Introduction

Echogenic intracardiac focus (EIF) is thought to represent mineralization or small deposits of calcium in the heart muscle. The incidence of EIF varies between 0.86% and 30% in fetuses1,2,3. Racial difference affects the frequency of EIF in the hearts of second-trimester fetuses; Asian patients are more likely to exhibit EIF than patients of other ethnic groups3,4. EIF itself does not affect health or heart function; typically, EIF on prenatal sonography is absent by the third trimester.

Isolated EIF is associated with high-risk populations with a 4.8-fold increase in relative risk for trisomy 21 (T21)5, the likelihood ratio for T21 is significant for any marker (isolated, multiple or combined with anomaly)6, and the prevalence of EIF is higher in T21 than in chromosomally normal fetuses7. Therefore, most researchers agree that in cases without other clinical implications and a negative cell-free DNA screening, or a negative first- or second-trimester screening result, isolated EIF appears to be a benign variant and no further evaluation is required4,8,9,10,11,12. However, it is noteworthy that the studies above focused on the relationship between EIF and whole chromosome aneuploidy.

Hay et al. reported that 27.56% of clinically significant chromosomal abnormalities may be missed for patients with advanced maternal age, abnormal maternal serum screen, family history, or soft ultrasound markers including EIF13. Our previous studies suggested an association between pathogenic copy number variation (CNV) and fetal ultrasound soft marker; however, the samples contained fetuses with other soft markers (e.g., thickened nuchal fold, echogenic bowel, and mild ventriculomegaly)14. The research results of Shaffer et al. showed that clinically significant genomic alterations were identified in fetuses with soft markers (2/77, 2.6%), but there were only 6 cases with EIF in the samples, and no clinically significant genomic alterations were detected15. Another study from China showed that there were 3 samples with pathologic CNV in 143 fetuses with isolated EIF. However, we reanalyzed the pathogenicity of the three CNVs and found that only one CNV was pathogenic (Patient No. 2). In the other two samples with CNV, one was female carrier of Steroid sulphatase deficiency and the other was variants of uncertain significance (VOUS). In addition, the sample size of this study was small16. Hitherto, large-sample studies pertaining to the estimation of association between isolated EIF and segmental aneuploidie are scarce.

Although using isolated EIF as a basis for deciding to offer amniocentesis will result in more fetal losses than cases of detected T21 and will decrease the prenatal detection of fetuses with T2117, it is undeniable that the presence of isolated EIF increases the incidence of invasive procedures substantially6,18. At present, high-resolution chromosome detection techniques such as chromosomal microarray analysis (CMA) and copy number variation sequencing (CNV-seq) are widely used in fetal chromosome detection in China19,20,21,22,23,24. CNV-seq provide a more sensitive, accurate, and affordable approach to assess genome-wide CNVs compared with CMA24,25,26. The purpose of this study was to investigate whether isolated EIF indicates an increased risk of segmental aneuploidies and whether prenatal screening or prenatal diagnosis including segmental aneuploidies analysis should be recommended.

Results

CNV-seq analysis of the 1,099 fetuses with isolated EIF

CNV-seq analysis was performed successfully on amniotic fluid samples from 1,099 fetuses with isolated EIF. The mean gestational age and maternal age at the time of amniocentesis were 25.6 weeks (range 19.0–36.4) and 27.5 years (range 17.0–34.0), respectively. Among the 1,099 fetuses with isolated EIF, 763 (69.4%) fetuses had EIF in the left ventricle, 62 (5.6%) in the right ventricle, and 274 (25.0%) in both ventricles; we found 662 fetuses (60.2%) with a single EIF and 437 (39.8%) with multiple EIFs. Seven samples (0.6%) exhibited abnormal segmental aneuploidies (including pathogenic CNV (pCNV) and likely pathogenic CNV (lpCNV)), and CNVs in eight fetuses (0.7%) were classified as VOUS. All seven fetuses with abnormal segmental aneuploidies had multiple EIFs; the incidence of abnormal segmental aneuploidies in multiple EIFs was 1.6% (7/437), while no clinically significant segmental aneuploidies were found in 662 fetuses with a single EIF. The incidence of abnormal segmental aneuploidies in the fetuses with multiple EIFs was higher than those with single EIF (P-fisher = 0.002) (Table 1). Among these seven cases with abnormal segmental aneuploidies, five (71.4%) had submicroscopic CNVs (smaller than 5 Mb). The CNV-seq results of the fetuses with isolated EIF are summarized in Table 2.

Table 1 Distribution of the chromosomal abnormalities in the fetuses.
Full size table
Table 2 CNV-seq findings of 1,099 fetuses with isolated EIF and follow-up outcome.
Full size table

CNV-seq analysis of the control group samples

The control group contained 5,616 samples; the mean gestational age and maternal age at the time of amniocentesis were 21.6 weeks (range 16.0–35.6) and 27.6 years (range 15.0–34.0), respectively. Furthermore, 64 samples (1.1%) had abnormal segmental aneuploidies, and the CNVs in 72 fetuses (1.3%) were classified as VOUS. Statistical difference was not indicated in the incidence of abnormal segmental aneuploidies between the case group and control group (x2 = 2.220, P = 0.136); in addition, statistical difference was not indicated in the VOUS incidence between the two groups (x2 = 2.397, P = 0.122) (Table 1).

Discussion

Chromosomal abnormalities occur in approximately 1 in 150 live births; autosomal trisomies are the most common aneuploidies, and T21 is the most common of these, with a prevalence of approximately 1 in 700–800 live births27,28,29. Most of our study participants received cell-free DNA screening, first-or second-trimester screening before amniocentesis. The detection of whole chromosome aneuploidy in amniotic fluid samples cannot objectively reflect its true incidence. Therefore, we have not compared and analyzed the results of whole chromosome aneuploidy in this study.

In our samples, the majority of EIF were unilateral; among the cardiac chambers affected, the left ventricle was the most frequent (69.4%), which was similar to that reported by other authors2,30. Some studies have shown that no chromosomal karyotype abnormalities were found in fetuses with multiple EIFs31,32. However, Towner et al. suggested that finding multiple EIFs was a stronger predictor of T21 than that described for a single EIF33. Our results indicated that all seven fetuses with abnormal segmental aneuploidies had multiple EIFs, and the incidence of abnormal segmental aneuploidies in multiple EIFs samples was 1.6% (7/437). However, no segmental aneuploidies of definite clinical significance were found in 662 fetuses with a single EIF. Therefore, fetuses with multiple EIFs have a higher incidence of abnormal segmental aneuploidies than those with single EIF (P-fisher = 0.002), i.e., multiple EIFs have more clinical significance than single EIF.

Among the seven cases of lpCNVs, two fetuses were diagnosed with a known syndrome, 16p13.11 recurrent microdeletion, which is a clinically heterogeneous disease; common phenotypes include mental retardation, epilepsy, and multiple congenital anomalies. It appears that this may be a susceptibility locus for neurocognitive disease, where 16p13.11 deletion is insufficient to cause the phenotype34. The parents of one fetus (EIF636) decided to continue the pregnancy, and our follow-up data showed that the infant demonstrated normal developmental milestones at 5 months. The parents of the other fetus (EIF638) opted for pregnancy termination. With regard to the other five fetuses with lpCNV: EIF44 (Fig. 1), EIF361, EIF364, EIF575, and EIF827, the follow-up results indicated normal development at 10 months, pregnancy termination, ankyloglossia, pregnancy termination, and pregnancy termination, respectively.

Figure 1
figure 1

Secondary confirmation of chromosomal abnormalities detected by CNV-seq. CNV-seq profiles of EIF44 (A), data are plotted as copy number (Y-axis) versus 20 kb chromosomal read bins (X-axis). The mean copy number along the length of each chromosome is indicated by the blue line. Confirmatory aCGH profiles is shown at the lower part (B).

Full size image

Among the eight samples with VOUS, the parents of only one fetus (EIF176) opted for pregnancy termination. One fetus (EIF103) was delivered at 34.1 weeks and pediatric examinations at 6 months indicated delayed development; however, because the pregnant woman refused to complete a full trio analysis, the origin and pathogenicity of the CNV in this case cannot be identified. Additionally, because of premature birth, we cannot determine the exact cause of the child's growth retardation. The remaining six fetuses were delivered at full term and exhibited normal development within a few months after birth. However, owing to the age of the babies, we could not assess their prognosis fully and accurately; therefore, further follow-up is required.

In our study, apart from whole chromosome aneuploidy, seven cases of abnormal segmental aneuploidies were detected, and the incidence of lpCNVs was 1/157 (7/1,099), which was similar to that reported by other authors16. Significant difference was not indicated in the detection rate of abnormal segmental aneuploidie between fetuses with EIF and those with normal ultrasound results (P = 0.136). In addition, although fetuses with multiple EIFs have higher incidence of abnormal segmental aneuploidies than those with single EIF, significant difference was not indicated in the detection rate of abnormal segmental aneuploidie between fetuses with multiple EIFs and those with normal ultrasound results (P = 0.387). Therefore, isolated EIF does not increase the risk of abnormal segmental aneuploidie in the fetus. In clinical work, if a fetus with isolated EIF has no other risk factors leading to segmental aneuploidy, clinicians should theoretically only focus on the risk of the common whole chromosome aneuploidy. Therefore, for fetuses with isolated EIF, if no other high-risk factors exist leading to segmental aneuploidies, screening methods such as cell-free DNA screening or first-or second-trimester screening are recommended to help assess the risk of fetuses suffering from the common whole chromosome aneuploidy; however, interventional prenatal diagnosis is not recommended as the first choice.

Meanwhile, the fragments of five among seven abnormal segmental aneuploidies were less than 5 Mb, which is typically undetectable by karyotype analysis. Combined with previous research data13,35, the proportion of submicrostructural abnormalities in fetal chromosomal aberrations is not low. In clinical practice, the presence of soft markers increased the incidence of invasive procedures substantially. Previous studies have shown that in patients at lower levels of T21 risk, the rate of amniocentesis was significantly higher following disclosure of isolated EIF when compared with pregnancies without EIF at similar risk levels18. Therefore, if pregnant women with fetal EIF, especially those with multiple EIFs, voluntarily chose invasive prenatal diagnosis, they should be recommended high resolution methods such as CNV-seq or CMA, rather than karyotype analysis.

In conclusion, our data indicated that the incidence of abnormal segmental aneuploidies among fetuses with isolated EIF was 1/157. Isolated EIF did not increase the risk of fetal segmental aneuploidies.

Materials and methods

Clinical data

The current study was a prospective case–control study; all pregnant women in the case group and control group were younger than 35 years old at the expected date of childbirth, and all samples were from pregnant women who received prenatal diagnosis in West China Second University Hospital of Sichuan University from February 2017 to December 2018. The study participants in the case group were fetuses that were diagnosed with isolated EIF based on ultrasonogram findings by two experienced ultrasonographers; none of the fetuses indicated any structural abnormalities or other soft markers by ultrasonogram through all the pregnancy stages. Samples without fetal ultrasound abnormalities but received prenatal diagnosis for other reasons (serological screening high-risk, voluntary request) were set as controls. The study was approved by the Medical Ethics Committee of West China Second University Hospital of Sichuan University, and we confirmed that all experiments were performed in accordance with relevant guidelines and regulations. All women underwent genetic counseling and signed an informed consent that explained the possible pathogenic significance of their condition; they voluntarily requested amniocentesis and fetal CNV-seq testing.

DNA extraction and detection of maternal cell contamination

According to routine operation specifications, 20 ml amniotic fluid was extracted and placed into four sterile centrifuge tubes. CNV-seq and quantitative fluorescence PCR (QF-PCR) were performed on two tubes, and the remaining two tubes were stored at 2–8 °C. DNA was extracted from the amniotic fluid using a DNeasy Blood and Tissue Kit (QIAGEN, Germany), according to the manufacturer’s standard extraction procedures. The quality and concentration of DNA were assessed with the NanoDrop 1000 (Thermo Fisher Scientific, USA). QF-PCR detection was performed using 21 trisomy/sex chromosome/polyploid and 18 trisomy/13 trisomy/polyploid detection kits (DAAN Gene, China), according to the manufacturer’s instructions. If the results of QF-PCR indicated that there were maternal cells in the samples, CNV-seq and QF-PCR were performed on the spare samples after cell culture.

CNV-seq

The DNA library was obtained using the Chromosome CNV Detection kit (Berry Genomics, China) and was subsequently sequenced on the Illumina Nextseq 500 sequencing platform (Illumina, United States) using the Nextseq 500 High Output kit (Illumina, United States). Finally, we compared the reads obtained by sequencing with the human reference genome and performed bioinformatics analysis to obtain the genomic copy number information of the samples as previously described14. In this study, the pathogenicity of CNVs > 100 kb was analyzed. By searching the DGV (https://dgv.tcag.ca/), DECIPHER (https://decipher.sanger.ac.uk/), ClinGen (https://www.clinicalgenome.org/), OMIM (https://omim.org/), and PubMed (https://www.ncbi.nlm.nih.gov/pubmed) databases, the pathogenicity of the CNVs were preliminarily classified into five categories: pCNV, lpCNV, VOUS, likely benign CNV (lbCNV), or benign CNV (bCNV). When lpCNV or VOUS was identified in the amniotic fluid samples, we recommended that the biological parents of the fetus underwent CNV-seq (using peripheral blood samples) to determine the origin of the CNV of the fetus. Then, combined with parents' CNV-seq results and phenotypes, we reclassified the pathogenicity of the CNV. The DNA extraction and CNV-seq methods of the peripheral blood samples were performed as described for the amniotic fluid samples.

Confirmatory testing of CNVs

Whole chromosome aneuploidies of 13, 18, 21, X, and Y were confirmed by QF-PCR, while pCNV, lpCNV or VOUS were confirmed using array-based comparative genomic hybridization (aCGH) or a second repeat of CNV-seq. aCGH was performed using the CGX v2 Oligo aCGH Kit (Agilent Technologies, USA). The microarray was scanned using the Agilent SureScan Microarray Scanner (Agilent, USA). Data were extracted using the Agilent CytoGenomics sofware (Agilent, USA) and analyzed using the Genoglyphix Analysis sofware (Perkin Elmer, USA)14.

Informed consent

We confirm that informed consent was obtained from all participants and/or their legal guardians.