Introduction

PGS is an early method to detect aneuploidies in preimplantation embryos. PGS can be offered to patients with an increased risk of having a higher percentage of chromosomally abnormal embryos, improving the reproductive outcome. By transferring only chromosomally normal embryos the miscarriage rates are expected to be lower, since about 60–70 % of miscarriages in the first trimester result from chromosomal abnormalities in the fetus [1].

Some studies clearly show that twins following SET can be the result of a concurrent natural conception [2, 3] and an incidence as high as 1 in 5 twins has been reported [4].

We describe a rare case of a 36-year-old infertile patient with recurrent implantation failures. PGS was performed and an euploid blastocyst transferred. However, a spontaneous abortion occurred after 13 weeks of gestation, ascertained to be 10 at ultrasound examination. Cytogenetic analysis revealed a 45,X karyotype. DNA fingerprinting analysis of POC and the euploid embryo transferred excluded a self-relationship confirming the occurrence of a spontaneous conception in concomitance with SET.

Materials and methods

A 36-year old woman experiencing 6 years infertility with a diagnosis of severe endometriosis was referred to Humanitas Fertility Center. She had a miscarriage at 8 weeks after an intrauterine insemination (IUI), other 2 negative IUI cycles and 3 negative in vitro fertilization (IVF) cycles with the transfer of 11 cleavage stage embryos from fresh and frozen cycles performed in other centers. A normal uterine cavity was confirmed at hysteroscopy while a transvaginal ultrasound imaging showed a 18 mm small endometrial cyst of the right ovary and 12 mm recto-vaginal lesion. Despite a high follicle-stimulating hormone (FSH) level of 14 IU/L, ovarian reserve assessed to be normal with an anti-Müllerian hormone (AMH) level of 3.8 ng/ml and the existence of 10 antral follicles. Her partner sperm analysis showed 12 million total motile sperm with 2 % normal forms and 80 % vitality. After counseling and obtaining informed consent, the patient was enrolled for a first attempt intracytoplasmic sperm injection (ICSI) with PGS cycle. Controlled ovarian hyper-stimulation was started with 225 IU recombinant FSH (rFSH) in a gonadotropin-releasing hormone (GnRH) antagonist analog cycle. Recombinant human chorionic gonadotropin (rhCG) (250 mcg) was administered on day 13 of the induction cycle after a total rFSH dose of 3325 IU.

A transvaginal ultrasound guided under deep sedation oocyte retrieval was performed 36 h after the rhCG trigger and 9 metaphase II oocytes were retrieved. In 8 metaphase II oocytes ICSI was performed and placed in an incubator with an integrated time-lapse system, where the embryos were cultured individually in 25 μl microwells and monitored for 24 h. At 16–18 h post-ICSI, the 8 oocytes were assessed for the presence of 2 pronuclei and cultured until blastocyst stage. Three expanded blastocysts underwent biopsy of TE cells day 5 and 4 day 6 as previously described [5], 1 embryo degenerated after development to morula. Blastocyst quality was assessed immediately before TE biopsy [6], and blastocysts vitrified according to the protocol described by Kuwayama et al. [7]. Of the 7 biopsied and cryopreserved blastocysts, 4 were euploid. TE biopsies were analysed by qPCR based comprehensive chromosome diagnosis [8, 9]. This method has been extensively validated for chromosome copy number analysis of TE biopsies in preclinical [8, 9] and clinical studies [10, 11] and was reported to have a recognizable error rate of 0.2 % in clinical pregnancies [12].

The embryo transfer was scheduled on a spontaneous cycle. The patient was monitored from day 6 of the cycle by ultrasound and a urinary luteinizing hormone (LH) kit. On cycle day 15 an endometrial thickness of 12.2 mm with 2 follicles of 18 and 22 mm were observed with a positive LH determination. Seven days after LH surge a single euploid blastocyst (embryo number 3) was selected for transfer based on morphological score, warmed and cultured at 37 °C until transfer, performed under ultrasound guidance.

A positive hCG determination was performed 14 days after the embryo transfer (1765 IU) and a 6 weeks ultrasound showed an intrauterine gestation with an embryo with normal fetal heart rate.

A diagnosis of internal abortion was performed at 13 weeks with a 10 weeks embryo development and after the diagnosis the patients was submitted to aspiration and curettage with selection of villous materials immediately sent to our Cytogenetic Laboratory.

Placental villi were then further selected under the inverted microscope and cytogenetic analysis performed by direct preparation after 24 h incubation [13]. This is the method of choice in our laboratory because it avoids the growth of normal cells from possible endometrial contamination [1]. Parental karyotyping was performed according to standard procedures.

FISH on 1000 interphase nuclei using Vysis CEPX (DXZ1) Spectrum Green Probe (Abbott Molecular, Abbott Park, Illinois, US), specific for chromosome X centromere was conducted following the manufacturer’s recommendations.

Array-CGH study was performed on the DNA extracted from uncultured villi using the QIAamp Mini Kit (QIAGEN®, Hilden, Germany) and hybridized to the SurePrint G3 Cancer CGH + SNP 4 × 180 K microarray platform (Agilent Technologies, Santa Clara, CA, USA). Data were analysed with Agilent CytoGenomics 3.0.1.1 software.

Since the couple did not abstain from having sexual intercourses during the embryo transfer period, DNA fingerprinting analysis of excess embryonic and POC DNA was carried out to rule out the hypothesis of a spontaneous conception rather than a false PGS negative result. A set of 40 previously published SNPs with high heterozygosity and low allele frequency variation were utilized [14]. These SNPs were originally characterized against ~2070 individuals from 40 populations. The estimated probability that siblings will have identical genotypes is 0.5 raised to the power of the number of SNPs used, since all of these markers have alleles close to 50:50 in the population. This technique has been validated by Scott and colleagues [15] as the first seamless and inexpensive trophectoderm qPCR-based DNA fingerprinting technology allowing the unequivocal discrimination of sibling human embryos and foetuses. Briefly, for either the multiplex amplification PCR product residual from the TE biopsy or the purified genomic DNA of the POC, real-time PCR for allelic discrimination was performed in duplicate reactions for each of the individual 40 loci. Each reaction used the individual TaqMan SNP Genotyping Assays, TaqMan Genotyping Master Mix (Applied Biosystems Inc.), a 5 μl reaction volume, a 384-well plate, and a Quant Studio sequence detection system, as recommended by the supplier (Applied Biosystems Inc.). Informative SNP genotypes were compared between samples to determine the levels of similarity and standard deviations from the mean were calculated.

Results and discussion

The couple’s karyotype was normal. Cytogenetic analysis on 20 Q-banded metaphases obtained from the POC showed a 45,X karyotype. This result was confirmed by FISH analysis showing one signal for CEPX in 94.9 % of the nuclei. In 5.1 % of nuclei 2 signals were present, probably due to a small maternal cell contamination that is difficult to avoid in a POC sample even after placenta villi selection. The 45,X karyotype observed in the metaphases obtained after 24 h of incubation excluded a mosaic karyotype in the embryo.

Array-CGH analysis performed on DNA from uncultured villi confirmed chromosome X monosomy.

Allelic discrimination analysis of 40 SNPs of the excess DNA from all the euploid blastocysts produced during the present IVF cycle, revealed that none of them had a genotype consistent with the POC. In particular, locus dropout (no amplification of either allele) was observed in only 4 assays of embryo number 3, with an overall call rate of 97.5 % (156/160) (Table 1). All the euploid embryos showed a similarity rate to POC below 80 %, clearly consistent with a sibling genotype and a non-self-relationship (Table 2). This analysis confirmed the hypothesis of a spontaneous conception in the course of the natural frozen embryo transfer cycle.

Table 1 Genotyping data of 40 SNPs in technical duplicates used to determine genetic similarity between the product of conception and euploid IVF embryos. 1 and 2 represents the two alleles for each SNP
Full size table
Table 2 Genetic similarity of the 3 euploid embryos obtained during the IVF cycle with the genomic DNA from the aneuploid POC
Full size table

To our knowledge this the only case of PGS and euploid SET resulted in a 45,X POC as consequence of a natural pregnancy. A similar event has been reported in a large study about twin births after SET where the authors demonstrated that about 1 in 5 twins is due to a concurrent spontaneous conception [4]. Another report described a case of dizygotic twins of different-sex delivered after SET demonstrated to be the consequence of a natural fertilization [2].

Conclusions

Our case report demonstrates the occurrence of an aneuploid spontaneous conception, resulted in a miscarriage, in the course of a natural SET cycle. This case highlights the importance of avoiding intercourses and natural transfer cycles for couples with reproductive genetic risk, such as those undergoing preimplantation genetic diagnosis (PGD)/PGS cycles, in order not only to reduce the incidence of multiple pregnancies but especially to avoid the risk of aneuploid natural conceptions.

Moreover, cytogenetic analysis of the POCs and fingerprinting of the children born after PGS are strongly recommended.

Furthermore, with this study we confirm that parental DNA information is not necessary since discrimination was achieved based on a clear genotypic heterogeneity among samples, without incorporating selection of informative SNPs from parental genotype. Therefore, the additional time and expense of evaluating parental DNA could be avoided with this methodology as also previously reported [15].