Preimplantation genetic testing for structural rearrangements by genome-wide SNP genotyping and haplotype analysis: a prospective multicenter clinical study

Summary

Background

Preimplantation genetic testing for chromosomal structural rearrangements (PGT-SR) has been widely utilized to select euploid embryos in patients carrying balanced chromosomal rearrangements (BCRs) by chromosome copy number analysis. However, reliable and extensively validated PGT-SR methods for selecting embryos without BCRs in large-cohort studies are lacking.

Methods

In this prospective, multicenter, cohort study, carriers with BCRs undergoing PGT-SR were recruited across 12 academic fertility centers within China. PGT-SR was performed using genome-wide SNP genotyping and haplotyping approach. Parental haplotypes were phased by available genotypes from a close relative or an unbalanced embryo. The karyotypes of embryos were inferred from the haplotypes. Only a single embryo was transferred in each cycle.

Findings

Between April 2018 and March 2023, 1298 carriers we randomly enrolled. A total of 7867 blastocysts from 1603 PGT-SR cycles were biopsied, in which 7750 (98.51%) were successfully genotyped and analyzed. Overall, 75.98% (1218/1603) of cycles obtained euploid embryos and 53.15% (852/1603) generated non-carrier embryos. The proportion of carrier and non-carrier embryos was similar in different subgroups. A total of 1030 non-carrier and 439 carrier embryos were transferred, 817 healthy babies were delivered cumulatively. Our results demonstrate that SNP-haplotyping method is highly accurate (sensitivity 95% CI: 98.34%–100%, specificity 95% CI: 96.63%–100%, respectively), and can be applied universally to different BCR types. Moreover, the clinical outcomes were comparable between the carrier and non-carrier embryo groups.

Interpretation

This study demonstrates the effectiveness of preimplantation genetic genome-wide SNP-genotyping and haplotyping method, resulting in the delivery of more babies with a normal karyotype.

Funding

This study was funded by the National Key Research and Development Program of China (2022YFC2703200, 2021YFC2700600, 2021YFC2700500), National Natural Science Foundation of China (82201807, 82171639, 82071717). Shanghai Science and Technology Innovation Action Plan Program (18411953800), and the Municipal Human Resources Development Program for Outstanding Young Talents in Medical and Health Sciences in Shanghai (2022YQ075).

Research in context.

Evidence before this study

Balanced chromosomal rearrangements (BCRs) can result in infertility, recurrent miscarriages, and affected offspring. Preimplantation genetic testing for structural rearrangements (PGT-SR) has been widely applied in BCR carriers to select euploid embryos for transfer. However, a limitation of routine testing techniques is that they cannot distinguish embryos with normal karyotypes from those with BCR karyotypes, resulting in passing on of the BCRs to their offspring. Our team and other research groups’ previous studies verified that genome-wide genotyping and haplotype analysis can accurately detect chromosomal numerical abnormalities and structural rearrangements simultaneously in one test. The methodology has been initially established. However, these studies included limited sample sizes, and its value in routine clinical settings has not been extensively assessed in a large unselected cohort of BCR carriers. Additionally, whether the technology can improve clinical outcomes, such as the birth of more babies with a normal karyotype, remains unknown, and the comparison of clinical pregnancy outcomes between the carrier and non-carrier embryo groups has not been reported yet.

Added value of this study

In this study, we conducted a prospective multicenter study across 12 academic fertility centers within China, with 1298 BCR carrier couples enrolled between April 2018 and March 2023 (final follow-up September 2023). We assessed the clinical characteristics of a group of individuals carrying BCRs, including hormone level and fertility, and investigated the proportion of inherited BCRs was near 70%. Additionally, we detected the proportion of euploid embryos among different types of rearrangements and further found that the distribution of carrier and non-carrier embryos was similar in different subgroups. By the end of the study, a total of 1030 non-carrier and 439 carrier embryos were transferred, 817 healthy babies were delivered cumulatively. Our results demonstrate that SNP-haplotyping method is highly accurate (95% CI: 98.59%–100%) and can be applied universally to different types of BCRs. An increased number of offspring (73.07%, 597/817) with normal karyotypes have been delivered, avoiding inheriting parental BCRs. Moreover, we revealed the clinical pregnancy outcomes between the carrier and non-carrier embryo groups was comparable.

Implications of all the available evidence

Evidence to date demonstrates the feasibility of preimplantation genetic haplotyping approach in detecting chromosomal imbalances as well as distinguishing BCR carrier embryos from non-carrier embryos in one test, universally across different types of BCR carriers. This method provides a strategy for BCR carriers to avoid the transmission of BCRs to their offspring, reducing the same associated risks of infertility when reaching reproductive age. Overall, this allowed 69.95% (852/1218) of PGT-SR cycles with euploid embryos to selectively transfer embryos without BCRs. All these results strongly indicate for the use of genetic counseling of carrier couples in clinical practice.

Introduction

Balanced chromosomal rearrangements (BCRs) can result in infertility, recurrent miscarriages, and abnormal offspring with congenital malformations owing to the unbalanced gametes formed during meiosis.^1,2 Most carriers of BCRs are usually phenotypically normal, yet a small subgroup of carriers are at risk for neurodevelopmental or other conditions, which may be caused by the disruption of specific functional gene structures or topologically associated domains (TADs) related to breakpoints.3, 4, 5 Balanced translocations are the most common type of chromosomal rearrangements followed by inversions, and a large cohort study in Denmark⁶ showed that the prevalence of balanced translocations is approximately 2.66/1000 in 34,910 newborns. This percentage can increase up to 4.08% in couples with recurrent miscarriages.⁷ BCRs seriously affect human reproductive and mental health.

Preimplantation genetic testing (PGT) was developed to select unaffected embryos before they were transferred back to the uterus.⁸ Over the past several decades, genetic techniques have been developed for the diagnosis of a wide range of indications, including single-gene diseases, structural chromosomal rearrangements, and chromosomal aneuploidy.⁹ Preimplantation genetic testing for structural rearrangement (PGT-SR) is implemented to screen genetically euploid embryos without chromosomal imbalances during in vitro fertilization (IVF) treatment cycles. PGT-SR was initially performed using fluorescent in situ hybridization (FISH) to test for unbalanced chromosomal aberrations, but some randomized controlled trials revealed that the FISH technology was inefficient, having no effect on the delivery rate for the limited number of chromosomes detected.10, 11, 12 With the refinement of techniques for blastocyst culture, biopsy and the growing use of comprehensive chromosome screening (CCS) techniques that can concurrently detect chromosomal imbalances and aneuploidy, including techniques such as array-comparative genomic hybridization (CGH), single nucleotide polymorphism (SNP)-array, and next generation sequencing (NGS), clinical delivery rate for BCR carriers have been greatly improved.13, 14, 15, 16 However, a key limitation of commonly used microarray and low-pass NGS in clinical labs is that they cannot distinguish embryos with normal karyotypes from those with BCR karyotypes. And many couples would like to have the opportunity of selecting a normal rather than balanced rearranged embryo for transfer, so that their children would not have the same fertility troubles when they reach reproductive age.

In recent years, a few methods have been reported to detect chromosomal imbalances as well as distinguishing BCR carrier embryos from non-carrier embryos.17, 18, 19, 20, 21 Some of these approaches usually initially identify rearrangement breakpoints by massively parallel sequencing or long-range sequencing, after which the embryo testing was performed. These methods are technically complex and also require breakpoint identification of rearrangements at nucleotide-level resolution, whereby errors may arise when the breakpoints are located in highly repetitive or complex genomic regions.^22,23 In contrast, our team^24,25 and other research groups’ studies,^26,27 verified that genome-wide genotyping and haplotype phasing analysis can accurately detect chromosomal numerical abnormalities and structural rearrangements simultaneously in one test and is universal for different types of BCR carriers, avoiding the need for the development of patient specific testing. The methodology has been initially established in previous work. However, these studies included limited sample sizes, and its value in routine clinical settings has not been extensively assessed in a large unselected cohort of BCR carriers. Additionally, whether the technology can improve clinical outcomes, such as the birth of more babies with a normal karyotype, remains unknown, and the comparison of clinical pregnancy outcomes between the carrier and non-carrier embryo groups has not been reported yet.

In this study, to further validate the effectiveness of preimplantation genetic haplotyping approach in clinical practice, we conducted a prospective multicenter study across a number of academic fertility centers within China, with 1298 BCR carrier couples between April 2018 and March 2023 (final follow-up September 2023). PGT-SR was performed using microarray-based genome-wide SNP genotyping and haplotype analysis. We assessed the clinical characteristics of a group of individuals carrying BCRs, including hormone level and fertility, and investigated the proportion of inherited BCRs. We detected the proportion of euploid embryos among different types of rearrangements and analyzed the distribution of carrier and non-carrier embryos within various subgroups. Moreover, we evaluated the accuracy of our method by comparing the karyotypes obtained from amniocentesis or umbilical cord blood with the embryo results, and assessed the proportion of newborns with normal karyotype among all newborns. Additionally, we compared the clinical pregnancy outcomes of frozen-thawed embryo transfer (FET) between the carrier and non-carrier embryo groups. The results demonstrate that SNP-haplotyping method is highly effective (95% CI: 98.59%–100%) and can be applied universally to different types of BCRs. Furthermore, the clinical outcomes of BCR carriers undergoing PGT-SR can be improved, more infants with a normal karyotype are born.

Methods

Study design

We conducted a prospective multicenter clinical study involving couples undergoing PGT, in which one partner was a BCR carrier, in order to validate the effectiveness of genetic haplotype phasing methods for detecting embryos prior to transfer.²⁸ Karyotyping of the carriers' parents was also recommended to determine whether the BCRs were inherited or de novo. The study was approved by the Ethics Committee for Human Subject Research of the Obstetrics and Gynecology Hospital of Fudan University, and other participating clinical centers in China (2018-22-X1). All the enrolled couples provided written informed consent before participation. A total of 60 months were required to complete the study after starting up (30 months' enrollment period, 30 months’ treatment, and follow up period).

Participant eligibility

Inclusion criteria

All the subjects included in this study were karyotyped from peripheral blood samples. The inclusion criteria for this study were that one spouse was a BCR carrier, and was preparing for PGT-SR treatment. BCRs included chromosomal reciprocal translocations, Robertsonian translocations, pathogenic chromosomal inversions, and insertion translocations. In addition, the women should have at least one D5/6 blastocyst for biopsy.

Exclusion criteria

Participants were not eligible for study participation if both the couple were BCR carriers, or one of the spouses was a BCR carrier and the couple was at high risk for transmitting a genetic disease. Patients were also excluded if women had congenital or acquired uterine malformations, such as a uterine congenital malformation (uterus unicornate, bicornate, or duplex); untreated uterine septum, adenomyosis, submucous myoma, or endometrial polyp(s); or with history of intrauterine adhesions, or medical conditions that contraindicated assisted reproductive technology or pregnancy, such as poorly controlled Type I or Type II diabetes; undiagnosed liver disease or dysfunction (based on serum liver enzyme testing); renal disease or abnormal serum renal function; significant anemia; history of deep venous thrombosis, pulmonary embolus, or cerebrovascular accident; uncontrolled hypertension, known symptomatic heart disease; history of or suspected cervical carcinoma, endometrial carcinoma, or breast carcinoma; undiagnosed vaginal bleeding. In addition, subjects using donor semen/oocytes or carrying polymorphic chromosomal inversions were also excluded.

Blastocyst culture, biopsy and amplification

The oocytes were inseminated by intracytoplasmic sperm microinjection approximately 4–6 h after follicular aspiration. Embryos were cultured in sequential medium with 5% carbon dioxide in air. Blastomere formation (cleavage rate) was observed 72 h after fertilization, and high-quality cleaving embryos were cultured to the blastocyst stage.

For embryos at the blastocyst stage, three to five trophectoderm cells were removed from each blastocyst on Day 5 or 6 post insemination. The biopsied cells were placed into PCR tubes with an alkaline denaturation buffer or phosphate buffer solution as previously described.²⁴ Whole genome amplification (WGA) was performed using multiple displacement amplification (MDA). Isothermal DNA amplification with phi 29 DNA polymerase was performed according to the manufacturer's protocol (Repli-g single cell kit, QIAGEN GmbH, Hilden, Germany). The isothermal amplification was performed at 30 °C for 8 h and the reaction was stopped by incubation at 65 °C for 3 min. Then the products would be stored at −20 °C.

Chromosome copy number analysis and haplotype phasing

The WGA products and genomic DNA of carrier couples, and the carriers' relevant parent if available, were processed by high throughput SNP microarray (Human Karyomapping-12; Asian Screening Array-24 v1.0; Illumina, San Diego, CA, USA) according to the manufacturer's instructions, and then scanned using the iScan Bead Array Reader. The microarray used should ideally include a greater number of high heterozygosity SNP loci, the haplotype phasing process used in this study has been described previously.^24,25 Briefly, SNP calling was performed and the informative SNPs was defined, then haplotypes across the genome in carrier couples and all embryos were phased with a reference. Either one parent or an unbalanced embryo could be used for haplotype phasing. The genome-wide recombination events can be identified accurately. Based on haplotype phasing principles, it was programmed in Practical Extraction and Reporting Language (Perl). For the analysis of aneuploidy or unbalanced rearrangements, the SNP B allele frequency and log R ratio was processed, and the algorithm was based on cnvPartition as reported previously.^29,30 In this research, the embryos with more than 80% mosaicism of whole or segmental chromosomes were classified as aneuploidy, and those with moderate mosaicism less than 50% could be considered for transfer when no euploid was available.

To detect balanced rearrangements, the regions of 2 Mb flanking the rearrangement breakpoints were analyzed emphatically. Relatively accurate breakpoint positions are usually identified by chromosomal loss or gain of unbalanced embryos; here, the breakpoints from the peripheral blood karyotype were used when no unbalanced embryo was identified, then the haplotype region used for linkage analysis extended to 5–10 Mb according to the resolution of the karyotyping. When the two haplotypes in the breakpoint regions were linked to derivative chromosomes, the embryo was diagnosed as an embryo with BCRs karyotype. On the other hand, when the two haplotypes in the breakpoint regions were linked to normal chromosomes, the embryo was diagnosed as an embryo with a normal karyotype. In addition, for about half of the embryos diagnosed with euploid or moderate mosaicism by SNP microarray in our study, the results would be validated by another NGS method (DA8600 platform, Basecare, China). Sequencing data were processed using Ion Reporter™ Software (Thermo Fisher Scientific, USA).

Blastocyst selection and transfer

Blastocysts were vitrified after biopsy and sealed using high-security vitrification straws. Genetic counseling regarding their PGT-SR results would be offered to all couples, after which the order of embryo transfer was then determined by the karyotype results and morphology of embryos. After warming and dilution, blastocysts were cultured in the medium for 1–2 h. Blastocysts with normal/balanced karyotypes were transferred through a catheter guided by transabdominal ultrasound. Single blastocyst thawing and transfer to each patient with well-cryopreserved embryos was recommended. Human chorionic gonadotropin hormone and ultrasound examinations were used to confirm normal intrauterine gestation after the transfer.

Pregnancy evaluation and follow-up

Serum quantitative human chorionic gonadotropin (hcg) will be measured to determine pregnancy 2 weeks after embryo transfer. If a biochemical pregnancy has been achieved, transvaginal ultrasound scan will be performed 28–35 days after embryo transfer to evaluate for clinical pregnancy. If a clinical pregnancy has been achieved, ultrasound scan will be repeated at 11 weeks gestation to confirm ongoing pregnancy. The subsequent follow-ups will be implemented every two or three months and continue until termination or delivery.

At 17–24 weeks gestation, the subjects will be followed up for prenatal diagnosis and the karyotyping results of amniotic fluid if available. If the subject unfortunately has a miscarriage during the follow-up, the karyotyping or CMA results of the embryonic tissue should be obtained if available. If the subject did not undergo prenatal diagnosis due to concerns about the risks of amniocentesis surgery, the subject will be suggested to collect neonatal cord blood at the time of delivery for karyotype analysis. At delivery, the delivery information (including gestational age, delivery mode, and/or delivery complications), neonatal information (including gender, birth weight, birth defects) will be obtained by reviewing obstetric and neonatal medical records.

Study outcomes

The primary outcome was the accuracy of embryo PGT results after frozen-thawed blastocyst transfer, which was confirmed via cytogenetic analysis of amniotic fluid cells in the second trimester of pregnancy or umbilical cord blood at birth. Another primary outcome was the proportion of newborns with a normal karyotype among all newborns. The pre-specified secondary outcomes included biochemical pregnancy, clinical pregnancy, pregnancy loss, neonatal congenital anomalies. Biochemical pregnancy was defined as a human chorionic gonadotropin level >10 mIU per milliliter, as measured at 14 days after embryo transfer. Clinical pregnancy was defined as the presence of a gestational sac in the uterine cavity at 35 days after embryo transfer, as detected through ultrasonography. Pregnancy loss was defined as pregnancies that eventuate in a spontaneous abortion or therapeutic abortion that occurred throughout pregnancy. A live birth was defined as delivery of any viable infant at 28 weeks or more of gestation. All pregnancy and neonatal outcomes were obtained through review of medical records. In addition, the semen quality between the spouse of the female carrier group and the male carrier group, and the ovarian reserve between female carrier and normal female, including antral follicle count and Anti-Müllerian hormone, was compared.

Statistical analysis

According to the preliminary experiment, we calculated the sample size based on the estimated sensitivity and the specificity was 0.95, and we set the allowable error (L) as 0.05 and α as 0.01. The prevalence of embryos with normal karyotype is 0.5 in all euploid embryos, and we estimated that the cumulative live birth rate would be about 50% after PGT treatment.31, 32, 33 Assuming a 50% rate of withdrawal for invasive prenatal diagnosis, the overall calculated number of participants required was 1265.

We included outcome data from the PGT controlled ovarian hyperstimulation (COH) cycle and follow-up evaluations. Continuous data were expressed as mean (SD), and between-group differences were tested by T test or Wilcoxon rank sum test according to the non-normality of the variables. Categorical variables data were represented as percentages, differences in these variables were assessed by χ² analysis, with Fisher's exact test for expected frequencies less than five, the relative risk and 95% CIs were calculated. Subgroup analysis was performed according to the chromosomal rearrangement types. Statistical significance was set at P < 0.05. The analyses were performed using the SPSS Statistics software version 18.0 (IBM SPSS, Armonk, NY, USA). This research was registered at the Chinese Clinical Trial Registry (number ChiCTR-1800015863).

Role of funders

The funders played no role in the design of the research, the collection of data, the analysis of data, the interpretation of results, the writing of the paper and the decision to submit the manuscript for publication.

Results

Participants and baseline characteristics

Patient enrollment began in April 2018, a total of 60 months were required to complete the study after starting up, 30 months' enrollment period, 30 months’ treatment, and follow up period. A total of 1298 carrier couples were enrolled, 42 were ineligible due to a failure in haplotype phasing for lacking an available reference. Finally, 1256 (640 female and 616 male carriers) were included, comprising 1017 with reciprocal translocations, 206 with Robertsonian translocations, 28 with inversions and 5 with insertion translocations (Fig. 1). Nearly 70% of BCRs were inherited from parents and the others were de novo. The baseline characteristics of all couples are shown in Table 1, the semen quality of the spouse in the female carrier group was superior to that of the male carrier group. However, the other semen characteristics were comparable between the two groups. In addition, the ovarian reserve between female carriers and normal females, including antral follicle count and Anti-Müllerian hormone, was similar.

Fig. 1.

Study enrollment and outcomes. Legend: A total of 1298 carrier couples were enrolled, 42 were ineligible due to a failure in haplotype phasing for lacking an available reference. Finally, 1256 were included, including 1017 with reciprocal translocations, 206 with Robertsonian translocations, 28 with inversions, and 5 with insertion translocations. Blastocyst biopsies were performed on 7867 blastocysts derived from 1603 PGT-SR cycles, a total of 2676 were euploid and further analyzed by genome-wide haplotyping, in which 1367 were non-carrier embryos and 1307 were carrier embryos. The percentage of carrier and non-carrier embryos in the different rearrangement type groups showed no significant difference (P > 0.05). In total, 1030 non-carrier and 439 carrier transferred embryos were transferred, resulting in 921 clinical pregnancies. Between the two groups, there was no significant difference in biochemical pregnancy rate per transfer, clinical pregnancy and miscarriage rates. To date, 817 live infants have been born. A total of 415 pregnancies underwent prenatal amniocentesis or newborn umbilical cord blood karyotype analysis, proving that the sensitivity and specificity of our proposed method to be 100% (95% CI: 98.59%–100% and 95% CI: 96.63%–100%, respectively). Of all the infants, 597 babies (73.07%, 597/817) are with a normal karyotype, nearly three times as many as carrier babies (26.93%, 220/817). COH = controlled ovarian hyperstimulation.

Table 1.

Characteristics of the participants at baseline.^a

Characteristic	Female BCRs	Male BCRs	P
	Carrier group (n = 640)	Carrier group (n = 616)
Age--yr
Female	31.0 ± 3.6	30.5 ± 3.8	NS
Male	32.1 ± 3.8	31.5 ± 3.8	NS
Body-mass indexb	23.0 ± 3.3	22.6 ± 3.2	NS
Family history
Inherited from parents	72.2%	71.1%	NS
De novo aberration	27.8%	28.9%
Ultrasonographic findings
Antral follicle count in both ovaries	16.8 ± 6.9	16.9 ± 6.5	NS
Endometrial thickness (mm)	8.7 ± 2.0	8.6 ± 2.2	NS
Sperm parameters
Semen volume (ml)	3.5 ± 1.5	3.6 ± 2.9	NS
Sperm concentration (×106/ml)	57.6 ± 41.2	42.6 ± 37.5	<0.001
Sperm progressive motility (%)	43.2 ± 17.4	34.2 ± 19.5	<0.001
Total sperm (106/ejaculate)	196.3 ± 157.2	150.8 ± 168.1	<0.001
Total progressively motile sperm (106/ejaculate)	90.98 ± 86.1	63.68 ± 89.6	<0.001
Laboratory tests
Follicle-stimulating hormone (IU/liter)	6.4 ± 1.9	6.4 ± 2.0	NS
Luteinizing hormone (IU/liter)	5.3 ± 3.3	5.5 ± 3.6	NS
Estradiol (pg/ml)	42.9 ± 25.3	42.5 ± 26.1	NS
Total testosterone (ng/ml)	0.4 ± 0.2	0.4 ± 0.2	NS
Prolactin (ng/ml)	17.7 ± 10.3	17.7 ± 9.7	NS
Anti-müllerian hormone (ng/ml)	4.5 ± 3.0	4.8 ± 3.3	NS

Statistical significance was set at P < 0.05. The italics in table indicate a statistical difference between the two groups.

^aPlus-minus values are means ± SD.

^bThe body-mass index is the weight in kilograms divided by the square of the height in meters.

Primary outcomes

The accuracy of embryo PGT results

Chromosomal aneuploidy screening results. In this study, blastocyst biopsy was performed on 7867 blastocysts derived from 1603 PGT-SR cycles (264 couples undergoing two or more cycles, Fig. 1). Comprehensive chromosome screening with microarray was conducted and showed that 2892 embryos (37.32%, 2892/7750) were rearrangement-related imbalance, 1195 (15.42%, 1195/7750) were de novo additional aneuploidies (including 401 chromosome mosaicism) and the other 987 (12.74%, 987/7750) had both imbalance and additional aneuploidies. The rate of abnormal embryos in reciprocal carriers (69.06%, 4420/6400) was higher than those in Robertsonian translocation (48.21%, 552/1145), inversion carriers (52.74%, 77/146) and insertion translocation carriers (42.37%, 25/59) (Table 2 and Fig. 2A).

Table 2.

The PGT-SR results of tested blastocysts.

Rearrangement type	Unbalanced rearrangements	De novo aneuploidiesa	Complex abnormalitiesb	Non-carrier embryos	Carrier embryos	Total embryos
Reciprocal translocation	2657 (41.52%)	881 (13.77%)	882 (13.78%)	1029 (16.08%)	951 (14.86%)	6400
Robertsonian translocation	209 (18.25%)	249 (21.75%)	94 (8.21%)	289 (25.24%)	304 (26.55%)	1145
Inversion	17 (11.64%)	56 (38.36%)	4 (2.74%)	35 (23.97%)	34 (23.29%)	146
Insertion translocationc	9 (15.25%)	9 (15.25%)	7 (11.86%)	14 (23.73%)	20 (33.90%)	59
Total	2892 (37.32%)	1195 (15.42%)	987 (12.74%)	1367 (17.64%)	1309 (16.89%)	7750

^aThese de novo aneuploidies included 401 mosaic embryos with whole or segmental chromosomes.

^bThe complex abnormalities result was defined as a combination of unbalanced rearrangements and one or more of the following features: monosomy, trisomy, segmental aneuploidy, or chromosomal mosaic.

^cFor the small sample size in insert translocation subgroup, bias of carrier and non-carrier distribution was inevitable compared to the theoretical 50:50.

Fig. 2.

The distribution of detected embryos in PGT-SR cycles. Legend: A represents the distribution of PGT-SR results of all tested blastocysts, including unbalanced rearrangements, de novo aneuploidies, complex abnormalities, non-carrier embryos, and carrier embryos; B represents the distribution of euploid embryos in PGT-SR cycles; C represents the distribution of non-carrier embryos in PGT-SR cycles; D represents the distribution of carrier embryos in PGT-SR cycles; PGT-SR = preimplantation genetic testing for structural rearrangements. In this study, only 5 insertional translocation carriers were included, the sample size was small and the results regarding the distribution of detected embryos in PGT-SR cycles may be biased, therefore this date was not shown in Fig. 2B–D.

An average of 24.02% (385/1603) of COH cycles failed to have a transfer because all of the embryos biopsied were either unbalanced or had additional aneuploidies, the rate was 26.11%, 15.27%, 10.34%, 16.67% in reciprocal carriers, Robertsonian translocation, inversion carriers and insertion translocation carriers, respectively. 16% (201/1256) of patients ultimately did not obtain transferable embryos despite undergoing multiple PGT-SR cycles. Table 3 shows the distributions of euploid embryos in COH cycles (Fig. 2B) and patients. In addition, among all these detected embryos, whole genome uniparental disomy (UPD) and triploid accounted for nearly 0.36% (28/7750) and 0.66% (51/7750), respectively. In addition, eight cryptic reciprocal translocations in which both the translocation segments were beyond the resolution of karyotype analysis and five insertion translocations were recruited (Supplemental Figs. S1 and S2 and Supplemental Tables S1 and S2). A total of 1559 embryos diagnosed with euploid or moderate mosaicism by SNP microarray platform in our study were validated by another NGS method, with the results showing near 80% (1196/1559) of embryos being consistent. In the remaining 20% of embryos, either the proportion of mosaicism (10.01%, 156/1559) or the mosaic chromosomes (13.27%, 207/1559) differed (Supplemental Fig.S3).

Table 3.

The distribution of embryos in PGT-SR cycles and patients.

Embryos subgroups	No embryo	One embryo	Two embryos	Three or more embryos	Total
Euploid embryos
The number of PGT-SR cycles
Total	385 (24.02%)	501 (31.25%)	318 (19.84%)	399 (24.89%)	1603
Reciprocal translocation	341 (26.11%)	421 (32.24%)	255 (19.53%)	289 (22.13%)	1306
Robertsonian translocation	40 (15.27%)	70 (26.73%)	55 (20.99%)	97 (37.02%)	262
Inversion	3 (10.34%)	7 (24.14%)	7 (24.14%)	12 (41.38%)	29
Insertion translocation	1 (16.67%)	3 (50.00%)	1 (16.67%)	1 (16.67%)	6
The numbers of patients
Total	201 (16.00%)	344 (27.39%)	282 (22.45%)	429 (34.16%)	1256
Reciprocal translocation	181 (17.80%)	297 (29.20%)	225 (22.12%)	314 (30.88%)	1017
Robertsonian translocation	16 (7.77%)	39 (18.93%)	49 (23.79%)	102 (49.51%)	206
Inversion	3 (10.71%)	6 (21.43%)	7 (25.00%)	12 (42.86%)	28
Insertion translocation	1 (20.00%)	2 (40.00%)	1 (20.00%)	1 (20.00%)	5
Non-carrier embryos
The number of PGT-SR cycles
Total	751 (46.85%)	499 (31.13%)	243 (15.16%)	110 (6.86%)	1603
Reciprocal translocation	634 (48.55%)	409 (31.32%)	189 (14.47%)	74 (5.67%)	1306
Robertsonian translocation	104 (39.69%)	79 (30.15%)	47 (17.94%)	32 (12.21%)	262
Inversion	10 (34.48%)	10 (34.49%)	6 (20.69%)	3 (10.34%)	29
Insertion translocation	3 (50.00%)	1 (16.67%)	1 (16.67%)	1 (16.67%)	6
The numbers of patients
Total	471 (37.50%)	420 (33.44%)	227 (18.07%)	138 (10.99%)	1256
Reciprocal translocation	401 (39.43%)	343 (33.73%)	178 (17.50%)	95 (9.34%)	1017
Robertsonian translocation	58 (28.16%)	67 (32.52%)	43 (20.87%)	38 (18.45%)	206
Inversion	10 (35.71%)	9 (32.14%)	5 (17.86%)	4 (14.29%)	28
Insertion translocation	2 (40.00%)	1 (20.00%)	1 (20.00%)	1 (20.00%)	5
Carrier embryos
The number of PGT-SR cycles
Total	754 (47.04%)	529 (33.00%)	223 (13.91%)	97 (6.05%)	1603
Reciprocal translocation	659 (50.46%)	422 (32.31%)	163 (12.48%)	62 (4.75%)	1306
Robertsonian translocation	84 (32.06%)	95 (36.26%)	52 (19.85%)	31 (11.83%)	262
Inversion	10 (34.48%)	9 (31.03%)	7 (24.14%)	3 (10.34%)	29
Insertion translocation	1 (16.67%)	3 (50.00%)	1 (16.67%)	1 (16.67%)	6
The numbers of patients
Total	492 (39.17%)	420 (33.44%)	216 (17.20%)	128 (10.19%)	1256
Reciprocal translocation	432 (42.48%)	343 (33.72%)	156 (15.34%)	86 (8.46%)	1017
Robertsonian translocation	50 (24.27%)	66 (32.04%)	53 (25.73%)	37 (17.96%)	206
Inversion	9 (32.14%)	9 (32.14%)	7 (25.0%)	3 (10.71%)	28
Insertion translocation	1 (20.00%)	2 (40.00%)	0 (0.00%)	2 (40.00%)	5

Balanced chromosomal rearrangement analysis results. Among all the 7867 biopsied blastocysts, a total of 2676 were euploid and further analyzed by genome-wide haplotyping, in which 1367 were non-carrier embryos with normal karyotypes and 1307 were carrier embryos with rearranged karyotypes. The percentage of carrier and non-carrier embryos in the different rearrangement type groups showed no significant difference (P > 0.05), the distribution was unbiased compared to the theoretical 50:50 (Table 2). As shown, more than half of the cycles generated embryos with normal karyotype, the proportion was 51.45% (672/1306), 60.31% (158/262), 65.52% (19/29), and 50% (3/6) in reciprocal translocation, Robertsonian translocation, inversion and insertion translocation carrier couples (Table 3 and Fig. 2C), respectively. In addition, more than one in five couples (21.26%, 267/1256) underwent multiple ovulation cycles and 62.5% of couples (785/1256) acquired non-carrier embryos cumulatively, with a rate of up to 71.84% (148/206) in Robertsonian translocation carrier couples. Finally, 74.40% (785/1055) of couples with euploid embryos got the chance to transfer non-carrier embryos avoiding transmission of BCRs. The distributions of carrier embryos in COH cycles and patients are shown in Table 3 and Fig. 2D. In addition, 61 blastocysts from 42 patients failed to further distinguish balanced rearrangement because of the lack of an available reference sample, either an unbalanced embryo or one of carrier's parent. According to our data, 10.72% (140/1306) cycles in reciprocal translocation carrier couples had no available unbalanced embryos, and the rate was up to 37.79% (99/262) and 67.86% (19/28) in Robertsonian translocation and inversion carrier couples respectively (Supplemental Table S3). For the inherited rearrangement carriers, nearly all the embryos were successfully diagnosed in this study when we obtained the sample of one parent.

Clinical outcomes after frozen blastocyst transfer. For the patients (39.33%, 494/1256) with both carrier and non-carrier embryos, all of them chose to transfer non-carrier embryos. In addition, for the patients (21.50%, 270/1256) with only carrier embryos, nearly half of them chose to undergo embryo transfer directly, while the rest would choose to start another ovulation cycle. In total, 1030 non-carrier and 439 carrier embryos were transferred, resulting in 921 clinical pregnancies. To date, 817 live infants have been born from these pregnancies. The cumulative pregnancy for all cases was 65.05% (817/1256). A total of 415 pregnancies underwent prenatal amniocentesis or newborn umbilical cord blood karyotype analysis, all the fetal karyotype results were all concordant with PGT results (carrier or non-carrier), proving that the sensitivity and specificity of SNP-haplotyping method to be 100% (95% CI: 98.34%–100% and 95% CI: 96.63%–100%, respectively) (Fig. 1 and Supplemental Table S4).

The proportion of newborns with normal karyotype

Of all the infants, 597 babies (73.07%, 597/817) are with a normal karyotype, and 220 babies (26.93%, 220/817) are with a BCR karyotype, the number of normal babies is nearly three times that of carrier babies. Currently, only about 3% of patients have not yet undergone transfer or have given up the transfer process because they only have carrier embryos.

Secondary outcomes

Between carrier and non-carrier embryo groups, there was no significant difference in biochemical pregnancy rate per transfer (70.77% and 66.05% in the non-carrier and carrier groups, respectively, P = 0.07), clinical pregnancy (64.27% and 59.00% respectively, P = 0.06) and miscarriage rates (8.01% and 10.42% respectively, P = 0.44), as shown in Table 4 and Fig. 1. In addition, we observed a higher clinical pregnancy rate among non-carriers (64.27%) compared to carriers (59.00%) (P = 0.06). However, the difference did not reach statistical significance. Moreover, this research showed the proportion of embryos with moderate mosaicism (30–50%) was 10.47%. Among all these transferred embryos, 39 mosaic embryos were included, our results indicated that the clinical outcomes were comparable with those of completely euploid embryos (Supplemental Tables S5 and S6). Additionally, the incidence of congenital anomalies of gestation period and newborn did not differ significantly between the two groups (Supplemental Table S7).

Table 4.

The FET outcomes of carrier and non-carrier embryos.

Embryo subgroups	The number	Hcg negative	Biochemical pregnancy	Clinical pregnancy	Miscarriage
Non-carrier embryos
Reciprocal translocation	819	224	595	530	45
Robertsonian translocation	195	70	125	124	7
Other types	16	7	9	8	1
Total	1030	301 (29.22%)	729 (70.77%)	662 (64.27%)	53 (8.01%)
Carrier embryos
Reciprocal translocation	347	119	228	205	24
Robertsonian translocation	80	24	56	49	3
Other types	12	6	6	5	0
Total	439	149 (33.94%)	290 (66.05%)	259 (59.00%)	27 (10.42%)

Discussion

BCRs seriously affect human reproductive outcomes. Rearranged derivative chromosomes generate a high proportion of abnormal gametes during meiosis, resulting in miscarriage or infertility as a consequence of the deletion/duplication of chromosomal segments. Some rearrangements may disturb important functional genes and cause genetic disorders.3, 4, 5 Moreover, the clinical effect of structural rearrangements also incorporates regulatory domains of the human genome.34, 35, 36 In addition, a 27% morbidity risk of neurodevelopmental and/or neuropsychiatric disorders was observed in de novo BCR carriers, which was significantly higher than that in a matched control group.³⁷ Furthermore, when reciprocal translocation involves the X chromosome, the phenotype of balanced carriers is unpredictable owing to random inactivation of the X chromosome during the early stages of embryo development,^38,39 three carriers of this specific type were included in our study. Most couples seek help from assisted reproductive technologies in the hope of achieving a health pregnancy. For the PGT-SR treatment of BCR couples, one goal of the technique is to help couples achieve a successful healthy pregnancy, with the other benefit aimed towards avoiding vertical inheritance of BCRs to the next generations.

In recent years, a few techniques have been developed to distinguish between BCR carrier and non-carrier embryos.17, 18, 19, 20, 21 However, some methods cannot simultaneously detect chromosome numerical and structural anomalies, separate procedures are required. In addition, these methods require rearrangements breakpoint identification to a single base, which is complicated in methodology and may not perform well when the breakpoint is located in highly repetitive regions of the genome.^22,23 Robertsonian translocations may be not applicable, as the breakpoints are located in the centromere regions. In contrast, our previous researches and some teams’ studies verified that SNP array-based genome-wide haplotype phasing method could accurately detect chromosomal aneuploidies, imbalances and balanced rearrangements comprehensively in one test and was universal for different BCRs without individual patient design.24, 25, 26

To further validate the efficacy of preimplantation genetic haplotyping method and potential improvements in pregnancy outcomes, a prospective multicenter study, containing 1298 carrier couples undergoing PGT were recruited and approximately 8000 embryos were analyzed. We found that nearly half of the embryos analyzed were abnormal associated with rearrangements, approximately a quarter had de novo chromosomal abnormalities unrelated to rearrangements and the remaining were normal/balanced euploids. Among these euploid embryos, the proportions of carrier and non-carrier embryos were equivalent, showing no statistical difference in the different rearrangement subgroups. In this study, 53.15% of PGT cycles (852/1603) and 62.50% of couples (785/1256) generated non-carrier embryos, and 62.93% of these couples (494/785) obtained both carrier and non-carrier embryos. These couples were more likely to benefit from carrier embryo detection and clinical pregnancy outcomes might be changed. In this cohort, all these couples give priority to transfer non-carrier embryos. For the couples who only had carrier embryos or had non-carrier embryos but were not pregnant after transfer, they underwent either carrier embryo transfer or another ovulation cycle. Between the carrier and non-carrier embryo groups, no significant differences were observed in pregnancy outcomes. Furthermore, we noted a higher clinical pregnancy rate among non-carriers (64.27%) compared to carriers (59.00%) (P = 0.06), though the difference did not reach statistical significance. Perhaps, additional samples are needed to confirm this difference, and further investigation is required to understand its underlying mechanism. Several women were concerned about the risk of miscarriage or infection from amniocentesis, therefore only 415 pregnancies underwent prenatal diagnosis amniocentesis or newborn umbilical cord blood karyotype analysis. The PGT-SR results were concordant with the karyotyping results of the fetus after transfer, proving that the sensitivity and specificity of the method were both 100%. An increased number of offspring (73.07%, 597/817) with normal karyotypes have been delivered, avoiding inheriting parental BCRs, statistically different from 50:50 (P < 0.001). In addition, our research indicated that the rates of euploid embryos were 30.94%, 51.79%, and 47.26% in couples carrying reciprocal translocations, Robertsonian translocations and inversion carriers, respectively. These rates were significantly higher than the theoretical rates and consistent with previously reported findings.40, 41, 42, 43 Therefore, these results strongly indicate for the use of genetic counseling of carrier couples in clinical practice. What's more, embryos with moderate mosaicism were considered for transfer when no euploid was available in our study, 39 mosaic embryos were transferred finally, whereby the clinical outcomes were similar to those of completely euploid embryos and consistent with other reported studies.44, 45, 46 Increasing studies have suggested embryos displaying low range mosaicism show reproductive capabilities somewhat equivalent to uniformly euploid embryos. In addition, we compared the PGT results of SNP-array and NGS method with newborn/prenatal karyotyping, and found that there was no statistical difference owing to the limited number of mosaic embryos transferred.

This study has several strengths. Firstly, we comprehensively assessed the clinical characteristics of a group of individuals carrying BCRs, with nearly 70% of BCRs being inherited from parents. Additionally, we revealed no differences in hormone levels and antral follicle counts between female carriers and the normal female group. However, the semen quality of the spouse in the female carrier group was better than that of the male carrier group. Importantly, our results, based on the analysis of nearly eight thousand embryos, demonstrated the accuracy of the whole-genome SNP-haplotyping method, making the findings highly reliable, and the universal applicability of this method to various types of BCRs, such as reciprocal translocation, Robertsonian translocation, inversion, insertion translocation, and other complex rearrangements. Another advantage of our study is that the SNP-microarray platform used for chromosomal aneuploidy screening highlighted instances of whole genome UPD and triploid embryos, which accounted for more than 1% of all the diagnosed embryos analyzed. These embryos would result in adverse pregnancy outcomes after transfer, such as the complication of hydatidiform mole formation. In contrast, other routinely used CCS methods, including array-CGH or NGS, cannot detect these abnormalities. In addition, homologous recombination across the genome can be identified accurately, our results showed that in very few embryos (0.58%), crossover at 1 Mb distance to one of the breakpoints occurred, while crossovers in both breakpoints were not identified; this did not interfere with the results. Furthermore, some special BCR carriers were recruited in our study, including those with cryptic reciprocal translocation and insertion translocation. For cryptic reciprocal translocation carriers, the translocations were detected by both FISH and optical genome mapping.⁴⁷ While the testing of unbalanced chromosomal translocation fragments in biopsied embryonic cells is very difficult due to the small size of deletions/duplications. For these cases, prenatal diagnosis is highly recommended after embryo transfer.

Of course, one limitation of our study was that a reference was the prerequisite for haplotype phasing. In our study, a total of 42 carriers were unable to obtain their parents' karyotype results and did not have an available unbalanced embryo in PGT cycles, resulting in BCR detecting failure in their 61 euploid blastocysts. According to our results, about 10.72% and 37.79% cycles had no available unbalanced embryos in reciprocal translocation and Robertsonian translocation carrier couples respectively, and this rate was up to 67.86% in inversion carrier couples, therefore BCR detection in embryos of this subset of couples might not be conducted successfully if their grandparents' karyotype was also unknown. In general, inversions and insertional translocations presented greater detection challenges than apparently reciprocal translocations. This is because the unbalanced embryo is usually unavailable in these carriers. In contrast, for inherited BCR carriers, the detection success rate close to 100% when the DNA sample of one parent can be obtained. In our opinion, we recommend that all the carriers’ parents undergo karyotyping to confirm whether the rearrangements were inherited, especially for Robertsonian translocation, inversion and insertional translocation carriers. Additionally, the sample size of some subgroups is limited and a few patients have not yet undergone embryo transfer. One limitation is that the accuracy and the ratio of normal embryos to carrier embryos is evaluated until the end of the study and may be affected by the remaining embryos to be transferred.

In conclusion, we conducted a prospective multicenter study and demonstrated the feasibility of PGT-SR by genome-wide SNP genotyping and haplotyping method, universally across different types of BCR carriers. This method exhibits high accuracy and can improve the clinical outcomes of individuals carrying BCRs, resulting in more infants with a normal karyotype being born. This study strongly suggests a strategy for BCR carriers undergoing PGT-SR to prevent the transmission of BCRs to their offspring, thereby reducing the same associated risks of infertility when reaching reproductive age.

Contributors

CJX, Z-JC, and SZ proposed and designed the study. SZ, YG, CJX, and Z-JC were in charge of the study conduct. SZ, YG, XHW, QL, JCT, MG, XFL, JYL, XMT, HL, YS, WDH, XHT, CXL, JPW, HCL, JW, SYL, XYX, JQZ, WW, SSL, JO, QZZ, RTJ, YPZ, CMX, DL, JHY, XXS, and KWC enrolled participants. SZ and BL did the statistical analyses and prepared the tables with oversight by YG. SZ, CJX, and Z-JC drafted the manuscript. Z-JC and CJX had a primary responsibility for final content. All authors were involved in data collection, interpreted the data, provided critical input to the manuscript, and approved the final manuscript.

Data sharing statement

Investigators can request data sharing by emailing the corresponding author. A proposal with a detailed description of study objectives and a statistical analysis plan will be needed for assessment of requests. Additional materials might also be required during the process of assessment. Our steering committee established for this cohort will review and approve the request. An agreement on how to collaborate will be reached based on the overlaps and conflicts between the proposal and our ongoing efforts.

Declaration of interests

CJX reports grants from the Shanghai Science and Technology Innovation Action Plan Program during the conduct of the study. YG reports grants from the National Key Research and Development Program of China during the conduct of the study. SZ reports grants from the Shanghai Science and Technology Innovation Action Plan Program during the conduct of the study. BL reports employment with Basecare. All other authors declare no competing interests.