A question of reproduction is usually an academic matter in individuals with functional autosomal aneuploidy. But toward the milder end of the phenotypic range, social and emotional development may be such that forming a stable relationship is possible. Some who lack that degree of maturity may yet have a social freedom that opens the possibility of a sexual encounter. Either on their own behalf, or more likely through the agency of parents or other caregivers (whose agenda may include sterilization; see Karp, 1981; Wolf and Zarfas, 1982), such people may present to the genetic clinic. Ethical issues raised in this context are presented in Chapter 1.
The mosaic state may on occasion be associated with an apparently normal phenotype, or at least very close to normal. Some of these persons will have been identified by their having had a child with the trisomy for which they themselves have a low-level mosaicism. Some present with infertility. The theoretical risk will depend on the extent to which the gonad carries the aneuploid line.
In contrast, some structural imbalances are without discernible phenotypic effect. Deletion or duplication of a small segment of euchromatin, or the presence of an extra structurally abnormal chromosome (ESAC), is occasionally recognized fortuitously in normal and fertile individuals.
Parental Trisomy 8 Mosaicism
Mosaic trisomy 8 arises postzygotically from an initially normal conceptus (Robinson et al., 1999b). Habecker-Green et al. (1998) review reports of reproductive status in 46/47,8 individuals, and there is only a tiny number of cases, usually in persons in whom the diagnosis would not have been suspected clinically. They describe a woman having a history of four spontaneous losses, including a 46,XX fetal death at 27 weeks; her next pregnancy produced an apparently normal 46,XX daughter. Rauen et al. (2003) report a woman who presented a more typical clinical picture of trisomy 8 mosaicism having a 46,XX child (phenotypic abnormality in the child probably reflected paternal characteristics). Mercier and Bresson (1997) studied an otherwise healthy man with whom recurrent miscarriage was the presenting problem and whose peripheral blood karyotype was 47,XY,+8/46,XY. On FISH analysis of 25,000 spermatozoa, 398 (1.6%) showed disomy 8, which compared with a rate in control sperm of 0.2%. It is perhaps surprising that such a low level of disomic 8 sperm should be associated with a high miscarriage rate (always assuming that the link is causal and not coincidental). We have seen a somewhat similar case, a man of above average intelligence and excellent physical health, with infertility due to oligospermia, in whom low level trisomy 8 mosaicism was shown on two separate blood samplings.
Parental Trisomy 18 Mosaicism
This is extremely rarely recorded in adulthood, and several cases had phenotypic abnormality of greater or lesser degree (Satge et al., 1996; Ukita et al., 1997; Lim and Su, 1998). Some patients presented with a history of miscarriage, and three (one father, two mothers) had a child with trisomy 18. Because of the usual high rate of lethality of trisomy 18 in utero, the genetic risks in such persons would apply substantially to miscarriage. Bettio et al. (2003) report a woman of normal intelligence with 70% trisomic 18 cells on blood but none on fibroblast karyotyping, presenting with infertility. Ovarian biopsy showed 90% trisomic cells from right ovarian biopsies, and normal karyotype in left ovarian tissue. The gonad may be free of the trisomic line, as in the father of a normal daughter described in Lim and Su (1998). He was of normal intelligence and worked as a sales representative, and had “slightly unusual facial features.” The trisomic line was found only in blood (76%) and not in skin fibroblasts, and the disomic 18 rate in sperm was similar to that of a control.
Parental Trisomy 21 Mosaicism
In practice, it is usually only those with a low percentage of 21 cells who seek genetic advice. These people typically come to notice because they are studied as apparently normal parents of more than one child with Down syndrome (DS). The important factor, if it could only be known, is the degree to which the gonad comprises 46,N and 47,+21 cells. The trisomic cells produce disomic and normal gametes in equal proportion; of course, normal cells, other things being equal, give rise only to normal gametes. Thus, the proportion of abnormal gametes produced depends on the proportion of germ cells that are trisomic. In the limit, the gonad might be fully 47,+21. Any level of correlation between the degree of mosaicism in lymphocytes and gametes is not readily amenable to study. Familial trisomy 21 mosaicism is on record but is exceptional (Werner et al., 1982).
Maternal Trisomy 21
At female meiosis, the classical scenario is that the three homologs form either a bivalent and a univalent, or a trivalent (Fig. 13-1; Wallace and Hultén, 1983). If the former occurs, the bivalent may disjoin and segregate symmetrically, but the univalent passes at random to either daughter cell (1:1 <1 segregation). If the latter occurs, a trivalent may of itself set the stage for aberrant segregation (2:1 segregation). In either case, the result is disomic (24,+21) and normal (23,N) gametes in equal proportions. Speed (1984) has observed trivalents in about 40% of meiotic cells and a bivalent plus a univalent in the remaining 60%. An alternative scenario is that the “third” chromosome 21 separates prematurely into chromatids, and each chromatid then passes to a daughter cell (the oocyte, and the first polar body). Cozzi et al. (1999) provide direct evidence for this mechanism in the FISH study of unfertilized oocytes from a woman who was presumed to be a 47,+21/46 gonadal mosaic.
Figure 13-1. Possible synapsis of three no. 21 chromosomes: (a) as a trivalent and (b) as a bivalent and a univalent.
In a review of the literature, Rani et al. (1990) list 30 reports of pregnancy in DS women. The ratio of DS to normal offspring was 10:17 (there were three abortions), which is not significantly different from a 1:1 ratio but suggestive of a deficit in trisomic offspring. A reasonable interpretation is that 46,N and 47,21 conceptions occur with equal frequency, but loss of pregnancy is greater with the trisomic fetuses. About one-third of the 46,N offspring were nevertheless abnormal, which may have reflected paternal or environmental factors. Cunniff et al. (1991) noted a diminution in the number of oocytes in the ovaries of DS girls at the time of birth, which subsequently could be the cause of subfertility.
Paternal Trisomy 21
Schröder et al. (1971) observed in a study of male meiosis in trisomy 21 that spermatogenesis apparently can proceed normally. Two DS males have been documented as having fathered a child (Sheridan et al., 1989; Bobrow et al., 1992; Zühlke et al., 1994), and one other has been implicated (Thompson, 1962).
Parental Partial Aneuploidy (46,rea)
Uncommonly, the unbalanced and phenotypically abnormal carrier of a chromosomal rearrangement may be functionally fertile. The usual forms are deletion, duplication, and the derivative chromosome from an unbalanced translocation. Barber (2000) lists deletions and duplications recorded in the literature, with all but four of the autosomes represented at least once, and chromosomes 5 and 18 being the most numerous (Table 13-1).
A number of instances of familial partial autosomal monosomy due to transmission of a deleted chromosome are on record, and some representative ones are mentioned here. Keppen et al. (1992) describe mild to moderate retardation in a grandmother, mother, and child, each having a deletion of chromosome 5 short arm, p13.3–p14.3 (not with cri du chat syndrome). A severe form of cri du chat syndrome has been documented in mother and child (Martínez et al., 1993; and see p. 17), as well as a number of instances of three-generational transmission of mild forms (Church et al., 1995). Loss of a tumor suppressor gene, the APC gene, which is the basis of familial adenomatous polyposis, accompanied a del(5)(q22q23.2) in a retarded man, his aunt, and inferentially his retarded mother, causing polyposis in at least the man and his aunt; other family members carried a balanced insertional rearrangement (Cross et al., 1992). Pettenati et al. (1992)report deletion for the segment 8p23.1–pter, and provide a family photograph showing unremarkable physical appearances; the del(8p) children had learning and behavioral difficulties. Fukushima et al. (1987) describe a mother and son with a 13q deletion, both with retinoblastoma; curiously, only the child had developmental delay. Neavel and Soukup (1994) report mother-to-child transmission of the Jacobsen syndrome of terminal 11q deletion (p. 283). A few examples are known of inheritance from parent to child of the 18p syndrome (Rigola et al., 2001). Aviv et al. (1997) describe a family in which some individuals with a 45,der(22) karyotype carried a unique 2122 composite chromosome in which part of 22q had been translocated onto 22p and the segment 21q21–qter translocated onto the 22q breakpoint. These people had “poor social adjustment, behavior problems, and learning difficulties,” presumably due, in part at least, to monosomy for the tiny segment 21q11.1–q11.2. The well known chromosome 22q11 deletion can be transmitted from parent to child, with a different clinical picture in each (p. 289).
Table 13.1. Euchromatic Deletions or Duplications on Record as Having Been Transmitted from Parent to Child and Having an Associated (and Presumably Causally Associated) Abnormal Phenotype
With increasing sophistication in the detection of subtle karyotypic defects, this general problem is likely to present itself more frequently. For obvious reasons, inherited deletions are more likely to be small, and quite probably a number of cases remain yet to be discovered. Sanford Hanna et al. (2001)describe a very small distal 1q deletion, del(1)(q42.1q42.3), in a mother and son with rather minor physical anomalies, the mother requiring assisted living services and the son attending special education classes at the age of 13, working at approximately a third grade level. Some years earlier, such a small deletion might have been beyond resolution. Subtle subtelomeric rearrangements may be associated with a less abnormal phenotype, and more examples of familial transmission are likely to come to light (Baker et al., 2002b). In one family, carriers of an unbalanced nonreciprocal t(6q;16p) in at least two generations had an effective subtelomeric 6q monosomy, and a rather mild phenotype, although the child of one of them with partial trisomy 16 was very abnormal (Kraus et al., 2003b).
Transmitted duplications are slightly less often reported than are deletions. Glass et al. (1998) list parent-to-child transmission of duplications of 7p, 8p, 9p, and 14q, and they report their own case of a dup(2)(q11.2q21.1), due to an (8;2) insertion. The mother and daughter with the dup(2) state had minor facial dysmorphism, short stature, mild intellectual deficit, and psychiatric disease (schizoaffective disorder in the mother, paranoid psychotic state in the daughter). With an insertion such as this, the genotype is presumably that of a pure 2q partial trisomy. If the duplication is associated with a known syndrome, then parent and child both carry the same diagnosis, as Joyce et al. (1999) illustrate in a mother and daughter with Silver-Russell syndrome due to a dup(7)(p12.1p13). Delatycki et al. (1999) describe a father and daughter with an unbalanced translocation, der(20)t(6;20)(p23;p13), in which the major imbalance is the duplication for 6p23–pter, although there may also have been an effect due to 20p telomeric deletion. Interestingly, the der(20) father, who was nonmosaic at least on blood and skin fibroblast analysis, was relatively mildly affected intellectually and was able to work as an inventory controller and use a computer. The daughter, as is typical in 6p duplications, displayed global developmental delay of considerable degree, at least at the age of 4 years.
Wolff et al. (1991) and Moog et al. (1994) each describe a mother and child having an abnormal chromosome 18 with a direct duplication of 18p: the phenotypic normality in the former pair, compared with mild retardation in the latter two, may reflect a slight difference in the amount of duplicated material. The mother in Moog et al.'s study may have been somewhat protected by a minor (20%) 46,XX line. Three-generation transmission of a dup(7)(p13p12.2) is recorded in association with an unremarkable physical and mild functionally abnormal neurological phenotype (Schaefer et al., 1995). We have seen a family with inferentially four-generation transmission of a duplicated chromosome (band 10p14). One great-grandparent may have been a gonadal mosaic, the duplication being passed down in two separate branches of the family, causing considerable psychological morbidity (Voullaire et al., 2000a). Only when the cytogeneticist gathered three families with the same dup(10p) for a study was it realized that they were, in fact, connected.
Chromosome 8p is disproportionately represented, presumably a reflection of a less crucial complement of loci in this region. Dhooge et al. (1994) describe a mildly retarded mother with a duplication of segment 8p22–p23.1 (or possibly 8p21.3–p23.1), whose two children had the same dup(8). A review of direct duplications of this segment in Fan et al. (2001) records a handful of similar cases. A somewhat larger segment is involved in the case of Sklower Brooks et al. (1998) in which a man with dup 8p22–pter due to an unbalanced Y;8 translocation (Fig. 5-12) passed the same unbalanced complement to a child of his; he himself was academically limited (but employed as an elevator operator and in reception security for a large firm), and his son required speech therapy. A little further proximally down 8p, Moog et al. (2000) describe three retarded sisters and the retarded daughter of one of them with a dup(8)(p12p21.1).
Genomic imprinting was an important factor influencing the effect of a familial ins(2;6) (p22.2;q22.33q23.3) that had been transmitted through three generations (Temple et al., 1996; discussed also on p. 169). A child with transient neonatal diabetes had a duplication for 6q22–q33 due to the insertion, which she had inherited from her father. He and his mother also had the same unbalanced karyotype, but neither had any history to suggest neonatal diabetes. It was the imprinting effect due to paternal transmission that resulted in the child's medical condition.
The dup(15)(q11.2q13). This is a vexing chromosome. In most cases, important euchromatin is not involved, and it is to be regarded as a normal variant (see p. 242). But in others a true functional aneuploidy exists, with clinical consequences. The distinction between a harmless and a harmful dup(15)(q11.2q13) is to be made according to the absence or presence, respectively, of the Prader-Willi/Angelman critical region (PWS/ASCR). There is the added complication that this region of chromosome 15 is subject to imprinting, with phenotypic abnormality typically occurring in the setting of transmission from the mother (see also p. 290).
A particularly complex story is that of the family in Gurrieri et al. (1999). A child with pervasive developmental disorder, atypical autism, and epilepsy had a dup(15)(q11q13) with duplication from D15S11 to D15S113; his mother had a duplication, but not quite so extensive, not including the D15S11 locus (from D15S128 to D15S113), and the grandparents had normal chromosomes. Thus the first duplication had arisen on transmission from the normal grandfather, and then on transmission from the normal mother to the affected child, the duplication extended itself. The child's abnormality is likely due to aberrant expression of imprintable genes.
If a normal cell line coexists with an abnormal karyotype—in other words, there is a 46,(abn)/ 46,N mosaic state—a high risk for abnormal pregnancy may be implied, an essentially normal parental phenotype notwithstanding. Yip et al. (1996) describe a normal woman with 46,XX,der(6)t(6;8)(q27;q22.2)/46,XX who presented with recurrent miscarriage. A balanced 6;8 translocation had presumably arisen at a somatic mitosis, and subsequently an unbalanced der(6) cell line was generated. This cell line likely contributed to ovarian formation, and the miscarrying pregnancies were due to conceptions from 23,der(6) ova. If the abnormal cell line in the parent is predominant, their phenotype will be abnormal, as exemplified by the mother reported by Cox et al. (2002b) having the karyotype 46,XX,dup(7)(p15.3p22)/ 46,XX, and whose abnormal son had the duplication in nonmosaic state. Very rarely, there can be parental mosaicism with two abnormal cell lines having the opposite imbalance, and a child can be born with one of these imbalances in nonmosaic state (de Pater et al., 2003b).
Parental Partial Aneuploidy due to Extra Structurally Abnormal Chromosome (47,1ESAC)
Genetically Important Extra Structurally Abnormal Chromosomes
Extra structurally abnormal chromosomes (SACs) are frequently called “supernumerary marker chromosomes” (SMCs), but since this latter term implies that their origin is unknown we prefer the term ESAC. Transmission of an ESAC that contains active genetic material and leads to a functional partial aneuploidy is very rarely observed. Mosaic cases are recorded and may be associated with a reproductive risk if the abnormal cell line is represented in the gonad. For example, Rothenmund et al. (1997) identified a tiny ESAC in a father as derived from the pericentromeric region of chromosome 8; he had the mosaic karyotype 47,XY,+mar/46,XY, while his two mentally handicapped daughters had a (nonmosaic) karyotype 47,XX,+mar. Crolla et al. (1997) propose that der(22) ESACs can be sorted into those with and those without euchromatin, the former typically being with phenotypic consequence, and the latter having none. Urioste et al. (1994b) describe a familial unstable supernumerary chromosome with a mother and two daughters having variable manifestations of cat-eye syndrome due to 47,+der(22)/46,N. (Partial autosomal aneuploidy resulting from a ring chromosome is dealt with in Chapter 10.)
Crolla (1998) reviewed the published studies of ESACs from 1991 to 1998, numbering some 168 cases. Excluded from his summary were the well-recognized inv dup(15), the rea(22) of the cat-eye syndrome, and ESACs that can be identified with standard cytogenetic methodology, such as isochromosomes of 9p, 12p, and 18p. The list is exhaustive, and every chromosome other than no. 5 is represented (Masuno et al. , subsequently completed the list with their case of 47,XY,+mar(5)/46,XY). Many were associated with phenotypic abnormality, in some severe; some ascertained at prenatal diagnosis eventually proved to be normal. Given the wide range of cytogenetic heterogeneity, of course, it is completely unsurprising that there should be a wide phenotypic range, including normality. Each ESAC needs to be assessed on its merits.
Genetically Harmless Extra Structurally Abnormal Chromosomes
By circular reasoning, the ESAC is regarded as harmless if it is seen in a phenotypically normal individual; the inference thus being drawn that the ESAC contained no genetic material of consequence. This category of ESAC is also called a “B-chromosome.” These are found by chance in normal people or at prenatal diagnosis (Ridler et al., 1970; Tsukahara et al., 1986; and see Chapter 25). Many ESACs are very small and prone to loss during cell division. Consequently, mosaicism and, in the case of a transmitted chromosome, familial mosaicism are seen frequently (Chudley et al., 1983). The majority of these harmless ESACs comprise acrocentric short arm and pericentromeric material or other autosomal pericentromeric chromatin (Callen et al., 1992). Adhvaryu et al. (1998) describe a bisatellited 15-derived ESAC in only 2% of cells in a grandfather, presumed of postzygotic origin in him, whose daughter and a grandson were nonmosaic 47,+ESAC, with the chromosome evolving in another grandchild into a very small ring. All these people were phenotypically normal.
The ESAC would probably form a univalent at meiosis rather than synapsing with whatever chromosome it was derived from. Martin et al. (1986b) analyzed sperm chromosomes from two men who had a bisatellited ESAC. Slightly less than half the sperm were found to carry it, although the distribution did not differ significantly from 1:1. However, Cotter et al. (2000) showed a much reduced frequency of sperm carrying an ESAC—6% compared with a prior expect of 50%—in a phenotypically normal man with 47,XY,+del(15)(q11.2). Familial ESACs are characteristically maternally transmitted, which could reflect preferential exclusion of the marker in spermatogenesis or, in some, a reduced male fertility. Jaafar et al. (1994) studied a phenotypically normal man presenting with infertility who had a supernumerary bisatellited heterochromatic chromosome. In most spermatocytes, the ESAC was in close proximity to the X-Y bivalent, and this may have been the cause of the infertility.
Parental Trisomy 8 or Trisomy 18 Mosaicism
The very rare individual for whom genetic counseling may be a practical issue is likely to have a low-grade mosaicism, and the diagnosis may have been made because of reproductive loss. Recurrent miscarriage may reflect a higher fraction of trisomic tissue in the gonad, and the risks for a further pregnancy may be substantial. Wei et al. (2000) undertook preimplantation genetic diagnosis in one case, an infertile man with mosaic trisomy 18.
A separate issue relates to the possible risk for hematological malignancy in individuals with trisomy 8 mosaicism (Seghezzi et al., 1996; Brady et al., 2000). Trisomy 8 is well known as an acquired change in the cascade of carcinogenesis in myeloid neoplasias (particularly myeloid leukemia, myelodysplastic disease). It may be that constitutional trisomy 8 acts similarly, in this case as the “first hit,” in those individuals in whom the trisomic cell line includes the marrow (as, of course, it must in patients diagnosed on a peripheral blood karyotype). The level of risk for cancer predisposition is unknown. For the myeloid neoplasias, unlike many other cancers, little therapeutic advantage comes from earlier diagnosis, and the counselor will need to consider carefully how, or if, this issue may be raised.
Parental Trisomy 21 Mosaicism
Theoretically, the risk for having a child with (nonmosaic) Down syndrome is high—up to 50%. Presumably the risk is related to the proportion of gonadal cells that are trisomic, but this is not accessible information. The proportion seen on lymphocyte analysis offers no real help in this question. One point is clear: it is certainly appropriate to offer prenatal diagnosis. Indeed, preimplantation genetic diagnosis may be warranted, as Conn et al. (1999) attempted with a woman who presented with a history of three out of four pregnancies with trisomy 21, and who showed a nonmosaic normal karyotype on blood. Three of four unfertilized oocytes were hyperhaploid 21, and four out of six analyzable embryos were trisomic 21 (some mosaic). Transfer of the two chromosomally normal embryos was not successful.
Parental Down Syndrome
The risk of conceiving a trisomic 21 conceptus is presumably 50%, at least in the case of maternal DS. The other parent is likely to have a degree of mental deficiency, possibly due to a genetic factor. Prenatal diagnosis would of course detect those fetuses with trisomy 21, but the risk of a chromosomally normal fetus having a birth defect or being mentally handicapped could be as high as 30% (Rani et al., 1990). Prevention of pregnancy in those DS women at risk is regarded by many as advisable. In one confirmed case of paternity in a DS man, prenatal diagnosis (with a 46,XY result) was performed (Bobrow et al., 1992). We refer the reader to the discussion in Chapter 1 (p. 16) for a general review of the ethical is-sues in this setting.
Parental Partial Autosomal Trisomy or Monosomy
The risk for a child to have the same defect as the parent is 50%, or very close to it. A lesser risk applies in the case of mosaicism. In the case of a person of only low–normal/borderline intelligence who is functionally fertile, the issue of the risk and possible questions of prenatal diagnosis or sterilization will be difficult to raise (p. 16).
A phenotypically normal individual said to have a partial aneuploidy merits a cytogenetic reevaluation, including the application of high-resolution banding and possibly molecular genetics, to check for the possibility that the supposed aneuploidy is actually a balanced rearrangement that was only partially characterized cytogenetically.
As cytogenetic techniques have become refined, new “defects” are being identified which, upon family study, are revealed as being unusual but functionally balanced forms (Barber, 1994, 2000). For practical purposes, they can be regarded as variants that are rare or possibly even “private” to that family. The counselor must take care that apparently abnormal karyotypes are not overinterpreted. In a family study undertaken to clarify a particular unusual chromosome, family members should understand the reason for the study. If the “defect” is de novo in an abnormal child, it may well have been the cause of the abnormality; if a parent and other relatives carry the same chromosome (especially if transmitted through both males and females), its probable harmlessness is demonstrated.
The Harmless Extra Structurally Abnormal Chromosome
Each chromosome needs to be judged on its merits when a detailed cytogenetic assessment and parental karyotyping study have been undertaken. If an ESAC is judged to be harmless, then on that basis it would be immaterial whether a parent transmits it to an offspring. Nevertheless, it may be useful to know that, in a particular family, there has not been any associated phenotypic abnormality, a point that may be settled by doing a family study. This enables the counselor to offer firm reassurance that those who inherit it in the future will also likely be normal. A possible association with male infertility is mentioned above.