing for 24-chromosome aneuploidy screening of human preimplantation embryos. Mol Cytogen 2015;8:38. Handyside AH. 24-chromosome copy number analysis: a comparison of available technologies. Fertil Steril 2013;100:595–602. Sachdev NM, Maxwell SM, Besser AG, Grifo JA. Diagnosis and clinical management of embryonic mosaicism. Fertil Steril 2017;107:6–11. Ribustello LET, Lanzas J, Liu E, Munne S. Mosaic embryos, detected by NGS, are mostly classified as euploid by aCGH. Reprod Biomed Online 2016;33: 258–9. Coates A, Bankowski BJ, Kung A, Griffin DK, Munne S. Differences in pregnancy outcomes in donor egg frozen embryo transfer (FET) cycles following preimplantation genetic screening (PGS): a single center retrospective study. J Assist Reprod Genet 2017;34:71–8. Fragouli E, Alfarawati S, Spath K, Tarozzi N, Borini A, Wells D. The developmental potential of mosaic embryos. Fertil Steril 2015;104:e96.
VOL. 109 NO. 4 / APRIL 2018
Next generation sequencing for preimplantation genetic screening improves pregnancy outcomes compared with array comparative genomic hybridization in single thawed euploid embryo transfer cycles , Ph.D.,b Yael Kramer, M.S.,a Jenna Friedenthal, M.D.,a Susan M. Maxwell, M.D.,a Santiago Munne David H. McCulloh, Ph.D.,a Caroline McCaffrey, Ph.D.,a and James A. Grifo, M.D., Ph.D.a a
New York University Langone Fertility Center, New York, New York; and b Cooper Genomics, Livingston, New Jersey
Objective: To evaluate whether the use of next generation sequencing (NGS) for preimplantation genetic screening (PGS) in single thawed euploid embryo transfer (STEET) cycles improves pregnancy outcomes compared with array comparative genomic hybridization (aCGH). Design: Retrospective cohort study. Setting: Single university-based fertility center. Patient(s): A total of 916 STEET cycles from January 2014 to December 2016 were identified. Cases included 548 STEET cycles using NGS for PGS and controls included 368 STEET cycles using aCGH for PGS. Intervention(s): Patients having a STEET after undergoing IVF and PGS with either NGS or aCGH. Main Outcome Measure(s): Primary outcomes were implantation rate, ongoing pregnancy/live birth rate (OP/LBR), biochemical pregnancy rate (PR), and spontaneous abortion (SAB) rate. Result(s): The implantation rate was significantly higher in the NGS group compared with the aCGH group (71.6% vs. 64.6%). The OP/ LBR was also significantly higher in the NGS group (62% vs. 54.4%), and there were significantly more biochemical pregnancies in the aCGH group compared with the NGS group (15.1% vs. 8.7%). After adjustment for confounding variables with a multiple logistic regression analysis, OP/LBR remained significantly higher in the NGS group. The SAB rate was not significantly different in the NGS group compared with the aCGH group (12.4% vs. 12.7%). Conclusion(s): Preimplantation genetic screening using NGS significantly improves pregnancy outcomes versus PGS using aCGH in STEET cycles. Next-generation sequencing has the ability to identify and screen for embryos with reduced viability such as mosaic embryos and those with partial aneuploidies or triploidy. Pregnancy outcomes with NGS may be improved due to the exclusion of these abnormal embryos. (Fertil SterilÒ 2018;109:627–32. Ó2017 by American Society for Reproductive Medicine.) Key Words: Next generation sequencing, array comparative genomic hybridization, preimplantation genetic screening, mosaicism Discuss: You can discuss this article with its authors and other readers at https://www.fertstertdialog.com/users/16110-fertilityand-sterility/posts/28723-25156
Received October 15, 2017; revised November 30, 2017; accepted December 15, 2017; published online March 28, 2018. J.F has nothing to disclose. S.M.M. has nothing to disclose. S.M. is an employee of CooperGenomics. Y.K. has nothing to disclose. D.H.M. reports compensation from ReproART: Georgian-American Center for Reproductive Medicine, Tbilisi, Georgia, Biogenetics Corporation, Mountainside, NJ, and Sperm and Embryo Bank of New York, New York, NY, outside of the submitted work. C.M. has nothing to disclose. J.A.G. has nothing to disclose. Reprint requests: Jenna Friedenthal, M.D., Department of Obstetrics and Gynecology, New York University Langone Fertility Center, New York, New York 10016 (E-mail: [email protected]). Fertility and Sterility® Vol. 109, No. 4, April 2018 0015-0282/$36.00 Copyright ©2017 American Society for Reproductive Medicine, Published by Elsevier Inc. https://doi.org/10.1016/j.fertnstert.2017.12.017 VOL. 109 NO. 4 / APRIL 2018
627
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
C
hromosomal abnormalities are the most frequent cause of first trimester pregnancy losses, and account for >50% of spontaneous abortions (1–3). In patients with advanced maternal age or recurrent pregnancy loss, IVF with preimplantation genetic screening (PGS) was developed as a means to identify and exclude chromosomally abnormal embryos, thereby increasing implantation rates and decreasing spontaneous abortion (SAB) rates. Various genetic platforms are available for comprehensive chromosomal screening for PGS. The most used platforms are array comparative genomic hybridization (aCGH) (4–6), single nucleotide polymorphism array (7, 8), quantitative polymerase chain reaction (PCR) (9–11), and next generation sequencing (NGS) (12–14). Next generation sequencing is the newest platform for PGS, which performs high throughput and high resolution sequencing by synthesis. Table 1 summarizes the detection capability of the different platforms according to the published literature. All platforms can assess aneuploidy of full chromosomes with low error rates, but not all platforms can reliably detect unbalanced translocations, segmental aneuploidies, polyploidy, or mosaicism. Mosaicism is defined as the presence of two or more populations of cells, each with different genotypes, within the same embryo and results from mitotic errors occurring after fertilization. Next generation sequencing has gained in popularity due to its ability to identify unbalanced translocations, segmental aneuploidies, some triploidies (20), and lower levels of mosaicism than other techniques (17, 21). This may be clinically important because mosaic embryos seem to have impaired viability and a decreased potential to result in a live birth. Maxwell et al. (20) re-examined embryos that were diagnosed as euploid by aCGH and resulted in miscarriage. Using NGS, they found that 31.6% of these miscarriages were from mosaic embryos and 5.2% were from triploid embryos. Array CGH was unable to detect these chromosomal abnormalities in whole genome amplified DNA from the same trophectoderm biopsy samples. In a multicenter study, Munne et al. (17) found that mosaic embryos diagnosed with NGS had significantly lower implantation rates than euploid embryos. Complex mosaic embryos and mosaic embryos with 40%– 80% abnormal cells also had lower ongoing implantation rates than other types of mosaic embryos. It is estimated that 20%–30% of blastocysts undergoing PGS with NGS are diagnosed as mosaic.
The purpose of this study was to determine whether the use of NGS, with its ability to exclude more mosaic embryos, improves pregnancy outcomes compared with aCGH when performing single thawed euploid ETs (STEET). We hypothesized that the clinical implementation of NGS would increase implantation and ongoing pregnancy rates (PRs) and decrease SAB rates among patients undergoing IVF with PGS.
MATERIALS AND METHODS This study was approved by the Institutional Review Board at New York University School of Medicine (NYU IRB#1300389). This is a retrospective cohort study of all STEET cycles from January 2014 to December 2016 at a single large university-based fertility center. Cases included all STEET cycles using NGS for PGS and controls included all STEET cycles using aCGH for PGS. The primary outcomes were implantation rate, SAB rate, biochemical PR, and ongoing PR/live birth rate. The implantation rate was calculated as the number of gestational sacs visualized on transvaginal ultrasound divided by the total number of embryos transferred. The SAB rate was defined as a pregnancy failure after a previously documented gestational sac on transvaginal ultrasound divided by the total number of clinical pregnancies. Biochemical pregnancies were defined as a positive hCG level on cycle days 28–30 R5 mIU/mL followed by declining hCG levels before the development of a gestational sac on transvaginal ultrasound. The biochemical PR was calculated as the number of biochemical pregnancies per ET with a subsequent positive hCG. The ongoing PR/live birth rate was defined as the number of ongoing pregnancies after the presence of a fetal pole with fetal heart tones and/or live births divided by the total number of embryos transferred. Monozygotic twins resulting from the transfer of one embryo were counted as one implantation and one ongoing pregnancy or live birth. At our center, all physicians used aCGH in 2014. Starting in January of 2015, NGS became available for PGS; therefore, two-thirds of our physicians elected to use NGS exclusively for their PGS cycles. One-third of our physicians continued to use aCGH for all of their PGS cycles in 2015. As of 2016, all physicians in the practice used NGS for PGS. We performed a subanalysis comparing only PGS cycles in 2015 in an attempt to control for changes in laboratory practices over time.
TABLE 1 A comparison of current preimplantation genetic screening platforms for comprehensive chromosomal screening. Characteristics Total independent data signalsa (reads per sample) Resolution in million megabytes Misdiagnosis of aneuploidies (4, 9, 12, 13, 15) Unbalanced translocations (16) Partial aneuploidies Polyploidy Percent mosaicism detectable (17, 18, 19)
qPCR
aCGH
SNP array
High resolution NGS
96 20 1% No No No No
2,700 6 2% Yes Yes No 40%–60%
32,000 6 2% Yes Yes Yes No
700,000 3 0 Yes Yes Yes 20%–80%
Note: aCGH ¼ array comparative genomic hybridization; NGS ¼ next generation sequencing; qPCR ¼ quantitative polymerase chain reaction; SNP ¼ single nucleotide polymorphism. a Number of reads per run number of samples per run percent of reads lost ¼ number of reads per sample. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
628
VOL. 109 NO. 4 / APRIL 2018
Fertility and Sterility® Demographic and baseline clinical data were collected on all patients including age, gravidity, parity, day 2 E2 and FSH levels, and infertility diagnosis. Exclusion criteria included patients who underwent double ET, mosaic ET, transfer of an embryo with degraded DNA after PGS biopsy, oocyte thaws, embryo thaws for biopsy, or those with incomplete outcome data. In addition, in an attempt to control for variation in analytic processes between laboratories, exclusion criteria also excluded any embryos not analyzed by CooperGenomics. All patients underwent controlled ovarian hyperstimulation (COH) using recombinant FSH or a combination of recombinant FSH with hMGs. Luteinizing hormone suppression was performed using either a GnRH antagonist starting around cycle day 8 or a GnRH agonist in a long or short protocol. Ovulation was triggered with either 10,000 IU of hCG or a combination of 40 U of a GnRH agonist and 1,000 IU of hCG when lead follicles reached 18 mm. Oocytes were retrieved 35 hours later and fertilized using standard insemination when possible. Intracytoplasmic sperm injection (ICSI) was performed for semen total motile counts <2 million or for patients with a history of poor fertilization or sperm acquired by testicular biopsy. Embryos were cultured to the blastocyst stage and underwent assisted hatching followed by trophectoderm biopsy on days 5 or 6. The trophectoderm biopsy specimens were sent to CooperGenomics, where they were analyzed by either aCGH or NGS per previously published protocols (4, 20). Statistical analyses of categorical data were performed using Fisher's exact test, and continuous variables were analyzed using Student's t-test. P values < .05 were considered significant. For the primary outcomes, relative risk with 95% confidence intervals were reported. Multiple logistic regression analysis was performed on the primary outcomes to adjust for confounding variables. Adjusted percentages with 95% confidence intervals were reported. Statistical significance was determined if the 95% confi-
dence intervals did not overlap the mean of the comparison group.
RESULTS A total of 916 STEET cycles were performed at our center from January 2014 to December 2016. Of these, 368 STEET cycles had PGS of their embryos using aCGH and 548 had PGS with NGS. Twenty-seven transfer cycles were excluded due to incomplete demographic or baseline clinical data: 12 from the NGS cohort and 15 from the aCGH cohort. Twelve additional transfer cycles were excluded from the NGS group (5 for mosaic ET and 2 for PGS analysis at a laboratory other than CooperGenomics), leaving 524 transfer cycles available for analysis. In the aCGH group, 353 transfer cycles were available for analysis. Of the five NGS mosaic ETs excluded, two did not implant and the remaining three resulted in SABs. Demographic and baseline clinical data are provided in Table 2. There was no difference in age, gravidity, parity, day 2 E2 and FSH values, or infertility diagnosis between groups. The average age of patients in the study was 36.2 years for the aCGH group and 35.6 years for the NGS group (P¼ .06). Patients with PGS using NGS had significantly more embryos remaining in storage after ET than those using PGS with aCGH (3.2 vs. 2.2; P¼ .0001). However, when excluding mosaic embryos, patients using NGS for PGS had significantly fewer embryos remaining in storage than those who used aCGH (1.7 vs. 2.2; P¼ .0066). The results of the primary outcome measures before and after adjustment can be found in Table 3. The implantation rate was significantly higher in the NGS group compared with the aCGH group (71.6% vs. 64.6%; P¼ .03), but after adjusting for day 2 FSH, gravidity, and uterine factor infertility, there was not a significant difference. The aCGH cohort had a significantly higher rate of biochemical pregnancies (15.1% vs. 8.7%; P¼ .013), but again after adjustment for day 2
TABLE 2 Demographic and baseline clinical data of patients undergoing single thawed euploid ET. Characteristics Age (y) Gravity Parity Day 2 E2 Day 2 FSH No. of embryos remaining No. of embryos remaining (excluding mosaic embryos) Infertility diagnosis Diminished ovarian reserve Endometriosis Male factor PCOS Recurrent pregnancy loss Tubal disease Uterine Unexplained
aCGH (n [ 353)
NGS (n [ 529)
P value
36.24 4.79 1.38 1.76 0.31 0.58 41.75 14.06 6.80 2.64 2.23 3.02 2.23 3.02
35.64 4.58 1.23 1.56 0.29 0.56 43.44 12.81 6.68 2.55 3.22 3.28 1.74 2.25
.06 .19 .67 .07 .51 .0001 .0066
17.8% (63/353) 3.7% (13/353) 10.2% (36/353) 11.3% (40/353) 9.1% (32/353) 3.7% (13/353) 3.1% (11/353) 41.1% (145/371)
15.3% (81/529) 4.5% (24/529) 11% (58/529) 14.6% (77/529) 6.2% (33/529) 4.7% (25/529) 2.3% (12/529) 41.2% (218/529)
.35 .61 .74 .19 .12 .41 .52 1.00
Note: aCGH ¼ array comparative genomic hybridization; NGS ¼ next generation sequencing; PCOS ¼ polycystic ovary syndrome. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
VOL. 109 NO. 4 / APRIL 2018
629
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
TABLE 3 Pregnancy outcomes for patients undergoing single thawed euploid ET with preimplantation genetic screening using either array comparative genomic hybridization or next generation sequencing from 2014 to 2016 and adjusted values after multiple logistic regression. Characteristics IR (%) BPR (%) SABR (%) OPR/LBR (%)
aCGH (n [ 353)
NGS (n [ 529)
P value
RR with 95% CI
Adjusted aCGH with 95% CI
Adjusted NGS with 95% CI
64.6% (228/353) 15.1% (41/272) 12.7% (29/228) 54.4% (192/353)
71.6% (379/529) 8.7% (36/416) 12.4% (47/329) 62% (328/529)
.032 .013 .90 .026
0.9 (0.82–0.99) 1.7 (1.44–2.65) 0.89 (0.58–1.37) 0.88 (0.78–0.99)
64.7 (9.6–10.8)a 14.2 (8.9–17.7)b 12 (3.8–5.0)c 53.8 (5.9–6.0)d
71.8 (7.9–9.4)a 8.3 (5.4–12.4)b 11.7 (3.1–3.8)c 62.1 (4.8–4.9)d
Adjusted significant No No No Yes
Note: aCGH ¼ array comparative genomic hybridization; BPR ¼ biochemical pregnancy rate; CI ¼ confidence interval; IR ¼ implantation rate; OPR/LBR ¼ ongoing pregnancy rate/live birth rate; NGS ¼ next generation sequencing; SABR ¼ spontaneous abortion rate. a Adjusted for FSH, gravidity, uterine factor infertility. b Adjusted for day 2 E2 and FSH, male factor infertility. c Adjusted for parity and infertility diagnosis of endometriosis. d Adjusted for gravidity and infertility diagnosis of endometriosis. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
E2 and FSH and male factor infertility, there was not a significant difference in the biochemical PR. An infertility diagnosis of endometriosis and a lower parity were associated with a higher miscarriage rate. Before and after adjustment for these confounding variables, the SAB rate was not significantly different between groups. The ongoing PR/live birth rate was significantly higher in the NGS group (62% vs. 54.4%; P¼ .026), and it remained significantly different after adjustment for gravidity and the infertility diagnosis of endometriosis. Results of the subanalysis of STEET cycles performed in 2015 only are shown in Supplemental Table 1(available online). The PGS using NGS had a higher implantation rate (69% vs. 61.3%; P¼ .15) and ongoing PR/live birth rate (59.8% vs. 50.1%; P¼ .087) than PGS using aCGH, but these
were not statistically significant. The biochemical PR was significantly higher when aCGH was used compared with NGS (20.6% vs. 8.6%; P¼ .006). The SAB rate was not significantly different between NGS and aCGH cycles (11% vs. 12.6%; P¼ .84). Figure 1 illustrates the percentage of STEET cycles that reached the different stages of early pregnancy development in the aCGH and NGS groups. The percentage of STEET cycles with a positive hCG was similar between patients using aCGH and NGS (77.1% vs. 78.6%). However, in the aCGH group, 11.6% of patients had a biochemical pregnancy compared with 6.8% of those in the NGS group. In the NGS group, 65.4% of embryos transferred led to a pregnancy with fetal heart tones visualized, compared with 60.3% of embryos transferred in the aCGH group. Sixty-two percent of
FIGURE 1 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0
aCGH NGS
PosiƟve HCG
GestaƟonal Sac
Fetal Heartbeat
Ongoing Pregnancy/Live Birth
Percentage of ET cycles that reached the different stages of early pregnancy development in the array comparative genomic hybridization and next generation sequencing groups. Friedenthal. NGS increases ongoing PRs. Fertil Steril 2017.
630
VOL. 109 NO. 4 / APRIL 2018
Fertility and Sterility® pregnancies were ongoing or resulted in a live birth in the NGS group compared with 54.39% in the aCGH group.
DISCUSSION The clinical implementation of NGS for PGS resulted in significantly improved ongoing PR/live birth rates compared with aCGH. Most of the pregnancy losses in the aCGH group were biochemical pregnancies before the formation of a gestational sac. The number of miscarriages between groups was not significantly different. The improved PRs seen may be attributed to several advances provided by NGS. First, NGS protocols may offer enhanced detection of segmental aneuploidies. Some NGS protocols have been shown to detect partial chromosomal gains and losses as small as 1.8–3 Mb (22, 23). In addition, NGS has greater potential to detect mosaicism compared with aCGH to due to its increased dynamic range (21, 24). Various NGS protocols have been compared with aCGH for the detection of embryo aneuploidy. The NGS results have been found to be concordant with aCGH 100% of the time (12, 13, 23), but when NGS detected mosaicism some results were classified by aCGH as normal and some abnormal (25). Few studies, however, have compared pregnancy outcomes with the clinical implementation of NGS compared with aCGH for PGS. Yang et al. (6) performed a pilot randomized control study comparing NGS and aCGH for PGS. They similarly found that NGS was concordant with aCGH 100% of the time for 24-chromosome diagnosis and more precisely detected segmental changes when compared with aCGH. Implantation, clinical pregnancy, and ongoing PRs were higher with NGS, but this was not statistically significant, possibly due to a small sample size. Most patients undergoing PGS in this study underwent a double euploid ET. The outcomes, therefore, do not reflect the more common practice of single ET when using comprehensive chromosomal screening. Our results show that pregnancy outcomes are improved with PGS with NGS compared with aCGH for patients undergoing single euploid ETs. Coates et al. (26) performed a retrospective study comparing birth outcomes in patients using PGS versus no PGS in the donor egg population and found that PGS significantly improved live birth rates with double ETs, but birth rates were not significantly different when single ET was performed. In their study, they used aCGH and NGS for PGS, and they did not perform a head-to-head comparison of the two modalities. Based on our findings of improved ongoing PR/ live birth rates with NGS compared with aCGH, studies comparing PGS versus no PGS in the donor egg population should be repeated using NGS for PGS and single ET. It is important to consider the potential consequences of exclusion of mosaic embryos from transfer. Mosaic embryos seem to have reduced viability. Fragouli et al. (27) studied 43 embryos transferred with known diploid-aneuploid mosaicism. Of these, 62% did not implant, 12% led to SAB, and 26% led to ongoing pregnancies. The pregnancy outcomes for mosaic embryos in this cohort were significantly decreased compared with euploid embryos. Our patients VOL. 109 NO. 4 / APRIL 2018
undergoing PGS with NGS had significantly fewer embryos available for transfer due to the detection and exclusion of mosaic embryos. By excluding mosaic embryos from the NGS group, the SAB rates were not significantly different between patients using NGS or aCGH for PGS. There was, however, a decrease in the rate of biochemical pregnancies with NGS. This provides a potential explanation for the results seen in our study. NGS may have the ability to identify and screen out embryos with reduced implantation potential. Because the embryos excluded by NGS are lost before implantation, they are removed from the pool before they are able to miscarry, which could explain the minimal difference seen in SAB rates. Miscarriages after euploid ET and implantation may, therefore, be due to other factors unrelated to embryo ploidy, such as uterine factors or spontaneous errors occurring during fetal development. Despite their reduced pregnancy potential, mosaic embryos have been reported to result in live births. Greco et al. (18) published a case series in which six live births were reported in a group of patients undergoing single mosaic ET. In addition, Maxwell et al. (20) found in their reanalysis of trophectoderm biopsies with NGS that 15.8% of live births had resulted from mosaic embryos. In addition, a recent multicenter analysis by Munne et al. (17) found that 41% of mosaic embryos transferred in their study led to ongoing pregnancies. This suggests that some mosaic embryos have the capacity to produce live births, and it may be acceptable to transfer certain types of mosaic embryos after extensive genetic counseling (24). In recent studies, Munne et al. (17) and Fragouli et al. (27) further analyzed which mosaic embryos are more likely to implant. As described, Fragouli et al. (27) found that, among diploid-aneuploid mosaic embryos transferred, most did not implant. In addition, Munne et al. (17) determined that complex mosaic embryos have significantly lower implantation rates (10%) than either aneuploidy mosaic, double aneuploidy mosaic, or segmental mosaics (50%, 45%, and 41%, respectively). This study was performed at a single institution with PGS results obtained by a single genetics laboratory. Therefore, all embryos undergoing PGS were subjected to the same PGS platform protocols with uniform analysis of the PGS results. Another strength of this study was its large sample size. One limitation is that aCGH and NGS were used at different points in time at our facility (aCGH being the predominant form of PGS in 2014, with NGS used exclusively as of 2016). To attempt to control for this, we performed a subanalysis of pregnancy outcomes using STEET cycles performed in 2015 when both aCGH and NGS were used for PGS. The 2015 subanalysis showed a trend toward improved outcomes with NGS. However, these results were underpowered to detect a significant difference with a power of 39.5%. The greatest limitation of this study was its retrospective design with the potential for confounding variables. Multiple logistic regression analysis was performed to minimize these effects. Randomized control trials should be performed to compare pregnancy outcomes between NGS and aCGH for PGS. Ideally, these studies should be expanded to multiple centers, which would allow for a greater sample size with increased generalizability of the results. 631
ORIGINAL ARTICLE: ASSISTED REPRODUCTION In conclusion, in this retrospective cohort study, we found that the use of NGS for PGS is associated with a significant improvement in ongoing PR/live birth rate when compared with aCGH for PGS. This may be due to the improved ability of NGS to detect mosaicism, triploidy, and partial deletions/duplications. Mosaic embryos may have reduced implantation potential. Therefore, the exclusion of mosaic embryos with NGS may explain these results. Larger, randomized, prospective studies are needed to confirm these findings.
12.
13.
14.
15.
REFERENCES 1. 2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
632
Petracchi F, Colaci DS, Igarzabal L, Gadow E. Cytogenetic analysis of first trimester pregnancy loss. Int J Gynaecol Obstet 2009;104:243–4. Warburton D, Fraser FC. Spontaneous abortion risks in man: data from reproductive histories collected in a medical genetics unit. Am J Hum Genet 1964;16:1–25. Hassold T, Chen N, Funkhouser J, Jooss T, Manuel B, Matsuura J, et al. A cytogenetic study of 1000 spontaneous abortions. Ann Hum Genet 1980; 44(Pt 2):151–78. Gutierrez-Mateo C, Colls P, Sanchez-Garcia J, Escudero T, Prates R, Ketterson K, et al. Validation of microarray comparative genomic hybridization for comprehensive chromosome analysis of embryos. Fertil Steril 2011; 95:953–8. Capalbo A, Wright G, Elliott T, Ubaldi FM, Rienzi L, Nagy ZP. FISH reanalysis of inner cell mass and trophectoderm samples of previously array-CGH screened blastocysts shows high accuracy of diagnosis and no major diagnostic impact of mosaicism at the blastocyst stage. Hum Reprod 2013;28: 2298–307. Yang Z, Lin J, Zhang J, Fong WI, Li P, Zhao R, et al. Randomized comparison of next-generation sequencing and array comparative genomic hybridization for preimplantation genetic screening: a pilot study. BMC Med Genom 2015;8:30. Treff NR, Su J, Tao X, Levy B, Scott RT Jr. Accurate single cell 24 chromosome aneuploidy screening using whole genome amplification and single nucleotide polymorphism microarrays. Fertil Steril 2010;94:2017–21. Johnson DS, Gemelos G, Baner J, Ryan A, Cinnioglu C, Banjevic M, et al. Preclinical validation of a microarray method for full molecular karyotyping of blastomeres in a 24-h protocol. Hum Reprod (Oxf, Engl) 2010;25:1066–75. Treff NR, Tao X, Ferry KM, Su J, Taylor D, Scott RT Jr. Development and validation of an accurate quantitative real-time polymerase chain reactionbased assay for human blastocyst comprehensive chromosomal aneuploidy screening. Fertil Steril 2012;97:819–24. Scott RT Jr, Upham KM, Forman EJ, Hong KH, Scott KL, Taylor D, et al. Blastocyst biopsy with comprehensive chromosome screening and fresh embryo transfer significantly increases in vitro fertilization implantation and delivery rates: a randomized controlled trial. Fertil Steril 2013;100:697–703. Forman EJ, Hong KH, Ferry KM, Tao X, Taylor D, Levy B, et al. In vitro fertilization with single euploid blastocyst transfer: a randomized controlled trial. Fertil Steril 2013;100:100–7.e1.
16.
17.
18. 19.
20.
21.
22.
23. 24. 25.
26.
27.
Kung A, Munne S, Bankowski B, Coates A, Wells D. Validation of nextgeneration sequencing for comprehensive chromosome screening of embryos. Reprod Biomed Online 2015;31:760–9. Wells D, Kaur K, Grifo J, Glassner M, Taylor JC, Fragouli E, et al. Clinical utilisation of a rapid low-pass whole genome sequencing technique for the diagnosis of aneuploidy in human embryos prior to implantation. J Med Genet 2014;51:553–62. Fiorentino F, Biricik A, Bono S, Spizzichino L, Cotroneo E, Cottone G, et al. Development and validation of a next-generation sequencing-based protocol for 24-chromosome aneuploidy screening of embryos. Fertil Steril 2014; 101:1375–82. Scott RT Jr, Ferry K, Su J, Tao X, Scott K, Treff NR. Comprehensive chromosome screening is highly predictive of the reproductive potential of human embryos: a prospective, blinded, nonselection study. Fertil Steril 2012;97: 870–5. Yeboah E, Munne S, Escudero T, Wagner Coughlin C, Surrey M, Ghadir S, et al. Chromosome abnormalities found in >60,000 blastocysts tested via array CGH. Fertil Steril 2015;104:e275. Munne S, Blazek J, Large M, Martinez-Ortiz PA, Nisson H, Liu E, et al. Detailed investigation into the cytogenetic constitution and pregnancy outcome of replacing mosaic blastocysts detected with the use of high-resolution next-generation sequencing. Fertil Steril 2017;108:62– 71.e8. Greco E, Minasi MG, Fiorentino F. Healthy babies after intrauterine transfer of mosaic aneuploid blastocysts. N Engl J Med 2015;373:2089–90. Tormasi S, Capaldi R, Gouw F, Munne S, Wells D, Fragouli E. The use of next generation sequencing to determine the developmental potential of mosaic blastocysts. Fertil Steril 2016;105:e41. Maxwell SM, Colls P, Hodes-Wertz B, McCulloh DH, McCaffrey C, Wells D, et al. Why do euploid embryos miscarry? A case-control study comparing the rate of aneuploidy within presumed euploid embryos that resulted in miscarriage or live birth using next-generation sequencing. Fertil Steril 2016;106: 1414–9.e5. Munne S, Wells D. Detection of mosaicism at blastocyst stage with the use of high-resolution next-generation sequencing. Fertil Steril 2017;107:1085– 91. Zheng H, Jin H, Liu L, Liu J, Wang WH. Application of next-generation sequencing for 24-chromosome aneuploidy screening of human preimplantation embryos. Mol Cytogen 2015;8:38. Handyside AH. 24-chromosome copy number analysis: a comparison of available technologies. Fertil Steril 2013;100:595–602. Sachdev NM, Maxwell SM, Besser AG, Grifo JA. Diagnosis and clinical management of embryonic mosaicism. Fertil Steril 2017;107:6–11. Ribustello LET, Lanzas J, Liu E, Munne S. Mosaic embryos, detected by NGS, are mostly classified as euploid by aCGH. Reprod Biomed Online 2016;33: 258–9. Coates A, Bankowski BJ, Kung A, Griffin DK, Munne S. Differences in pregnancy outcomes in donor egg frozen embryo transfer (FET) cycles following preimplantation genetic screening (PGS): a single center retrospective study. J Assist Reprod Genet 2017;34:71–8. Fragouli E, Alfarawati S, Spath K, Tarozzi N, Borini A, Wells D. The developmental potential of mosaic embryos. Fertil Steril 2015;104:e96.
VOL. 109 NO. 4 / APRIL 2018