WiSSPers, November 1997, page 2

In Depth

Confined Placental Mosaicism
Richard M. Pauli, M.D., Ph.D.

Introduction
When an individual has two or more different genetic contents of cells, then we speak of mosaicism. We tend to think of the genetic content of all of the cells of an individual as being equal, constant and unvarying. In reality, differences in actual genetic content or functional genetic content are very common. More than half of the human population is functionally mosaic: women have, randomly, one or the other of their X chromosomes inactivated early in development resulting in effective mosaicism for all of the genes that are located on the X chromosome. Actual genetic mosaicism can be either somatic or germinal. Somatic mosaicism, mosaicism not affecting the germ cells, is exceedingly common — simple nevi or moles of the skin probably arise because of mutation resulting in a different genetic makeup of the cell that gives rise to these moles. A functionally more important example of somatic mosaicism is the gene changes that underlie most forms of cancer.

Germinal mosaicism, in contrast, includes two or more genetically distinct cell lineages including in the germ cells. This means that this kind of mosaicism can result in abnormal offspring of the affected individual.

Both somatic and germinal mosaicism can be of single genes or of whole chromosomes.

Confined Placental Mosaicism
A special kind of mosaicism was first recognized about 15 years ago (and became a hot topic with the development of chorionic villus sampling for prenatal diagnosis). Based on a few isolated instances in which chromosomal differences were seen when a baby and its placenta were both assessed, confined placental mosaicism was defined as any discrepancy in the chromosomal makeup of an embryo/fetus/infant compared with its placenta.

This phenomenon was identified as a critical determinant of survival in some aneuploid babies. It was discovered that in all liveborn babies with trisomy 13 or trisomy 18, there was, in the placenta, a normal cell line. That is, survival beyond the fetal period and resultant live birth only arises in these conditions when a spontaneous reversion to a normal chromosome makeup occurs by chance in the cell lineages that give rise to the placenta (‘rescue’ in an otherwise uniformly lethal condition). This observation helped to explain why many babies with these trisomies are miscarried or are stillborn while others are born alive. Presumably trisomy 13 or trisomy 18 results in serious abnormalities of placental function and these placental functional abnormalities are what result in intrauterine death.

Confined Placental Mosaicism and Prenatal Diagnosis
A thorny problem arose related to confined placental mosaicism with the development of chorionic villus sampling for prenatal diagnosis. Chorionic villus sampling at 10-12 weeks of gestation secures cells of the placenta. Initially it was assumed that, without exception, these cells of fetal origin would be genetically identical to the cells of the fetus per se. That proved not to be the case. In fact, in around 1% or so of chorionic villus samples a chromosomal abnormality will be detected that is not present in the fetus (and far less frequently will fail to detect an abnormality that is present in the fetus). These discrepancies are secondary to confined placental mosaicism. [Confined placental mosaicism, of course, can also be looked for in term placentae. In one such study where multiple samples were taken of term placentae, 6 of 124 were found to have a mosaic chromosome makeup. Other similar studies have reported frequencies of between 0.04% and 5% for such mosaicism. It seems, as expected, that the more carefully looked for, the higher the apparent frequency of placental mosaicism. Probably the 5% figure is a reasonable estimate of true frequency.]

In fact, chorionic villus biopsy allows sampling of two different tissue types. Direct chromosomal studies are of cytotrophoblast cells. Cultured cells for chromosomal investigation are of the mesenchymal cells of the placental stroma. So, confined placental mosaicism can be of different types.

Types of confined placental mosaicism
Different types of confined placental mosaicism are defined by whether the abnormality is apparent on direct CVS preparations (that is, those done without initial lengthy culture in the lab), on cultured cells from CVS, or both (type I, II and III, respectively). In turn, this reflects whether the chromosome abnormality is of the cytotrophoblast cells, the mesodermal villous stroma cells, or both. And, not surprisingly, these different types, when found by prenatal diagnosis, have different implications. Type I accounts for 45-55% of confined placental mosaicism, type II about 30-45% and type III 10-20%.

Type I almost never includes chromosomal abnormality in the fetus and almost always implies a good outcome of pregnancy. It is, then, mostly a troubling prenatal finding that creates anxiety until the cultured chromosome results are available.

Types II and III are more worrisome in terms of fetal outcome. In part, predictions appear to depend on which chromosome pairs are involved. If chromosome 13, 18, 21 or a sex chromosome is found in abnormal number, almost always this means that the fetus is also cytogenetically abnormal. In contrast, if chromosomes 2, 3, 7 or 8 are found in CVS mesenchyme they seem to have no effect. And finally, trisomy 16 when present only in the placenta seems to have a clear correlation with an increased risk of intrauterine growth retardation.

In fact, understanding the effects of confined placental mosaicism is an extraordinarily complicated topic. Effects may vary by timing of the chromosomal accident, by the type of confined placental mosaicism that results, by the proportion of normal and abnormal cells present, by which chromosome is involved, and so forth.

How bad effects might arise.
If the fetus is truly chromosomally normal (in all tissues and without mosaicism) then two primary mechanisms can result in harm. Confined placental mosaicism might cause the placenta itself to be poorly functional. This appears to be the most likely mechanism explaining intrauterine growth retardation in some instances of confined placental mosaicism.

Secondly, we have to consider how confined placental mosaicism could arise and how that might affect the genetic makeup of the fetus. First, a zygote might be chromosomally normal and then an accident of mitosis could result in a cell line with a chromosomal abnormality confined to the placenta. That mechanism will probably usually give rise to normal babies, unless the dysfunction of the placenta itself is so severe as to cause problems with survival or growth. Secondly, and more interesting, the zygote might be chromosomally abnormal but then a spontaneous ‘correction’ or ‘rescue’ results in a chromosomally normal cell line. Let’s suppose that this rescue-derived line then gives rise to all fetal tissues. Why should we expect the fetus to be anything but perfectly normal?

An example might best explain this. Let’s suppose that nondisjunction in an egg or sperm results in a newly fertilized zygote with three copies of chromosome 7 instead of the usual two. Two of these would come from one parent (let’s say for this example from the mother) and one from the other parent. Now let’s ‘correct’ this trisomy through a second chance event — loss of one of the three chromosome 7s. Two-thirds of the time one of the maternal 7s will be lost and everything will, presumably, be ok. However, one-third of the time the paternal 7 will be lost and what will remain are two maternal 7s. So, the developing embryo will have the right number of chromosome 7s, but both will be derived from one parent. This is termed uniparental disomy.

Uniparental disomy can have bad effects in two ways. First it makes more likely homozygosity (a double dose) of harmful recessive genes, since only one parent need be a carrier of a particular harmful gene. If this is a lethal recessive, then this might result in intrauterine death. Secondly, we now know that copies of genes from our mother and father function differently - a phenomenon call imprinting. This means that it does matter where the two chromosome copies come from. Imprinting problems secondary to uniparental disomy also could cause bad outcomes, including intrauterine death.

Confined placental mosaicism and intrauterine death.
Some investigators are convinced that confined placental mosaicism is an important, if under-appreciated, cause of both early and late pregnancy losses. That impression, however, is mostly based on selected and small studies. So, for example, Kalousek and Barrett show very preliminary evidence that suggests that pregnancy loss may be associated with confined placental mosaicism of chromosomes 2,3 9, 16 and 18 and that, at least for chromosomes 2 and 16 some of these losses may be stillbirths. Other uncontrolled case collections suggest a higher frequency of fetal death in confined placental mosaicism of chromosomes 3, 9, 14 and 16 in particular. In contrast, Leschot et al., in 108 sequential cases, found that there were only 2 spontaneous losses and 2 neonatal deaths — not significantly above what one might expect by chance. Likewise, Wolstenholme et al. conclude, based on more than 8000 CVS samples and 73 sequential cases of confined placental mosaicism, that their "data cannot exclude a small increase in the fetal loss rate among cases of cytogenetic abnormalities of the placenta, but do not support the high level of adverse outcomes reported in certain previous reports."

Rather than searching for bad outcomes among those with identified confined placental mosaicism, one could, of course, look for confined placental mosaicism in spontaneous abortions or stillbirths. But here, too, the data are very uncertain. Griffin et al. found discrepant placental karyotypes in 4 of 65 informative spontaneous abortions — a bit higher than would be expected, while Kalousek identified confined placental mosaicism in 11 of 54 spontaneous abortions — much greater than population expectations. It isn’t at all clear what accounts for this difference.

Nor have any good studies been done on stillborn infants. I have reviewed WiSSP data and have found two, and possibly three, instances of apparent confined placental mosaicism — one of chromosome 2, one of chromosome 6 and possibly one of chromosome 12. That is only 3 out of 1206 stillborns and only 3 out of 494 (0.6%) in whom successful cytogenetics was obtained. Even this latter proportion is no greater than that to be expected by chance. Thus, while a well designed and well controlled study of confined placental mosaicism in stillborns is clearly needed, data do not support incorporation of such studies into routine protocols, at least not yet.