In this chapter we consider parents who are themselves karyotypically normal and have had a child or a pregnancy that aborted with a full aneuploidy or a polyploidy. Thus we include the major trisomies (13, 18, 21) and sex chromosome aneuploidies (XXX, XXY, XYY, and 45,X) as well as less commonly seen autosomal aneuploidies and sex chromosome polysomies. The category of polyploidy is substantially devoted to triploidy. In the great majority, these defects arise from an abnormal event during meiosis or (in triploidy) at conception. In a few cases there is postzygotic generation of aneuploidy. Only in the case of parental gonadal mosaicism or in the hypothetical setting of an apparent predisposition to nondisjunction is there an increased risk of recurrence over and above that associated with a parental age effect (if there is one).
Full aneuploidy is presumed in the great majority to be the result of meiotic nondisjunction. A diminished degree of meiotic recombination is typically observed, and this has led Hassold and Sherman (2000) to propose a two-hit sequence, the first hit being a less well–tethered bivalent at meiosis I, and the second hit being a consequential aberrant distribution at meiotic metaphase. Meiotic nondisjunction can happen at any maternal age, but it is more frequent in older mothers. Alternatively, an abnormality has arisen in a premeiotic gametocyte, with the parent thus having a “wedge” of gonad that carries the abnormality (gonadal mosaicism). Such a parent would, of course, have an increased risk for only the one karyotypic defect. Finally, a small fraction of apparent full aneuploidy may be due to early mitotic nondisjunction in an initially 46,N conceptus with loss of the normal cell line or restriction to extraembryonic tissue of the normal cell line. For convenience, we note here also those forms of Down syndrome that are due to translocations (in most of these, however, the genetic imbalance is essentially the same as in the case of standard trisomy).
Trisomy 21 (Down Syndrome)
Down syndrome (DS) is the archetypal chromosome disorder. It was the first medical condition shown to result from a chromosome abnormality, in 1959. It has for many years been recognized as the most common single known cause of intellectual disability and has the highest incidence at birth of any chromosome abnormality. Every counselor can expect to deal frequently with problems relating to DS and should be familiar with its genetics.
The Genotype to the Phenotype. The DS phenotype—the characteristic facial appearance, body build, and mental defect—is due, in sum, to a triple amount of chromosome 21. Epstein (2002) provides a broad philosophical and historical as well as scientific review of the central role of the dosage effect in understanding the pathogenesis of the DS phenotype. Particular organ systems are particularly vulnerable, and Torfs and Christianson (1998) have identified characteristic malformations in a population study of nearly 3000 affected infants (Table 16-1). It is, in a sense, a “contiguous gene syndrome,” in which there is an additional dose of an en bloc set of genes. The entire chromosome was sequenced by 2000, and the gene complement turned out to be surprisingly low, only 225 loci in all (Hattori et al., 2000). This gene-sparseness is plausibly a factor in the survivability of the trisomic state.
It was logical that attempts be made to define those regions of the chromosome that might contribute predominantly to the DS phenotype—that is to say, to identify a DS critical region that might contain particular “DS genes.” Such an attempt is illustrated in Figure 16-1, with the region 21q22.13–q22.2 apparently having a major influence. Pritchard and Kola (1999) propose that there are only a few loci whose 150% amount is central in the pathogenesis of DS, and that certain loci may make specific contribution to certain components of the phenotype. Genes that have a role in the control of apoptosis might, for example, in an overactive state, trim too many cells and have a major effect in various tissue hypoplasias; a gene such as ETS2, at 21q22.3, could fit this description. Genes influencing brain growth are obvious candidates, and SIM2 (Drosophila single-minded 2 homolog), located in the DS critical region, is in this category. Ema et al. (1999) and Chrast et al. (2000) studied mice that overexpressed murine sim2, and certain abnormal behaviors in these mice may be due to this effect and perhaps have relevance to the human state. The neural cell adhesion molecule DSCAM, which maps to 21q22.2–q22.3, is strongly expressed (in the mouse) in those equivalent regions of the brain that are compromised in DS, and this is an attractive candidate as a contributor to the DS brain phenotype (Barlow et al., 2001). Other chromosome 21 genes may be of less critical effect when they are overfunctioning, and for some, perhaps many, it may make no difference at all.
Table 16.1. Some Malformations Frequently Observed in Down Syndrome
Shapiro (1997) has a somewhat different viewpoint, championing the “amplified developmental instability” hypothesis, and comments that “the search for a minimal region on chromosome 21 (the so-called DS critical region) responsible for producing DS has come full circle back to almost the entire chromosome.” In his view, a direct role for one or a few single loci with a one-on-one gene-to-phenotype relationship is simplistic: “traits that characterize DS are complex, and should be viewed and analyzed accordingly.” His general proposition is not unreasonable: that an excess of chromosome 21 coded gene products perturbs the functioning of the products of many loci, from all chromosomes, in all manner of developmental and physiological pathways. Attempting to draw together the two viewpoints, the gene dosage theory and the amplified developmental instability theory, we could suppose that the important genes—the DS loci— may have their pathogenic role in the modulation of layer upon layer upon layer of cellular interactions that lead, as the end result, to a phenotypic range that is clinically recognizable as DS. “Complex” may be too simple a word to describe this.
Figure 16-1. Phenotypic (trisomic) map of chromosome 21. Thick lines represent regions that must be trisomic to produce the particular trait. Thin-line regions may also contribute to that trait; the contribution of dotted-line regions is less clear. M, mild; P, profound. (From Korenberg et al., 1994, courtesy J. R. Korenberg and with the permission of the National Academy of Sciences of the U. S. A.).
What about the characteristic DS facies? Simply to observe other individuals is enough to convince one that development of the human face must be the most subtle and complex and precise process. How could we begin to understand why the DS face is different, and recognizably so? Among others, one contributor may be the failure of the facial musculature to divide into its proper various components in fetal development, and this might have, of itself, a distorting effect upon soft tissue formation of the face (Bersu, 1980). Which gene in triple dose, or which combination of genes, might lead to such a process? Listing the genes that are on chromosome 21 is merely another step on the way to knowing the means by which this trisomy causes this phenotype.
Different Cytogenetic Forms
Triplication is usually due to the presence of an extra chromosome 21 in all or most cells— in other words, standard trisomy 21 (Fig. 16-2). The disorder has a number of other cytogenetic forms, and Figure 16-3 depicts the proportions graphically. Differences in the source and nature of the genetic errors underlying these various forms require each to be considered separately.
Standard Trisomy 21 Down Syndrome
The great majority (about 95%) of DS is due to simple trisomy of chromosome 21. Around 90% reflects a maternal meiotic error (Yoon et al., 1996). Three-quarters of these maternal errors occur at meiosis I, and one-quarter, apparently at meiosis II, although the latter may actually have been set up at meiosis I. Meiotic I errors are associated with a reduced number of recombinations between the chromatids of the no. 21 bivalent or an actual absence of recombination. Particularly an absence of recombination (with no chiasma forming, thus an “achiasmate” bivalent) may lead to each homolog being able to segregate without reference to the other, and thus without the imperative to move symmetrically. The site of recombination may also be a factor: a recombination at one or other end of chromosome 21 (the telomeric third or the pericentromeric region) seems to endow a meiotic instability compared with crossovers in the middle of the chromosome (Savage Brown et al., 2000).
Among the small fraction (about 10%) due to paternal errors, the proportions due to meiotic I and meiotic II errors are nearly equal. No usefully discernible paternal age effect exists. As in the female, a reduced frequency of recombination observed in the meiotic I cases may underlie the cause of this male nondisjunction (Savage et al., 1998). Two unexplained observations concerning trisomy 21 due to paternal meiotic errors are these: this fraction is a little greater among prenatally (11%) than postnatally (7%) diagnosed cases; and there is an excess of males among the DS offspring (Muller et al., 2000). Apart from the majority due to meiotic nondisjunction in a parent, the remainder of standard trisomy 21, about 5%, is due to a mitotic nondisjunction.
Standard trisomy DS typically occurs as a sporadic event, and recurrences are rare. The causes of recurrence are: gonadal mosaicism, a parental predisposition to nondisjunction, and chance.
Figure 16-2. Karyotype of a child with standard trisomy 21.
Figure 16-3. Origins of trisomy 21 (percentages rounded). rob, Robertsonian.
Recurrence due to Mosaicism. A trisomy 21 cell population in a parent (gonadal or so-matic–gonadal mosaicism) is presumed to be an uncommon cause of the production of disomic 21 gametes, although perhaps less rare than originally thought (Bruyère et al., 2000; Kuo, 2002; Mahmood et al., 2000). Pangalos et al. (1992b) studied 22 families in which trisomy 21 had occurred more than once (in siblings and in second- and in third-degree relatives), applying DNA polymorphism analysis. Parental gonadal mosaicism was proposed as the cause of sibling recurrence in 5 of 13 families (about 40%); but other than these, chance alone was enough to explain the recurrences in most if not all families. James et al. (1998) studied four women, each of whom had had three trisomy 21 conceptions. Two of the mothers were under age 35 at the time of the trisomic conceptions, and they both showed a very low level mosaicism (0.5% and 4% on blood karyotyping). Neither had a DS phenotype. One of these mosaic mothers was shown to have originated as a trisomic conceptus, but an early mitotic loss of a chromosome 21 almost entirely corrected the karyotype. The other two women were older, and in their case it was more likely that the multiple trisomic conceptions occurred independently, as maternal age-related events. Sachs et al. (1990) followed 1211 pregnancies at prenatal diagnosis, subsequent to the occurrence of trisomy 21 in a previous pregnancy, and observed six recurrences (for a rate of 0.5%). In two of these instances, mosaicism was shown. One father karyotyped as 47,+21/46,N on skin analysis; and one mother showed trisomic cells in 3%, 14%, 44%, and 47% on culture of, respectively, blood and skin, and, in a more direct observation, in each ovary.
The gonad, or the gametes themselves, naturally provide the best evidence. Ovarian biopsy proved the point in a mother of three DS children (and one normal child) who typed 46,XX on peripheral blood, but in whom 8 out of 20 ovarian cells showed trisomy 21 (Tseng et al., 1994). Similar examples are on record. Nielsen et al. (1988) report a couple having had six documented pregnancies with standard trisomy 21, and five other unkaryotyped pregnancies ending in neonatal death or abortion. The mother typed 46,XX on peripheral blood, and 47,XX,+21/46,XX in ovarian somatic cells. Even if the oogonia were all or nearly all 47,+21, it remains perplexing that no known 46,N conception occurred. An in vitro fertilization (IVF) setting enabled analysis of the gametes themselves in a woman studied by Cozzi et al. (1999). She had had a normal child and a DS child at ages 29 and 32, and then had prenatal diagnoses of trisomy 21 at 32 and 36 years. No trisomic mosaicism was detected on peripheral lymphocyte analysis. At IVF, of seven embryos, four were trisomy 21 and one tetrasomy 21, with only two showing normal disomy 21. Four unfertilized oocytes were analyzed, and three had a supernumerary chromosome 21.1 As for the male, Hixon et al. (1998) analyzed sperm samples from 10 men who had fathered a DS child, the additional chromosome 21 having been demonstrated to be of paternal origin. None showed any increase in the fraction of sperm with disomy 21.
Recurrence due to Nondisjunctional Tendency. Do some (nonmosaic) individuals, for a certain biological reason, run an increased risk of producing a trisomic 21 conception? Is there a range of “meiotic robustness” in the population? This is a perfectly respectable concept, albeit one that remains quite hypothetical. If so, what possibilities might there be? Several theories for a general predisposition to aneuploidy have been put forward, and some of these are discussed on p. 48. In the specific case of trisomy 21, an additional theory has been proposed. An intriguing link exists between DS and Alzheimer's disease, and the microtubule might be the linking factor (Schupf et al., 2001). An onset of dementia in middle adult life, over and above the congenital mental defect, is a common observation in DS. One Alzheimer susceptibility locus, the β-amyloid precursor protein (APP) gene, is located on chromosome 21 and may be overexpressed in both conditions. A tendency toward chromosome 21 nondisjunction (mitotic in brain, meiotic in gonad) could be a unitary explanation (Geller and Potter, 1999; Migliore et al., 1999). Microtubules are the basic component of the meiotic and mitotic spindles; and the tau protein, aggregation of which within the brain is an Alzheimer hallmark, is a microtubule-associated protein.
While some of these various possibilities may be more plausible than others, they are all speculative, and we conclude that there is at present no routinely practicable basis enabling the counselor to identify ahead of time those parents whose risk is high and those whose risk is low to have a second pregnancy with trisomy 21.
Recurrence Risk Estimates after One Affected Child or Pregnancy. The earliest estimates of risk are from Penrose (1956),2 prior to the discovery of the chromosomal basis of DS, and Stene (1970). Penrose proposed the risk of recurrence to be “doubled, or perhaps nearly trebled” compared to the general population risk, irrespective of maternal age; while Stene derived a figure of 1% for mothers under age 30, with no difference for those over 30, at the time of birth of their DS child. More sophisticated analyses were subsequently enabled by the collection of amniocentesis data. Recently, D. Warburton and colleagues (Warburton et al., 2001, and pers. comm., 2002) have determined new estimates on the basis of an extensive data set, and the advice set out in the Genetic Counseling section below is based on this work. It does remain true that for younger mothers the risk is, in absolute terms, small.
Recurrence Risk Estimates after Two Affected Children/Pregnancies. When a couple have had two (or more) trisomic 21 conceptions, one has to assume that an increased risk applies to a subsequent pregnancy, quite possibly a substantial risk. The recurrence may well have been due to gonadal mosaicism. In a collaborative series from six Japanese clinics, Uehara et al. (1999c) record the exceptional case of a couple having had five successive pregnancies with trisomy 21 (one DS child, four diagnoses at amniocentesis). Both parents had normal karyotypes on blood and skin analysis. It would seem rather probable that one parent may have had fully trisomic gonadal tissue.
Occurrence Risk Estimates with Down Syndrome in a Second or Third-Degree Relative. More widely in the family, it appears that a history of standard trisomy DS in second- or third-degree relatives does not, in the main, imply an increased risk (Hook, 1992; Pangalos et al., 1992b). Berr et al. (1990) assessed 188 families in which a DS child had been born, and there were comparable numbers of DS cases among the second- and third-degree relatives and in the relatives of 185 control families.
Mosaic Down Syndrome
47,+21/46,N mosaicism accounts for about 2% of individuals with clinically diagnosed DS. Mosaicism results from a malsegregation of homologs, or an anaphase lag of one homolog, occurring postzygotically. Some individuals with mosaic DS arise from initially trisomic 21 zygotes, losing one of the no. 21s at anaphase lag (Fig. 2-8c). Others may arise from normal conceptuses, with nondisjunction producing 45,-21/47,+21/46,N mosaicism, with the 45,-21 line thereafter lost (Fig. 2-8a). Pangalos et al. (1994) studied 17 families in which there was a child with mosaic trisomy 21, and 10 children had three no. 21 alleles, indicating their origin from a trisomic conceptus. The chromosome 21 chromosome subsequently lost to enable formation of the 46,N cell line showed no predilection for being a maternal or paternal homolog. The remaining seven mosaics had no evidence of a “third allele,” and distinction in these between an initially 46,N or 47,+21 conception was not possible. Whatever the basis, for practical purposes counseling needs to proceed as though the child has standard trisomy 21, recognizing that this will overestimate the risk for some parents. Genetic counseling for the mosaic individuals themselves is covered on p. 216.
Isochromosome 21 Down Syndrome
After standard trisomy 21, this is the most common chromosomal category of DS. It has often been called a “21q21q Robertsonian translocation,” but in fact the two 21q components are usually identical and thus isochromosome is the more accurate term, and the karyotype is more accurately 46,i(21q) (Antonarakis et al., 1990; Shaffer et al., 1992; Robinson et al., 1994; Ruiz-Casares et al., 2001). Molecular studies suggest that many of these originate at an early postzygotic mitosis (Fig. 17-3), and this is consistent with the observation that the recurrence risk is low. In one series of 112 de novo “rob(21q; 21q)” probands, none of 130 full sibs and 34 half-sibs had DS (Steinberg et al., 1984). Nevertheless, three of the parents actually showed a low-grade mosaicism, and presumably their having had an affected child reflected that the 21q21q cell line was included in the gonad. A few examples of recurrence in subsequently born siblings are recorded, and parental gonadal mosaicism can be the basis of such recurrence (Sachs et al., 1990; Robinson et al., 1994). This is a point that Mark et al. (1977) directly proved in one case: a woman having sequential pregnancies with the karyotype 46,i(21q) herself typed 46,XX,i(21q)/+46,XX on ovarian fibroblast analysis (but 46,XX on blood). Hall (1985) offers the cautionary story of a mother given a low risk of recurrence, who went on to have a second affected child from a second marriage (on resampling of her, a single 46,XX,rob(21q;21q) cell was found in 100 cells analyzed). Given this desirability of distinguishing between postzygotic and meiotic mechanisms, with their differing counseling implications, Kovaleva and Shaffer (2003) advocate study with polymorphic markers in this (and other) “homologous Robertsonian translocations.” Where the DS child is a 46,i(21q)/ 46,N mosaic, the case for postzygotic formation of the isochromosome is, for practical purposes, secure.
Robertsonian Translocation Down Syndrome
Almost all translocation DS concerns a Robertsonian translocation (discussed in detail in Chapter 6). About one-quarter of Robertsonian translocation DS is familial and three-quarters are de novo (1% and 3% of all DS, respectively).
De Novo Robertsonian Translocation Down Syndrome. Both parents, by definition, have normal chromosomes. The abnormal chromosome may usually arise as a sporadic event in maternal meiosis I from a chromatid translocation (Petersen et al., 1991). Such mutational events are rare and in the great majority of families recurrences are not seen. But gonadal mosaicism remains a possibility. The so-called rob(21q;21q) is, in most cases at least, actually an isochromosome (see above).
Familial Robertsonian Translocation Down Syndrome. One or the other parent (almost always the mother) is a translocation heterozygote and has transmitted the translocation in an unbalanced state to the DS offspring.
Down Syndrome with Reciprocal Translocation
The DS phenotype is substantially due, as we note above, to a duplication of the chromosome segment 21q22.2–q22.3. In a gamete from the heterozygote, a reciprocal translocation involving chromosome 21 has the potential to produce a duplication for the DS critical region, whether from 2:2 or 3:1 meiotic segregation. The unbalanced adjacent-1 karyotype from the t(18q;21q) illustrated in Figure 4-15 (second row) is an example. Or, interchange trisomy 21 may result (Fig. 4-13). These translocation scenarios are extraordinarily rare, being the cause of less than 0.1% of DS. Scott et al. (1995) describe a child with DS from a maternal t(12;21)(p13.1;q22.2) that could only be identified with FISH, and Nadal et al. (1997) describe a similar case from a paternal t(15;21)(q26;q22.1). It is from studies such as these, of typical DS children with a partial trisomy, that phenotypic maps, as in Figure 16-1, can be drawn. Interchange trisomy 21 was reviewed by Dominguez et al. (2001), with a total of only 23 published families being accumulated.
Other Chromosomal Forms of Down Syndrome
A number of chromosomally distinct forms of DS result from specific structural changes to chromosome 21. The least rare of these is the terminal rearrangement that produces a mir-ror-image chromosome around the telomeric region (Pfeiffer and Loidl, 1982). The chromosome has two centromeres, one of which is usually inactive, and satellites on both ends. Such chromosomes are always the result of sporadic mutational events, possibly the result of a translocation between sister chromatids (Pangalos et al., 1992a). Down syndrome is seen occasionally in association with other aneuploidies, almost always a sex chromosome aneuploidy, such as 48,XYY,+21 and 46,X,+21; this is known as double aneuploidy. It is usually the result of a double event of nondisjunction resulting in one abnormal gamete; the circumstance that promoted the nondisjunction of one chromosome may have applied also to the other, or the fact of one nondisjunction having occurred may have set the stage for a second. Rather less likely is a scenario of separate events in gametogenesis in both parents.
“Interchromosomal effect” has been invoked in standard trisomy DS in the setting of a parental karyotypic abnormality not involving chromosome 21 (e.g., a 13;14 Robertsonian translocation or a reciprocal translocation). In other words, might it be that the rob(13;14) or the rcp in some way perturbed the distribution of the no. 21s? The answer in fact seems to be no, with the case for interchromosomal effect remaining tenuous at best, although a possible exception might be the Robertsonian translocation in the setting of oligospermia (see p. 136).
Trisomies 13 and 18 (Edwards Syndrome and Patau Syndrome)
These syndromes are much less frequent than DS (about 1 in 12,000 and 1 in 6000 live births for trisomies 13 and 18, respectively), and both show a maternal age effect. As with trisomy 21, correlative phenotypic mapping allows certain segments of chromosomes 13 and 18 to be implicated in the genesis of certain phenotypic traits observed in these syndromes (Tharapel et al., 1986; Epstein, 1993; Boghosian-Sell et al., 1994). On molecular studies in trisomy 18, over 90% reflect a maternal meiotic nondisjunction. Uniquely, nondisjunction is considered to happen most frequently at the second meiotic division, this division not taking place until the short period of time surrounding the process of fertilization (Bugge et al., 1998). A contrary view has emerged from Verlinsky's group; from the direct analysis of polar bodies, chromosome 18 meiosis I errors outnumbered those in meiosis II (Verlinsky and Kuliev, 2000; Verlinsky et al., 2001a). In close to 90% of trisomy 13 the additional chromosome is of maternal origin (Bugge et al., 2002).
Recurrence of trisomy 18 has been recorded in one or two single case reports, and one or two instances of recurrence, or none at all, have been seen in prenatal diagnostic series or retrospective surveys (Pauli et al., 1978; Ferguson-Smith, 1983; Stene et al., 1984; Baty et al., 1994; Uehara et al., 1999c). Baty et al. (1994) note a 39-year-old mother having had prenatal diagnosis of trisomy 18 at age 39 years, and a liveborn trisomic 13 infant at age 40 years. No case of trisomy 13 recurrence is recorded. It had seemed, on balance, that no increased recurrence risk existed. However, in Warburton's recent review, a real risk, albeit a small one, has emerged from the analysis of a very large body of prenatal diagnostic data. This is dealt with in more detail in the Genetic Counseling section below.
Other Autosomal Trisomy
It is extremely rare for any other autosomal trisomy to survive through to (or near to) term. About two dozen examples each of trisomy 9 and 22 are known, and nonmosaic trisomies 7, 8, 10, and 14 are represented by only one or two reports (Brizot et al., 2001; Schinzel, 2001; Tinkle et al., 2003). A unique mosaic case with two different aneuploidies is reported in Mielke et al. (1997): an infant born at 36 weeks gestation and surviving for 7 weeks had the karyotype 45,X/47,XX,10 at amniocentesis and on peripheral blood analysis. A 46,XX cotwin may have allowed maintenance of the pregnancy.
In contrast, trisomies are very common in miscarrying pregnancies, a matter dwelt upon in detail in Chapter 21. Robinson et al. (2001) considered that recurrences of trisomy (even of the same trisomy) in spontaneous abortion might represent no more than a common thing happening again, and this may often be the case. Nonetheless, Warburton's large review comprised sufficient data to demonstrate that a previously karyotyped trisomic spontaneous abortion indeed does increase the risk for a potentially viable trisomy at a subsequent prenatal diagnosis (see Genetic Counseling section), thus indicating, in some instances at least, that an individual predisposition may have been the cause.
Many autosomal monosomies are presumed to end in arrested growth in the first few mitoses, prior even to the time of implantation, with some possibly proceeding to the stage of occult abortion (p. 343). Their existence would be unknown, were it not for the window of observation afforded by preimplantation diagnosis. The single exception may be monosomy 21, although most earlier reports of monosomy 21 have since been reinterpreted as being due for the most part to an unbalanced translocation involving chromosome 21 (West and Allen, 1998). One more recently recorded case was identified at 17 weeks of pregnancy, going on to fetal death in utero early in the third trimester, although again the cytogenetic diagnosis was not beyond doubt (Chang et al., 2001; Phelan, 2002).
SEX CHROMOSOME ANEUPLOIDY
XXY (Klinefelter Syndrome), XXX, XYY
These aneuploidies occur at roughly similar frequencies, about 1 per 1000 of the appropriate sex. About 75% of XXX and about 40% of XXY Klinefelter syndrome (KS) is due to a maternal meiotic error. In three-quarters of each of these it is the first meiotic (MI) division that is involved, this MI group showing a maternal age effect. It is noteworthy that almost half of KS results from a paternal MI error (MacDonald et al., 1994). Fathers of paternally originating KS may have marginally elevated levels of disomic XY sperm in comparison with fathers of maternally originating cases, possibly reflecting an inherent tendency among a small minority of these men to produce aneuploid sperm (Eskenazi et al., 2002). In what may have been the only known example of a recurrence, Woods et al. (1997) report two XXY brothers. The karyotype in both reflected a paternal MI error, and a consanguineous background may have allowed expression in the father of a recessive “meiosis gene.” Manifestly, XYY of meiotic origin must be due to a paternal error, at MII. All three sex chromosomes aneuploidies can have a postzygotic mitotic generation, typically, in that case, being mosaic.
45,X Turner Syndrome
In about three-quarters of Turner syndrome (TS) it is the paternal X chromosome that is absent (Hassold et al., 1991; Uematsu et al., 2002). Mostly, the error is a meiotic one, and resides in paternal gametogenesis, possibly reflecting an absence of pairing along most of the X-Y bivalent with a consequential vulnerability in the process of disjunction (Jacobs et al., 1997). Fathers of nonmosaic 45,Xm Turner girls may be prone to produce sperm nullisomic for a sex chromosome. Martínez-Pasarell et al. (1999) analyzed sperm from four fathers and eight controls, and there was a slight increase in 24,XY sperm (0.22%) and nullisomic sperm (0.48%) in the fathers compared to the fractions in controls (0.11% and 0.32%, respectively). Alternatively, the loss may have occurred postzygotically, and the “45,X” child is actually a 45,X/46,XX mosaic, with a very low proportion of XX cells, but this is apparently a rare event (Jacobs et al., 1997). Uematsu et al. (2002) suggest that most TS may actually be due to a structurally abnormal gonosome (X or Y) having been generated in paternal meiosis, with a 46,X,abn(X) conception resulting. Subsequent mitotic loss of the abn(X) leaves a 45,X karyotype.
In the case of a postzygotic origin, if it could be presumed to have been an event that occurred at random in a single mitosis in the early embryo, the risk of recurrence would be regarded as not being raised at all. Kher et al. (1994) did, however, report a unique family with occurrence of 45,X/46,XX in sisters. In the literature review of these authors, they could find only one instance of 45,X recurrence in sisters. J. L. Halliday (pers. comm., 1995) records 52 prenatal diagnoses done on the basis of previous 45,X: there were no recurrences. One instance of 47,XXY in a subsequent pregnancy is not without interest, considering the sperm study mentioned above, but a single example should not be overinterpreted.
Polysomies such as XXXX, XXYY, XYYY, XXXY, XXXXX, and XXXXY are very rare. Successive nondisjunctions in one parent, the other contributing a single sex chromosome, is the mechanism in most if not all (Hassold et al., 1990; Deng et al., 1991). Apart from the extraordinary circumstance of (hypothetically) a familial tendency to mosaicism, these polysomies arise sporadically (Bergemann, 1962; Kher et al., 1994). Rare reports of coincidence with some other aneuploidy in the family may more likely reflect chance than a causal link (Court Brown et al., 1969).
The chromosome count in triploidy is 3n = 69, with a double (2n) chromosomal contribution to the conceptus from one parent (Fig. 16-4).
Triploidy can reflect diandry or digyny, with the double contribution coming from the father or mother, respectively (also referred to as types I and II triploidy) (Fig. 16-5). The great majority of triploid conceptions abort during the 10- to 20-week period, and of these, most are diandric. The differing clinical presentations of diandric and digynic triploidy are presumed to reflect the influence of differing imprinted states.
Figure 16-4. Karyotype of a 69,XXY triploid fetus (see also Fig. 21-3).
Diandry is practically always the consequence of two sperm simultaneously fertilizing the ovum (dispermy) (Zaragoza et al., 2000; McFadden et al., 2002).3The fundamental problem in this instance lies in the “zona reaction,” which is the response of the investing membrane of the ovum, the zona pellucida, to prevent further sperm from entering after the first has penetrated. A minority is due to fertilization with a diploid sperm, this diploid state having arisen from a complete nondisjunction in spermatogenesis. The fact of being diploid seems to cause little compromise of the fertilizing ability of the sperm.
Digyny is most commonly due to a diploid egg, which may be the result of nondisjunction of the entire chromosome set at either the first or the second meiotic division in oogenesis, of retention of a polar body, or of the fertilization of an ovulated primary oocyte. In the series of Zaragoza et al. (2000), meiotic errors accounted for the majority of digyny, with 67% being due to failure of MII and 22% due to MI failure. A rare cause may be the fusion of two eggs (which these authors whimsically call “dieggy”). Individual susceptibilities may exist, such as that exemplified in a woman coming to IVF who had had two previous triploid pregnancies and in whom triploidy was identified by preimplantation genetic diagnosis in 2 out of 13 (PGD) conceptions. In this case at least, a maternal MII error could be implicated (Pergament et al., 2000). Diploidy can be presumed to exist in the giant binucleate oocyte, and these visibly abnormal gametes have actually been shown at IVF to lead to a triploid embryo (Balakier et al., 2002; Rosenbusch et al., 2002).
Figure 16-5. The three major routes through which triploidy may arise. A complete failure of a meiotic division produces a diploid egg (left) or sperm (middle). Simultaneous fertilization by two sperm is dispermy (right).
Triploidy is not uncommon in early pregnancy (1%–3% of recognized conceptions), but about 99.99% are lost as first-trimester miscarriage or second-trimester fetal death in utero. Of all 16-week pregnancies, only 1 in 30,000 are estimated to be triploid, and at 20 weeks, only 1 in 250,000 (Snijders et al., 1995). There is a differential viability according to the origin being diandric or digynic (Zaragoza et al., 2000). Diandric triploids mostly abort in the 10- to 20-week period, the mean at 12 weeks, and, as noted above, these cases comprise the considerable majority of all triploidy. Actually, very early diandric abortions (before 6 weeks) are not molar, but with increasing gestational age, the classic placental phenotype of partial hydatidiform mole (p. 355) is more likely to be observed. The different causes of the doubling of the paternal chromosome set, as discussed above, do not correlate with there being a molar or a non-molar phenotype. The very few diandric triploid pregnancies that survive to the second trimester typically show partial hydatidiform mole; growth retardation is usual but not invariable (Daniel et al., 2001).
Dygynic triploids mostly abort early (mean 10 weeks), although those exceptional few that remain are able to continue through to the third trimester, when they actually come to outnumber diandric cases. These surviving digynic triploids develop as a severely growth-retarded fetus with marked head–body disproportion, the head being relatively large, and with an abnormally small and nonmolar placenta (Dietzsch et al., 1995; Miny et al., 1995; Daniel et al., 2001). In one case of a digynic 69,XXX triploid coexisting with a normal 46,XY twin, survival to 20 weeks (when selective feticide was done) may have been supported by the normal fetus (Gassner et al., 2003). Intrauterine survival may also be promoted if there is fetal–placental karyotypic discordance, with the placenta being diploid (Kennerknecht et al., 1993a). A theoretical explanation for the atypical survival invokes the state of maturity of the egg at fertilization, in relation to the integrity of the imprinting process (Zaragoza et al., 2000). Survival to the third trimester is associated almost invariably with perinatal death. Of those liveborn, hardly any digynic triploids survive for more than a month; there is one extraordinary instance of death not until 312 days (Sherard et al., 1986; Hasegawa et al., 1999).
Van de Laar et al. (2002) accumulated 25 cases from the literature and reported three of their own. These three came from a population catchment of 15 million over a 20-year period, attesting to the rarity of the condition. The triploid line typically reflects digyny, and the basic mechanism may be inclusion of the second polar body at a very early stage after conception of a diploid zygote. Similarly in diandric cases, the mechanism may be dispermy, but with one sperm pronucleus sequestered in the cytoplasm for a few divisions before being incorporated into the nucleus (Daniel et al., 2003c). These authors refer to “delayed digyny” and “delayed dispermy,” respectively, as the course of events whereby the extra pronucleus sits to one side, so to speak, while the diploid lineage is in the process of being established, and the pronucleus then being taken up into the nucleus of one blastomere to give rise to the triploid cell line. Survival of the affected fetus in utero is presumably promoted by the diploid cell line (Carakushansky et al., 1994). In most cases the triploid line is not seen on a blood analysis, and fibroblast culture is necessary. A single instance of a false-negative amniocentesis is to be noted (Flori et al., 2003).
Tetraploidy (4n = 92) in a term pregnancy is exceedingly rare, and survival in one apparently nonmosaic case to 26 months is unprecedented (Teyssier et al., 1997; Guc-Scekic et al., 2002). The usual mechanism may be a normal division of chromosomes but failure of cytoplasmic cleavage at the first division of the zygote; another possibility is dispermic fertilization of an ovum in which MI has failed. Mosaic diploidy/tetraploidy in a person has been described in association with severe mental defect, and may only be detectable on skin fibroblast study (Edwards et al., 1994). A complex case is that reported by Leonard and Tomkins (2002) of a retarded woman with body asymmetry and hypomelanosis of Ito, in whom some fibroblasts cultured from hypopigmented skin showed 92,XXXX, while others were 46,XX and 46,XX,t(1;6)(p32;q13), and 46,XX was found on blood analysis.
True diploid/tetraploid mosaicism may be frequent at the blastocyst stage of development, but either the abnormal embryos are cast off shortly thereafter or, especially if the proportion of tetraploid cells is small and the blastocyst is otherwise of good quality, the polyploid component may be confined to the trophoblast and in due course come to comprise a minor fraction of placenta (Clouston et al., 2002; Bielanska et al., 2002b). Possibly for this reason, tetraploidy can occasionally be seen at chorionic villus sampling (CVS) and an amniocentesis, reflecting a “normal” tetraploidy of part of the placenta, with the remaining extra-fetal and fetal tissues being karyotypically normal (Benkhalifa et al., 1993). Alternatively, tetraploidy at prenatal diagnosis may be artifactual.
The central requirement for accurate genetic advice in DS is knowledge of the chromosomal form in the affected family member. If a child diagnosed as having DS has died and no chromosome studies were performed, it may be reasonable to check for the possibility of a familial translocation in the consultand(s).
Previous Child with Standard Trisomy 21 (Including Mosaicism)
If the child has standard trisomy 21 or is a 47,+21/46 mosaic, it is unnecessary to study the parents' chromosomes. One can assume with considerable confidence, that they will type as 46,XX and 46,XY. The risks for recurrence of trisomy 21 in a subsequent amniocentesis, or occurrence of a different aneuploidy, with reference to Warburton's work (Warburton et al., 2001, and pers. comm., 2002), and as summarized in Table 16-2, are as follows.
1. For a mother under 30 years old having had a previous DS pregnancy, the risk for recurrence of trisomy 21 at amniocentesis is about eightfold the age-related risk. Thus, for a 25-year-old having had a previous DS child, her risk would go from about 1 in 930 (see Table 22-3, p. 366) to 1 in 115.
Table 16.2. Increases in Recurrence Risk, Compared with the Maternal Age-Related Baseline, and Measured at Amniocentesis, for Women Who Have had a Previous Trisomic Pregnancy. Data of D. Warburton (pers. comm., 2002).4
2. A woman in the 30 age bracket may have a lesser increase in risk, but an increase nevertheless, on the order of a doubling. So, for example, a 37-year-old would go from the standard age-related risk of about 1 in 150 at amniocentesis (Table 22-3) to 1 in 75.
3. The risk of recurrence at birth will be a little less, reflecting a greater likelihood for natural loss of a trisomic pregnancy in the period following the time of amniocentesis.
4. The risk for a different viable trisomy (amniocentesis diagnosis) is double the age-related risk, for both younger and older women. Since the baseline figures are very low (Table 22-4, p. 367), the risk figure is still a low one.
In any event, regardless of the exact figure, the practical point is that the risk for a recurrence of DS is comfortingly low, only approaching the 1% mark by the midthirties. Nevertheless, most couples seek the reassurance of prenatal diagnosis in pregnancies after having had a child with DS. Elkins et al. (1986b)observe that some of these parents declare they would not abort a trisomy 21 fetus, and the counselor needs to be sensitive to possible ambivalent feelings of the parents in this setting.
Two Previous Trisomic 21 Conceptions
One can only offer an educated guess that the risk for a third trisomic conception will be substantial. A skin biopsy study would be largely academic. If gonadal mosaicism (rather than de novo recurrence) is the cause, a considerable fraction of whichever gonad it is must be involved, since two separate samplings have already come from this fraction. A risk in the range 10%–20% may be a fair figure to offer. Preimplantation genetic diagnosis would have an obvious attraction.
Isochromosome 21 Down Syndrome
From the 0/164 fraction among siblings of de novo isochromosome 21q DS in Steinberg et al.'s series (1984), the risk for recurrence is presumed to be small. Nevertheless, three parents (3%) in this series were demonstrably mosaic, and there is a handful of recurrences otherwise on record; a cautious stance is thus prudent. A risk figure in the region of 2% may be a reasonable one to offer. For the mosaic case, 46,i(21q)/46,N, no increased risk would apply, on the assumption of postzygotic generation of the isochromosome.
Previous Child with Robertsonian Translocation Down Syndrome
Obviously, distinction between de novo and familial forms of translocation DS is crucial; this distinction is made by chromosomal studies of the parents. For the de novo translocation, a recurrence risk figure of <1% is applicable (Gardner and Veale, 1974). In the case of familial Robertsonian translocation DS, the genetic risk for the female carrier is substantial. The risk of having a liveborn child with translocation DS is about 10%, while the likelihood of detecting translocation trisomy 21 at amniocentesis is about 15%. For the male carrier, the risk of having a child with translocation DS is small, about 1% (see Chapter 6).
Previous Child with Non-Robertsonian Translocation Down Syndrome
In the rare instance that translocation DS is associated with a familial reciprocal translocation, the principles presented in Chapter 4 are to be followed.
Previous Child with Other Chromosomal Category of Down Syndrome
For sporadic structural changes such as the terminal rearrangements, the risks are presumed to be very low (less than 0.5%). For the double aneuploidies, there is no evidence to suggest that the risks are any different from the recurrence risks for standard trisomic DS.
Wider Family History of Down Syndrome
There is no conclusive evidence of an increased risk for second- and third-degree relatives of individuals with standard trisomic DS themselves to have offspring with the condition. The appropriate action in the setting of a family history of DS is to determine whether the affected member has standard trisomy 21. If this is so, the family may be reassured that there is no discernibly increased risk. This advice could also reasonably be offered if a single case was associated with older maternal age. If the karyotype of the index case is unknown and the mother was younger, the small possibility of a familial translocation may be checked by chromosome study of the counselee.
Trisomy 21 in Products of Conception
The finding of trisomy 21 in products of conception after spontaneous abortion (in those centers where this testing may be done) presents a problem. Should this, for genetic counseling risk assessment, be regarded as equivalent to having had a child with DS? From about 10 weeks gestation through to term, about a third of trisomic 21 conceptions are lost (p. 364), and it may be stochastic events in utero rather than intrinsic genetic differences that distinguish those that abort and those that survive. It may be prudent to err on the side of caution and provide a risk figure as though the abortion had been a liveborn child.
Trisomy 13 and Trisomy 18, and Other Autosomal Trisomy
Recurrence of trisomy 13 or 18 is almost unknown, with zero being the most common percentage figure in formal series. Nevertheless, in Warburton's data (Table 16-2), a true increase did emerge, about a doubling, albeit that the absolute figure will still be small, given the very small baseline age-related figures (Table 22-4, p. 367). There is also a small increased risk for a different potentially viable trisomy. Similarly, in the case of a previous pregnancy with some other type of autosomal trisomy (typically identified in products of conception following spontaneous abortion), there is an increased risk, albeit small in absolute terms, for a potentially viable trisomy at the time of prenatal diagnosis. Thus it appears that there are some women who are at increased risk for meiotic errors in general, compared to other women of the same age. Be that as it may, many cases of recurrent trisomy in older women, whether in spontaneous abortions or in live-births, must represent merely the high risk of a trisomic conception related to advanced maternal age.
XXX, XXY, XYY, 45,X, Other Sex Chromosome Aneuploidy
There is no firm evidence (and indeed little soft evidence) that a recurrence risk above the age-specific figure exists.6 Prenatal diagnosis is discretionary.
Diandric triploidy associated with partial hydatidiform mole has a 1%–1.5% risk of recurrence; we discuss this in more detail on p. 359. As noted in the Biology section, some women may have a predisposition for digynic triploidy. However, the level of risk for recurrence of triploidy or occurrence of an aneuploidy must usually be small, since in the series of Robinson et al. (2001), no increased risk was discernible for women having had more than one previous spontaneous abortion due to triploidy (or aneuploidy) to have yet another chromosomally abnormal pregnancy. Prenatal karyotyping and/or early pregnancy ultrasonography may reasonably be offered.
True (that is, not artifactual) tetraploidy is too rare for a clear picture to have emerged. Sporadic occurrence would seem very probable.
1. Two of the no. 21 chromosomes had identical haplotypes, indicating that the mother's mosaicism was due to postzygotic error in an initially normal 46,XX conception (Fig. 2-8a).
2. His paper was entitled “Some notes on heredity counselling,” and he also referred to “genetical counselling,” one of the first uses of this expression.
3. Dispermy could be deduced simply from the cytogenetic analysis in the case reported in Lim et al. (2003), the man carrying a translocation 46,XY,t(2;6)(p12;q24). The 69,XXY mole had both the balanced translocation and one unbalanced form, reflecting fertilization with one sperm from alternate segregation and the other from adjacent-1.
4. The data for trisomies 13 and 18 have not been separated, on the assumption of a probable similar underlying mechanism, and given the practical fact of the small numbers of the respective numerators.
5. This multiple applies to a subsequent pregnancy whenthe mother is still under age >30. For a subsequent pregnancy at maternal age 30, the increase in the appropriate age-related risk is less, at 4.6×.
6. But note the recent evidence from Warburton that anincreased risk for subsequent aneuploidy may actually apply in the case of XXX and XXY (Table 16-2).