Human conception and pregnancy is both a vulnerable and a robust process. It is vulnerable in that a large proportion of all conceptions are chromosomally abnormal, with the great majority of such pregnancies aborting. It is robust in that more than 99% of the time a term pregnancy results in a chromosomally normal baby. Unbalanced chromosomal abnormalities are seen in less than 1% of newborns (Table 1-3). But the economic cost of chromosomally abnormal conceptions is not horrendous; it is measured largely in terms of miscarriage, seen or unseen. Relatively speaking, the occasional chromosomally abnormal child is an exceptional outcome—the tip of an iceberg (Fig. 21-1).
Most of this chromosomal vulnerability lies in the process of producing eggs and sperm. Meiosis hangs, literally and figuratively, upon tender filaments, and often the meiotic chromosomes are incorrectly distributed to the daughter cells. Indeed, humans are more prone to produce aneuploid germ cells than any other species studied (McFadden and Friedman, 1997). (Actually, Warburton suggests that this error rate conveyed an evolutionary advantage in former times: miscarriages due to aneuploidy led to a wider spacing of offspring, enhancing their survival to an age to be able to mate.) The group particularly likely to produce abnormal gametes is carriers of balanced chromosome rearrangements, and much of this book is devoted to that fact.
Advances in reproductive technology now enable many otherwise infertile couples to have children. Translocation carriers may have recourse to preimplantation genetic diagnosis (PGD) as a means of improving their chances of achieving a successful pregnancy (Chapter 24). In the case of men with poor sperm production, intracytoplasmic sperm injection (ICSI) at in vitro fertilization (IVF) is a means of getting a single sperm into an egg. Success with IVF is not necessarily easy to achieve, nor is it a certain outcome, and counselors dealing with infertile couples need a particular awareness of the psychological and practical difficulties they may face (Boivin et al., 2001). A “failed embryo transfer” following IVF may be considered a form of pregnancy loss not unlike that of the natural miscarriage of a wanted pregnancy.
BIOLOGY
Gametic Cytogenetics
Many more sperm are made than eggs, by orders of magnitude, and logically one might expect a higher standard of meiotic fidelity in the scarcer gamete (Hunt and Hassold, 2002). But in fact it is the other way round, and so it is the egg that commands most of our attention in terms of the practical relevance of gametic chromosomal pathology.
|
|
|
Figure 21-1. The iceberg of chromosomal pregnancy loss. |
Oocytes
In vitro fertilization has become widely applied in the management of infertility, and one consequential benefit of this has been the access afforded to study of the oocyte. Many eggs sampled prove to be surplus to the requirements of the couple, and are often willingly donated for research. Overall, a fifth to a quarter are cytogenetically abnormal. Pellestor et al. (2002) report their own experience with 1397 analyzable oocytes, and critically review the literature. Most of the abnormalities are accounted for by hyperhaploidy with a 24chromosome count (an additional double-chromatid or single-chromatid chromosome), hypohaploidy with a 22-chromosome count (a missing homolog) or a 23-chromosome count (homolog represented by only a single-chromatid chromosome), and diploidy with 46 chromosomes. These three categories of abnormality might produce, respectively, zygotes with trisomy, monosomy, and triploidy. The aneuploidy rates correlate with increasing maternal age, most notably in the single-chromatid nondisjunction category (Pellestor et al., 2003). One particular type of abnormal egg, the giant binucleate oocyte, is typically diploid (Balakier et al., 2002; Rosenbusch et al., 2002).
Another window to oocyte karyotyping is to use the polar body as its countertype surrogate, an approach that may be applied for clinical diagnostic purposes (Munné et al., 1995; Verlinsky and Kuliev, 2000). Both the polar body and the oocyte were assessed in the study of Mahmood et al. (2000), in which 127 oocytes from 72 IVF patients (average age 33 years) were subjected to 7-chromosome FISH analysis. According to their scoring criteria, 9 of the 127 oocytes (7%) were hyperhaploid (disomies 13, 16, 18, 21, X), with chromosome 21 the extra one in five oocytes. Since two of the women may have been (surprisingly) gonadal mosaics, it would not necessarily be valid to double this figure to account for hypohaploid states.
All these data are necessarily influenced by the source of the material: the ova come mostly from women being treated for infertility, who are typically of an older childbearing age, and some of the ova may be of poor quality, although the distribution of abnormality does not differ whether the infertility is due to the female or to the male (Pellestor et al., 2002). What does seem clear is that in the vicinity of 20%–25% of eggs from this population are chromosomally abnormal, about half of these manifesting as full aneuploidy. Maternal age is a most important correlate, and this link is well illustrated in the work of Battaglia et al. (1996; and see p. 32), which shows how the structural integrity of the oocyte's meiotic apparatus declines as a woman gets older.
Sperm
The gamete whose chromosomes are most readily accessible to analysis is the sperm. The earlier sperm karyotyping studies used the “humster” (human sperm + hamster ovum pseudofertilization) test. Guttenbach et al. (1997) reviewed the findings from eight research groups in the field, combining a total of over 20,000 sperm karyotypes from healthy donor men using the humster test. From this accumulated experience, the conclusion is that around 10% of sperm are chromosomally abnormal.
Aneuploidy is observed in 1%–3%, and about another 5%–10% of sperm have structural chromosome abnormalities, many of which were presumed to have arisen during spermatogenesis as an immediate postmeiotic event.
More recently, fluorescence in situ hybridization (FISH) analysis has enabled chromosome counts to be made on very large numbers of sperm. This approach bypasses any question that the humster test might have selected against abnormal sperm, although in fact it appears that this did not happen (Van Hummelen et al., 1996). Shi and Martin (2000c) have reviewed the published experience, observing that over 5,000,000 sperm from about 500 normal men have been analyzed in a number of laboratories around the world, using one-, two-, or three-probe FISH. Considerable variation existed between subjects, which was probably a biological effect; interlaboratory variation was also noted, presumably reflecting local methodological differences. The average disomy rate for each of the autosomes ranges around the 0.1%–0.2% mark. The figure for chromosome 14 being a little above, at 0.4%, may be artifactual, while higher levels in the G-group chromosomes, nos. 21 and 22, are more likely to reflect reality. X+Y disomy is observed in 0.25%. No ethnic differences have come to light, at least in a comparison of Caucasian and Chinese sperm chromosomes (Shi and Martin, 2000b). Neither is there any correlation with paternal age, or at least none that is consistently observed, except possibly with respect to XY disomy (p. 364).
Fathers of Aneuploid Children. There are very few data on 46,XY men who have actually fathered children with chromosomal abnormalities. Hixon et al. (1998)studied 10 fathers of paternal-error Down syndrome children (the error at meiosis I in three, meiosis II in six, postzygotic in one) and found no differences: these fathers had a mean of 0.15% disomic 21 sperm, vs. 0.17% in the controls. Similarly, Blanco et al. (1998) studied a group of 15 fathers of children with trisomy 21, and overall, the fraction of disomic 21 sperm was little different from a control group: 0.31% vs. 0.37%. However, of the total of 25 fathers from the two studies, three stood out with twice the level of sperm disomy 21; and two of these also showed an increase in sperm disomic for chromosomes 13 and 22 (Soares et al., 2001a). As for monosomy X Turner syndrome (TS), three-quar-ters of which may be the consequence of nullisomy X in the sperm, four fathers of TS daughters reported in Martínez-Pasarell et al. (1999) and in Soares et al. (2001b) had increased levels of sperm with a sex chromosome aneuploidy, as well as with disomies 13, 21, and 22. Thus, it is tempting to suppose that a minority of normally fertile men may be predisposed to meiotic errors at spermatogenesis, whether generalized or restricted to one chromosome; but the data are as yet too insubstantial to make a firm statement in this regard.
Whether or not meiotic recombination occurs may be an important factor influencing the integrity of the disjunctional process (a point well established in female meiosis). Savage et al. (1998) demonstrated that among paternally derived trisomy 21 cases, the rate of chromosome 21 recombination was only about half the expected level. Likewise, Hassold et al. (1991) showed that recombination was considerably reduced in the pseudoautosomal region of males with 47,XXY Klinefelter syndrome, in whom the causative meiotic error had been shown to be paternal. The conceptions in these cases had been, presumably, 24,XY sperm + 23,X ovum. Shi et al. (2001b) provide corroborative support for this concept in their analysis of sperm from a normal male who was usefully heterozygous for two markers in the pseudoautosomal region (DXYS15 and STS). Like any normal man, he produced some disomic 24,XY sperm among the normal 23,X and 23,Y sperm. In a comparison of the normal and disomic sperm in this subject, recombination occurred between the two markers in 38% of the former and in only 25% of the latter.
Sperm karyotyping studies in men who themselves have an abnormal karyotype, whether balanced (e.g., a reciprocal translocation) or unbalanced (e.g., 47,XYY), are dealt with in the chapters on the particular abnormality.
Cytogenetics of the Very Early Conceptus
An aneuploid gamete (nullisomic or disomic) will lead to an aneuploid conceptus (monosomic or trisomic). A diploid gamete, combining with a normal gamete, will give rise to a triploid conceptus. On simple arithmetic, given that around 20% of oocytes may be aneuploid and 10% of sperm are abnormal, and supposing equal fertilizing/fertilizable capacity, the expectation is that about 30% of conceptions will be chromosomally unbalanced. On top of this, dispermy (two sperm fertilizing the one ovum) can cause triploidy. An abnormal postzygotic cell division can give rise to mosaicism, and this may be a common event (discussed below). A very rare event is the generation of uniparental disomy (UPD) due to two gametes being coincidentally nullisomic and disomic for the same chromosome. These several possibilities all add up to a substantial potential for chromosome abnormality in the very early conceptus, in the first days of existence.
The development of the technology of PGD in association with IVF has offered a much clearer view of the frequency of chromosomal abnormalities in the zygote and in the first 2–3 days post-conception.1 Admittedly, couples presenting for PGD will not, in the main, be a true representation of all couples. One category of patient will, however, be close to “chromosomally typical”: otherwise normally fertile women who are heterozygotes for a Mendelian condition presenting for diagnosis of embryonic sex or for specific mutation testing. These embryos offer the best insight to the true background rate of chromosomal abnormality, with respect to the maternal age groups involved. In assessing some or all of chromosomes 13, 16, 18, 21, 22, X, and Y, Pellicer et al. (1999), studied 10 Mendelian heterozygous women of mean age 34 years, range 30–36. These women had a total of 12 abnormal embryos out of 62 tested (19%), but a considerably higher figure (46%) was observed in a group of older mothers presenting for the same reason (see also Table 21-1).
The “atypical” group of patients presenting to the IVF clinic are, of course, a population of clinical interest, and thus the observations gained from study of them, however unrepresentative they might possibly be of the general population, are very germane to the agenda of the counselor. Numerous studies confirm that a high fraction of unused IVF embryos are karyotypically abnormal, the ranges observed being from 30% to 65% (Wilton, 2002). Phenotypically abnormal embryos with multinucleate blastomeres are very likely to be chromosomally abnormal (Kligman et al., 1996). In a study of day-3 embryos, Bahçe et al. (1999) looked at aneuploidies not typically associated with clinical miscarriage or with term pregnancy, and showed that trisomies 1, 15, and 17 occur not infrequently. Voullaire et al. (2000b) applied the technique of comparative genomic hybridization (CGH) to the study of embryos that had surpassed the 5-year statutory storage limit in the state of Victoria, Australia, comprising 12 embryos from eight women aged 26–33 years. Comparative genomic hybridization has the advantage of enabling detection of aneuploidy for all chromosomes. Upon thawing and culturing, three of these twelve embryos were scored as being of good morphology, three as average, and six as poor. Three out of three good-morphology embryos were chromosomally abnormal in one or more blastomeres, and likewise two out of three average-morphology embryos and four out of six poor-morphology embryos; thus only a quarter of embryos were normal. Several of the aneuploidies were in mosaic state. (On five-color FISH, only three of the nine abnormal embryos detected by CGH would have been recognized as abnormal.)
The short period during which the cleavage embryo advances through the morula stage and into the blastocyst may be a major bottleneck during which the development of many chromosomally abnormal pre-embryos arrests, with most monosomies and many trisomies being lethal, although the figures do vary among laboratories (Clouston et al., 2002; Rubio et al., 2003). Extensive mosaicism may also prevent progression into the blastocyst (Bielanska et al., 2002a). Those abnormalities that survive may then be able to continue through until the next major period during which selection pressure applies, later in the first trimester and the early second trimester, when pregnancy loss is recognizable as “clinical miscarriage” (see below).
Mosaicism of the Pre-Embryo
Mosaicism of the very early embryo has been one of the more startling discoveries to emerge from the PGD laboratory, although there is always the caveat that it would be unsafe to draw too many conclusions from these laboratory observations concerning the risk for mosaicism in chromosomally normal zygotes conceived naturally. Many IVF embryos are aneuploid or diploid/haploid mosaics, and even in normal-appearing embryos the fraction, analyzing a limited number of chromosomes, ranges from 17% to 43% (Wilton, 2002). Iwarsson et al. (1999), applying FISH for chromosomes 15, 16, 17, 18, X, and Y, found as many as 72% of good-morphology freeze-thawed embryos to be chromosomally abnormal, with 57% being diploid mosaics. The first few postzygotic divisions may be especially susceptible to nondisjunction, at least under the conditions obtaining at IVF, and especially so with a zygote that is karyotypically abnormal at conception. The first two divisions (going from 1 to 4 cells) may be the most hazardous. When the mosaicism is extensive (several different karyotypes), these embryos may be referred to as being chromosomally “chaotic.” As noted above, using CGH methodology to check on every chromosome, Voullaire et al. (2000b) showed that a majority of surplus embryos of IVF patients were mosaic. A CGH study by Blennow et al. (2001) on embryos from translocation carriers that had been diagnosed as aneuploid on PGD demonstrated in some that every cell could be different—the absolute maximum mosaicism.
Further insight is afforded in following development for a few more days, into the blastocyst stage. Magli et al. (2000) tracked the cytogenetic progression in vitro of a number of known aneuploid embryos. Having been identified at PGD on day 3, they were then allowed to proceed for 2 more days through the morula and blastocyst stages, at which point the cells of the inner cell mass (which give rise to the embryo proper; Color Fig. 25-1, 2, see separate color insert) were separated and analyzed by FISH. One embryo, for example, with monosomy 18 on PGD, went on to yield 46 analyzable cells from its inner cell mass, of which 22 were monosomic 18, and 24 were normal. Another embryo that was monosomic for both X and 18 at PGD showed four cell lines in its inner cell mass: normal, monosomy X and 18, just monosomy 18, and “complex.” Others exhibited similar scenarios. An attractive hypothesis had been that the abnormal cells in a mosaic conceptus would often be sequestered to the trophoblast, and thus be out of harm's way, in the developing chorionic villi, but Magli et al. show that this is not necessarily the case, at least in vitro. Rather, these karyotypically compromised blastocysts are typically abnormal throughout.
Cytogenetics of Spontaneous Abortion and Later Pregnancy Loss
Nonimplantation and Occult Abortion
Although the natural in vivo circumstance might differ from the observations in vitro, it seems nevertheless a fair assumption that a substantial fraction of human conceptions have a lethal genetic burden and will not implant. It becomes a semantic question whether the existence of a nonimplanting morula or blastocyst could be described as a pregnancy, and whether its loss could be considered an abortion. Transient implantation may be associated with little or no perturbation of the menstrual cycle, although the woman may fleetingly feel pregnant as a hormonal response is briefly elicited. This is occult abortion (Miller et al., 1980). Monosomy, or extensive mosaicism, may be lethal even before the morula converts into the blastocyst, at least on in vitro observation, but some will reach the blastocyst stage (Sandalinas et al., 2001; Clouston et al., 2002; Rubio et al., 2003). Some trisomies impart early lethality. Trisomy 1 may exist in a small fraction of day-3 embryos in an IVF population, and yet is almost unknown in an established pregnancy. Most trisomy 17 is apparently lost in similar circumstances, with a fall in its incidence from 12% at day 3 to only 0.2% in spontaneous abortions (Bahçe et al., 1999; Dunn et al., 2001). The frequency and range of aneuploidies seen in blastocysts is similar to that seen at the stage of late first-trimester miscarriage, and this is now the next major period during which selective pressure is exerted.
Clinical Miscarriage
At the late first trimester, we have a clearer idea of how many conceptuses are chromosomally abnormal and what the abnormalities are. Of all recognized pregnancies (recognized in the traditional way, that is), about 10%–15% end in clinical miscarriage, or spontaneous abortion, mostly toward the end of the first trimester. If the products of conception are successfully cultured and karyotyped, in most studies a little over a half of abortuses are shown to have a chromosome abnormality, the fraction being somewhat greater in miscarriage occurring before 10 weeks gestation (McFadden and Friedman, 1997; Fritz et al., 2001b). A more efficient and probably more accurate approach is to sample chorionic villi transabdominally, following the diagnosis of inevitable abortion on ultrasound, and by this means abnormality rates above 60% are observed (Sánchez et al., 1999).
The technique of CGH allows analysis without the prerequisite of successful culturing, and Fritz et al. (2001b) report a 72% aneuploidy rate from a base material of 57 specimens that had previously failed to culture, using extracted DNA from stored frozen tissue. They note that previous studies having the highest culture success showed the highest aneuploidy rate, indicating that a majority of failed cultures are aneuploid. Their proposition that the aneuploidy incidence is underestimated is well taken, and their suggested figure of “nearly 70%” as the contribution of chromosome abnormality to the totality of first-trimester pregnancy loss has merit. When the cytogenetic focus is specifically on those cases of missed abortion in which a severely growth-disorganized embryo is actually observed through an endoscope prior to operative evacuation, the proportion with a chromosome abnormality does reach 70% (Philipp and Kalousek, 2002).
|
|
|
Figure 21-2. Chromosomal findings in products of conception from spontaneous abortion (after Kajii et al., 1980). |
Trisomies account for about 60% of all cytogenetic abnormalities identified at spontaneous abortion (Fig. 21-2). The most commonly seen abnormal karyotypes are trisomy 16, monosomy X, and triploidy. Monosomy X and triploidy account for approximately 20% and 15% of all abnormalities, respectively, and as many as 1% of all human conceptions may have trisomy 16 (Benn, 1998). Double trisomy is infrequent, this being an observation in 0.7% of Reddy's (1997) series. Structural rearrangements constitute most of the remainder.
The origin of the abnormality is, in most, an error at maternal meiosis I, and this includes most of the major trisomies: trisomies 13, 14, 16, 21, and 22, with trisomy 18 being a possible exception (p. 256). Robinson et al. (1999b) analyzed the originating status of certain of the less studied karyotypes: trisomies for chromosomes 2, 4 through 10, 12, 15, 17, and 20. About three-quarters showed three alleles for the trisomic chromosome, thus confirming a meiotic origin. Most of the remainder are presumed to have been somatic errors; some might have been mosaic, but not detected as such. Trisomy 8 is unusual, in that all cases are due to a meiotic error, which stands in contrast to somatic errors being almost entirely the basis of mosaic trisomy 8 that is diagnosed postnatally. Uniparental disomy appears not to be a causative factor in miscarriage (Smith et al., 1998b; Robinson et al., 1999b).
Phenotypes of the Embryo/Fetus. An embryo or fetus may or may not be identifiable in the products of conception collected at the time of spontaneous abortion due to chromosomal abnormality. Severe growth disorganization can be graded according to whether there is complete absence of any detectable embryonic parts (empty or anembryonic sac, or blighted ovum), a tiny nubbin of tissue without recognizable embryonic landmarks, or an embryo in which cephalic and caudal poles can be distinguished (Philipp and Kalousek, 2002). The triploid embryo in Figure 21-3 has a very distorted but recognizable face, trunk, and limbs. Warburton et al. (1991) provide a graphic catalog of embryonic/fetal phenotypes from their material of about 1300 karyotypically abnormal spontaneous abortuses collected over a 12-year period in New York state. What actually leads to expulsion of the conceptus from the uterus may be the declining vascular and endocrine function of the placental tissue, with decidual necrosis (that is, death of tissue) finally causing uterine irritation and contraction (Rushton, 1981). The underlying process of decline, or at least a contributory factor, may be accelerated apoptosis: Qumsiyeh et al. (2000) observed a higher apoptotic index in villi of the abortus with an abnormal versus a normal karyotype.
If an abnormal twin dies, the normal twin may ensure continuation of the pregnancy, and only a parchment-like vestige (fetus papyraceous) remains, preserved in the uterus along with the normal twin. A “vanishing twin” has plausibly been proposed in the study of a pregnancy in which two cell lines were identified at chorionic villus sampling (CVS), 46,XX and 47,XY,+9. Amniocentesis gave a 46,+XX result, and a normal girl was subsequently born. Analysis of a fibrotic area of the placenta gave the same two karyotypes, 46,XX and 47,XY,+9 (Falik-Borenstein et al., 1994). The likely explanation is that a 47,XY,+9 co-conceptus died, and the fibrotic placental tissue was the only remnant.
Fetal Death in Utero, Perinatal Death
Concerning mid-trimester loss, which, coming between miscarriage and stillbirth, may be referred to as “fetal death in utero,” chromosome abnormality may be present in about a half, although at this stage it is the “viable” rather than the invariably lethal aneuploidies that come to light (Howarth et al., 2002). Similarly, a fraction of pregnancies going through to term or at least to the third trimester, but with the baby stillborn or dying in the early neonatal period (perinatal death), are due to chromosomal abnormality, whether full or partial aneuploidy; a common representative of this group is trisomy 18. Among liveborn babies, only 1 in 250 hasan unbalanced chromosome abnormality on standard karyotyping (Table 1-3). Thus, there has been a very effective selection against those conceptions that were abnormal (Fig. 21-4).
Recurrent Abortion
Do some couples who are themselves karyotypically normal miscarry because of a predisposition to produce aneuploid conceptions? Ulm (1999) reviews theoretical possibilities through which such a risk might apply: recessive genes, parental chromosome abnormality, gonadal mosaicism, satellite association, and maternal age. A case exists also for the basis of some recurrent aneuploid miscarriages being simply one of randomness, in the setting of a high background rate of aneuploidy in humans, with increasing maternal age being the only clear predisposing factor. A common event is common, and not uncommonly it may happen more than once.
Addressing this question, Robinson et al. (2001) provided data from a study group of 54 couples having had 2–4 spontaneous known aneuploid/polyploid abortions, 122 abortions in total. The mean maternal age was 38 years. The distribution and frequency of karyotypes essentially did not differ from the data of a comparison group of 307 single miscarriage cases. Upon stratification for maternal age, a similarity of proportions of karyotypes in recurrent versus single abortions within each age group held up, as it did with an earlier analysis by Hassold et al. (1993). Trisomies in the Robinson study accounted for 80%, and the four most commonly represented abnormalities were trisomies 15, 16, and 21 (14%, 15%, and 8% respectively) and triploidy (13%). A few couples had a repeat of the same trisomy, but these observations were not made more often than might have been expected by chance. In those cases in which the parental origin of the trisomy could be determined, the great majority, around 90%, were maternal. And yet, although the supposition of randomness and maternal age as the important contributors to recurrent miscarriage is given considerable weight by these data, to the contrary, a small effect indicating a predisposition to aneuploidy recurrence actually was discernible in Warburton's review of a very large body of prenatal diagnostic data (see Table 16-2).
|
|
|
Figure 21-3. A triploid (69,XXY) embryo. The face has no landmarks other than eyes and a single opening. The anterior trunk is open, with the heart and liver visible. Spontaneous abortion occurred at 18 weeks gestation, but the length is that of 6 to 7 weeks gestation. The disrupted tissue at the neck was the site of biopsy for the cytogenetic analysis. |
Some further clarification emerged from studies from Pellicer's group, in which the end point of observation was FISH analysis of the embryo at PGD, from a patient population having previously experienced multiple miscarriage. In an initial report, nine couples were studied, the women of ages 30–35 years who had had on average four previous miscarriages and had presented for IVF with ICSI and PGD as a means to choose chromosomally normal embryos (Pellicer et al., 1999). Of 72 embryos biopsied at day 3 for karyotyping, 66 were analyzable, and in 35 (53%) an abnormality was found using FISH for chromosomes 13, 16, 18, 21, 22, X, and Y. A control group of similar age (10 women, ages 30–36) had an abnormality rate of only 19%, but a group of six older women (ages 37–41) had a comparable degree of abnormality (46%) to the miscarrying patients. (The controls were drawn from women of presumed normal fertility presenting for PGD for Mendelian indications.) The chromosomal abnormalities included nullisomies, monosomies, trisomies, and tetrasomies for some of the five autosomes and for 45,X, 47,XXX and 47,XYY. As with the observations of Robinson et al. above, there was no apparent tendency for the same aneuploidy to recur. A tentative conclusion is that a fraction of recurrent abortion may be explained by a predisposition to recurrent aneuploidy, and this effect is more apparent in younger women. Further work by Pellicer's group (Rubio et al., 2003) has continued to support these findings, and the figures set out in Table 21-1 show notably higher frequencies of aneuploidies for the tested chromosomes in embryos of those women suffering recurrent miscarriage.
|
|
||||||||||||||||||||||||||||||||||||||||||||
|
Figure 21-4. The frequency of chromosome abnormalities at gametogenesis and during pregnancy, demonstrating the effectiveness of selection against aneuploid states. The figures given for gametes through to embryos are very approximate, and considerable individual variation is probable. The oocyte percentage varies considerably according to maternal age. The day-3 embryo percentage, drawn from IVF data, may exaggerate the true picture in vivo. The figures for fetal and newborn abnormality are quite accurate. |
||||||||||||||||||||||||||||||||||||||||||||
|
Table 21.1. Frequencies of Aneuploidies for Certain Chromosomes in a Cohort of Women Having Had Recurrent Miscarriage (467–559 Embryos Analyzed), Compared with a Presumed Normally Fertile Cohort (104–202 Embryos) |
||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||
It may seem counter-intuitive, but for couples who have suffered multiple miscarriage, an aneuploid abortion indicates a better chance for a normal live birth in a subsequent pregnancy than when a miscarriage is euploid (Ogasawara et al., 2000; Carp et al., 2001). Aneuploidy of the abortus is less often observed in couples who have had a large number of miscarriages, sometimes into double figures, than among those who have had fewer. The probable reason is that a chromosomally normal miscarriage reflects an underlying maternal factor that will apply to all pregnancies, whereas aneuploidy at least offers the hope that better fortune might attend the next ovulation.
For the small group of people who are heterozygous for a chromosomal rearrangement, pregnancy loss may, of course, occur with a much higher frequency, and this briefly stated fact is the basis of much of what is written in this book. In about 0.5% of couples who have had more than two spontaneous abortions, and in a purely obstetric-gynecologic referral base in which a karyotype is done as a first-up investigation, one or the other member of the couple is a translocation or inversion carrier (Simpson et al., 1989). Somewhat higher fractions (2%–5%) reported from other studies may reflect a degree of patient selection. These chromosomal rearrangements are typically of sufficient size to be readily detectable at standard karyotyping; subtle subtelomeric translocations appear not to have a particular association with recurrent miscarriage (Benzacken et al., 2002; Fan and Zhang, 2002).2 An actual example of a chromosomally unbalanced pregnancy leading to spontaneous abortion in the first trimester is shown in Figure 4-10; this was the third miscarriage out of three pregnancies for the couple, the wife being a t(13;16) carrier.
A Rare Complexity
The “jumping translocation” describes the circumstance of one chromosome being the donor of a segment that translocates to more than one other chromosome (see p. 271). This process was proposed as the basis of miscarriage in two cases studied by Levy et al. (2000). In one of these, for example, the conceptus was initially 46,XX,der(15)t(1;15)(q10;q10). A second line arose, with the 1q part of the der(15) being replaced by an additional chromosome 15 which then generated an i(15q), along with (presumably independently) trisomy 7. Five further lines then budded off, all with considerable degrees of imbalance, the pregnancy eventually terminating in first-trimester abortion.
Cytogenetics of Infertility
Infertility is defined as the inability to achieve conception, or the inability to sustain a pregnancy through to live birth (the latter known also as “infecundity”). Certainly it is common, affecting about 15% of couples. It is worth emphasizing that infertility is to be seen in the context of the couple, not necessarily of the individuals separately. An oligospermic man may be fertile with a “superfertile” female partner, but not with a woman of average fertility, for example. Many causes of infertility exist that involve the male (Skakkebæk et al., 1994) or the female (Healy et al., 1994) partner; a fraction of these cases is presumed to be genetically determined (Layman, 2002), with demonstrable chromosomal causes being a minority. Sex chromosomal defects include XXY and XXY/XY in the male, typically presenting with azoospermia and occasionally severe oligospermia, and Turner syndrome and its variants in the female. The common Yq microdeletion is dealt with below. The XX male and XY female are rare (Chapter 18). Autosomal abnormalities are infrequently seen as a cause of infertility. The reciprocal translocation (especially when an acrocentric is involved) and the inversion may be associated, though infrequently, with severe hypospermatogenesis and moderate to severe oligospermia (Chapters 4 and 8). Robertsonian translocations are occasionally associated with infertility in the male or, less often, the female (Chapter 6). Translocation between a sex chromosome and an autosome is a rarely identified cause of infertility (Chapter 5). Complex rearrangements (Chapter 11) and rings (Chapter 10) typically present an insurmountable obstacle to cell division in the spermatocyte, resulting in azoospermia; oogenesis is apparently more robust.
The frequency of karyotypic abnormality in couples with infertility depends considerably upon the criteria of ascertainment, and quite wide ranges of figures have been produced. Couples presenting to ICSI programs might be supposed to manifest a male factor infertility; but van der Ven et al. (1998) were surprised to discover that female partners had about as many chromosomal abnormalities (X aneuploidy, reciprocal and Robertsonian translocations, inversions) as did the males in a series of 305 couples presenting for ICSI; the experience of Meschede et al. (1998a) was not dissimilar. A large French study (Gekas et al., 2001) brought together all the ICSI programs in France over a 3-year period and included some 3208 individuals, 2196 men and 1012 women, who had come forward as candidate couples for ICSI. Gynecologic causes of infertility had been excluded. Each individual had at least 20 metaphases examined. Sex chromosome mosaicism at a level of <10% was categorized as minor, and normal variant chromosomes were (naturally) ignored. In the men, 6% showed a chromosomal abnormality, and in the women (excluding probably insignificant minor sex chromosome mosaicism), 2%. This latter figure may again seem surprising, since most of the male partners had abnormal andrology. Certainly, in the fraction of their patients with simple “fertilization failure,” cause unexplained, the proportion of women with a chromosome abnormality was greater, at 6.4%; but it remains that 2.6% of the women had an abnormal karyotype even when an explanation for infertility had been determined in the male partner. Some support comes from a study in which an age-matched control population had less low-level 45,X/46,XX mosaicism (45,X in at least two cells) than women presenting for an ICSI procedure (Morel et al., 2002b). The abnormalities in the Gekas study included numerical and structural sex chromosome abnormalities, reciprocal and Robertsonian translocations, inversions, and other structural abnormality (Table 21-2). This group compared their own data with 10 other similar series, and their figures of about 6% and 5% for male and female karyotypic abnormality are close to the averages of about 5% and 4%, respectively. The figure is higher (16%) in men presenting with azoospermia. Considering just translocations, and in relation to the nature of the infertility, Stern et al. (1999) noted the rate of balanced rcp and rob carriers to be 3% in 219 couples (both partners tested) who had failed more than 10 embryo transfers, and 9% in 130 couples who had three or more consecutive first trimester abortions. In one couple from the latter group, both were translocation carriers (Fig. 11-6).
Factors in the Female
Fertility in the 46,XX female begins to fall in the mid-thirties, paralleling both the increase in risk for trisomy and an increasing failure in IVF implantation from this age (Spandorfer et al., 2000). An important age-related factor may be a decline in the functional competence of the meiotic spindle, which compromises chromosomal distribution and leads to the generation of aneuploid gametes (p. 32).
Various sex chromosomal abnormal states, mostly mosaic and containing a 45,X cell line, account for, or are at least associated with, a number of cases of female infertility; autosomal abnormalities are less frequent. In some, the infertility is primary (there has never been a period of fertility), and in others it is secondary (following a previous fertile period). In a Malaysian study, Ten et al. (1990) karyotyped 117 women with primary amenorrhea who had previously been investigated for other causes, and one-third had a sex chromosome abnormality. They were classified as follows: X aneuploidies (8%), X structural abnormalities (7%), presence of a Y (14%), and presence of a gonosomal marker chromosome (2%). Six women were mosaic, all having a 45,X cell line. Secondary infertility may be due to premature ovarian failure, and Devi and Benn (1999) studied 30 women with unexplained secondary amenorrhea under the age of 40 years. Four (13%) had chromosomal abnormalities: an Xq isochromosome, Turner syndrome mosaicism (45,X/46,XX), an X-Y translocation, and an X-autosome translocation. A Robertsonian translocation may be an uncommon association (Kawano et al., 1998), as may the 47,XXX state (Itu et al., 1990). Kuo (2002) systematically studied 1010 couples having had recurrent abortion and found two women with low-grade trisomy 21 mosaicism, each of whom had had at least one trisomy 21 abortus. In women who have suffered recurrent failure of IVF treatment, in spite of having had a considerable number of embryo transfers, the rate of non-mosaic autosomal translocation carriers in one series of 65 women who had >15 failed transfers was 8%, two being sisters with the same translocation (Raziel et al., 2002). (Compare with the 3% translocation figure based upon rather larger numbers, and testing both of the couple, in Stern et al., 1999, noted above.)
|
Table 21.2. Chromosome Abnormalities in a Series of Candidate Couples for Intracytoplasmic Sperm Injection |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||
As mentioned above, a small fraction of female partners of oligospermic men are found to have low-level sex chromosomal mosaicism. According to the experience of Sonntag et al. (2001), this does not compromise the course and outcome of ICSI.
Failure of the meiotic apparatus, with no formation of the first polar body, may be a rare cause of female infertility and be due to an autosomal recessive gene (Neal et al., 2002; Schmiady and Neitzel, 2002). A particular association of premature ovarian failure with the fragile X premutation heterozygote is mentioned on p. 223.
Factors in the Male
Fertility is not necessarily synonymous with normospermia and, as mentioned above, a man with oligospermia3 may be fertile with a woman of “superfertility” (Krausz and McElreavey, 2001).
Nevertheless, much couple infertility is associated with diminished sperm production in the male, and a fraction of this is associated with an abnormal karyotype (Table 21-2) (Gekas et al., 2001). In men presenting with azoospermia or oligospermia, numerical and structural gonosomal abnormalities (mostly XXY and Y rearrangements) and structural autosomal abnormalities (mostly reciprocal and Robertsonian translocations) are identified in 3%–13% (De Braekeleer and Dao, 1991; Meschede et al., 1998a; Stuppia et al., 1998; van der Ven et al., 1998; Causio et al., 1999; Dohle et al., 2002). Rare observations include the Y;autosome translocation (p. 112), the “XX male” (p. 297), and the small isodicentric 15 (Eggermann et al., 2002).
Translocation Carriers. In the setting of a balanced rearrangement, gametogenesis in the male heterozygote appears more vulnerable than in the female to the complexities imposed by a chromosomal abnormality, and infertility occasionally results. An important element in this male vulnerability may be the integrity at meiosis of the X-Y bivalent, synapsing and recombining at the pseudoautosomal regions at the tips of Xp and Yp (Hale, 1994). Unpaired autosomal segments, particularly of acrocentric autosomes, might disturb this integrity, leading to disruption of spermatogenesis (Guichaoua et al., 1990). Another element may be impaired synapsis of homologous segments in the normal and the rearranged chromosomes, which of itself prevents further progress in gametogenesis, and spermatogenesis may be more sensitive to this obstacle than oogenesis (Hale, 1994). A compromised testicular environment due to the presence of a translocation may also of itself predispose to the production of diploid sperm (Egozcue et al., 2000).
Yq Microdeletions. An important fraction of male infertility resides in a Y chromosome microdeletion, with particular reference to the AZF (azoospermia factor) regions in Yq11, in which certain spermatogenesis factors have their loci (Fig. 5-1). The fraction varies according to patient selection, and when other causes of oligospermia/azoospermia have been excluded, the proportion due to AZF deletion reaches 10%–20%. There is a large body of literature on this subject, reviewed in Foresta et al. (2001). While the initial discovery was made by cytogeneticists (Tiepolo and Zuffardi, 1976), these Y-deletions are mostly not detectable cytogenetically and are routinely analyzed using molecular methodology. There are three main AZF regions, a, b, and c, and deletions in one or more region can impair spermatogenesis, or lead to its complete failure. The most commonly seen deletion involves the AZFc region, in Yq11.23, the causative mechanism being similar to that described in Chapter 17, with inappropriate apposition of “duplicons” (Kuroda-Kawaguchi et al., 2001). AZFc contains the DAZ(deleted in azoospermia) multi-gene family, the products of these duplicated loci being important (but not necessarily crucial) spermatogenesis factors. As a rule, AZFa or AZFb deletions are more severe in their effects than AZFc. Different causes for disordered spermatogenesis may coexist in an individual, and Jaruzelska et al. (2001) point to the need for cytogenetic studies, bringing to attention cases in which 45,X/46,XY mosaicism may have had an additive effect along with an AZFc deletion.
A male child conceived from a father with a constitutional Yq microdeletion would very likely have similar infertility (although, as noted below, some men with a Yq deletion may retain fertility). Komori et al. (2001) formally demonstrated that a man with a del(Yq) on blood karyotyping could transmit the deletion, in showing the deletion actually to be present in sperm, as did de Vries et al. (2001c) in all of seven infertile men with deletion of the DAZ gene cluster. The observation of the same deletion in the sons of men who had conceived via ICSI confirms the reality of vertical transmission (Cram et al., 2000). The reduction in fertility may be relative, at least for AZFc deletions, and at a younger age, and perhaps with a partner of “excellent” fertility, a man with a deletion may father children with no obvious difficulty (Krausz and McElreavey, 2001). Chang et al. (1999) report the example of an azoospermic 63-year-old man with a DAZ deletion who had been fertile in his younger days, having had five children from when he was 25 to 38 years of age. His four sons all had the deletion, and the three of them tested (ages 24–37) were oligospermic or azoospermic.
Normal Karyotype. Among infertile men whose karyotype is normal and whose sperm count is abnormally low, there is an increase in the sperm aneuploidy/diploidy rate, with the sex chromosomes being the most prone to exhibit disomy (Shi and Martin, 2001). This effect is more apparent in those men with severe oligospermia and in those aged 40 and over (Asada et al., 2000). Vegetti et al. (2000) assessed the influence of sperm count and motility and showed that both these indices correlate with the frequency of sperm disomy, testing chromosomes 13, 18, 21, X, and Y. The observations at testicular biopsy in men with severe oligoasthenozoospermia support this interpretation, with univalents or oligochiasmatic and achiasmatic bivalents being frequently seen (Egozcue et al., 2000). As for men with actual azoospermia, in whom sperm can be obtained only by testicular or epididymal biopsy or aspiration, the early data on fairly small numbers also show elevated disomy rates for some autosomes and the X and Y chromosomes (Martin et al., 2000b; Burrello et al., 2002; Mateizel et al., 2002; Palermo et al., 2002).
Considerations Relating to in Vitro Fertilization
Treatment with Intracytoplasmic Sperm Injection. In theory, given the increased level of sperm aneuploidy in men with oligospermia, an increase might be expected in the numbers of aneuploid conceptions from ICSI. It is true that sperm abnormality rates correlate inversely with ICSI pregnancy success (Calogero et al., 2001). An aneuploid sperm does not necessarily look abnormal to the embryologist, and so choosing a normal-looking spermatozoon will not guarantee euploidy (Ryu et al., 2001). Some studies have shown astonishingly high fractions of sperm chromosomal abnormality, especially with severe oligospermia, but other studies have not, and the reasons for the differences are controversial (Pang et al., 1999; Griffin et al., 2003; Silber et al., 2003). Equally, it is possible that ICSI failure in severely oligospermic men may be due to other than chromosomal sperm-related factors.
Looking at the outcomes in embryos, no obvious chromosomal differences exist between standard IVF and ICSI-IVF (Munné et al., 1998b). A comparison may be made among those couples having ICSI-IVF, according to the degree of the paternal spermatogenic defect. Silber et al. (2003) studied the differences between men with oligospermia who used ICSI with ejaculated sperm, and those with nonobstructive azoospermia needing testicular sperm extraction (TESE). The oligospermic men had embryos with a certain distribution of chromosomal anomalies, according to FISH analysis for six or eight chromosomes (42% normal, 26% aneuploid, 27% mosaic, 5% polyploid). The aneuploidy rate was actually somewhat less in embryos from TESE, at 17% (the mean maternal age being a little younger in this group). But the striking difference in the TESE group was in the large fraction of chaotic mosaic embryos—53%. These authors make the case that mosaicism of the embryo may therefore reflect male factors, noting the important role of the spermatozoon-derived centrosome in controlling the first few mitotic divisions of the embryo.
In terms of outcomes in the newborn, the interpretations are, for the most part, reassuring. In the very large follow-up material of Bonduelle et al. (2002a), comprising nearly 3000 children, these authors discerned no significant differences between ICSI and standard IVF pregnancies with respect to malformations in general and most types of chromosome abnormality (prenatal diagnosis having been done in about half of these pregnancies) in particular. The level of autosomal trisomies was no higher than expected on the basis of the maternal age profile.
In two categories, however, a small increase was observed. A further report from this group (Bonduelle et al., 2002b) documented an 11year experience comprising 1586 ICSI pregnancies in which prenatal diagnosis had been done, and de novo structural aberrations and sex chromosome anomalies were seen in 1.6% (cf. 0.5% in the general population). These comprised 10 sex chromosomal (XXY, XXX, XYY, and X mosaicisms) and 15 autosomal anomalies; for the latter group eight were numerical (mostly trisomies 18 and 21) and seven were structural (mostly apparently balanced translocations). As Bonduelle et al. emphasize, all the gonosomal cases involved the father being severely oligospermic, and this male factor, rather than the ICSI procedure itself, may have been the basis for the increase; the abnormality rate (gonosomal and autosomal) in children of men with sperm counts >20 million/ml was only 0.24%. In similar work comparing the rates according to semen examination, with respect to chromosomally normal parents, Gianaroli et al. (2000) noted that aneuploidy occurred more often in those with severe indices, again implicating the parental factor rather than the ICSI procedure. We may conclude that an additional risk of chromosomal abnormality for children conceived from ICSI is small but not negligible, about 1% above baseline, and prenatal diagnosis may appropriately be offered, more especially in the case of men with very low sperm counts.
Karyotyping of the oligospermic man should be done before proceeding to ICSI (Bonduelle et al., 2002b). Bofinger et al. (1999) provided ICSI to a couple in which the husband had severe oligospermia and the wife was of older childbearing age. At amniocentesis, on the grounds of the mother's age, a 45,X/46,X,r(Y) chromosome constitution was discovered, and belatedly, the same karyotype was found in the father. The experience of Veld et al. (1997) is equally telling, concerning two men who, having suffered reproductive misfortune following ICSI, turned out to have a Robertsonian translocation.
Epigenetic Effects. Fertilization in vitro occurs in an artificial environment. It may be that the delicate interplay in the epigenetic reprogramming of chromosomes according to parent of origin is vulnerable in this artificial setting (De Rycke et al., 2002), and the question arises whether children born from IVF could be at increased risk for an imprinting disorder. This does indeed appear to be the case with respect to Beckwith-Wiedemann syndrome due to epigenetic error (p. 330), perhaps more so in the case of ICSI having been employed, and the risk may be several-fold that of the general population (DeBaun et al., 2003; Maher et al., 2003). The case for Angelman syndrome is more tenuous, with essentially anecdotal reports by early 2003 of just three affected children born from ICSI conceptions (Cox et al., 2002a; Ørstavik et al., 2003). Nevertheless, the fact that these three cases were all in the category of epigenetic error raises a valid concern, given the rarity of this type. Equally, the statistical weight of the thousands of unaffected IVF children is not to be discounted, and this points to a low level of risk, if risk there be.
A Rare Complexity
A syndrome of infertility associated with “largeheaded sperm” is described in Benzacken et al. (2001a). Polyploidy may be the explanation here. Benzacken et al. studied infertile brothers with oligoasthenoteratozoospermia, in whom half of all sperm had the large-head phenotype. On FISH analysis (X, Y, 18) in one brother all sperm cells were diploid or polyploid (3n, 4n, and >4n). The basic fault may lie in a failure of the cell to cleave at the two meiotic telophases and, with brothers affected, a genetic cause may reasonably be presumed. Similar cases were reported by Devillard et al. (2002) and Lewis-Jones et al. (2003). In the latter study, three men had complete teratozoospermia (all sperm with abnormal forms, such as double heads, large heads, multiple tails), and the frequency of chromosomal abnormality was very high, up to 100%.
Cytogenetics of Hydatidiform Mole
Hydatidiform mole is an abnormal pregnancy that can be considered, in a sense, a male chromosomal disorder. The pathology and genetics are reviewed in Petignat et al. (2003). Typically, there is either a completely paternal karyotypic origin (two haploid paternal sets, 2n = 46) or an additional male haploid set (two paternal and one maternal haploid sets, 3n = 69). The chorionic villi undergo a degenerative change, forming fluid-filled sacs (hence hydatidiform; mole means mass). The characteristic appearance has long been recognized (Fig. 21-5). The phenotype is marked (complete mole) when the genetic origin is completely paternal, and attenuated (partial mole) in the presence of a maternal haploid contribution. A specific genetic cause in rare cases may be maternal homozygosity for a predisposing gene.
|
|
|
Figure 21-5. The appearance of hydatidiform mole, probably the first recorded depiction (Baillie, 1799). |
Complete Mole
The complete form usually has the karyotype 46,XX looking, at first, like a normal female karyotype. In most cases this is due to a doubling of the chromosomal complement of a (normal) 23,X sperm, while a minority are dispermic (Lawler et al., 1991). Consequently, the mole's nuclear genome is of entirely paternal origin: a total uniparental paternal disomy (uniparental diploidy). Moles due to doubling of a sperm complement are entirely homozygous— in other words, they have a complete uniparental isodisomy. The aberrant imprinting accompanying this state is the presumed basis for the severe derangement of fetal–placental development (Ariel et al., 1994). Perhaps the sperm's doubling represents an attempt, albeit a forlorn one, to correct the situation when it enters an “effectively empty egg,” that is, an ovum in which the integrity of the nuclear structure has broken down. This nuclear failure occurs more often at the beginning and end of reproductive life in the female: complete mole is more common in the early teenager and in women in their 40s (Bagshawe and Lawler, 1982). Some diandric triploid molar pregnancies, when presenting early in pregnancy (before 8 weeks), may present a complete molar phenotype, rather than the partial mole usually observed (see below) (Zaragoza et al., 2000).4
The complete mole usually presents with vaginal bleeding, and ultrasonography shows a vesicular pattern of the placenta, reflecting the swollen villi. There is a widespread and marked hyperplasia of the trophoblast. When diagnosis is made early and curettage performed, some nonchorionic elements (yolk sac, capillaries, amnion) may be identifiable, as well as embryonic parts (Zaragoza et al., 1997; Petignat et al., 2003). The incidence of complete mole is about 1 in 2000 diagnosed pregnancies, although regional and ethnic variations exist (Jeffers et al., 1993; Palmer, 1994). In Japan, the incidence has been falling, from about 1 in 2500 in the 1970s to 1 in 1500 by the late 1990s (Matsui et al., 2003).
There is a small but significant risk of recurrence, and there are several recorded observations of women having had three or more complete moles. Recurrences can be of either kind of mole, complete or partial. For example, Kircheisen and Schroeder-Kurth (1991) report three sisters having had seven molar pregnancies between them, both complete and partial, and none with a normal child. Repeating mole is unusual in being diploid and biparental. Maternal homozygosity for a recessive gene may be the basis, and the effect may be compromise of the normal process of establishing a maternal imprint in an otherwise chromosomally normal conception (Helwani et al 1999; Moglabey et al., 1999; Fisher et al., 2000; Judson et al., 2002).
Partial Mole
An additional paternal haploid chromosome set is the basis of most cases of partial mole. This is triploidy, 69,XXX or 69,XXY, which typically is the result of a normal ovum fertilized either with two sperm (dispermy) or with a diploid sperm (Zaragoza et al., 2000) (and see p. 258). Recurrences are on record, and a possible explanation is a genetically determined weakness in the zona pellucida of the ovum, which should act (the zona reaction) to prevent more than one sperm penetrating. The double paternal contribution is referred to as “diandry” (type I triploidy). Partial moles typically present as threatened, incomplete, or missed abortion, during the late first or early second trimester, the mean at 12 weeks. There is hydatidiform change of some villi, and the placenta is abnormally large. It is underdiagnosed and may occur in as many as 1 in 700 pregnancies (a figure Jeffers et al. [1993] derive from a review of all 2251 spontaneous abortions occurring in the catchment population of a Dublin hospital over a 3-year-period, during which there were 19,457 recorded pregnancies). Fetal development in the very few cases proceeding far enough for this to be assessed is characterized by a relatively normal growth pattern (McFadden et al., 2002). If the triploidy is confined to the placenta, it is possible for the pregnancy to proceed successfully to the birth of a 46,N child (Sarno et al., 1993).
Uncommonly, partial mole has a normal diploid karyotype, with biparental inheritance. At least some of these cases may be due to the woman being homozygous for a recessive im-printing-control gene, as discussed above.
GENETIC COUNSELING
Recurrent Miscarriage
People who have had one or perhaps two miscarriages generally do not come to a genetic clinic and do not have cytogenetic analysis of the products of conception, nor would it be appropriate for them to do so. Their physician or obstetrician will have advised them that this loss is very likely part of the 15% or so of all pregnancies that miscarry, and the chance of a successful pregnancy in the future is good. But having had three miscarriages requires investigation. To use the jargon, such couples have had “multiple abortions” or “recurrent miscarriage” (or to put it in Latin, abortus habitualis). The usual gynecological investigations and a chromosome analysis of the couple should be done at this point. If a chromosomal rearrangement is identified, this is probably the underlying cause (Gadow et al., 1991); but the possibility of a fortuitous discovery is not to be discounted. The precise nature of the rearrangement (consult the appropriate chapter), the reproductive history of any others in the family who have it, and the presence or absence of gynecological pathology allow one to judge its role in the etiology of the abortions. In the case of recurrent abortion due to a parent being a translocation carrier, Munné et al. (2000b) report that PGD can very substantially reduce the incidence of abortion (p. 383), and if access to this technology is available, “translocation couples” may wish to consider this option.
The majority of couples will have a normal karyotype, 46,XX and 46,XY. In most centers, cytogenetic analysis of abortus material (an expensive and time-consuming procedure) is not routinely done, and so chromosomal normality or abnormality usually cannot be demonstrated. Some workers have argued that this policy should shift, and Stephenson et al. (2002) speak of “this unfortunate omission” compromising the management of couples presenting with recurrent miscarriage. Since the discovery of an aneuploidy can avoid the necessity for further investigation, they argue that routine karyotyping would actually be cost-effective and have the further benefit of helping couples understand and thus come to terms with their reproductive failure, as Sánchez et al. (1999) have also suggested. Jobanputra et al. (2002) propose that a FISH panel for chromosomes 13, 15, 16, 18, 21, 22, X, and Y applied to uncultured abortus tissue would enable a relatively inexpensive screen, with culture proceeding only on those with an apparently normal result. A miscarriage due to aneuploidy actually implies a better prognosis for a subsequent pregnancy than if the abortus is euploid. Even so, knowing that a series of miscarriages was due to a series of aneuploidies might be not be easy for couples to accommodate, and Robinson et al. (2001) comment that “counseling a couple who have experienced multiple spontaneous abortions with chromosomal abnormality is difficult, because the more losses the couple have experienced, the less likely either the involved couple or the physician will feel comfortable attributing the spontaneous abortions to just ‘bad luck’.” Yet, this may indeed be the true state of affairs, at least in many couples; for others, there may be a predisposition to aneuploidy (see Tables 16-2 and 21-1). In vitro fertilization with PGD might benefit some women who have had several miscarriages.
Fetal Death in Utero
Pregnancy loss in mid-trimester is less frequent than in the first trimester, and some may thus see a lower threshold for karyotyping the products of conception. In this case, Howarth et al. (2002) propose offering CVS or amniocentesis, rather than attempting culture of fetal tissue post-delivery, in order to improve the chances of getting a definite chromosomal result.
Women of Older Childbearing Age
Maternal age is an important factor in recurrent miscarriage. The meiotic apparatus of the oocyte deteriorates with age; returning to Figure 2-7, the reader can marvel at the disposition of the chromosomes in the eggs of the older women and appreciate how perfectly plausible it is that egg after egg could be aneuploid. The evidence from IVF points to a sharply increasing likelihood for aneuploid conception in women of older childbearing age (Verlinsky et al., 1999). Some instances of reproductive “bad luck” can seem like very bad luck, but, as mentioned above, this may still be the true explanation. Ulm (1999) describes her experience dealing with two couples having had several losses, expressing her and their frustration at not being able to provide a satisfactory explanation or to offer a precise recurrence risk. In the first case, following the birth of a normal child at the mother's age of 34, the next four pregnancies, in her late thirties, miscarried or were terminated. The first was not tested, and the next three were trisomic 22, 21, and 14, respectively. The sixth pregnancy, at age 40, tested 47,XX,+15 on direct CVS but normal on long-term CVS and at amniocentesis, with a biparental chromosome 15 pattern. The infant was developing normally as an 8-month-old. The second case concerned a woman who had three miscarriages at age 36–37 years which karyotyped as trisomies 22, 16, and 21, respectively.
We are not much able to ease Ulm's frustration, other than to point again to the high background rate of human aneuploidy and to the major influence of maternal age. It is not easy to distinguish, just on the basis of history, which older couples are destined to experience reproductive misfortune. The counselor needs to recognize that many in this situation will go on to have successful pregnancies, but should retain some reservation that the risk may be significant, and perhaps substantial, for women who are getting into their late thirties or forties. For some women with the concern that their reproductive years may be limited, IVF with PGD might warrant consideration to ensure that only a normal embryo (if available) is being transferred (Pellicer et al., 1999).
Infertility
Infertility is common and, in Western countries at least, about 15% of couples wishing to have a child are affected (Foresta et al., 2002). Intrinsic fertility cannot be restored in men with persistent azoospermia associated with seminiferous tubule failure or in women who have had ovarian failure. The counselor will need to understand how disappointing and indeed devastating this may be to some couples (sometimes one of them more than the other) and be prepared to deal with this. Those for whom assisted reproductive technology may offer hope need to be made aware that this is not necessarily an easy path and that success cannot be guaranteed.
Among the catalog of investigative tests available, a karyotype is well up on the list. A grouping of experts from the Italian professional community has addressed the question of what tests should be done and when (Foresta et al., 2002). They propose that karyotyping be done routinely in men with azoospermia and oligospermia, and in the United Kingdom, karyotyping of men presenting for ICSI is “commonplace” (Griffin et al., 2003). Yq microdeletions should be checked for in men with nonobstructive azoospermia and severe oligospermia, but this is unlikely to be the cause in lesser degrees of oligospermia (>5–10 million/ml) (Foresta et al., 2002; Quilter et al., 2003). At present, sperm karyotyping is a discretionary investigation (if available). Karyotyping should be routine in women presenting with primary ovarian dysfunction or recurrent miscarriage. These investigators propose that fragile X premutation analysis be considered in women with premature ovarian failure; informed consent is required given the other implications of making this diagnosis.
Parental Chromosomes Abnormal
If a chromosomal defect is discovered, this at least provides an explanation for the infertility and (according to the exact nature of the defect) may prevent the disappointment of undergoing pointless further investigation. In some patients artificial reproductive technology may enable a normal/balanced gamete to be identified and retrieved and used at IVF. Where this is impossible, artificial insemination or IVF using donor gametes offers an entrée to parenthood and may enable one of the couple to be a genetic parent.
Women. In women with a sex chromosomal abnormality having oocyte donation, endocrinological management may be necessary to “prime” the reproductive tract (Devroey et al., 1988). In some cases the woman's own mother, with whom of course she shares half her genes, has been the donor. Artificially stimulated ovulation has been attempted in one case of a chromosomal state associated with secondary amenorrhea. Causio et al. (2000) describe a 29-year-old woman with a 46,X,t(X;16) karyotype who had undergone premature ovarian failure, and in whom ovulation was then achieved by treatment with gonadotrophin releasing hormone and follicle stimulating hormone (GnRH and FSH). But no pregnancy resulted.
Men. In men with complete spermatogenic arrest, gamete donation may be considered. In those in whom the chromosome defect leads to oligospermia, rather than complete failure of spermatogenesis, IVF with ICSI is a possible means to achieve pregnancy, and PGD will often be appropriate. Otherwise, given the small increased risk for gonosomal aneuploidy following ICSI, a subsequent conventional prenatal diagnosis may appropriately be offered. Translocations and other rearrangements need to be assessed on their merits. An example concerning an oligospermic man with a t(Y;18) is considered on p. 113. A small (but growing) number of cases of fatherhood in men with Klinefelter syndrome have resulted from ICSI, including one pregnancy in which one of triplet fetuses was 47,XXY (see p. 198). Rare sex chromosome abnormalities are judged individually. For example, a sperm study in the case of a man with sex chromosomal mosaicism (45,X/47,XYY), which gave normal findings, was instrumental in a decision not to have preimplantation diagnosis following an ICSI conception (see p. 204) (Dale et al., 2002). In the case of a Yq microdeletion, couples choosing the option of IVF with ICSI should know that a male child would be predicted to have, very probably, the same type of infertility (Foresta et al., 2001). Some might consider having PGD to ensure having a daughter; although Kim et al. (1998)comment that “interestingly, after genetic counseling, the decision to proceed with ICSI for the overwhelming majority of couples remains unchanged.” Nap et al. (1999) assessed 28 such couples and interviewed the 10 counselors who had seen them in six clinics in The Netherlands and in Belgium. A considerable majority of couples (79%) chose to continue with plans for ICSI, with only a few choosing donor insemination (7%) or opting out altogether (14%).
Parental Chromosomes Normal
When the male has oligospermia and if IVF with ICSI is to be attempted, a very slightly increased risk for de novo structural aberration or gonosomal aneuploidy may be presumed, as discussed above. At present writing, sperm karyotyping is not routinely practiced as a basis for informing genetic counseling, although this is an area in which research is continuing and opinions are evolving (Griffin et al., 2003). It may be prudent to offer prenatal diagnosis for an ICSI-produced pregnancy. However, given the immense investment couples will have made to achieve an ICSI pregnancy, they may have reservations about proceeding to an invasive prenatal diagnostic procedure, even while being aware of a possibly increased genetic risk with ICSI. In this context, the data from Aytoz et al. (1998) offer reassurance. These Belgian workers compared outcomes in 576 ICSI pregnancies having amniocentesis (singletons) or CVS (twins), with 540 control ICSI singleton and twin pregnancies not having prenatal diagnosis. The fetal loss rates, and certain obstetric indices (preterm delivery, low birth weight) did not differ significantly between the two groups. Further, the data did not tend toward a greater risk for miscarriage in the prenatal diagnosis patients: the odds ratios for fetal loss were 0.86 (amniocentesis) and 0.47 (CVS) compared to the controls. Never-theless, in a German population, Meschede et al. (1998b) report that only 17% of a cohort of 107 women having undergone ICSI chose subsequent amniocentesis (or fetal blood sampling), the great majority preferring noninvasive ultrasonography or serum screening. This preference was more marked in those who had had genetic counseling prior to entering the ICSI program. In contrast, an Italian clinic recounted an opposite figure, with 86% choosing invasive prenatal diagnosis (and 100% choosing ultrasound screening); these workers could see “no logical explanation for the great difference” (Monni et al., 1999).
Recent reports about an increased risk for an imprinting disorder (Beckwith-Wiedemann syndrome very probably; Angelman syndrome possibly) in IVF-conceived children raise a question of whether it is prudent to advise couples of these concerns, according to the accumulating understanding of these risks. Counselors working in the IVF clinic will want to maintain a watching brief.
Chromosomally abnormal children following pregnancy by donor insemination. If a pregnancy achieved through gamete donation turns out to be chromosomally abnormal, should that donor continue to be used? Kuller et al. (2001) surveyed a number of reproductive endocrinologists and obstetrical geneticists to determine the current practice, with particular reference to trisomy 21 and monosomy X. It was clear that no consistent policy was followed. For chromosomal abnormalities generally regarded as being sporadic (or where any predisposition might reside in the recipient rather than the donor), it would seem unnecessary to remove that donor from the panel.
In Vitro Fertilization and Multiple Pregnancy. Twinning and multiple pregnancies are common in IVF, for the obvious reason that often more than one embryo is transferred following IVF, this being a standard policy to improve the chances for a successful implantation. These twins will, naturally, be dizygous. The conservative number of transfers is two (the ideal is one), so that if both embryos do implant, no more than twins will result (unless a single embryo might go on to produce monozygous twins5). Some clinics transfer more, sometimes for the simple economic reason that if couples can only afford one IVF cycle and transfer, using three (or even more) embryos increases the chance of pregnancy. The disadvantage is, of course, that if most or all of the embryos implant, a high multiple pregnancy results. Whatever the risk for aneuploidy,6 that risk will apply to each embryo individually, thus increasing the overall risk that at least one might be chromosomally abnormal. If both abnormal and normal fetuses are present and diagnosed at subsequent amniocentesis, selective feticide of the chromosomally abnormal fetus may be chosen. For a lethal aneuploidy (trisomy 13, trisomy 18), the parents may opt to continue the pregnancy in the expectation that the abnormal fetus will die (Sebire et al., 1997). One of the claims made for PGD is that single embryo transfers can be done with a better expectation of success.
Hydatidiform Mole
Berkowitz et al. (1994) report a 27-year experience in following 1205 women having had a complete mole and 149 having had a partial mole (not distinguishing between diandric and digynic partial moles). Overall, the risk figure for recurrence is 1%–1.5%, and recurrence can be either of the same or of the other type. Ultrasonographic surveillance is advisable in a future pregnancy. In Berkowitz et al.'s (1994) series, 24 women having previously had two molar pregnancies had yet another, for a risk figure of 20% to have a third mole. A few of the single cases, but probably most of the multiple recurrences, may be due to maternal homozygosity for an autosomal recessive predisposing gene (Judson et al., 2002). These repeating cases typically show biparental inheritance, in contradistinction to the androgenetic basis of the majority of moles. Fisher et al. (2000) suggest that parental origin is worth establishing in those couples considering IVF with ICSI as a means to diminish the risk in a subsequent pregnancy, since such an approach would be futile if the mole had been biparental.
Hydatidiform mole may exist as a twin pregnancy, along with a normal co-twin (Petignat et al., 2003). Most such pregnancies (if not otherwise terminated at the time of first-trimester diagnosis) will end in second-trimester loss, with death in utero of the normal fetus. But 40% will have a successful outcome, with live birth of the normal twin at ≥32 weeks gestation. There seems to be only a fairly small chance of serious obstetric complication in these women (Sebire et al., 2002).
A major aspect of management is that the mole may undergo neoplastic transformation (gestational trophoblastic disease). With complete mole, the risk for the development of invasive mole is about 15%, and for the more dangerous choriocarcinoma, it is 3%. The risks are much less with partial mole, the respective figures being 0.5% and 0.1% (Seckl et al., 2000).
Notes
1. Strictly speaking, in utero life is divided into three periods: pre-embryonic (the first 2 weeks), up to formation of the primitive streak; embryonic (to the end of the 8th week), when the body form and organs are constructed; and fetal (from 8 weeks to term), characterized by growth and changes in proportion rather than the appearance of new features. Often the word fetal is used loosely to refer to the entire period, and in IVF parlance (and in the present discussion) the word embryo is routinely applied in reference to the conceptus in the early cleavage stage during the first few days. Conceptus, in theory, applies to any stage, but it generally refers to early pregnancy. The conceptus at the one-cell stage—the fertilized egg—is the zygote.
2. An intriguing report of what might be termed acrocentric promiscuity awaits further study. In a cohort of fifty miscarrying couples, Cockwell et al. (2003)made an unexpected discovery of rearrangement within the centromeres and pericentromeric regions of acrocentric chromosomes in four individuals, and they suggested that this might predispose to malsegregation with consequential whole chromosome aneuploidy.
3. Oligospermia is defined as a sperm count of 20 mil-lion/ml. Oligoasthenoteratozoospermia includes the observations of poor motility (astheno) and an increased fraction of abnormal forms (terato). Severe oligospermia is a count of <2 million/ml, moderate oligospermia is 2–5 mil-lion/ml, and the mild is from 5–20 million/ml. Azoospermia is absence of sperm. A distinction is to be made between obstructive and nonobstructive azoospermia; in the latter, the primary fault is a severe defect of spermatogenesis.
4. A unique case is on record of confined placental mosaicism for molar and normal tissue, the infant being normal; photographs of the placenta give an obvious visual illustration of the mosaicism (Makrydimas et al., 2002). An androgenetic lineage arose postzygotically from a normal 46,XX conception.
5. Monozygous twinning is slightly but definitely increased in IVF pregnancies, and this may be due to an immaturity of the zona pellucida. The process of ovulation stimulation will, naturally, bring to the fore some ova that would otherwise not have been quite ready to be released. Some particular quality of the “hatching hole” in the zona of the resulting embryo may influence the blastocyst during its extrusion, in such a way that the inner cell mass is induced to split into two (Da Costa et al., 2001; Sheen et al., 2001).
6. If donor eggs are used, it is the age of the donating woman that counts in determining the age-related aneuploidy risk.
