Introduction

Noninvasive prenatal testing (NIPT), which has excellent sensitivity and specificity for conventional trisomy, is widely used in clinical applications1,2. With the rapid development of sequencing and bioinformatics analysis, multiple studies3,4,5,6,7,8have indicated that expanding NIPT can be used to detect sex chromosome abnormalities (SCAs), rare chromosome aneuploidies (RCAs), and microdeletion/microduplication syndromes (MMSs). However, according to many reports in the literature, the positive predictive value of expanded NIPT for RCAs is only 4–7%9. Cell-free fetal DNA originates from the placenta rather than the fetus, and confined placental mosaicism (CPM) is the main reason for false positives. In addition, RCAs may increase the risk of feto-placental disease, suggesting that they be classified as high-risk pregnancies, which would be beneficial for pregnancy management10. Trisomy rescue in early embryonic development, which may lead to uniparental disomy (UPD), may be associated with chromosomal imprinting disorders in chromosomes 6, 7, 11, 14, 15 and 2011. The American College of Medical Genetics (ACMG) and Genomics recommends that when prenatal diagnosis is detected with imprinted chromosomal abnormalities via chorionic villus sampling (CVS) or amniocentesis, it is important that the possibility of UPD be ruled out by MS-MLPA12. For chromosomes containing gene imprinting, trisomy 7 is most commonly detected by NIPT, but no UPD7 is detected by prenatal diagnosis13,14. For other chromosomes, few case reports exist15,16.

Genomic imprinting constitutes a classical epigenetic phenomenon, encompassing the transcriptional silencing of one parental gene allele, with profound implications extending to both fetal growth and placental development17. Many neurological developmental abnormalities are difficult to detect via ultrasound examination during early pregnancy. We retrospectively studied the prenatal diagnosis and pregnancy outcomes of 9 patients whose NIPT results revealed chromosomal abnormalities related to imprinting genes and explored the application of NIPT in imprinting-related diseases.

Materials and methods

Study subjects

This retrospective study included pregnant women with chromosomal aberrations of chromosomes 6, 7, 11, 14, 15 and 20 identified via NIPT. From January 4, 2019, to December 31, 2023, a total of 9 pregnant women were recruited from 69,605. The average age of the recruited pregnant women was 31.4 years (ranging from 23 to 38), and the average gestational age was 15.25 weeks (ranging from 12 to 20 weeks). Among them, the serological screening results of 6 pregnant women indicated high risk, whereas the other 3 did not undergo serological screening (Table 1). All pregnant women received detailed genetic counselling and prenatal diagnosis. Subsequently, karyotyping analysis, CNV-seq, CMA, FISH or MS-MLPA were performed on the amniotic fluid samples. We collected the intrauterine phenotypes via ultrasound and followed them up until the induction of labor or one year after birth. This study was approved by the Medical Ethics Committee of Guangdong Women and Children’s Hospital and Health Institute (No. 202301382), and all research was performed in accordance with relevant guidelines/regulations. Informed consent was obtained from the participants.

Table 1 Patient clinical characteristics and molecular testing results of amniotic fluid
Full size table

NIPT

Peripheral blood (10 µl) samples from each pregnant woman were collected and stored in EDTA tubes and processed according to the following procedures after collection within 8 h. Afterward, cfDNA extraction, library construction, quality control, and pooling were performed via JingXin Fetal Chromosome Aneuploidy (T21, T18, and T13) Testing Kits (CFDA registration permit No. 0153400300). The JingXin BioelectronSeq 4000 System (CFDA registration permit NO. 20153400309), a semiconductor sequence, was used for DNA sequencing. The sequencing reads were filtered and aligned to the human reference genome (hg19). Fetal and maternal chromosome copy number variations (CNVs) were classified via our modified Sturfer’s z-score method as described previously. Chromosomes with a z score between − 3 and 3 were defined as disomic, whereas chromosomes with a z score ≥ 3 were defined as trisomic.

Karyotyping analysis

Amniotic fluid cells were aseptically inoculated into two independent cell culture flasks and cultured in a 37 °C, 5% CO2 incubator for 6‒8 days, depending on their growth status. Chromosome preparation was performed via conventional methods. Karyotypes were scanned via an automated Leica machine, and only well-distributed banded clones with chromosomes having no or relatively few (≤ 5) overlapping were selected for analysis.

CNV-seq

Genomic DNA (10–40 ng) was fragmented and used to construct DNA libraries via adapter ligation and PCR amplification. DNA libraries were sequenced on the NextSeq CN500 platform (Berry Genomics, Beijing, China). A total of 2.5–3.5 million uniquely mapped reads were aligned to the human reference genome (hg19) via the Burrows–Wheeler algorithm18. Reads were processed, and CNVs were evaluated via an in-house pipeline via read counts on the basis of a smoothness model (Berry Genomics, Beijing, China) according to previous methods19. Chromosomal variants were analysed by querying databases such as DECIPHER (https://www.deciphergenomics.org/),OMIM(https://www.omim.org/),ClinGen(https://www.clinicalgenome.org/) and DGV (http://dgv.tcag.ca/dgv/app/home). The pathogenicity of the candidate CNVs was assessed according to the guidelines outlined by the American College of Medical Genetics (ACMG) for interpretation of constitutional copy-number variants20.

CMA

DNA was extracted from amniotic fluid cells via the QIAamp DNA Blood Mini Kit (No. 51106; Qiagen, Germany). An Affymetrix CytoScan 750 K chip (Thermo Fisher Company) was used. The extracted DNA was digested, ligated, PCR amplified, purified, fragmented, and labelled for hybridization according to the instructions. After washing and staining, the chip was scanned, and the data were collected and analysed via ChAS 3.0 software. The copy number variants (CNVs) and region of homozygosity (ROH) status were analysed by setting thresholds of 100 kb for CNV and 5 Mb for ROH.

FISH

FISH was performed on 5 ml of uncultured amniotic fluid. A prenatal chromosome detection kit from Abbott (CEP 15:D15Z1 15p11.2, CEP 16:D16Z3 16q11.2, LSI 15:D15S11 15q11-q13) was used for hybridization, washing, restaining, and placing the slides under a fluorescence microscope according to the instructions in the manual. More than 100 cells were counted.

Methylation-Specific MLPA (MS-MLPA)

MS-MLPA analysis was performed via the SALSA MLPA Probemix ME028 Prader-Willi/Angelman reagent kit (MRC-Holland, Amsterdam, The Netherlands) according to the manufacturer’s instructions. The SALSA MLPA Probemix ME028 contains 49 MLPA probes with amplification products between 129 and 481 nucleotides. Thirty-six probes are target probes, of which 8 probes contain a HhaI recognition site and provide information on the methylation status of the 15q11 chromosomal region. All probes present will also provide information on copy number changes in the analysed sample. Eleven reference probes that are not affected by HhaI digestion are included, and genes located outside the 15q11 region are detected. Data analysis in combination with the appropriate lot-specific MLPA Coffalyser sheet proceeded in Coffalyser. Net software.

Results

There were 9 pregnant women (0.0129%, 9/69605) with chromosomal aberrations in chromosomes 6, 7, 11, 14 or 15 according to NIPT in our cohort. All pregnant women received at least one prenatal diagnostic test. Six fetuses (6 out of 9) were diagnosed with regional abnormalities of Imprinting Disease. The most commonly diagnosed syndrome was 15q11-q13 duplication syndrome (3 out of 6), followed by mosaic trisomy 7 (2 out of 6) and Temple syndrome (1 out of 6). The other three fetuses (3 out of 9) were diagnosed with absence of heterozygosity (AOH). After genetic counselling, 4 pregnant women (4 out of 9) chose induced labor, 3 pregnant women (3 out of 9) chose spontaneous labor, and 2 pregnant women (2 out of 9) chose cesarean section (Table 1). All of the couples were nonconsanguineous.

P1-P3 (15q11-q13 duplication syndrome)

P1 (G1P0 A0) was a 27-year-old woman with a 20-week gestation. Serological screening indicated a high risk of trisomy 21 (1/751). At 20 weeks of gestation, NIPT revealed that there was a high risk for trisomy 15 (z score, 8.467). Prenatal ultrasound revealed a ventricular septal defect. Amniotic fluid was then taken to test for CMA, FISH and MS-MLPA. First, CMA and FISH suggested mosaic duplication of chromosome 15 (arr(15)×2–3, Fig. 1A; nuc ish (D15Z1 × 3, D16Z3 × 2)[65/100], Fig. 1B). MS-MLPA revealed maternal mosaic duplication of chromosome 15 (rsa[GRCh37] 15q11q13(19000000_33599999)×2–3, Fig. 1C, D). In summary, the fetus of P1 was diagnosed with 15q11-q13 duplication syndrome (maternal interstitial). The family finally chose induced labor.

P2 (G3P1 A1) was a 35-year-old woman with a 16-week gestation. Prenatal ultrasound revealed a retromembrane hematoma and an umbilical cord around the neck. NIPT plus reported that there was a high risk for duplication of 15q11.2q12 (6 Mb). CMA revealed 4 copies of the 15q11q13 region (arr[GRCh37] 15q11.2q13.1(22770422_28943029)x4, Fig. 1E). MS-MLPA revealed four maternal copy number duplication of 15q11q13 (rsa[GRCh37] 15q11q13(19000000_33599999)×4, Fig. 1F, G). The fetus of P2 was diagnosed with four copy number duplication of 15q11q13, this abnormality usually occurs in the form of isodicentric 15 chromosome syndrome, and the family ultimately chose induced labor.

P3 (G1P0 A0) was a 23-year-old woman with a 14-week gestation. Serological screening indicated that individual indicators were abnormal and that the PAPPAMOM value was low. NIPT reported that there was a high risk for duplication of 15q11.2q12 (6.5 Mb). Prenatal ultrasound revealed bilateral choroidal cysts, and the fetal nasal bones were not visible. The G-band karyotype of the amniotic fluid was normal. CNV-seq and FISH revealed 3 copies of the 15q11q13 region (seq[GRCh37]dup(15)(q11.2q13.1), nuc ish (D15S11 × 3, D15Z1 × 2)[100], Fig. 1H, I). MS-MLPA revealed paternal duplication of 15q11q13(rsa[GRCh37] 15q11q13(19000000_33599999)×3, Fig. 1J, K). After genetic counselling, the family chose to continue the pregnancy and had a cesarean section at 37+3 weeks. One month and 10 days after the baby was born, a child health examination was conducted. The child’s height was 54 cm, her weight was 3.9 kg, and her head circumference was 35.7 cm. The physical examination showed normal results.

Fig. 1
figure 1

Results of genetic tests in P1, P2 and P3. A CMA plot of the P1 fetus revealed mosaic duplication of chromosome 15: arr (15)×3 [55%]. B FISH results of the P1 fetus revealed mosaic duplication of chromosome 15: nuc ish (D15Z1 × 3, D16Z3 × 2)[65/100]. MLPA analysis (ME028) of the 15q11 chromosomal region in the P1 fetus revealed that the C copy number ratio was approximately 1.2, and D a duplication is maternally inherited the ratios of imprinted methylation probes are expected to be ~ 0.7 E CMA plot of P2 fetus showing duplication of 15q11q13: arr[GRCh37] 15q11.2q13.1(22770422_28943029)x4. MLPA (ME028) revealed that the 15q11 chromosomal region in P2: F copy number ratio was approximately 2, and G the imprinted methylation probes ratio was 0.8 H CNV-seq plot of P3: 6.5 Mb duplication of 15q11.2q12. I FISH results of the P3 fetus revealed duplication of the 15q11q13 region: nuc ish (D15S11 × 3, D15Z1 × 2)[100]. MLPA (ME028) revealed that the 15q11 chromosomal region in P3:J copy number ratio was approximately 1.6, and K the duplication is paternally inherited the ratios of imprinted methylation probes are expected to be ~ 0.3.

Full size image

P4-P7

P4 (G2P0 A1) was a 29-year-old woman with a 13-week gestation. She experienced a biochemical pregnancy. Serological screening indicated a high risk of trisomy 21 (1:200). NIPT reported that there was a high risk for duplication of 2q35q37.3 and deletion of 11q24.3q25. The G-band karyotype of the amniotic fluid was normal (Fig. 2A). CMA of the amniotic fluid revealed maternal loss of heterozygosity in the 11q24.3q25 region arr[GRCh37] 11q24.3q25(128375224_134930689)x2 hmz, whereas P4 was normal (Fig. 2B, C). 11q24.3q25 is located outside of the imprinted gene region on chromosome 11. After extensive genetic counselling, the family chose to continue the pregnancy and had a cesarean section at 40+2 weeks. Postnatal follow-up was normal. The child underwent a health examination 7 months and 6 days after birth. The body weight was recorded as 7.7 kg, the height was 71.9 cm, and the head circumference was 42 cm. Pharyngeal congestion was observed during the examination. No abnormalities were detected in other tests, and there were no apparent issues with the development of the nervous system.

P5 (G4P1 A1E1) was a 38-year-old woman with a 16-week gestation. This pregnant woman was of advanced pregnancy age and gave up serological screening. NIPT reported that there was a high risk for trisomy 7 (z score, 16.835). CMA of the amniotic fluid revealed mosaic duplication of chromosome 7: arr(7)×2–3 (Fig. 2D). The fetus at P5 was diagnosed with mosaic trisomy 7. The family finally chose induced labor.

P6 (G3P1 A1) was a 31-year-old woman with a 14-week gestation. Serological screening indicated a high risk of trisomy 18 (1:894). NIPT reported that there was a high risk for trisomy 7 (z score, 10.139). CMA of the amniotic fluid indicated mosaic duplication of chromosome 7: arr(7)×2–3 (Fig. 2E, G). The CMA result of P6 was normal (Fig. 2F). The fetus at P6 was diagnosed with mosaic trisomy 7. The family finally chose induced labor.

P7 (G4P1 A2) was a 32-year-old woman with a 13-week gestation. Serological screening indicated a high risk of trisomy 21 (1:73). NIPT reported that there was a high risk for seq[GRCh37] dup(14)(q21.1q32.33) (z score, 5. 775, Fig. 2H). After detailed consultation with the geneticist, the pregnant woman underwent invasive diagnosis at 18 gestational weeks. CMA revealed two regions associated with the loss of heterozygosity on chromosome 14: arr[hg19]14q11.2q13.3 (20520197_37350812)x2 hmz (16.8 Mb) and 14q24.2q32.33 (70758466_104 240618)x2 hmz (33.5 Mb) (Fig. 2I, J). The fetus at P7 was diagnosed with Temple syndrome. After providing informed consent, the pregnant woman chose to continue gestation, and fetal growth and development were monitored via regular ultrasound. Pregnant women underwent ultrasound Doppler tests at 20, 25, 35, 37, and 39 weeks. The foetus was found to experience intrauterine growth retardation (IUGR) in late pregnancy. At 38+3 weeks of gestation, the pregnant woman delivered a small for gestational age infant with a birth weight of only 1.78 kg vaginally. Neonatal cardiac ultrasound revealed a patent foramen ovale, and no abnormalities were detected in the brain, kidney, ureter, or lung diaphragm via ultrasound. During the neonatal period, the main symptoms were feeding difficulties and hypotonia. When the child was one year old, brain magnetic resonance imaging (MRI) did not reveal any abnormalities, but the child was diagnosed with protein-energy malnutrition and delayed growth and development by a pediatrician.

Fig. 2
figure 2

Results of genetic tests in P4, P5, P6 and P7. A. The G-band karyotype of the P4 fetus was normal. B and C The CMA results of the foetus indicated maternal loss of heterozygosity in the 11q24.3q25 region: arr[GRCh37] 11q24.3q25(128375224_134930689)x2 hmz, and P4 was normal. D CMA result of P5 fetus indicating mosaic trisomy 7: arr(7)×2–3. E and G. CMA plot of P6 foetus showing mosaic duplication of chromosome 7: arr(7)×2–3. The UPDtool indicates that UPD7 is excluded, as the F CMA plot of P6 was normal. H NIPT plot of P7: seq[GRCh37] dup(14)(q21.1q32.33). I and J CMA plot of P7 fetus: arr[GRCh37] 14q11.2q13.3 (20520197_37350812)x2 hmz and 14q24.2q32.33(70758466_ 104240618)x2 hmz.

Full size image

P8-P9

P8 (G3P1 A1) was a 38-year-old woman with a 12-week gestation. This pregnant woman was of advanced pregnancy age, had placental limitations and gave up serological screening. NIPT reported that there was a high risk for seq[GRCh37] del(6)(q25.1q27) (z score, −10.38). The G-band karyotype of the amniotic fluid was normal (Fig. 3A). CMA of the amniotic fluid revealed maternal loss of heterozygosity at 6q25.2q27: arr[GRCh37] 6q25.2q27(152902809_170896644)x2 hmz, whereas the CAM result of P8 was normal (Fig. B, C). 6q24-related transient neonatal diabetes mellitus (6q24-TNDM) is caused by the overexpression of imprinted genes at 6q24 (PLAGL1 and HYMAI). Typically, only the paternal alleles of PLAGL1 and HYMAIare expressed. Paternal UPD of chromosome 6 leads to 6q24-TNDM21. The clinical significance of maternal UPD of 6q25.2q27 is unclear. After genetic counselling, the family finally chose to continue gestation. P8 subsequently gave birth to a healthy newborn, with an Apgar score of 9–10–10.

P9 (G3P2 A0) was a 30-year-old woman with a 20-week gestation. History of allogeneic blood transfusion after the first delivery 7 years prior. NIPT reported that there was a high risk for seq[GRCh37] del(14)(q21.2-q32.11) (z score, −7.379). CMA of the amniotic fluid revealed a loss of heterozygosity at 15q23q25.3: arr[GRCh37]15q23q25.3(70673857_86592054)×2 hmz (Fig. D, E). The MS-MLPA results were negative (Fig. 3F, G). The region of loss of heterozygosity was not in the imprinted gene region of chromosome 15. It does not induce imprinting disorders, but it may heighten the risk of recessive genetic diseases. After providing full informed consent, the family finally chose to continue gestation. P9 delivered a healthy baby and followed up normally after birth.

Fig. 3
figure 3

Results of genetic tests in P8 and P9. A. The G-band karyotype of the P8 fetus was normal. B and C CMA of the amniotic fluid indicated maternal loss of heterozygosity at 6q25.2q27: arr[GRCh37] 6q25.2q27(152902809_170896644)x2 hmz. The CAM result of P8 was normal. D and E The CMA results of the P9 fetus indicated loss of heterozygosity at 15q23q25.3: arr[GRCh37]15q23q25.3(70673857_86592054)×2 hmz. MLPA (ME028) revealed that the 15q11 chromosomal region in P9:F copy number ratio was approximately 1, and the G methylation ratio was normal.

Full size image

Pregnancy outcomes

In the three patients diagnosed with 15q11-q13 duplication syndrome, two maternal duplications (P1, P2) opted to terminate their pregnancies. The other paternal duplication (P3) chose to continue the pregnancy and was followed up to one year of age with normal development. All patients (P5, P6) with Mosaic trisomy 7 chose induced labour. One patient (P7) who was diagnosed with Temple syndrome chose to continue the pregnancy, and growth restriction occurred both before and after birth. Three patients (P4, P8, and P9) whose LOH did not affect imprinted genes were followed up normally after birth.

Discussion

NIPT is more popular than traditional serological screening because of its higher sensitivity and specificity in screening for trisomy 21, 18, and 13, effectively reducing the number of invasive diagnostic procedures6,8. A survey of Chinese specialists working in prenatal diagnosis revealed that more than 80% of specialists expressed support for expanding NIPT to more conditions (SCAs, RCAs and MMSs) other than common trisomies22. However, many scholars believe that RCAs and MMSs detected by extended NIPT have limited clinical utility and uncertainties regarding PPV and NPV. In particular, studies on RCAs and MMSs associated with disrupted imprinted gene expression are currently insufficient. Owing to the complexity of the pathogenesis of these diseases, they also pose great challenges in clinical management and genetic counselling. Few studies have focused on non-invasive prenatal testing in imprinting disorders, possibly due to its low incidence. The survey by Ting Hu et al. showed that 50 cases of Regions of homozygosity/uniparental disomy(ROH/UPD) were detected in 158,824 pregnant women, of which only 5 cases were related to imprinting diseases23. In a review of studies of noninvasive testing for trisomy 7, it was also pointed out that cases of UPD7 are scarce24.Therefore, most cases of imprinting diseases confirmed by prenatal diagnosis after noninvasive detection are reported as case reports25,26. Our results also indicate that the case of imprinting diseases is scarce. Nevertheless, the chromosomal abnormalities associated with imprinting diseases detected by NIPT, we still need to be alert to the possibility of imprinting diseases. When NIPT finds an MMS that may be associated with imprinting disease, it needs to be further verified by combining ultrasound testing, karyotype analysis, CMA, MS-MLPA and other prenatal diagnostic techniques. Clarifying the parental origin of RCAs and MMSs is highly important for assessing the prognosis of the fetus and may also alleviate the anxiety of the family during this process.

The clinical phenotypes of 15q11-q13 duplication syndrome are variable, duplications of maternal origin are usually associated with a more severe phenotype, whereas duplications of paternal origin are milder and may be asymptomatic in some patients. Duplications of maternal origin usually have developmental delay, intellectual disability, behavioral problems, and seizures, and are associated with a higher risk of ASD. The phenotypes are closely related to the overexpression of multiple genes within this region, such as UBE3 A, SNRPN, GABRB3, GABRA5, GABRG3, HERC2, CYFIP1, NIPA1, NIPA2, and TUBGCP521,27. UBE3 A plays an important role in neurodevelopment and synaptic function, and overexpression of UBE3 A may lead to behavioral problems (such as social impairment, and stereotyped behavior) and intellectual disability that are characteristic of ASD. SNRPN abnormal expression may lead to delayed motor and language development. GABRB3, GABRA5 and GABRG3 are major inhibitory neurotransmitters that play a key role in the development and function of the nervous system. Aberrant expression of HERC2 may be associated with behavioral problems, such as self-injury-related behavior. Overexpression of CYFIP1 may contribute to the behavioral problems (e.g., social impairment, stereotyped behavior) and intellectual disability that are characteristic of ASD. Abnormal expression of these genes, NIPA1 and NIPA2, may affect the development and function of the nervous system, leading to behavioral problems and intellectual disability. Abnormal TUBGCP5 may be associated with behavioral problems such as ADHD.

According to Maria Bisba et al., the intrauterine phenotype of duplication in this region is rare and not well-defined, and there may be some cases presenting with intrauterine growth retardation or congenital heart disease28. The relationship between these phenotypes and gene in this region is not clear. It may be related to the maternal inheritance of UBE3 A overexpression and CYFIP1 and NIPA1overexpression leading to intrauterine growth retardation or neurological abnormalities. According to the literature review, the phenotype of four copies in this region is more severe than three copies. These include more severe seizures and behavioral problems. Although the test results of P1–P3 were all 15q duplications, the mechanisms of the three cases were different. CMA, CNV-seq and FISH were used to clarify the existence of duplications and copy numbers. MS-MLPA methylation levels were used to analyse the parental origin of duplications. P1 is a maternal interstitial 15q11.2-q13.1 duplication that typically includes one extra copy of 15q11.2-q13.1 within chromosome 15, resulting in trisomy for 15q11.2-q13.1 (~ 20%). P2 is suspected to be a maternal isodicentric 15q11.2-q13.1 supernumerary chromosome, typically comprising two extra copies of 15q11.2-q13.1 and resulting in tetrasomy for 15q11.2-q13.1 (~ 80%). Maternal 15q duplication syndrome is characterized by hypotonia and motor delays, intellectual disability, autism spectrum disorder, and epilepsy, including infantile spasms. The phenotype of four copies in this region is more severe. Patients with a maternal isodicentric 15q11.2-q13.1 duplication are typically more severely affected than those with an interstitial duplication29,30. P3 was a paternal interstitial 15q11.2-q13.1 duplication. The pathogenicity of paternal 15q11.2-q13.1 duplication is unclear30,31. This is also the theoretical basis for P3 to make different choices.

Trisomy 7 is the most commonly observed type of RCA detected via expanded NIPT24,32,33. In prenatal or postnatal diagnosis, chromosome 7 abnormalities are usually mosaic trisomy 7. Trisomy 7 usually leads to spontaneous abortion in early pregnancy13. In our previous studies13, we reported that most of the trisomy 7 cases detected by NIPT were placental chimerism. When mosaic trisomy 7 is detected by NIPT, placental mosaicism or fetal mosaicism should be confirmed for the next step. Trisomy 7 can result in the loss of one copy of extra chromosome 7 through trisomic rescue, which may lead to UPD14. Silver-Russell syndrome (SRS), which features severe intrauterine and postnatal growth retardation, is associated with maternal UPD7 in approximately 10% of cases34,35. The reported clinical phenotypes of mosaic trisomy 7 differ substantially. Prenatal diagnosis of mosaic trisomy 7 in a foetus can reveal heart and kidney abnormalities on ultrasound examination in late pregnancy32. It has also been reported that mosaic trisomy 7 case follow-up within three months of delivery is normal33. Therefore, it is difficult to judge the risk for prenatally diagnosed patients. In our cohort, CMA confirmed low-proportion mosaicism trisomy 7 in P5 and P6. Fetal ultrasound examination revealed no abnormalities. Both families chose to terminate their pregnancies.

Temple syndrome (TS14) is an imprinting disorder caused by molecular disruption of the imprinted region in 14q32 (deletions, maternal UPD, loss of methylation)36,37. The most common symptoms are IUGR, hypotonia, precocious puberty, small for gestational age birth, tube feeding and psychobehavioral problems38. TS14, SRS and Prader‒Willi syndrome share many features, including fetal and postnatal growth failure, feeding difficulty, and muscular hypotonia, and often lead to misdiagnosis in clinical management39,40,41. At present, TS14 is challenging to diagnose prenatally because of a lack of precise and well-characterized fetal phenotypes. The main prenatal phenotype of P7 was a smaller abdominal circumference than the gestational age (−2SD). Interestingly, NIPT of P7 revealed a high risk of trisomy 14, whereas CMA indicated maternal UPD (14). Trisomy rescue is the main mechanism that may be involved in the occurrence of maternal UPD (14). An association between confined placental T14 mosaicism and fetal maternal UPD (14) has been reported42,43. The NIPT and CMA of P7 may provide arguments in favour of this mechanism.

In this study, patients presented a comparatively favourable prognosis; pregnant women all opted to proceed with their pregnancies, and subsequent postnatal follow-ups revealed normal outcomes. Conversely, when the prognosis of the disease is relatively poor or uncertain, most pregnant women opt to terminate their pregnancies. Clarifying the parental origin of RCAs and MMSs is highly important for patients making decisions.

Conclusion

The widespread use of NIPT in prenatal screening provides more opportunities to detect RCAs and MMSs in mid-pregnancy. Owing to the complexity of the pathogenic mechanism of imprinting-related diseases, clinicians often need to combine multiple detection technologies to accurately judge chromosomal abnormalities and their parental origins. The combination of NIPT and other prenatal diagnostic technologies can help increase the possibility of detecting imprinting-related diseases with no phenotype or a late phenotype in utero.