Chromosome Abnormalities and Genetic Counseling , 3rd Edition

11.Complex Rearrangements

Complex chromosomal rearrangements (CCR) occurring in phenotypically normal persons are rare. Batanian and Eswara surveyed the literature in 1998 and listed a total of 115 cases. Familial cases comprise a minority, with only about 35 examples recorded (Berend et al., 2002b). Three or more chromosomes are involved, and a considerable variety of rearrangements are possible. Translocation segments may involve distal segments, as in the usual reciprocal translocation, or interstitial segments, as in the insertion. An inversion and a translocation, for example, may coexist on the same chromosome. In the phenotypically normal person, the rearrangement is taken to be balanced.

BIOLOGY

Three Major Categories of Complex Chromosome Rearrangement

Three major categories of CCR are recognized (Kausch et al., 1988). The most common category is the three-way exchange, in which three segments from three chromosomes break off, translocate, and unite (Fig. 11-1).¹ Most three-way CCRs are familial and are usually transmitted through the mother; although in one of the largest kindreds on record, showing five-generation transmission, three (great)grandfa-thers must have been CCR heterozygotes (Farrell et al., 1994). The second category consists of more complicated CCRs, encompassing a wide theoretical range, but there are not many actual cases. The third category, the simplest CCR, is the double two-way exchange, in which there is a coincidence of two separate simple reciprocal translocations. In a sense, the double two-way exchange is not a true CCR and might as well be described as a double or a multiple rearrangement (Kausch et al., 1988; Creasy, 1989; Phelan et al., 1990).

An apparently balanced karyotype may be associated with a normal or an abnormal phenotype. If the individual is phenotypically normal, the chromosome rearrangement is assumed to be truly balanced. These cases are often familial. In the phenotypically abnormal individual, presumably some submicroscopic imbalance or other genetic defect exists. Such a case was proven in a CCR studied by Brandt et al. (1997), in a child with tricho-rhino-phalangeal syndrome (TRPS) (see p. 282). The child had a de novo apparently balanced t(7;13;8)(p21;q21;q24.1), but using FISH probes for the TRPS critical region at 8q24.1, a 3 Mb deletion was revealed. Kaiser-Rogers et al. (2000) describe two similar patients, and they use the expression “partially cryptic” CCRs. These cases characteristically involve a de novo chromosome abnormality. CCRs typically originate in male gametogenesis (Röthlisberger et al., 1999; Kaiser-Rogers et al., 2000).

MEIOSIS

The carrier of a CCR has a risk for an abnormal conception due to either malsegregation of the derivative chromosomes or the generation of a recombinant chromosome. Malsegregation follows the general principles as set forth for the simple translocation, but naturally the range of unbalanced combinations is greater. For the three-way CCR, the broad categories of malsegregation are 3:3 and 4:2, and (theoretically) 5:1 and 6:0. Recombination is rare indeed, and only eight examples are recorded (Berend et al., 2002b).

Three-way Complex Chromosomal Rearrangements

At meiosis in the three-way CCR heterozygote, the expectation is that the chromosomes involved in the rearrangement will come together and form a multivalent (Saadallah and Hultén, 1985; Fig. 11-2). Consider how meiosis would proceed in the rcp(2;18;11) translocation illustrated in Figure 11-1. In theory, a hexavalent configuration would allow full synapsis of homologous segments (Fig. 11-3). If disjunction were then symmetric (3:3), up to 20 possible gametic combinations could occur. The two arising from alternate segregation (arrows in Fig. 11-3) would be the only ones to be balanced; the remaining 18 would be unbalanced to a greater or lesser degree. Were asymmetric segregation (4:2, 5:1, 6:0) to occur, a great variety of extremely unbalanced gametes would result. However, it may be that in a number of families a tendency to favor symmetric alternate segregation and a combination of very early lethality of severely unbalanced conceptuses implies a fair prospect for achieving a normal pregnancy (Walker and Bocian, 1987). An excess of heterozygotes has been noted among the balanced female offspring (Batista et al., 1994).

The risk of having a pregnancy that would go to term but produce an abnormal child reflects the nature of the rearrangement—that is, whether there are possible chromosomal combinations that would lead to aneuploidy for a survivable amount of genetic material. Thus, considering the preceding rcp(2;18;11) example, three unbalanced combinations, one 3:3 and two 4:2, might be expected to be viable (Fig. 11-4). In reviewing 29 families with a CCR, Batista et al. (1994) determined that an abnormal live birth is most commonly (78%) due to 3:3 adjacent-1 segregation, followed by 4:2 segregation. Recombination would add yetfurther possibility of imbalance, but, as mentioned above, this is very rarely seen.

The concept of adjacent-1 and adjacent-2 segregations can be applied in the setting of the CCR, in the case of 3:3 segregations. Thus, the segregant gamete shown at left in Figure 11-4, having one of each chromosome pair represented (one of each centromere), would reflect adjacent-1 segregation. An example of 3:3 adjacent-2 segregation is given by Xu et al. (1997). A mother had the karyotype 46,XX,t(5;16;22), and cytogenetic analysis of her morphologically abnormal fetus following intrauterine death at 16 weeks gestation showed 46,XY,der(5),der(16),t(5;16;22),22. In this case, the abnormal ovum would have had one chromosome 5, two chromosome 16s (one normal, one the derivative), and lacked a chromosome 22.

4:2 segregation particularly characterizesCCRs in which an acrocentric chromosome is a component. Schwinger et al. (1975) reported a mother of two children with typical Down syndrome, who herself had a three-way 7;21;11 CCR. The affected children had an interchange trisomy 21, in that they had, in addition to the maternal translocation pattern, a second intact chromosome 21. Fuster et al. (1997) give an example of a 4:2 malsegregant, diagnosed at chorionic villus sampling, in which it took a chromosomal in situ suppression (CISS) hybridization study to recognize that there was a three-way paternal 2;22;11 CCR. The fetal karyotype could then be interpreted as 47,2,der(2),der(22)t(2;22;11)(q13; q11.2;q23). The parents continued the pregnancy, and the retarded and abnormal child had a double partial trisomy: a duplication of the segments 11q23–qter and 22pter–q11.2. The couple had previously had one normal child, and three miscarriages.

Exceptional Complex Chromosomal Rearrangements

More complex rearrangements imply an even greater potential range of abnormal gametes. Kausch et al. (1988) calculated a minimum of 70 possible unbalanced gametes due to 4:4, 5:3, 6:2, and 7:1 segregations from an octavalent in the case of a woman with a five-breakpoint CCR with translocations of chromosomes 1, 2, 5, and 11 and an inversion of chromosome 1 who had presented with three first-trimester miscarriages. Van der Burgt et al. (1992) report a similarly complex de novo balanced CCR (chromosomes 5, 11, 12, 16; five breakpoints in all) in a mother who had had one miscarriage, one 46,XY child, the index abnormal child, and, as a quite unexpected outcome, a de novo 45,rob(13q14q) at prenatal diagnosis in her fourth pregnancy.

The CCR shown in Figure 11-5 with six breakpoints in five chromosomes offers useful illustration (Bass et al., 1985). The woman who carried this rearrangement had four pregnancies, only one of which miscarried, and two produced offspring with a balanced constitution, though different in each child and different from their mother! Recombination involving the centric segment of chromosome 1 led to a daughter receiving a rec(1)—we might also call it a “new der(1)”—with just the 6p segment being translocated, and a son with a different rec(1) having just the 7q segment. A son and a grandson had unbalanced karyotypes which were different, but each led to partial 7q trisomy. Readers who relish esoteric puzzles may wish to refer to the original paper. In a similar vein, Madan et al. (1997) describe a mother with a familial four-breakpoint t(2;3;8) in which the 2q translocated segment had split, with 2q23–q33 going to the der(3) and 2q33–qter to the der(8). Her child's karyotype was a simple 46,t(2;3) with a “new der(3)”; the simplification resulted from recombination at maternal meiosis between her der(3) and normal chromosome 2.

Vying for the title of the most complicated familial CCR in the world are the cases described by Röthlisberger et al. (1999) and Berend et al. (2002b). In the former, a father carried a rcp(6;7;18;21)(q22 & q25;q21.3, q31.1 & q32.1;p11.21 & q21.3;q21.3), de novo in himself. As this unofficial nomenclature attempts to indicate, there were eight breakpoints altogether, two in chromosomes 6 and 18, three in 7, and one in 21. Flourescence in situ hybridization and spectral karyotyping were needed to clarify the detail of the rearrangement. Three different recombinant forms were passed to his three children: a rec(7), a rec(21), and a rec(18). The child with the rec(21) had a balanced karyotype, and he has become a balanced carrier for a simple translocation, 46,t(7;21)(q21.3;q21.3). The other two have partial trisomies, for 6q and 7q. In the study by Berend et al., the mother of an abnormal child with an apparently balanced three-way t(3p;8q;16q) CCR herself proved to carry a t(3;8;16) in which the der(3) had an exchange in each arm, the 3q being involved in an exchange with 5q, and there was also an inv(2). Seven breakpoints were identified in all, including the inv(2), and the authors' nomenclature was 46,XX,inv(2)(p23q37.1),der(3) t(3;8;16)(p24.2;q22.2;q22)t(3;5)(q21;q13.3), der(5)t(3;5),der(8)t(3;8;16),der(16)t(3;8;16).

A subtler example is the case in Gibson et al. (1997) of a mother with a five-break rearrangement, de novo in herself, in which two small interstitial segments in 2q and 5q, and two terminal segments in 1q and 5q, exchanged position. Her abnormal child was initially thought to have a possibly unbalanced t(1q;5q), but further analysis revealed his wonderfully complicated true karyotype, and the reader may care to draw an ideogram and then compare it with the original paper:

46,XY,der(1),der(5),t(1;5;2)(1pter → 1q42.3::5q23.2 → 5qter;5pter → 5q21.2::2q33 → 2q35::1q42.3 → 1qter;2pter → 2q33::5q21.2 → 5q23.2::2q35 → 2qter)mat.

Some rearrangements are so subtle that standard FISH cannot delineate the picture, and further sophisticated methodologies must be used. Chen et al. (1997)studied a mother with a karyotype of 46,XX,der(1),der(4) who transmitted a der(1) to her abnormal son. At first glance, this might have seemed a simple reciprocal translocation, involving the exchange of small distal 1p and 4q segments. The FISH picture, in which whole chromosome paints were used, presented more questions than answers, showing only a subterminal gap on the der(4), with absence of any chromosome 1 color staining. It took microdissection— scratching off parts of the der(4) from the glass slide and using these fragments to generate probes—to figure out exactly what had happened. The 1p translocated segment (1p36.13–pter) had united with chromosome 4 at 4q33, and then the combined 1p + 4q segment (1p36.33 → p36.13::4q33 →q31.3) underwent a paracentric inversion. The der(1) had only a single breakpoint, and so the child's karyotype could be written quite simply, as 46,XY,der(1)t(1;4)(p36.13;q33), in contrast to the mother's 46,XX,der(1)t(1;4)(4qter → 4q33::1p36.13 → 1qter),der(4)t(1;4) inv(4)(4pter → 4q31.3::1p36.33 → 1p36.13::4q33 → 4q31.3::1p36.33 → 1pter). For the record, although this is a de novo case, we note the 15-break CCR described in a retarded boy by Houge et al. (2003).

Double Two-way Complex Chromosomal Rearrangements

Presumably, two separate and independently operating quadrivalents form (Bowser-Riley et al., 1988). Burns et al. (1986) recorded sperm karyotypes in a man with a double two-way CCR 46,XY,t(5;11)(p13;q23.2),t(7;14)(q11.23; q24.1), whose wife had had four miscarriages, a child with cri du chat syndrome, and a normal son carrying the t(7;14). Only 4 of 23 sperm analyzed had an overall balanced complement, and the majority (13) had adjacent-1 segregants for one or other translocation. Another five showed 3:1 and (a unique observation) one sperm showed 4:0 segregation.²

We referred on p. 86 to a couple in which each member had a simple reciprocal translocation, both happening to involve chromosome 7 (7p in one, 7q in the other). It is a useful exercise to imagine how the chromosomes might be transmitted in this family. In theory, the couple could have a child with a double two-way CCR who would have a combination of their own karyotypes. Providing fertility were not compromised, this child of theirs in generation II could then have two types of balanced progeny in generation III: one with the t(7;11), and the other with the t(7;22), as in the couple of generation I. We set out this scenario in Figure 11-6. No offspring with a normal karyotype could be produced in generation III, unless recombination between the two der(7) chromosomes were to restore a normal chromosome 7.

A real example of this sort of complexity across generations, but even more complex, is given by Zahed et al. (1998). A grandfather had two separate translocations: a simple translocation rcp(1;8)(p31;q21.1), and an insertional translocation ins(9;8)(q34;p23.1pter). Thus, he had two abnormal no. 8 chromosomes: one having a segment from distal 1p attached at 8q21.1, and the other having a deletion at 8p23.1. He had a daughter and a son to whom he transmitted a rec(8), the same rec(8) to each, in balanced state. This rec(8) had a deletion at its p extremity and a 1p translocated segment on its q extremity. Presumably, his two abnormal no. 8 chromosomes had recombined in meiosis at a point somewhere between the p23.1 and q21.1 breakpoints. His daughter in turn had two children, and in each of them she restored, by recombination, the grandpaternal chromosomes: the del(8p) in one child, and the der(8)t(1p;8q) in the other. Both children had an unbalanced state, but different in each. One had a straightforward del(8)(p23.1) karyotype, and thus a partial 8p monosomy. The other had the grandpaternal simple rcp(1;8), and would otherwise have been normal; but in addition she inherited the ins(9), which conferred a partial 8p trisomy. The reader may care to draw the chromosomes of the three generations from this description, and check back to Figure 2 in the original paper of Zahed et al. (1998).

Effect on Fertility

In several complex rearrangements, in the female at least, gametogenesis can accommodate itself to the complexity thrust upon it, and the heterozygote may be fertile and have pregnancies that produce phenotypically normal children. However, the rule of the greater vulnerability of spermatogenesis to chromosomal complexity seems to apply particularly in the situation of the CCR, and the male heterozygote is often sterile because of spermatogenic arrest or is subfertile (Rodriguez et al., 1985; Saadallah and Hultén, 1985). The involvement of an acrocentric chromosome in the CCR may particularly predispose to this male sterility (Gabriel-Robez et al., 1986).

The Cryptic Complex Chromosome Rearrangement

Upon detailed cytomolecular study, a CCR may be shown to have a greater number of breakpoints than had originally been appreciated (Batista et al., 1994). An apparently simple translocation may actually harbor a more complex rearrangement, not detectable on routine cytogenetics but requiring FISH and molecular analysis for its demonstration. At most, this phenomenon is likely to be infrequent among all translocations, and it may in fact be rare. The example from Chen et al. (1997) discussed above is one example.

Another is that of Wagstaff and Hemann (1995). They describe a phenotypically normal father and his two abnormal children, the father and son having an apparently balanced 46,XY,t(3;9)(p11;p23) and the daughter apparently 46,XX. On FISH and DNA studies, they were able to show that the father had a tiny segment of chromatin from the breakpoint in 9p23 removed and inserted into the long arm of a chromosome 8 (Fig. 11-7). At meiosis, it may have been that a quadrivalent formed from the no. 3 and no. 9 elements, while the two no. 8 homologs synapsed independently as a bivalent. On this interpretation, the two children reflect alternate segregation of the no. 3 and no. 9 elements. With respect to the chromosome 8s, the t(3;9) son inherited his father's normal homolog, and so the lack of the 9p23 segment was not corrected, while the “46,XX” daughter received the chromosome 8 with the 9p23 insertion. Thus, the son has a del(9)(p23) and the daughter, a dup(9)(p23). A somewhat different scenario is that presented in Aboura et al. (2003). Here, a mother and her infant son, the latter with minor dysmorphism and abnormal functional neurology, appeared at first to have the same simple t(3q;22q) translocation. FISH analysis showed this to be a t(3;22;9)(q22;q12;q34.1).Yet finer analysis using a probe to the ABL locus on 9q34.1 revealed a very small deletion at this site on the der(9) of the proband, but not in his mother, and neither in a carrier sister. The deletion was presumed de novo, arising during maternal meiosis. These scenarios raise pressing questions: do some other apparently balanced simple reciprocal translocations have a cryptic complex rearrangement; and how often does a de novo deletion occur on the background of a parental balanced rearrangement (p. 94)?

GENETIC COUNSELING

The male CCR heterozygote who is not otherwise known to be fertile should have a semen analysis to check whether sperm are being produced. For the heterozygote (male or female) who is fertile, a conceptus having either a normal chromosome constitution or the same balanced CCR as the parent would be expected to produce a normal child. But a high proportion of conceptions have an unbalanced karyotype. Madan et al. (1997) have determined empiric risk estimates. Overall, the risk for spontaneous abortion is 50%, and the risk for a liveborn abnormal child is 20%. The level of risk is related to the mode of ascertain-ment—whether through the birth of abnormal infants, multiple miscarriage, male infertility with abnormal spermatogenesis, or fortu-itously—and to the family history. If multiple miscarriage has been the pattern in the family in the past, it is likely to continue to be so. In such cases, it may be that all unbalanced forms would lead to miscarriage (Creasy, 1989). If abnormal infants have been born, carriers are likely to have a high risk for the same thing to happen again.

For the three-way CCR, it is generally justifiable to advise that, sooner or later, a normal outcome could reasonably be expected. Thus, the couple may be willing to make continued attempts until a successful pregnancy is achieved. As always, the pedigree should be studied, in order to understand what might be the particular pattern of meiotic behavior with that CCR. If the reproductive history is very unpromising, optimism may need to be guarded, and the reality of a low chance for a normal child faced (Evans et al., 1984). Intuitively, the likelihood for a successful pregnancy would be less for an “exceptional CCR.” Preimplantation diagnosis might be seen as particularly appropriate in these cases, but the complexity of the rearrangement and making the distinction between balanced and unbalanced forms would be challenging and possibly insurmountable.

Bowser-Riley et al. (1988) review the specific case of the double two-way translocation and propose that the risk of having an abnormal child would be approximately the sum of the figures derived separately for each rcp. They acknowledge that this might be an overestimate due to nonviability of doubly imbalanced combinations, although each on its own might be viable.

As for prenatal diagnosis, some patients may prefer initially to rely on first-trimester ultrasonography, declining chorionic villus sampling, and leaving early abortion to happen naturally. An unfortunate miscarriage history might well lead to a heightened sensitivity to the small risk associated with prenatal diagnosis. Others may prefer the early information that a chorionic villus sampling could provide. If the pregnancy continues normally on ultrasound into the second trimester, a judgment can be made whether this of itself would be sufficiently reassuring (perhaps in the setting of all unbalanced forms being very unbalanced) or whether amniocentesis would in fact be desirable. On several levels, each case will have to be assessed on its merits. The CCR will need to be very carefully characterized cytogenetically in the parent and the fetus to ensure accurate prenatal diagnosis.

The same balanced state identified at prenatal diagnosis raises the same questions, but more pointedly, as in the simple reciprocal translocation (p. 94). By way of example is the CCR 46,XX,t(5;16;10;18)(q13;q22;q11.2; q21) identified at routine prenatal diagnosis in a woman having a history of recurrent miscarriage, reported by Lee et al. (2002), with the same karyotype then being shown in herself. Normal ultrasonography was encouraging, and the pregnancy was continued; at age 2 years, the child was normal. But this fortunate outcome could not have been guaranteed.

Advice in the case of a de novo CCR discovered at prenatal diagnosis is given on p. 417.

Notes

1. In the I.S.C.N. description of the karyotype, the order of chromosomes in the three-way CCR is as follows: firstly, the lowest number (or X) chromosome; secondly, the chromosome which receives a segment from the first; and lastly, the chromosome donating a segment to the first listed chromosome. Thus, the karyotype for the CCR shown in Fig. 11-1 is written 46,XX,t(2;18;11) (q13;q21.1;p15.3).

2. This case is instructive in illustrating the point that different rcps can have different meiotic 11) segregants but only 30% of the rcp(7;14) showed alternate segregation.