A retrospective study was performed from January 2016 to December 2020. In total, seven cases carrying 21q21.1–21.2 deletions or duplications were selected from 11,867 pregnant women who had indications (e.g., abnormal non-invasive prenatal testing (NIPT) or fetal imaging) and underwent invasive diagnostic testing via amniocentesis or cordocentesis at the Prenatal Diagnosis Center of Obstetrics and Gynecology, Southwest Hospital. Informed consent for invasive prenatal diagnosis was obtained from the parents before detection. This research was approved by the Ethics Committee of Southwest Hospital, Third Military Medical University (Army Medical University). Six fetuses from six unrelated Chinese families were identified as carrying 21q21.1–q21.2 duplications, as their pedigree verification information was collected, and they were classified as pedigrees 1, 2, 3, 4, 5, and 6. In addition, a fetus carrying a 21q21.1–q21.2 deletion and its family members were labeled as pedigree 7. The pregnant women in these seven pedigrees did not have pregnancy complications and denied any related family history.

Pedigrees 1–6: The maternal age at the time of amniocentesis was between 25 and 32 years, and a gestational age ranging from 18 + 2 to 25 weeks. The pregnant woman in pedigree 3 chose amniocentesis because of pulmonary sequestration of the fetus examined by ultrasound. The others all chose amniocentesis because NIPT screening showed an abnormality on chromosome 21.

Pedigree 7: A 22-year-old woman (gravida 4, para 1) was subjected to cordocentesis at 26 + 5 gestational weeks because the bilateral ventricle of the fetus had widened, as tested by ultrasound examination (left: 14 mm, right: 14 mm).

Approximately 0.5 ml of each peripheral blood sample and 0.4 ml of each umbilical cord blood sample were inoculated into a T-cell culture medium (BAIDI, China) and incubated at 37°C, for 3 days. Approximately 20 ml of each amniotic fluid sample was inoculated into an amniotic fluid medium (BIO-AMF™-2, BI, China) and incubated at 37°C, with 5% CO₂, for 7–10 days. Chromosomal karyotyping was performed according to the standard protocol using G-banding at a 400-banded (amniotic fluid samples) or 550-banded (blood samples) resolution, and karyotypes were described according to the International System for Human Cytogenetic Nomenclature 2016 (ISCN 2016) criteria (Stevens-Kroef et al., 2017).

Uncultured amniotic fluid samples (10 ml per fetus), umbilical cord blood samples (600 µL per fetus), and peripheral blood (600 µL per person) of the pedigree members were collected, and DNA was extracted using the TIANamp Genomic DNA Kit (TIANGEN, China). The Infinium Global Screening Array (Illumina, San Diego, CA, United States) contains approximately 700,000 genome-wide tag SNPs. Genomic DNA was hybridized to an Infinium Global Screening Array as reported previously (Srebniak et al., 2011). The array was scanned with the iScan array scanning system (Illumina, San Diego, CA, United States). Molecular karyotype analysis was performed using GenomeStudio V2011.1 software (Illumina, San Diego, CA, United States), which was used for annotation. Copy number variations (CNVs) that were larger than 100 kb or affected more than 50 markers were considered and were annotated based on the GRCh37 (hg19) genome. CNVs were evaluated according to the guidelines (Richards et al., 2015; Riggs et al., 2020), scientific literature, and publicly available databases as follows: DGV (http://dgv.tcag.ca/dgv/app/home), OMIM (http://www.ncbi.nlm.nih.gov/omim), gnomAD (http://gnomad-sg.org/), DECIPHER (http://decipher.sanger.ac.uk), dbVar (http://www.ncbi.nlm.nih.gov/dbvar), ClinVar(http://www.ncbi.nlm.nih.gov/clinvar), ClinGene (https://www.ncbi.nlm.nih.gov/projects/dbvar/clingen/), and Pubmed (https://www.ncbi.nlm.nih.gov/pubmed/). Benign or likely benign CNVs were not reported.

Ultrasound results of the second and third trimesters of pregnancy were collected. Postnatal clinical follow-up assessments via telephone were performed from 6 months to 3 years after birth. After obtaining their parents’ informed consent, the child’s healthcare data were collected to assess developmental details. General child healthcare was carried out by professional doctors according to the World Health Organization’s physical and mental development table for infants aged 0–3 years. Child healthcare in tertiary hospitals was performed according to the Denver Developmental Screening Test (Wijedasa, 2012).

Amniotic fluid samples, umbilical cord blood samples, and peripheral blood samples of family members were subjected to conventional karyotyping because balanced rearrangements will escape SNP array detection (Levy and Wapner, 2018).

Pedigrees 1–6: The conventional G-banding analysis showed that the karyotypes of the fetuses and their parents were normal.

Pedigree 7: Although typical karyotypic analysis by G-banding might be able to delineate chromosomal aberrations greater than 5–10 Mb in size (Shaffer and Bejjani, 2004), the 8.7 Mb deletion of 21q21.1–21.2 was not identified in our study. Owing to the small size of chromosome 21, the deletion region could only be identified above 700-banded resolution, whereas the conventional amniotic karyotyping could only achieve 550-banded resolution at most. Therefore, the fetus (III:2, Figure 1), its elder brother (III:1, Figure 1), mother (II:2, Figure 1), and maternal grandmother (I:2, Figure 1) had normal karyotypes. Its father (II:1, Figure 1) and maternal grandfather (I:1, Figure 1) also had normal karyotypes.

Pedigrees 1–6: The fetuses of six unrelated pedigrees carried the 21q21.1–q21.2 duplications, which were inherited from their mothers, and with the same coordinates and lengths as those of their mothers. The smallest duplication length was 1 Mb (chr21:20,195,657–21,199,532, hg19 build), and the largest was 2.7 Mb (chr21:23,573,580–26,310,725, hg19 build). The duplicated regions in pedigrees 1, 2, 3, 5, and 6 did not contain any protein-coding gene, and only the duplication in pedigree 4 contained the NCAM2 gene (Table 1; Figure 2; Supplementary Figure S1). All fathers were also tested with SNP arrays, and the results were negative.

Pedigree 7: SNP array results showed that the fetus (III:2, Figure 1) carried an 8.7 Mb deletion of chromosome 21q21.1–21.2 (chr21:16,767,983–25,441,375, hg19 build; Figure 2; Supplementary Figure S1), and the pedigree analysis found that the CNV was inherited from the mother with a normal phenotype (II:2, Figure 1). To obtain more genetic evidence, the elder brother and maternal grandparents of the fetus were also tested with SNP arrays. The extended analysis of the pedigree revealed that two other healthy members also carried the deletion, the elder brother, 3 years of age (III:1, Figure 1), and maternal grandmother, 41 years of age (I:2, Figure 1). In addition, the results of others (II:1 and I:1, Figure 1) in this pedigree were normal. Otherwise, the elder brother was found to carry another deletion, which was located on 5p15.33 (chr5:19:38,139–1,124,703, hg19 build) and was proven to be a de novo variation of uncertain significance (VUS) mutation. No other significant CNVs were found among the seven pedigrees.

Pedigrees 1–6: No abnormalities were found during the second and third trimesters of pregnancy, except for the fetuses in pedigree 3 with pulmonary isolation. Three boys (fetuses of pedigrees 1, 3, and 6) and three girls (fetuses of pedigrees 2, 4, and 5) were born at full-term delivery. Now, the youngest individual is 2.5 years of age, the oldest is 3.5 years of age, and none of them show signs of developmental delay or intellectual disability based on child’s healthcare examination (Table 2).

Pedigree 7: Ultrasound and MRI examinations were performed at 32 weeks of gestation, and no further widening of the lateral ventricles was observed in both examinations (left: 14 mm, right: 14 mm, examined by ultrasound; left: 12.5 mm, right: 13.6 mm, examined by MRI). After genetic counseling, the pregnant woman and her husband chose to continue the pregnancy. A healthy boy was born by natural delivery at 39 gestational weeks, without any special facial features. The boy is 8 months old currently and does not have any abnormal phenotypes; moreover, the details of the child’s healthcare examination were normal (Table 2). His elder brother is 3 years of age and also does not show developmental delay or intellectual disability.