Abstract
To evaluate the proportion of chromosomal abnormalities in recurrent pregnancy loss (RPL) assisted by array comparative genomic hybridization (aCGH) bright out with higher detection rate, more accuracy, and less sample failure as compared with conventional cytogenetic analysis. In this study, product of conception samples with abnormal ultrasonogram (USG) findings of the fetus and clinical history of RPL were first processed for karyotyping and fluorescence in situ hybridization (FISH) analysis. Normal results given by karyotype and FISH samples with major anomalies detected by ultrasound with RPL were divided into six groups and aCGH was performed to detect the gain or loss and copy number variations (CNVs) of a particular gene present in chromosomal segments. Among a total of 300 products of conception samples, 100 abnormal samples were identified either by karyotype (n = 70) or by FISH (n = 30). From the remaining 200 samples, 5 showed the presence of maternal cell contamination excluded. aCGH analysis revealed (n = 195) that 74 (38%) samples with CNVs and 2 samples with variants of unknown clinical significance were clinically associated with the clinical findings and 121(62%) samples showed no change in CNVs. The most frequent CNVs were loss of chromosome regions at 2q33.1, 7q11.21, 15q11.1, 16p11.2, Xp22.33, and Yp11.32. CNVs at arr[GRCh37]7p22.3,p21.2(830852-15124702)×1,7q34q36.3(141464180-158909738)×3, 14.2Mbp deletion of 7p22.3p21.2 (SUN1 gene) and 17.4Mbp duplication of 7q34q36.3 (KCNH2, CNTNAP2, and SHH genes) were found in one sample, and CNVs at arr[GRCh37]8p22.2q22.3(86326349-105509986)×1 and 2.48Mbp deletion of 8p22.2q22.3 (GRHL1 gene) were found in another sample.
Lay summary
Recurrent pregnancy loss is considered as two or more consecutive pregnancy losses. Fetal birth defects are thoroughly associated with chromosomal (thread-like structures containing packaged genetic material) abnormalities, which are the underlying causes of pregnancy loss. The evaluation of chromosomal abnormalities is required to diagnose pregnancy loss to improve the prognosis of future pregnancies. The largest proportion of chromosomal abnormalities was observed in the fetal tissue that remains in the uterus after pregnancy. We analyzed 300 retained fetal tissue samples and implicated different methods to recognize the structural abnormalities in the chromosomes. Moreover, simultaneously detect the expression of thousands of genes from fetal tissue. A clinical indication and their association of chromosomal abnormalities were obtained in 38% of cases with assorted fetal ultrasound defects in multiple pregnancy losses and two samples with a variety of unknown clinical indications. It revealed that chromosomal alteration in fetal birth defects is responsible for multiple pregnancy losses.
Introduction
Approximately 10–15% of all clinically recognized pregnancies result in miscarriages, with the majority of losses occurring in the first trimester and about 1% of a couple experiencing recurrent pregnancy loss (RPL) (Viaggi et al. 2013). Chromosomal abnormality has been reported in about 50–60% of spontaneous abortions. The 80–90% rates of early gestation fetal death because of chromosomal abnormalities are numerical such as autosomal abnormalities (>50%), monosomy X (20%), triploidy (15%), and tetraploidy (5%) (Reddy et al. 2012).
In an era of advances in ultrasonography, detection of fetal anomalies becomes precise and provides more information about fetal development. Patients with ultrasound findings of lethal congenital anomalies like cystic hygroma, severe cardiac anomalies, open neural tube defects, and associated secondary soft markers like ventriculomegaly, echogenic bowel, isolated cardiac echogenic foci, and bilateral cyst are frequently referred for chromosomal analysis, which showed a clinical association between abnormal karyotype findings with ultrasound anomalies (Daniel et al. 2003). This chromosomal analysis in patients with abnormal ultrasound and pregnancy loss is important to understand possible chromosomal rearrangement that may lead to RPL or fetuses with congenital abnormalities and provide important information on recurrent risk for future pregnancy.
Classical cytogenetic investigation of products of conception (POC) requires viable tissue but up to 49% of cases undergo culture failure. Moreover, poor chromosome morphology and maternal cell contamination (MCC) can give false results (Baxter & Adayapalam 2013). Standard GTG-banded chromosomal karyotype at 450–600 band resolution can detect only whole chromosome aneuploidy and balance translocation >5 Mb deletion and duplication (Lomax et al. 2000). Fluorescence in situ hybridization (FISH) can detect smaller size abnormalities in the range of 50–150 kb, but it is limited by probe availability (Harris et al. 2011). Array comparative genomic hybridization (aCGH) does not require live cells; it can be used on DNA from tissue samples. It is a high-resolution genome analysis technique to detect genomic imbalances. The deletions (loss) and duplications (gain) result in a change in the number of copies of the genome known as copy number variation (CNVs). aCGH increases the finding capacity of genetic alteration which is responsible for fetal death by providing coverage of the entire genome at higher resolution by detection limits as small as 50–100 kb deletions and duplications (Strassberg et al. 2011).
However, many studies (Daniel et al. 2003, Tyreman et al. 2009, Shaffer et al. 2012, De Wit et al. 2014) have focused specifically on the use of microarray analysis in fetuses with abnormal ultrasound findings; there is little information available on the association of ultrasound anomalies causative submicroscopic CNVs (Tyreman et al. 2009, De Wit et al. 2014). Hence, the current study focused on correlating the chromosomal abnormalities in POC samples using aCGH with the ultrasound anomalies detected in RPL cases.
Methods
Sample collection
POC samples were collected from different hospitals after medically terminated pregnancies and pregnancies that ended in spontaneous fetal death with a history of RPL. Written consent was taken during sample collection, and POC samples were collected in a sterile container with normal saline and antibiotics. All POC samples were divided into six groups (Table 1) based on anomalies detected in ultrasound and history of RPL. Human ethical committee approval for this study was obtained from Gujarat University, Ahmedabad, Gujarat, India vide approval no. GU/IEC/01/2018.
Groups on the basis of USG findings with RPL and results of aCGH. Results are presented as n (%).
| Groups | Anomalies with RPL | Samples | MCC, n | Results of aCGH | ||
|---|---|---|---|---|---|---|
| Yes* | No | Normal | Abnormal | |||
| I | CHD | 23 (11.5%) | 1 | 22 | 8 (36.3%) | 14 (63.6%) |
| II | CNS anomalies | 10 (5%) | 1 | 9 | 4 (44.4%) | 5 (55.5%) |
| III | MUS anomalies | 12 (6%) | 0 | 12 | 5 (41.6%) | 7 (58.3%) |
| IV | IN/BF | 30 (15%) | 1 | 29 | 12 (41.3%) | 17 (58.6%) |
| V | IUGR | 9 (4.5%) | 0 | 9 | 7 (77.7%) | 2 (22.2%) |
| VI | Other US anomalies | 116 (58%) | 2 | 114 | 85 (74.5%) | 29 (25.4%) |
| All anomalies | 200 (100%) | 5 | 195 | 121 (62%) | 74 (38%) | |
*aCGH not performed.
aCGH, array comparative genomic hybridization; CHD, congenital heart defect; CNS, central nervous system; IN/BF, increased nuchal translucency/body fluids; IUGR, intrauterine growth retardation; MCC, maternal cell contamination; MUS, musculoskeletal; US, ultrasound.
Chromosome analysis
Karyotyping from POC samples using GTG banding was performed by standard protocol (Levy et al. 2009). Nutrient Mixture F-10 Ham complete medium (HiMedia, AL083A) was used for POC sample culture with sodium bicarbonate, 25 mM HEPES buffer, l-glutamine, 20 mL fetal calf serum (HiMedia, RM9955), and 1 mL antibiotics (penicillin–streptomycin) (HiMedia, A001). Cultured cells were harvested using 30 µL colchicine (HiMedia, TCL062) for 3 h, followed by hypotonic treatment of 0.56% KCL solution for 15 min and fixation using standard 3:1 methanol:acetic acid fixative. Standard GTG banding technique (Francke & Oliver 1978) was performed with slight modification. Microscopic examination of 20 metaphases for each case was carried out using the software Applied Spectral Imaging (ASI) in the Olympus BX53 microscope.
Array comparative genomic hybridization
Genomic DNA was extracted from tissue samples using a QIAamp DNA Mini kit (Qiagen, ID 56304). MCC was assessed using information derived from the increased repeat number of the variable number tandem repeat (VNTR) locus of mother and fetus DNA. Samples with significant MCC were excluded from this study.
aCGH was performed using two different arrays.
(i) 4 × 180K SurePrint G3 CGH+SNP Microarray, containing approximately 120,000 CGH probes and 60,000 SNP probes in the Agilent SureScan Microarray Scanner Bundle instrument. Genotypes on this array were measured using one SNP probe per SNP, providing ∼ 5–10 Mb resolution for loss of heterozygosity (LOH)/uniparental disomy (UPD) detection across the entire genome. The CNVs/microdeletion were based on the overall median probe spacing with 25 kb in Refseq genes (Sureprint G3 Human CGH+SNP Microarray, Agilent Technologies). Probes vary in oligonucleotide size to represent areas of interest (25–85 base pairs).
(ii) 8 × 60K SurePrint G3 Human CGH Microarray, containing approximately 60,000 oligos covering the whole genome. The report represents only the CNVs based on the overall median probe spacing with 41 kb in Refseq genes (SurePrint G3 Human CGH Microarray, Agilent Technologies). The genomic backbone resolution of this array is 200 kb which is capable to detect larger than 400 kb.
The level of resolution was determined by considering both probe size and the genomic distance between DNA probes. Genomic DNA from the tissue samples and a control sample are differentially labeled fluorescent dyes and hybridized to the oligos.
Statistical analysis
Microarray slides were scanned with a DNA Microarray Scanner (Agilent Technologies). Data were received using the Agilent Feature Extraction software and Agilent CytoGenomics software. Data were analyzed using quantitative imaging methods and analytical software (Agilent CytoGenomics Software, Agilent Technologies) to assist in identifying each targeted DNA sequence as a loss of copy number (deletion), a gain of copy number (duplication), or a normal copy number. CNVs were detected using the ADM-2 algorithm with filters of the minimal size of 200 kb in the region, >5 Mb of copy number LOH. Genomic positions were estimated using the human genomic reference sequence GRCh37.
Results
In the present study, a total of 300 POC samples from a female with a history of RPL showing abnormal ultrasound were first used to detect chromosomal abnormalities by karyotyping. In total, 190 (63%) samples were successfully cultured for chromosomal analyses, whilst cells from 110 (36.7%) samples failed to proliferate. Chromosomal analysis of the 190 samples revealed that 120 (63%) samples showed normal karyotype and 70 (37%) samples showed abnormal karyotype. The samples that could not be cultured (110) were further considered for the FISH study to detect the most common anomalies, that is, abnormalities in chromosomes 13, 18, 21, X, and Y. The results showed that 80 (73%) samples were normal and 30 (27%) samples showed aneuploidy. The chromosomal abnormalities found in 100 (33%) samples by both karyotype (70) and FISH (30) were not considered for further aCGH analysis. The 200 samples diagnosed as cytogenetically normal in karyotype or FISH and samples with no results, only for chromosomes 13, 18, 21, X, and Y normal in FISH were further used for microarray (aCGH) analysis. Of these 200 samples, 5 samples were excluded due to MCC (Fig. 1).
The remaining 195 POC samples from females having RPL with no maternal cell contamination (NMCC) were divided into six groups according to the results obtained from ultrasound. They were Group I: congenital heart defect (CHD) (11.5%), Group II: central nervous system anomalies (CNS Anom.) (5%), Group III: musculoskeletal anomalies (MUS. Anom.) (6%), Group IV: increased nuchal translucency/body fluids (IN/BF) (15%), Group V: intrauterine growth retardation (IUGR) (4.5%), and Group VI: other ultrasound anomalies (Oth. Anom.) (58%) (Table 1).
Overall aCGH observation depicted that the normal results were obtained in 121 (62%) out of 195 POC samples. From these 195 POC samples, the abnormal aCGH results (i.e. pathogenic, likely pathogenic, and variants of uknown significance) were obtained in 74 (38%) samples. Of these 74 abnormal aCGH samples, 30 (40.5%) samples showed single aneuploidy except for trisomy of chromosomes 13, 18, 21, X, and Y which was excluded during the FISH study, and 44 (59.5%) POC samples showed structural chromosomal aberrations (Fig. 1; Table 2).
Aneuploidy detected in POC samples by aCGH.
| Groups | Trisomy 7 | Trisomy 9 | Trisomy 16 | Trisomy 22 | Triploidy | Total |
|---|---|---|---|---|---|---|
| I | 2 | 1 | 1 | 3 | 2 | 9 (30.0%) |
| II | – | 2 | 1 | 1 | 1 | 5 (16.6%) |
| III | 1 | – | – | – | 1 | 2 (06.6%) |
| IV | 1 | - | 1 | 2 | 1 | 5 (16.6%) |
| V | – | 1 | – | – | – | 1 (03.3%) |
| VI | 3 | 1 | 1 | 2 | 1 | 8 (26.6%) |
| Total | 7 (23.3%) | 5 (16.7%) | 4 (13.3%) | 8 (26.7%) | 6 (20.0%) | 30 (100.0%) |
Out of the 121 normal aCGH samples, the group-wise distribution was as follows - CHD: 8 (6.6%); CNS Anom.: 4 (3.3%); MUS. Anom: 5 (4.1%; IN/BF: 12 (10%); IUGR: 7 (5.8%) and Oth. + RPL: 85 (70.2%) (Fig. 2A). On the other hand, the group-wise distribution of the 74 abnormal aCGH was - CHD: 14 (19%); CNS Anom.: 5 (6.8%); MUS. Anom.: 7 (9.5%);, IN/BF: 17 (23%); IUGR: 2 (2.7%) and Oth. + RPL: 29 (39%) (Fig. 2B).
(A) Groupwise normal results of aCGH. (B) Groupwise abnormal results of aCGH.
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
(A) Groupwise normal results of aCGH. (B) Groupwise abnormal results of aCGH.
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
(A) Groupwise normal results of aCGH. (B) Groupwise abnormal results of aCGH.
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
Out of the 74 abnormal aCGH samples, 30 samples (40.5%) showed aneuploidy. Among these, 30 aneuploidies, 24 samples showed trisomy in the chromosome (Trisomy 7 (23.3%), Trisomy 9 (16.7%), Trisomy 16 (13.3%), and Trisomy 22 (26.7%)) whereas, 6 (20%) samples showed triploidy (Table 2). The remaining 44 abnormal aCGH out of the 74 abnormal aCGH samples showed well-documented microdeletion and microduplication in various CNV regions (Table 3). Conversely, two samples (S-278 and S-251) showed novel CNVs and were considered a variant of unknown significance (VOUS) (Table 4).In a case of RPL with increased nuchal translucency detected by ultrasound, sample results revealed deletion at 7p22.3 and duplication at 7q34 in chromosome 7 (Fig. 3a), involvement of gene at 7p22.3 deletion with SUN1 gene is clinically associated with RPL (Fig. 3b), involvement of genes at 7q34 duplication with CNTNAP2, KCNH2, and SHH genes associated with fetal death (Fig. 3c). Another case of RPL with three prior losses, sample results with deletion at 8q22.2 in chromosome 8 (Fig. 4a), involvement of genes at 8q22.2 deletion with GRHL2 gene is clinically associated with neural tube defect (Fig. 4b).
(A) A case with involvement of deletion at 7p22.3 and duplication at 7q34 in chromosome 7. (B) A case with involvement of genes at 7p22.3 deletion. The SUN1 gene is clinically associated with RPL (OMIM). (C) A case with involvement of genes at 7p22.3 deletion. The SUN1 gene is clinically associated with RPL (OMIM).
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
(A) A case with involvement of deletion at 7p22.3 and duplication at 7q34 in chromosome 7. (B) A case with involvement of genes at 7p22.3 deletion. The SUN1 gene is clinically associated with RPL (OMIM). (C) A case with involvement of genes at 7p22.3 deletion. The SUN1 gene is clinically associated with RPL (OMIM).
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
(A) A case with involvement of deletion at 7p22.3 and duplication at 7q34 in chromosome 7. (B) A case with involvement of genes at 7p22.3 deletion. The SUN1 gene is clinically associated with RPL (OMIM). (C) A case with involvement of genes at 7p22.3 deletion. The SUN1 gene is clinically associated with RPL (OMIM).
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
(A) A case with involvement of deletion at 8q22.2 in chromosome 8. (B) A case with involvement of gene at 8q22.2 deletion. The GRHL2 gene is clinically associated with neural tube defect (OMIM).
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
(A) A case with involvement of deletion at 8q22.2 in chromosome 8. (B) A case with involvement of gene at 8q22.2 deletion. The GRHL2 gene is clinically associated with neural tube defect (OMIM).
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
(A) A case with involvement of deletion at 8q22.2 in chromosome 8. (B) A case with involvement of gene at 8q22.2 deletion. The GRHL2 gene is clinically associated with neural tube defect (OMIM).
Citation: Reproduction and Fertility 4, 2; 10.1530/RAF-22-0092
List of samples found with frequent CNV region of microdeletion, microduplication, and the major genes involved.
| Groups | Sample ID | Microarray results | Genes | |||
|---|---|---|---|---|---|---|
| ISCN format | CNV region | Gain/loss | Copy number, kb | |||
| I | S-27 | arr[GRCh37]7q11.21(62,516,153-64,500,004)×3 | 7q11.21 | Gain | 1,984 | ELN, FKBP6, GTF2I |
| S-85 | arr[GRCh37]15q11.1q11.2(20,408,283-22,586,951)×1 | 15q11.1 | Loss | 2,179 | BCL8, HERC2P3, GOLGA6L6 | |
| S-105 | arr[GRCh37]Xp22.33(428231-2362192)×1 | Xp22.33 | Loss | 1,934 | SHOX, TBL1Y, CSF2RA, PRKX | |
| Yp11.32(378231-2312192)×1 | Yp11.32 | Loss | 1,934 | SHOX, TBL1Y, CSF2RA, PRKY | ||
| S-172 | arr[GRCh37]8p23.1(11,373,941-12,241,152)×1 | 8p23.1 | Loss | 1,388 | GATA4, BLK, FDFT1, CTSB | |
| 6p21.33p21.32(31,847,522-33,235,568)×1 | 6p21.33 | Loss | 0,867 | CYP21A2 | ||
| S-206 | arr[GRCh37]17p12(14,111,772-15,442,066)×3 | 17p12 | Gain | 1,330 | SCO2, COX10, PMP22, HS3ST3BI | |
| III | S-12 | arr[GRCh37]22q11.1q13.33(17,096,855-51,178,264)×3 | 22q11.1 | Gain | 34,081 | TBX1 |
| S-30 | arr[GRCh37]4p16.1(7,032,501-8,856,535)×3 | 4p16.1 | Gain | 1,824 | HMX1 | |
| Xp22.33(298,292-2,656,392)×1 | Xp22.33 | Loss | 2,358 | SHOX,TBL1Y, CSF2RA, PRKX | ||
| Yp11.32p11.31(248,292-2,606,392)×1 | Yp11.32 | Loss | 2,358 | SHOX, TBL1Y, CSF2RA, PRKY | ||
| S-97 | arr[GRCh37]22q11.22(23,056,562-23,152,475)×3 | 22q11.2 | Gain | 0,096 | TBX1 | |
| S-152 | arr[GRCh37]10q11.22(47011584-47604877)×1 | 10q11.22 | Loss | 0,593 | PPYR1, ANXA8, ANXA8L1, CHAT | |
| S-200 | arr[GRCh37]17p13.1p12(9,408,448-12,453,402)×3 | 17p13.1 | Gain | 3,045 | PAFAH1B1 | |
| IV | S-35 | arr[GRCh37]2q37.3(242,930,600-242,948,040)×1 | 2q37.3 | Loss | 1,172 | HDAC4 |
| S-61 | arr[GRCh37]22q11.21(18,919,942-21,440,514)×3 | 22q11.21 | Gain | 2,521 | TBX1 | |
| S-82 | arr[GRCh37]2p16.3(50,108,925-52,639,734)×3 | 2p16.3 | Gain | 2,531 | PEX13 | |
| S-99 | arr[GRCh37]4p16.1(7,032,501-8,868,815)×3 | 4p16.1 | Gain | 1,836 | HMX1 | |
| S-126 | arr[GRCh37]15q11.1q11.2(20,575,646-22,509,254)×1 | 15q11.1 | Loss | 1,934 | BCL8, HERC2P3, GOLGA6L6 | |
| S-155 | arr[GRCh37]Yp11.2(6433155-9073289)×1 | Yp11.2 | Loss | 2,640 | ZFY,PAR1 | |
| S-235 | arr[GRCh37]2p16.1p15(60,932,053-62,727,947)×1 | 2p16.1 | Loss | 1,796 | REL | |
| 16q22.1(68,550,119-70,689,998)×1 | 16q22.1 | Loss | 2,140 | HYDIN | ||
| 17q22q23.2(56,504,240-59,001,799)×1 | 17q22 | Loss | 2,498 | SEPT4, TEX14, TUBD1, PPMID | ||
| S-240 | arr[GRCh37]1p32.3(50,987,414-53,427,232)×1 | 1p32.3 | Loss | 2,440 | NFIA, CASP8, CASP10, BMPR2 | |
| 2q33.1q33.2(201,726,125-204,493,161)×1 | 2q33.1 | Loss | 2,767 | NSD2, LETM1, MSX1 | ||
| 4p14(38,858,095-41,087,376)×1 | 4p14 | Loss | 2,229 | |||
| S-278 | arr[GRCh37]7p22.3p21.2(830852-15124702)×1 | 7p22.3 | Loss | 14,494 | SUN1 | |
| 7q34q36.3(141464180-158909738)×3 | 7q34 | Gain | 7,446 | KCNH2, CNTNAP2, SHH | ||
| S-282 | arr[GRCh37]4q35.2(189,374,480-190,469,337)×1 | 4q35.2 | Loss | 1,095 | SORBS2, HELT, FAT | |
| 15q11.1q11.2(20,102,541-22,509,254)×3 | 15q11.1 | Gain | 2,406 | BCL8, HERC2P3, GOLGA6L6 | ||
| S-289 | arr[GRCh37]10p11.3p11(327273-24599490)×1 | 10p11.3 | Loss | 24,272 | GZMM, BSG, HCN2 | |
| 19q13.1q13.43(32879003-59022461)×1 | 19q13.1 | Loss | 26,143 | RGS9BP, SLC7A9, LRP3 | ||
| S-299 | arr[GRCh37]Xp22.33(298,292-2,362,192)×1 | Xp22.33 | Loss | 2,064 | SHOX, TBL1Y, CSF2RA, PRKX | |
| Yp11.32(248,292-2,312,192)×1 | Yp11.32 | Loss | 2,064 | SHOX, TBL1Y, CSF2RA, PRKY | ||
| V | S-215 | arr[GRCh37]10q26.3(131,266,049-135,211,969)×3 | 10q26.3 | Gain | 3,946 | FGFR2 |
| VI | S-15 | arr[GRCh37]6p22.2p22.1(26,124,966-27,951,976)×3 | 6p22.2 | Gain | 1,827 | HIST1H1E, HIST1H3B, HIST1H2BK |
| 7q11.21(64,691,936-66,824,455)×1 | 7q11.21 | Loss | 2,133 | ELN, FKBP6, GTF2I | ||
| S-21 | arr[GRCh37]Xp22.33(298,292-2,362,192)×1 | Xp22.33 | Loss | 2,595 | SHOX, TBL1Y, CSF2RA, PRKX | |
| S-32 | arr[GRCh37]15q11.1q11.2(20,575,646-22,509,254)×1 | 15q11.1 | Loss | 1,934 | BCL8, HERC2P3, GOLGA6L6 | |
| S-44 | arr[GRCh37]2q33.1q33.2(201,588,352-204,631,833)×1 | 2q33.1 | Loss | 3,043 | CASP8, CASP10, BMPR2 | |
| 3p21.31(47,084,157-50,323,638)×1 | 3p21.31 | Loss | 3,239 | SETD2, CSPG5, PTH1R, SMARCC1 | ||
| 7q11.21q21.11(64,021,202-77,541,026)×1 | 7q11.21 | Loss | 13,520 | ELN, FKBP6, GTF2I | ||
| 7q21.31q22.3(97,495,807-105,356,164)×1 | 7q21.31 | Loss | 7,860 | SHFM1, EPS15L1 | ||
| 12q24.23q24.31(120,670,596-124,028,560)×1 | 12q24.23 | Loss | 3,358 | KDM2B, SETD1B | ||
| 19p13.3p12(1,905,897-21,884,318)×1 | 19p13.3 | Loss | 19,978 | MYB, BCL11A, KLF1 | ||
| 20p11.22p11.21(21,494,433-24,990,081)×3 | 20p11.22 | Gain | 3,358 | JAG1 | ||
| S-50 | arr[GRCh37]4q12q13.2(59,369,165-68,913,048)×3 | 4q12 | Gain | 9,544 | EPHA5, GNRHR, LPHN3 | |
| S-52 | arr[GRCh37]7q11.21q11.23(63,108,972-76,787,351)×1 | 7q11.21 | Loss | 13,678 | ELN, FKBP6, GTF2I | |
| Xp22.33(61091-2475574)×1 | Xp22.33 | Loss | 2,414 | SHOX, TBL1Y, CSF2RA, PRKX | ||
| Yp11.32(11091-2425574)×1 | Yp11.32 | Loss | 2,414 | SHOX, TBL1Y, CSF2RA, PRKY | ||
| S-65 | arr[GRCh37]2p25.3p25.2(827606-6135163)×3 | 2p25.3 | Gain | 5,308 | LETM1, WHSC1, HDAC4, PTPN11 | |
| 16p13.3p13.2(5916887-7964338)×3 | 16p13.3 | Gain | 2,047 | CREBBP | ||
| S-74 | arr[GRCh37]9q34.11q34.3(132995601-140954147)×3 | 9q34.11 | Gain | 7,959 | TSC1,TTF1, RXRA, NCS1, ASSI, ABL1 | |
| 16p13.3(96766-589853)×1 | 16p13 | Loss | 1,493 | HBA1, HBA2, MPG, HBX, HBM, HBQ1 | ||
| S-77 | arr[GRCh37]5q13.1q13.2(68393966-70754834)×1 | 5q13.1 | Loss | 2,361 | SMN1, SMN2, MARVELD2, NAIP5 | |
| S-80 | arr[GRCh37]16p11.2(32066967-33773163)×1 | 16p11.2 | Loss | 1,706 | HERC2P4, TP53TG3 | |
| S-92 | arr[GRCh37]4q35.2(187,370,897-190,896,674)×3 | 4q35.2 | Gain | 3,526 | HPGD, TRAPPC11, CYP4V2, TLR3 | |
| S-96 | arr[GRCh37]16p11.2(32,624,578-33,604,468)×1 | 16p11.2 | Loss | 1,980 | HERC2P4, TP53TG3 | |
| S-138 | arr[GRCh37]4q32.3q35.2(164,775,281-190,896,674)×3 | 4q32.3 | Gain | 26,121 | ANP32C, KLHL2, SC4MOL, CPE, TLLI | |
| 13q31.3q34(92,416,837-115,059,020)×1 | 13q31.3 | Loss | 22,642 | GPC5, GPC6, DCT, SOX21, ABCC4 | ||
| S-157 | arr[GRCh37]2p25.3p11.1(39,193-19,815,738)×3 | 2p25.3 | Gain | 91,777 | FAM11OC, H3YL1, ACP1 | |
| 2q11.2q37.3(97,260,682-243,041,364)×3 | 2q11.2 | Gain | 14,578 | KIAA1310, ERIL5, LMAN2L | ||
| 22q11.1q13.33(17,528,442-49,752,893)×3 | 22q11.1 | Gain | 32,224 | CECR7, IL17RA, TBX1, CECR6 | ||
| S-179 | arr[GRCh37]8q11.21q11.23(50,558,946-52,753,592)×3 | 8q11.21 | Gain | 2,195 | PCMTD1 | |
| S-186 | arr[GRCh37]5q13.1q13.2(68393966-70754834)×1 | 5q13.1 | Loss | 2,361 | MARVELD2, SMN1, SMN2 | |
| S-215 | arr[GRCh37]22q11.21(18,919,942-21,440,514)×3 | 22q11.21 | Gain | 2,521 | PRODH, DGCR2, TSSK2, SLC25A1, CLTCL1, HIRA, UFD1L, CLDN5, SEPT5, GP1BB, TBX1, GNB1L | |
| S-236 | arr[GRCh37]2p16.1p15(60,932,053-62,727,947)×1 | 2p16.1 | Loss | 1,796 | REL, PEX13, XPO1, CCT4, COMMD1 | |
| 16q22.1(68,550,119-70,689,998)×1 | 16q22.1 | Loss | 2,140 | |||
| S-251 | arr[GRCh37]8p22.2q22.3(86326349-105509986)×1 | 8q22.2 | Loss | 2,480 | GRHL2 | |
| S-267 | arr[GRCh37]8p23.1(11,373,941-12,241,152)×1 | 8p23.1 | Loss | 1,388 | EHMT2,ZBTB12 | |
| 6p21.33p21.32(31,847,522-33,235,568)×1 | 6p21.33 | Loss | 1,867 | BLK, GATA4, FDFT1,CTSB | ||
| S-288 | arr[GRCh37]17q22q23.2(56,504,240-59,001,799)×1 | 17q22 | Loss | 2,498 | ZFP90, CDH3, CDH1, HAS3, CIRH1A, SNTB2,VPS4A,PDF, COG8, TERF2, CYB5B | |
Novel CNVs detected by aCGH which are clinically classified as variant of unknown significance (VOUS).
| Groups | aCGH results | ||||
|---|---|---|---|---|---|
| ISCN format | CNV region | Gain/loss | Copy number (kb) | Genes involved | |
| IV | arr[GRCh37]7p22.3p21.2(830852-15124702)×1 | 7p22.3 | Loss | 14,494 | SUN1 |
| 7q34q36.3(141464180-158909738)×3 | 7q34 | Gain | 7446 | KCNH2, CNTNAP2, SHH | |
| VI | arr[GRCh37]8p22.2q22.3(86326349-105509986)×1 | 8q22.2 | Loss | 2480 | GRHL2 |
Discussion
Various genetic methods other than the conventional G-banding karyotyping have been proposed for the cytogenetic analysis of POC, including FISH and DNA-based analysis such as aCGH, quantitative fluorescent polymerase chain reaction, and multiplex ligation-dependent probe amplification (Van den Berg et al. 2012). Each of these methods, however, has some drawbacks and is insufficient when used in isolation. In the conventional G-banding method, a relatively high rate of culture failure, overgrowth of maternal cells, and limited banding resolution are major shortcomings (Van den Berg et al. 2012). Culture failure may frequently occur in spontaneously discharged specimens because of poor viability. Previous reports reported that cytogenetic results are obtained in only 45–65% of stillbirths because of culture failure (Raca et al. 2009). Because microarray does not require live cells, it can be used on DNA from macerated tissues, which will increase the yield of finding a genetic cause of fetal death. Studies on first-trimester spontaneous abortion tissue with aCGH revealed abnormalities not commonly seen by karyotype, including a high rate of double aneuploidy and autosomal monosomy (Schaeffer et al. 2004, Benkhalifa et al. 2005).
In the present study, overall aCGH observation depicted that the normal results were obtained in 121 (62%) out of 195 POC samples. In the study of Ozawa et al., aCGH analysis was successful in all 15 cases and 10 cases had abnormal results: gain in copy number (n = 7) and loss in copy number (n = 3). Most of them were estimated to be whole chromosome aneuploidy, whereas one case was compatible with microdeletion. (Ozawa et al. 2016). Two cases were suspected to be male diploid contaminated by maternal DNA or triploid because of the unsatisfactory signal patterns on X/Y chromosomes. Two of three cases with normal female DNA patterns were identified to be contaminated with maternal DNA by the additional analysis of short tandem repeats.
MCC is a major troublesome obstacle to cytogenetic analysis of POC. To minimize contamination, it is extremely important to completely separate chorionic villi from maternal tissues (Lathi & Milki 2002). In spontaneously discharged POC, however, accurate identification of chorionic villi is often difficult due to extensive tissue degeneration. In the present study, 5 out of 200 samples (2.5%) had MCC. In the study of Ozawa et al., 2 of 15 miscarriage specimens were suspected to be contaminated with maternal DNA by GDA analysis, and the additional STR analysis revealed that 2 of 3 normal female cases contained mostly maternal DNA. Thus, the confirmatory assay should be additionally performed to evaluate maternal contamination, especially in the specimens with normal female array results. DNA extracted for aCGH is also applicable for STR analysis (Ozawa et al. 2016).
Genomic microarray has proven successful in providing genetic information in cases of pregnancy loss when cell culture failure makes cytogenetic evaluation difficult. aCGH has been successfully applied to the study of samples from early pregnancy loss (before 20 weeks of gestation). Benkhalifa et al. studied cases of spontaneous abortions that failed to grow in culture and detected abnormalities in more than half of the samples (15/26, 57%) (Benkhalifa et al. 2005). Schaeffer used aCGH to study 41 samples from POC that were previously analyzed by G-banding. They were able to confirm all abnormalities seen by classic cytogenetics and detect new abnormalities in 9.8% of the cases that were not seen by karyotype analysis (Schaeffer et al. 2004). Warren et al. examined CNVs in 35 women who experienced pregnancy loss between 10 and 20 weeks of gestation using aCGH. Of the 35 fetal deaths, 9 had a normal karyotype and 26 did not undergo karyotype evaluation. DNA was successfully isolated in 30 of 35 (86%) cases. aCGH detected de novo CNVs in 13% of cases where routine cytogenetic testing was normal or not performed (Warren et al. 2009).
Although array-CGH has several advantages compared to the conventional G-banding analysis, it is more costly and incapable of finding some polyploidies such as 69,XXX, 92,XXXX, 92,XXYY, and balanced translocations (Van den Berg et al. 2012, Dhillon et al. 2014). Polyploidies cause about 5–10% of all first-trimester miscarriages (Van den Berg 2012). Balanced abnormalities are not the cause of miscarriage, but the findings suggest that if a parent could be a balanced carrier then the results of parental karyotyping become useful information for future pregnancies. Therefore, the traditional G-banding analysis should be primarily performed for the specimens collected by dilation and curettage for the identification of balanced chromosomal translocation derived from parents in the setting of RPL. Other than that if the quality of specimens is poor for the G-banding analysis due to spontaneous discharge, array-CGH supplemented with STR analysis can be a powerful technique for cytogenetic evaluation of miscarriage, since STR analysis allows for the detection of triploidy and part of tetraploidy as well as maternal contamination (Ozawa et al. 2016).
Previously described difficulties with detecting triploidy on SNP+CGH arrays were surmounted through the creation of histograms of B allele frequencies for all SNP array loci. Histograms from diploid samples have peaks at 0, 0.5, and 1, whereas triploid samples show peaks at 0, 0.33, 0.67, and 1, thus distinguishing diploid from triploid. Therefore, chromosomal microarrays that include SNP analysis can detect almost all clinically significant non-mosaic, and some mosaic, chromosome abnormalities, including polyploidy (Bug et al. 2014).
In the study of Rosenfeld et al., aCGH was performed successfully on 515 samples out of 535 fetal demise specimens of any gestational age and a subset of 107 specimens underwent additional single nucleotide polymorphism (SNP) analysis. Out of them, normal karyotype was observed in 288 samples (55.9%), abnormal karyotype in 22 samples (4.3%), and failed karyotype in 205 samples (39.8%). Out of the 288 normal karyotype samples, aCGH showed a normal array in 255 samples (88.5%), a pathogenic array in 20 samples (6.9%), VOUS in 11 samples (3.8%), and an abnormal SNP array in 2 samples out of 48 samples (4.2%). Therefore, the detection rate was 11.4% (Rosenfeld et al. 2015).
Fewer studies of the role of microarray in the evaluation of stillbirth have been performed. Raca et al. observed that aCGH detected 2 abnormalities among 15 tested stillbirths (trisomy 21 and an unbalanced translocation between chromosomes 3 and 10), for an overall detection rate of 13% in stillborns with malformations (Raca et al. 2009).
Although chromosomal microarrays offer higher success rates than traditional cytogenetic analysis, they can still fail. One study of stillbirths encountered 12.6% microarray failure, significantly less than the 29.5% of karyotypes failing. Rosenfeld et al. showed an 8.3% aCGH failure rate, due to poor DNA quality or MCC which was higher than our study (2.5%) (Rosenfeld et al. 2015).
Conclusion
Conventional cytogenetic analysis with G-banding is the primary testing for RPL for the evaluation of inherited chromosomal balanced translocation. But apart from that, the increased yield of diagnostic utility, aCGH testing makes it the preferable method for delineating chromosomal etiologies of pregnancy loss at any gestational age. It is a powerful molecular cytogenetic technique for quickly scanning through an entire genome for imbalances. It allows to perform chromosome analysis on DNA extracted from direct fetal tissue without the need for live dividing cells. It detects almost all abnormalities found by traditional cytogenetic analysis, while chromosomal abnormalities that were missed by karyotypingwere detected by aCGH. The increased detection rate makes aCGH a better diagnosis technique, particularly in RPL cases.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This study did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
Author contribution statement
KG: Data acquisition, data analysis, manuscript preparation, manuscript review; AP: Study design, literature search, data analysis, data acquisition, manuscript preparation, manuscript review; BP: Literature search, manuscript review; SC: Conception, study design, literature search, manuscript review; DJ: Conception, study design, literature search, manuscript editing, manuscript review.
References
Baxter L & & Adayapalam N 2013 A comparative study of standard cytogenetic evaluation and molecular karyotyping for products of conception. Diagnostic Molecular Pathology 22 228–235. (https://doi.org/10.1097/PDM.0b013e31829265ab)
Benkhalifa M, Kasakyan S, Clement P, Baldi M, Tachdjian G, Demirol A, Gurgan T, Fiorentino F, Mohammed M & & Qumsiyeh MB 2005 Array comparative genomic hybridization profiling of first-trimester spontaneous abortions that fail to grow in vitro. Prenatal Diagnosis 25 894–900. (https://doi.org/10.1002/pd.1230)
Bug S, Solfrank B, Schmitz F, Pricelius J, Stecher M, Craig A, Botcherby M & & Nevinny-Stickel-Hinzpeter C 2014 Diagnostic utility of novel combined arrays for genome-wide simultaneous detection of aneuploidy and uniparental isodisomy in losses of pregnancy. Molecular Cytogenetics 7 43. (https://doi.org/10.1186/1755-8166-7-43)
Daniel A, Athayde N, Ogle R, George AM, Michael J, Pertile MD, Bryan J, Jammu V & & Trudinger BJ 2003 Prospective ranking of the sonographic markers for aneuploidy: data of 2143 prenatal cytogenetic diagnoses referred for abnormalities on ultrasound. Australian and New Zealand Journal of Obstetrics and Gynaecology 43 16–26. (https://doi.org/10.1046/j.0004-8666.2003.00025.x)
De Wit MC, Srebniak MI, Govaerts LC, Van Opstal D, Galjaard RJ & & Go AT 2014 Additional value of prenatal genomic array testing in fetuses with isolated structural ultrasound abnormalities and a normal karyotype: a systematic review of the literature. Ultrasound in Obstetrics and Gynecology 43 139–146. (https://doi.org/10.1002/uog.12575)
Dhillon RK, Hillman SC, Morris RK, McMullan D, Williams D, Coomarasamy A & & Kilby MD 2014 Additional information from chromosomal microarray analysis (CMA) over conventional karyotyping when diagnosing chromosomal abnormalities in miscarriage: a systematic review and meta-analysis. BJOG 121 11–21. (https://doi.org/10.1111/1471-0528.12382)
Francke U & & Oliver N 1978 Quantitative analysis of high-resolution trypsin-Giemsa bands on human prometaphase chromosomes. Human Genetics 45 137–165. (https://doi.org/10.1007/BF00286957)
Harris RA, Ferrari F & & Ben-Shachar S 2011 Genome-wide array-based copy number profiling in human placentas from unexplained stillbirths. Prenatal Diagnosis 31 932– 944. (https://doi.org/10.1002/pd.2817)
Lathi RB & & Milki AA 2002 Tissue sampling technique affects accuracy of karyotype from missed abortions. Journal of Assisted Reproduction and Genetics 19 536–538. (https://doi.org/10.1023/a:1020916003666)
Levy B, Jobanputra V & & Warburton D 2009 Preparation and culture of products of conception and other solid tissues for chromosome analysis. Current Protocols in Human Genetics 60 8.5.1–8.5.11. (https://doi.org/10.1002/0471142905.hg0805s60)
Lomax B, Tang S, Separovic E, Phillips D, Hillard E, Thomson T & & Kalousek DK 2000 Comparative genomic hybridization in combination with flow cytometry improves results of cytogenetic analysis of spontaneous abortions. American Journal of Human Genetics 66 1516–1521. (https://doi.org/10.1086/302878)
Ozawa N, Sago H, Matsuoka K, Maruyama T, Migita O, Aizu Y & & Inazawa J 2016 Cytogenetic analysis of spontaneously discharged products of conception by array-based comparative genomic hybridization. SpringerPlus 5 874. (https://doi.org/10.1186/s40064-016-2594-6)
Raca G, Artzer A & & Thorson L 2009 Array-based comparative genomic hybridization (ACGH) in the genetic evaluation of stillbirth. American Journal of Medical Genetics. Part A 149A 2437–2443. (https://doi.org/10.1002/ajmg.a.33083)
Reddy UM, Page GP & & Saade GR 2012 The role of DNA microarrays in the evaluation of fetal death. Prenatal Diagnosis 32 371–375. (https://doi.org/10.1002/pd.3825)
Rosenfeld JA, Tucker ME, Escobar LF, Neill NJ, Torchia BS, McDaniel LD, Schultz RA, Chong K & & Chitayat D 2015 Diagnostic utility of microarray testing in pregnancy loss. Ultrasound in Obstetrics and Gynecology 46 478–486. (https://doi.org/10.1002/uog.14866)
Schaeffer AJ, Chung J, Heretis K, Wong A, Ledbetter DH & & Lese Martin C 2004 Comparative genomic hybridization-array analysis enhances the detection of aneuploidies and submicroscopic imbalances in spontaneous miscarriages. American Journal of Human Genetics 74 1168–1174. (https://doi.org/10.1086/421250)
Shaffer LG, Rosenfeld JA, Dabell MP, Coppinger J, Bandholz AM, Ellison JW, Ravnan JB, Torchia BS, Ballif BC & & Fisher AJ 2012 Detection rates of clinically significant genomic alterations by microarray analysis for specific anomalies detected by ultrasound. Prenatal Diagnosis 32 986–995. (https://doi.org/10.1002/pd.3943)
Strassberg M, Fruhman G & & Van den Veyver IB 2011 Copy-number changes in prenatal diagnosis. Expert Review of Molecular Diagnostics 11 579–592. (https://doi.org/10.1586/erm.11.43)
Tyreman M, Abbott KM, Willatt LR, Nash R, Lees C, Whittaker J & & Simonic I 2009 High resolution array analysis: diagnosing pregnancies with abnormal ultrasound findings. Journal of Medical Genetics 46 531–541. (https://doi.org/10.1136/jmg.2008.065482)
Van den Berg MM, van Maarle MC, van Wely M & & Goddijn M 2012 Genetics of early miscarriage. Biochimica et Biophysica Acta 1822 1951–1959. (https://doi.org/10.1016/j.bbadis.2012.07.001)
Viaggi CD, Cavani S, Malacarne M, Floriddia F, Zerega G, Baldo C, Mogni M, Castagnetta M, Piombo G, Coviello DA, et al.2013 First-trimester euploid miscarriages analysed by array-CGH. Journal of Applied Genetics 54 353–359. (https://doi.org/10.1007/s13353-013-0157-x)
Warren JE, Turok DK, Maxwell TM, Brothman AR & & Silver RM 2009 Array comparative genomic hybridization for genetic evaluation of fetal loss between 10 and 20 weeks of gestation. Obstetrics and Gynecology 114 1093–1102. (https://doi.org/10.1097/AOG.0b013e3181bc6ab0)
