Structural rearrangements are often divided into two general categories, balanced and unbalanced. Balanced rearrangements contain no net loss or gain of genetic information and the individuals who carry them are generally phenotypically normal. In contrast, additional and/or missing genetic material is present in individuals who carry unbalanced rearrangements. Just as modifications in the amount of the various ingredients added to any recipe cause changes in the final product, deviation from the normal disomic genetic complement results in a clinically affected individual.

While it is easy to define balanced and unbalanced rearrangements, distinguishing between a truly balanced and an unbalanced rearrangement using traditional cytogenetic techniques is often impossible. The maximum level of resolution obtained using standard microscopy of G-banded prometaphase chromosomes is reported to be 3–5 megabases or 3–5  ×  10⁶ base pairs. This number will vary, however, ­depending on the quality of the chromosome preparations and the skill of the cytogeneticist examining the karyogram(s). The ability to resolve or identify a rearrangement will also be influenced by the degree to which the banding pattern, overall size, and centromere location of an involved chromosome is altered. Obviously, the more apparent the change, the more likely it is to be detected. A number of molecular cytogenetic techniques such as fluorescence in situ hybridization (FISH), 24-color karyotyping, and chromosome microarray analysis are currently being used to detect submicroscopic or otherwise cryptic rearrangements that cannot be detected using traditional cytogenetics (see Chaps. 17 and 18). Recent studies suggest that chromosome imbalances can be found in an additional 10–14% of patients using chromosome microarray analysis [35–37]. Additionally, chromosome microarray analysis has demonstrated that as many as 40% of phenotypically abnormal individuals with apparently balanced simple chromosome rearrangements detected by traditional karyotyping actually contain submicroscopic gains and/or losses at or near one or more rearrangement breakpoints, while others have been found to have cryptic imbalances unrelated to their identified balanced rearrangements [38–41].

Once a structural chromosome rearrangement is detected, regardless of whether it is balanced or unbalanced, the subsequent steps to take depend on the type of specimen that was analyzed. For prenatal samples or those obtained from a child, parental karyotypes or molecular cytogenetic studies should be obtained whenever possible to assess whether the rearrangement has been inherited or represents a de novo mutation. If neither parent is found to be a carrier of the rearrangement, the most likely scenario is that it represents a de novo abnormality rather than an inherited one. Non-paternity should be considered. Since the possibility of gonadal mosaicism can never be excluded, this family would be given a very low risk of having another child with the same structural abnormality. Prenatal testing would also be offered for all future pregnancies.

In contrast to the very low recurrence risk quoted to a couple with a child or pregnancy carrying a de novo rearrangement, the risk of chromosomally abnormal conceptions for an adult who carries a balanced structural rearrangement is much higher. In fact, for some familial rearrangements, the risk can approach 50%, and for very rare carriers of a homologous Robertsonian translocation, the risk for an abnormal conception is 100%. It is therefore imperative that these ­families are identified so that they can be given accurate genetic counseling regarding their reproductive risks and options. In situations where a familial rearrangement is identified, it must be remembered that it is not just the immediate family but distant relatives as well who may be at risk for having children with unbalanced karyotypes and associated mental and/or physical abnormalities. By systematically karyotyping the appropriate individuals in each generation, all those with elevated reproductive risks can be identified and appropriately counseled regarding their risks and options. Although there has been some debate regarding the appropriateness of karyotyping the phenotypically normal minors of balanced carriers, 50% of whom would be expected to be balanced carriers themselves, there is a consensus that these children should be referred for appropriate genetic counseling when they reach reproductive age.

The situation becomes a bit more complex when chromosome analysis of a bone marrow or tumor specimen results in an apparently balanced rearrangement, not associated with any particular neoplasm, in all cells examined. In these cases, it is imperative to ascertain whether such a rearrangement represents a patient-specific acquired change (which can then be monitored during treatment, remission, relapse, or any change in disease aggression) or a constitutional abnormality present from birth.

The reasons for this are twofold. First of all, from the point of view of the physician treating the patient, the presence of any acquired cytogenetic change is significant (see Chaps. 15 and 16). Equally as important in the long term, however, is establishing whether the chromosome rearrangement is familial and therefore has potential reproductive consequences for the extended family. While the potential familial reproductive issues are understandably not the primary concern of oncology patients nor their physicians, given the critical nature of their disease and the fact that cancer patients are often well beyond childbearing age themselves, this is an issue that should not be overlooked. Genetic counseling is covered in detail in Chap. 21.

Every chromosome rearrangement was at one time a new or de novo rearrangement that carried the risks associated with an undefined entity. Children who carry unbalanced rearrangements, regardless of whether they represent new mutations or an unbalanced form of a familial rearrangement, almost inevitably demonstrate an abnormal phenotype. An imbalance is an imbalance regardless of how it arose.

In contrast, accurate predictions regarding the phenotype of a child or fetus that carries an apparently balanced de novo chromosome rearrangement are more difficult to make. In this situation, it is not known what has occurred at the molecular level within the rearrangement, and there are no family members with the rearrangement from whom inferences can be made. The risk for an abnormal phenotype is therefore always higher for an individual with an apparently balanced de novo rearrangement than for an individual who has inherited a similar rearrangement from a normal parent. Obviously, these individuals also carry a significantly higher risk for phenotypic abnormalities than their chromosomally normal counterparts. Several population studies have shown, for example, that the incidence of de novo apparently balanced rearrangements among the mentally retarded is approximately seven times that reported in newborns [42]. Apparently balanced de novo rearrangements detected at amniocentesis have also been associated with a risk for congenital abnormalities that is two- to threefold that observed within the general population [1].

A number of different mechanisms are thought to be responsible for the abnormal phenotypes observed in children with apparently balanced de novo rearrangements. One possibility is that the translocation is not truly balanced. As discussed earlier, structural rearrangements that appear balanced at the microscopic level may actually contain large duplications and/or deletions at the molecular level. Another possibility is that the rearrangement is “balanced,” but a break has occurred within a critical gene or its surrounding regulatory sequences such that the gene product or its expression is altered. This scenario has been demonstrated in several patients with Duchenne muscular dystrophy, for example [43]. A position effect, in which the expression of a specific gene or group of genes is altered when the chromosome segment containing them is moved to a different location, could also result in an abnormal phenotype. Such an effect has been demonstrated in several X;autosome translocation chromosomes in which inactivation seems to spread from the inactive X chromosome into neighboring autosomal segments. This phenomenon has been documented in drosophila and plants as well. Finally, the possibility that an individual’s abnormal phenotype may be completely unrelated to his or her rearrangement must always be examined. Other nonchromosomal genetic disorders, prenatal exposures, birth trauma, non-paternity, etc., must all be considered.

Balanced structural rearrangements may pass through multiple generations of a family without detection. When these families are ascertained, it is usually due to the presence of infertility, multiple spontaneous pregnancy losses, and/or clinically abnormal family members (Fig. 9.2). Meiotic events that result in cytogenetically unbalanced conceptions can explain the presence of all three occurrences within these families.

During normal meiosis, homologous chromosomes pair utilizing a mechanism of formation thought to depend, at least in part, upon interactions between their shared sequences. Under normal circumstances, all 23 pairs of homologous chromosomes align themselves to form 23 paired linear structures or bivalents that later separate and migrate to independent daughter cells (see Chap. 2). In cells carrying structurally rearranged chromosomes, pairing cannot occur in a simple linear fashion. Instead, complex pairing configurations are formed in an attempt to maximize pairing between homologous regions that now differ with regard to their chromosomal location and/or orientation (see the later sections “Deletions,” “Duplications,” “Inversions,” “Reciprocal Autosomal Translocations,” and “Insertions”). Chromosome malsegregation and/or particular recombination events within these complex configurations can then lead to unbalanced conceptions, many of which never implant or are spontaneously lost during gestation.

Cytogeneticists are frequently asked to make predictions regarding a balanced carrier’s risk of producing an abnormal liveborn child. While this is a legitimate question, it is in practice very difficult to answer accurately. One source of difficulty is the fact that, with very few exceptions, each family’s rearrangement is unique. Therefore, unless a family is large and accurate information regarding the reproductive history and phenotype of each family member is available, typically no empiric data are available from which to obtain risk values. A second source of difficulty one encounters in assessing the reproductive risks associated with a particular balanced rearrangement is the breadth and complexity of the variables involved.

One important factor that is considered when assessing the reproductive risks of a carrier parent is the extent of imbalance demonstrated by the potential segregants. In general, the smaller the imbalance, the less severe the phenotype, and the more likely the survival. An additional rule of thumb is that the presence of excess genetic material is less deleterious than the absence of genetic material. Another variable to be considered is the quality of the genetic information involved. Some chromosomes, such as 16 and 19, are infrequently involved in unbalanced structural rearrangements. Presumably, this occurs because of the importance of maintaining a critical dosage for a gene or group of genes on these chromosomes. Conversely, imbalances involving other chromosomes such as 13, 18, 21, X, and Y appear to be more easily tolerated. In fact, a complete trisomy involving any of these chromosomes is survivable.

Each family’s reproductive history can also provide important clues regarding the most likely outcome for an unbalanced pregnancy. As might be expected, those families or individuals who have had a liveborn child or children with congenital abnormalities, especially when an unbalanced form of the familial rearrangement has been documented, are at highest risk for having unbalanced offspring. In families or individuals in whom multiple spontaneous abortions and/or infertility are noted, the risk for liveborn unbalanced offspring would be expected to be lower. In these families, it is assumed that the unbalanced conceptions are being lost very early as unrecognized pregnancies (infertility) or later during gestation. Interestingly, the sex of the carrier parent, in some cases, also influences the risk of having unbalanced offspring. In situations where a sex bias does exist, the female carrier invariably possesses the higher risk. Why male carriers appear to produce fewer unbalanced offspring than their female counterparts is not known. Perhaps fewer unbalanced segregants form during spermatogenesis relative to oögenesis, and/or the selective pressure against unbalanced gametes is greater in the male, and/or imprinting effects may cause the unbalanced embryos of male carriers to be less viable than those of their female counterparts. Male infertility may also play role [26, 44] (see Chap. 11).

On rare occasions, an abnormal phenotype is observed in an apparently balanced carrier of a familial rearrangement. While some of these cases may simply represent coincidental events, other possible explanations exist as well. Very rarely, abnormal offspring resulting from uniparental disomy, or the inheritance of both homologous chromosomes from a single parent, has been documented in the offspring of balanced translocation carriers [45] (see Chap. 20). Incomplete transmission of a partially cryptic rearrangement has also been observed in the abnormal offspring of a phenotypically normal carrier parent. Wagstaff and Herman, for example, describe a family in which an apparently balanced (3;9) translocation was thought to be segregating [46]. After the birth of two phenotypically abnormal offspring with apparently balanced karyotypes, molecular analysis demonstrated that the father’s apparently balanced (3;9) translocation was actually a more complex rearrangement involving a cryptic insertion of chromosome 9 material into chromosome 8. Abnormal segregation of this complex rearrangement led to a cryptic deletion of chromosome 9 material in one sibling and a duplication of the same material in the other.

Phenotypic discrepancies between child and parent may also be explained by the presence of a recessive allele that is inherited from a chromosomally normal parent. While the parent is phenotypically normal due to the presence of a complementary normal allele on the homologous chromosome, the abnormal allele can be expressed in the offspring, who has no normal allele. The affected child inherits two mutant alleles; one mutant allele is inherited secondary to the balanced chromosome rearrangement, while the other is inherited from the cytogenetically normal parent (Fig. 9.3). In a slight variation of this theme, the second inherited hit or mutation in the affected child is nonallelic but presumably functions within a biological pathway identical to or related to the first hit. The idea is that each parent carries a single mutation that is benign or causes only mild clinical manifestations of disease. When both mutations are inherited together, however, the involved pathway(s) is sufficiently altered to cause disease. Alternatively, one predisposing mutation could be inherited, while the other occurs de novo.

The two-hit scenario described earlier has also recently been invoked to provide one possible explanation for the variable expressivity and decreased penetrance observed in association with many microdeletions and microduplications. This is especially true of some of the smaller copy number changes that are currently being identified by chromosome microarray analysis [47, 48].

Autosomal deletions that can be detected by traditional, “high-resolution,” or molecular cytogenetic methods produce monosomies that are generally associated with significant pathology. Some exceptions, however, do exist. Loss of the short arm material from acrocentric chromosomes during the formation of Robertsonian translocations, for example, has no impact on phenotype. Similarly, the striking size variation of heterochromatic regions in normal individuals suggests that loss of some, if not all, of this material is insignificant. There have even been reports of “benign” deletions in regions that are considered euchromatic. Gardner and Sutherland catalog deletions of this type in bands 2p11.2-p12 and 2q13-q14.1; 3p25.3-pter; 5p14; 8p23.1-pter and 8q24.13-q24.22; 9p21.2-p22.1; 11p12; 13q21; 16q21 and 16q13-q22; and 18p11.2-pter [49]. Close examination of these regions reveals that many are relatively gene poor and therefore less likely to contain a dosage-sensitive gene. For example, the 11 Mb of genomic material within the 5p14 band and the 7 Mb of genomic material within the 11p12 band contain only eight and nine genes, respectively.

Among deletions of pathological significance, classic cytogenetic deletions that can be detected by routine methodology tend to be larger and associated with major malformations. Generally, large deletions have a more significant impact on phenotype and survival than smaller ones. The nature of the deleted material, however, also plays an important role in determining whether a specific deletion is viable. Thus, deletions of large segments of the short arms of chromosomes 4 and 5, and of the entire short arm of chromosome 18, are recurrent abnormalities among infants with major malformations, while deletions of similar size involving the short arms of chromosomes 17 and 19 are rarely, if ever, seen in liveborns [50].

Classic deletions have traditionally been described as either terminal (Fig. 9.4) or interstitial (Fig. 9.5) based on chromosome banding patterns. A deletion is considered “­terminal” if there is no discernable material beyond the site of initial breakage. Conversely, interstitial deletions have a proximal breakpoint, missing material, and a more distal breakpoint beyond which the chromosome continues with a normal banding pattern to its terminus. All stable chromosomes have telomeres comprised of the human consensus telomere sequence (TTAGGG)_n. Chromosomes with apparent terminal deletions are no exception and are assumed to have acquired “new” telomeres following the deletion event.

Several mechanisms for acquiring or retaining a telomere have now been documented among chromosome deletions. One mechanism referred to as telomere healing involves the addition of a new (TTAGGG)_n sequence at or near the deletion breakpoint [51–53]. In these cases, a telomerase recognition site in the vicinity of the deletion breakpoint is bound by the enzyme telomerase, which synthesizes a completely new telomere. These therefore represent true terminal deletions. Other chromosomes with apparent terminal deletions have been shown to actually represent derivative chromosomes that have acquired their subtelomeric and telomeric regions from another chromosome secondary to a translocation event. These translocation or “telomere capture” events are hypothesized to occur secondary to homologous recombination mediated by regions of shared homology that exist within the deleted chromosome and the subtelomeric region of a separate chromosome [54, 55]. Still other deletions appear to be terminal by traditional cytogenetic analysis but have been shown by molecular analysic analysis to be interstitial. It is estimated that 7–25% of apparent terminal deletions fall into this category [56–58]. Because a chromosome with an interstitial deletion retains its original telomere, there is no reason to synthesize or acquire a new one.

The use of “high-resolution” banding and molecular cytogenetic techniques has led to the identification of another class of cytogenetic abnormality variously referred to as chromosomal microdeletions, contiguous gene syndromes, and segmental aneusomy syndromes (SAS). These ­abnormalities are mostly very small interstitial deletions, often at or below the resolution of microscopic analysis, that recur with appreciable frequency and are associated with distinct clinical phenotypes. The term, “microdeletion,” is descriptive but fails to include the minority category of “microduplications” (e.g., CMT1A; see also section “Duplications,” later) and the variable etiologies for some of the disorders. The term “contiguous gene syndromes” was introduced in 1986 to describe the involvement of multiple contiguous genes in the production of a clinical phenotype [59]. While this terminology remains appropriate for some of the disorders in this new category, others are actually single gene disorders, or the result of imprinting defects or uniparental disomy (see Chap. 20). In an effort to more accurately characterize the pathogenesis of these disorders, the term segmental aneusomy syndrome was proposed to imply that the phenotype is the result of “inappropriate dosage for a critical gene(s) within a genomic segment” [60].

Williams syndrome is one example of an SAS that results from a small deletion [61]. These patients typically carry a  ∼1.5 Mb deletion within the proximal long arm of chromosome 7 that encompasses approximately 30 different known and predicted genes. At least some of these genes appear to be responsible for the cardiovascular abnormalities, growth and developmental delays, infantile hypercalcemia, and dysmorphic facial features that are associated with Williams syndrome. Deletion of the elastin gene (ELN), for example, has been implicated in the cardiovascular abnormalities. This gene is also presumed to play a causative role in some of the other features associated with this syndrome including renal artery stenosis, hypertension, hoarse voice, premature sagging of the skin, and perhaps some of the facial features. Similarly, loss of LIM-kinase 1 (LIMK1), a novel kinase expressed in the brain, is predicted to explain some of the cognitive abnormalities in these patients. Presumably, some, or all, of the remaining genes identified within the common Williams syndrome deletion also contribute to the physical features associated with this contiguous gene syndrome.

Molecular studies of the Williams syndrome deletions have revealed the presence of flanking low-copy repeat (LCR) sequences at the common breakpoint sites. These LCR sequences appear to provide recombination sites for unequal meiotic and mitotic exchange events that produce the recurring Williams syndrome deletions [62–64]. In some cases, these unequal exchange events seem to be promoted by the presence of heterozygosity for a submicroscopic paracentric inversion that spans the same low-copy repeat sequences that mediate the common 1.5 Mb Williams syndrome deletion [65–68]. It is estimated that the risk of having a child with a Williams syndrome deletion is approximately fivefold higher for an individual who is heterozygous for this inversion when compared to an individual who does not carry it. However, despite their elevated relative risk, it is important to note their absolute risk remains low at approximately 1 in 1,750. Several studies have now demonstrated that, at least in some cases, these inversions increase the probability of a rearrangement by producing better LCR substrates for recombination. In the case of the recurring 15q13.3, 16p12.1, and 17q21.31 microdeletions, for example, the associated inversion polymorphism increases the probability of an unequal exchange by changing the directional relationship between the mediating LCRs [48, 69, 70]. Without the inversion polymorphism, the LCRs are in an inverted orientation and unequal exchange is not promoted. With the inversion, however, the LCRs are placed in a direct orientation and as such are ideally suited for the NAHR events between homologous chromosomes and sister chromatids that are primarily responsible for these deletions (Fig. 9.1a, b). In other cases, the inversion may improve upon the involved LCR substrates by increasing their length. It is also possible that the inversion loop that forms to maximize homologous chromosome pairing in the heterozygous parent renders the paired chromosomes more susceptible to unequal crossing-over (see section “Inversions” and Fig. 9.12 later).

As noted for Williams syndrome, flanking LCR sequences have also been found at the deletion sites of several other SASs. Recombination events localized to these LCR sequences appear to account for the size consistency and the frequency of the deletions associated with these disorders as well. A partial listing of classic cytogenetic deletion or SASs can be found in Table 9.1. Given the recent widespread use of chromosome microarray technology in many research and clinical cytogenetics laboratories, the number of microdeletion and microduplication syndromes being identified and characterized is growing at a rapid pace. For an extensive list of chromosome abnormalities that includes some of the more recently indentified syndromes and their associated phenotypes, see the Websites for Wellcome Trust Sanger Institute “DECIPHER,” the European Cytogeneticists Association Register of Unbalanced Chromosome Aberrations (ECARUCA), and Unique – The Rare Chromosome Disorder Support Group [71–73].

In contrast to the size consistency and recurrent use of specific LCR sequences documented among many of the interstitial SAS deletions, other deletions appear to have multiple independent breakpoints and vary considerably in size. This size variability has been noted in association with multiple deletions including those that involve the short arms of chromosomes 1, 4, and 5 [28, 74, 75]. Nonhomologous end joining (NHEJ) and fork stalling template switching (FoSTeS), rather than nonallelic homologous recombination (NAHR), are believed to be responsible for these nonrecurring rearrangements [7, 76, 77].

The term “duplication” as applied to chromosome abnormalities implies the presence of an extra copy of a genomic segment resulting in a partial trisomy. A duplication can take many forms. It can be present in an individual as a “pure duplication,” uncomplicated by other imbalances (Fig. 9.6), or in combination with a deletion or some other rearrangement. Examples of some types of rearrangements that involve duplications include isochromosomes, dicentrics, derivatives, recombinants, rings, and markers. The origins and behavior of these abnormal chromosomes are discussed elsewhere in this chapter.

Tandem duplications represent a contiguous doubling of a chromosomal segment. The extra material can be oriented in the same direction as the original (a direct duplication) or in opposition (an inverted duplication). Most cytogenetically detectable tandem duplications in humans appear to be direct [78].

Autosomal duplications produce partial trisomies and associated phenotypic abnormalities. As mentioned in the introduction, the phenotypes associated with duplications are typically less severe than those associated with comparable deletions. Relatively few duplications, however, have occurred with sufficient frequency or been associated with such a strikingly characteristic phenotype that they have been recognized as defined clinical syndromes (Table 9.2). A few cases of distal 3q duplication have been reported in patients with features similar to Cornelia de Lange syndrome. However, these patients also have additional abnormalities not usually associated with the syndrome [50]. Paternally derived duplications of distal 11p have also been associated, in some cases, with Beckwith-Wiedemann syndrome [79]. More intriguing, and perhaps more significant, is the emerging recognition of recurring duplications that involve the same genomic segments that are associated with some of the established microdeletion syndromes. These complementary microduplication/microdeletion syndromes are thought to represent the reciprocal products of recurring unequal exchange events that are mediated by flanking homologous low-copy repeat (LCR) sequences. The causative unequal exchange events can occur following misalignment of either sister chromatids or homologs as shown in Fig. 9.1.

One of the first complementary microduplication syndromes identified involves the Prader-Willi and Angelman syndrome (PWS/AS) region within the proximal long arm of chromosome 15 (Fig. 9.7). These duplications are mediated by the same LCRs and encompass the same loci that are deleted in PWS/AS. The clinical significance of these particular duplications was initially difficult to assess because many affected patients had phenotypically normal relatives with the same apparent duplication. Molecular studies have now provided us with an explanation for the apparent absence of a genotype-phenotype correlation in these families. With few exceptions, the clinically affected individual(s) within these families demonstrate mental retardation, decreased motor coordination, autism spectrum disorder, and mild to no dysmorphic features and carry a maternally derived ­duplication. In contrast, the duplicated chromosome in the normal relatives of these patients is typically paternally inherited. These data suggest that imprinting within the PWS/AS region is responsible not only for the phenotypic differences we observe with maternal versus paternal deletions but also for the presence or absence of a clinical phenotype in patients with a duplication (see Chap. 20).

In addition to the PWS/AS region, complementary microdeletion-microduplication syndromes have also been documented for the Williams, Smith-Magenis, 17q21.31, and DiGeorge syndrome critical regions, as well as for the hereditary neuropathy with liability to pressure palsies (HNLPP)/Charcot-Marie-Tooth type I region (Table 9.2). More recently, chromosome microarray analysis has resulted in the identification of additional previously unrecognized complementary recurring microdeletions and duplications involving 1q21.1, 3q29, 15q13.3, 16p11.2, 16p13.11, 17q12, distal 22q11.2, and multiple other regions [48]. For an extensive list of chromosome abnormalities that includes some of the more recently indentified microduplication and microdeletion syndromes and their associated phenotypes, see the Websites listed earlier in the section “Deletions” [71–73].

Inversions are intrachromosomal rearrangements formed when a chromosome breaks in two places and the material between the two breakpoints reverses orientation. Inversions can be of two types: pericentric or paracentric. In pericentric inversions, the breakpoints lie on either side of the centromere and formation of the inversion often changes the chromosome arm ratio (centromere position) and alters the banding pattern of the chromosome (Figs. 9.8 and 9.9). Paracentric inversions, on the other hand, have both breakpoints on the same side of the centromere, or within a single chromosome arm (see Chap. 10, Fig. 10.6). In paracentric inversions, the centromere position does not change, and the only clue to their presence is an alteration in the chromosome banding pattern. Prior to the development of banding techniques, the existence of paracentric inversions was theorized but could not be proven.

In those studies in which parents of a proband with a unique inversion have been karyotyped, the inversion is found in a parent as often as 85–90% of the time [80–82]. Most inversions of both types therefore appear to be inherited. Additionally, among the handful of apparent recurring inversions studied thus far, most have not been formed from multiple independent events but instead have been inherited from a single common distant ancestor. Examples include inv(3)(p25q21), inv(5)(p13q13), inv(8)(p23q22), inv(10)(p11.2q21.2), and inv(11)(q21q23). The recurrent variant inv(2)(p12q13), however, appears to be one of the few exceptions (Fig. 9.9). This inversion has occurred in the human genome on multiple occasions, and its recurring formation appears to occur secondary to nonallelic homologous recombination (NAHR) [83, 84].