About 10-15% of pregnancies fail after conception has been recognized, amounting to 23 million losses a year. Chromosomal anomalies underlie many embryonic and fetal losses, but their exact frequency and localization to the embryo or placenta are still unclear. A new study published in Nature Medicine reports on a chromosomal analysis of over 1,700 spontaneous early miscarriages.
The most common period of pregnancy loss is before the ninth week, though many may occur earlier and pass unrecognized. While about 11% of women have at least one miscarriage, the proportion goes down with two or three, at 2% and <0.7. respectively.
With parental age, it becomes more likely that one or both parents will be infertile, while older mothers have a higher risk of aneuploidy. Moreover, aneuploidy occurs at a high rate at two stages of the reproductive process.
The first is at meiosis, preceding the final maturation of the oocyte if chromosomes fail to segregate after replication. The second is during the initial cleavage of the zygote and early embryo before implantation due to the high rate of DNA replication.
Earlier research has shown that the high rate of chromosome instability (CIN) in early embryos is not reflected in samples taken at birth, indicating that most such embryos are naturally weeded out by implantation failure and early pregnancy losses. CIN causes embryonic mosaicism, but cells with aneuploid genomes may be selectively lost before implantation occurs.
Multiple mechanisms have been proposed for this, including the restoration of trisomy or monosomy to euploidy or by cellular fragmentation with the extrusion of abnormal blastomeres during blastocyst development. Another possible pathway is restricting the abnormal cells or the rescue mechanisms to a particular region in space and/or time, accounting for confined placental mosaicism (CPM).
Importantly, CPM is the chief reason for false positives on non-invasive prenatal testing (NIPT) and thus promotes unnecessary invasive fetal testing on a normal fetus. This is the case with over seven out of ten large non-inherited copy number variations (CNVs) found to be localized to the placenta.
While the number and type of genotypic abnormalities in mid-and late pregnancy losses have been described, along with those in fetuses aborted by parental choice due to phenotypic abnormalities, the current study focuses on first-trimester miscarriages.
What did the study show?
Of the over 1,700 pregnancies that were miscarried, half the aborted embryos or products of conception (POCs) had normal karyotypes. The group with chromosomal anomalies were older, with a mean maternal age of 29 years vs 28 years for the normal-karyotype group and a mean paternal age of 31 years vs 30 years. The odds of such anomalies increased independently with increasing maternal and paternal age.
Further analysis of the 94-genome subset with normal parental and POC karyotypes was conducted using genotyping techniques based on the identification of genome-wide single-nucleotide polymorphisms (SNPs). Both extra-embryonic mesoderm and chorionic villi, representing embryonic and trophoblastic tissues, respectively, were studied.
Using a method called “genome haplarithmisis”, haplarithms were produced from the bulk DNA genotype data. This allowed the scientists to identify the occurrence of meiotic errors, and further to meiosis I and meiosis II, respectively.
In this subset, over a third (35%) showed chromosome abnormalities despite a normal karyotype. If the karyotypes are also included, this would indicate that almost 68% of the entire sample showed genomic aberrations. This exceeds earlier reported rates using other criteria for chromosomal anomalies.
The genomic abnormalities observed in sporadic and recurrent pregnancy losses were different, with higher copy number aberrations per POC and, more numerical abnormalities for recurrent vs fewer copy number abnormalities and more segmental abnormalities for sporadic losses, respectively.
Chromosomes 7 and 16 were most heavily affected, corroborating earlier studies, with 80% of the aberrations being maternal in origin. Chromosome 7 showed mitotic errors in 5/6 samples, but meiotic aberrations occurred with chromosome 16. Trisomy 16 is rarely detected by NIPT at 11-12 weeks, compared to pre-implantation genetic testing (PGT) on very early embryos; trisomy 16 may reduce survival odds.
In viable gestations, that is pregnancies that end with a live birth, mosaicism (occurrence of normal and abnormal cells in the same organism) is commonly found only in chorionic villi, mostly as confined placental mosaicism. In contrast, the current sample showed a higher rate of mosaicism (58%), mostly autosomal, in the extraembryonic mesoderm vs 33% of chorionic villi.
What are the implications?
The study emphasizes the need for comprehensive genotyping rather than mere karyotyping in early pregnancy loss to understand how such miscarriages occur and possible management approaches. With genomic haplarithmisis, using both embryonic and trophoblastic tissue, the current study had a markedly higher yield (67%) of genomic aberrations compared to a combined yield of 54% in recent large studies of chromosomally anomalous fetuses.
This was matched only by a recent study that used high-density SNP arrays. “Thus, haplarithmisis gives a superior yield over karyotyping or standard microarray approach.”
Not only does it avoid missing aberrations due to the use of single POC tissues, but it does not require cell culturing, with its high failure rates of up to 42%, depending on the laboratory. Thirdly, it can identify the parent and the phase of DNA replication (meiosis or mitosis) at which the aberration arises and detect contamination by maternal cells, unlike conventional methods in routine use.
“There is an emerging need for prospective clinical studies comparing genome haplarithmisis with conventional methods to evaluate its clinical implementation and cost-effectiveness for management of pregnancy loss.”
The second implication of these findings is that genomic aberrations appear to emerge in the inner cell mass prior to implantation and predispose to miscarriage, persisting in the extra-embryonic mesoderm if not in the chorionic villi. Abnormal cells are tolerated in the placenta better than in fetal tissues, as indicated by the consistently low rates of fetal mosaicism observed on NIPT.
This has implications for assisted reproductive techniques (ART), where fertilized embryos are transferred post-NIPT. NIPT interpretation becomes more difficult since mosaicism currently impacts the odds of embryo transfer. However, several studies indicate that most observed mosaicism in the POCs disappears or is not detected with advancing gestation.
This underlines the need for more accurate detection of the origin of genomic anomalies to rule out fetal abnormalities, thus preventing the requirement for invasive fetal testing. Mitotic errors are likely to be restricted to CPM without affecting gestational viability, unlike meiotic errors. Cell-free DNA may help detect the presence of mosaicism in fetal as well as placental tissues.